Cover Image

PAPERBACK
$125.25



View/Hide Left Panel
Click for next page ( 167


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 166
PREDICTING THE DISPERSION AND FATE OF CONTAMINATED MARINE SEDIMENTS Y. Peter Sheng University of Florida ABSTRACT In order to select the proper remedial action agement strategy to clean up a contaminated marine is essential to be able to predict the dispersion and fate of contaminated marine sediments both under existing condi- tions and following a variety of proposed remedial actions. This paper reviews our current understanding and predictive ability of the dominant processes controlling the dispersion and fate of contaminated marine sediments. While it is pos- sible to predict the circulation and wave fields and turbu- lent mixing in marine environments, predictive ability is lacking for the other sediment dispersion processes. In par- ticular, due to the lack of reliable and comprehensive field data, much of our understanding of erosion/resuspension and deposition processes has been obtained from limited labora- tory studies which contain many simplifying empirical assump- tions. Extrapolation of these empirical process models to field application requires excessive amount of data for cali- bratton. Further field-based research is urgently needed to advance our understanding of erosion, deposition, and floccu- lation processes. and man- site. it INTRODUCTION Marine sediments are the sinks to a variety of contaminants (e.g., heavy metals and toxic chemicals) and nutrients (e.g., phosphorus and nitrogen) from industrial, agricultural, and municipal discharges. These contaminants, both in particulate and dissolved forms, may sub- sequently reenter the water column due to resuspension of sediments, while the dissolved form may also be diffused into the water column. Once in the water column, contaminants may be transported by the three- dimensional turbulent flow field away from the contaminated site while adsorbed onto the fine sediment particles ~ _ _ Thus' in order to quanti- cactve~y assess One tong-term rate or contaminants at a contaminated marine site (e.g., New Bedford Harbor, Massachusetts and Puget Sound, Washington), it is essential to be able to perform a mass balance study for the fine sediments within the water body. Basically, this means 166

OCR for page 166
167 that we must be able to monitor or predict the long-term dispersion of sediments and contaminants within a large water body. Moreover, in order to assess the feasibility of a proposed remedial action (e.g., dredging or capping), it is necessary to be able to predict the impact of such action on the long-term fate of contaminants. Due to the complexity of the sediment/contaminant dispersion pro- cesses and the tremendous difficulties and costs of comprehensive field measurements, long-term monitoring at contaminated marine sites is usually not done. Hence, it is extremely important to develop predic- tive capabilities of sediment/contaminant dispersion processes. This paper provides a brief review on our current understanding of the var- ious processes controlling sediment dispersion. The difficulties of comprehensive field studies and the deficiencies of some laboratory studies are discussed. The use of comprehensive models to assist the quantification of sediment dispersion processes is illustrated with a brief discussion of the laboratory studies of erosion and deposition using rotating annul). Uncertainties in model parameters are discussed throughout the paper. Recommendations for further research are given at the end. SEDIMENT DISPERSION PROCESSES As shown in Figure 1, the dominant sediment dispersion processes in marine environments include advection, turbulent mixing, flocculation and settling, erosion/resuspension, and deposition (Sheng, 1986a, 1987). Mathematically, a mass balance study for sediments within a water body as shown in Figure 1 is equivalent to solving the following mass conservation equation for suspended sediment concentration: aC + aUc + auC + 8(W+ws~c _ a (AH ~~) + ~ (AH~) (~) + a `4,aC, where C is the suspended sediment concentration, (u,v,w) is the three- dimensional flow velocities in the (x,y,z) directions, t is the time, Us is a settling velocity for sediment particles, and AH and ~ are the horizontal and vertical turbulent eddy diffusivities. Assump- tions have been made that sediment particles are of similar shapes and sufficiently small and uniform sizes such that they more or less follow the turbulent eddy motions. Assuming the three-dimensional flow field is known, the above equation can be solved in conjunction with the following boundary conditions:

OCR for page 166
168 WED RIVER LOADING WAVES ~ TIDES FRESH ~ ---- . WATER FIGURE 1 Dominant processes controlling sediment dispersion in an estuarine environment. ~ .__ ~ ~ ~ ~ ~ ~ . ~ , TURBULENT - A/ - - 1FLOCCULATION r MIXING N:N t SETTLING ~ \ ~ 7 EROSION . ~ ADVECTlON ~ ~ DEPOSltlON RESUSPENSION ~ ~ ~ ~, ~ ~ , 4 SEDIMENTS ~ BENTHOS TIDES Lo SALT WATER f -wsC + ~ ~z 0 @ z = ((x,y,t) f - -wsC + ~6 = D - E @ z = -h(x,y,t) (2) C = C(x,y,t) @ Lateral Boundaries where f represents a net vertical flux of sediments at the free sur- face, ((x,y,t), or the bottom, -h~x,y,t), D is the deposition, and E is the erosion or resuspension. Both D and E depend on hydrodynamic and sedimentary parameters, and may vary significantly with space and time. ADVECTION The advection process is governed by the flow field of (u,v,w), which is generally three-dimensional, time-dependent, and turbulent. In estuarine environments such as Puget Sound, New Bedford Harbor, and Chesapeake Bay, the flow field is driven by tide, wind, and density gradient. Although much research is still needed to understand the hydrodynamic processes, our current understanding on the estuarine circulation and advection process is much better compared to what we know about the other sediment dispersion processes. Numerous estuarine hydrodynamic and dispersion studies have been performed in the major estuaries in the United States and other coun- tries. The more recent studies have recognized the importance of the three-dimensional aspects of the flow field and their effects on the dispersion (e.g., Sheng, 1983; Byrne et al., 1987; Sheng, 1988a). The formation of turbidity maxima (e.g., Dyer, 1986) due to tidal residual circulation and formation of salt wedge, has also been the topic of num- erous studies. Recent reviews of available multidimensional numerical models of estuarine hydrodynamics can be found in Sheng (1986b) and Nihoul and Jamart (1987~. It should be noted that generalized curvi- linear grid models are presently being developed to simulate the long- term circulation in estuaries with complex bathymetry and geometry.

OCR for page 166
169 TURBULENT MIXING Turbulence in marine environments can be transported across the air-sea interface by wind-induced mixing and breaking of surface waves. It can also be generated due to breaking of internal waves, shearing motion at the bottom, and unstable stratification, etc. In shallow estuarine environments, turbulence may exist over the entire water column, including the bottom boundary layer and the surface boundary layer to enhance the mixing of various materials. Turbulent transport within the bottom boundary layer plays the important roles in affecting sediment erosion and deposition, while turbulent mixing throughout the water column can affect the floccu- lation and,settling of sediment particles . The transport of turbulence within a water body, however, is very complex and cannot be accurately described in terms of a simple "diffusion" process, which is appropri- ate for the laminar mixing. The accurate description of turbulent transport processes (produc- tion, advection, damping, diffusion, and dissipation, etc.) is of ut- most importance for describing the large-scale circulation as well as the boundary layer dynamics, which can affect the sediment dispersions. A review of the turbulence models suitable for estuarine applications can be found in Sheng (1986b). It is presently possible to simulate the turbulent transport in bot- tom boundary layers driven by current, wave, current with wave, and in the presence of vegetation canopy, density stratification, and rough- ness features. However, turbulent transport in the immediate vicinity of large bottom roughness features and in the near-field of a dredged material disposal plume is fully three- dimens tonal, highly random, and very complex, and hence still not well understood. FLOCCULATION AND SETTLING Marine sediments often contain particles of various sizes ranging from the submicron clay particles to the large floes or sand particles of hundreds of microns. Flocs are formed as the fine-grained (d < 60 ~m) sediment particles are brought into frequent collisions by the turbulent shearing motion and differential settling, and if there is sufficient ionic strength in the water to render the suspended sediment particles cohesive. Larger floes possess weaker strength and may be broken up by increasing turbulent shear. However, little is known quan- titatively about the flocculation process and its dependence on numer- ous physical and chemical properties of the sediment, the fluid, and the flow. While flocculation can lead to orders of magnitude increase in settling velocity of sediment particles, feeding activities of benthos can also produce larger particles and settling velocity. Furthermore, the settling velocity generally increases with the concentration of sus- pended sediment particles until the hindered settling starts to reduce the settling velocity (Krone, 1962~. At the present time, settling velocity of sediment particles is generally determined in laboratory

OCR for page 166
170 f and there exists no predictive model of settling velocity. settling velocity Ws in Eq. (l) is generally treated as a tuning parameter, with values reported in literature varying by more than two orders of magnitude. EROSION/RESUSPENSION AND DEPOSITION Hence, the Erosion or resuspension is the process by which the surfacial layer of sediments is removed from the bottom sediments due to increase in hydrodynamic stress and turbulent intensity and/or weakening of sedi- ment resistance. As such, the erosion process depends on the hydrody- n~mic forces as well as everything that affects the strength of bottom sediments. Deposition, on the other hand, is the process by which the suspended sediment particles arrive at the bottom. Thus, the deposi- tion process depends only on the hydrodynamic process and the proper- ties of suspended sediments. Comprehensive field data are needed in order to derive an erosion model and a deposition model for application with Eq. (l). Our understanding of the erosion and deposition processes, however, has been rather qualitative and limited due to the complexity of the ~ ~ ~ in Figure a well- a typical exchange processes at the sediment-water interface. As shown 2, a relatively clear water column over bottom sediments with defined sediment-water interface has been regarded by many as z(m) T ,.'oo 1 r 1~2 L FIGURE 2 Vertical distribution of flow and sediment in an idealized estuarine environment. 1 WAVES ~ ODES - I U AIR C; \ WATER COLUMN BOTTOM 80UNDARY LAYER (BBL) BED

OCR for page 166
171 z(m) WED rim (_ j ~ ~ ~ _ | SeDIMtNT ~1 STATIONARY ~ FLUID MUD A__.________ AIR / WATER COLUMN a___.___ BOTTOM BOUNDARY LAYER (SOL) \ CONSOLIDATING \ BED ~ i_ ~ PERMANENT BED FIGURE 3 Vertical distribution of flow and sediment in a more realistic estuarine environment. situation for studying erosion and deposition. In realistic estuarine environments, however, relatively high-concentration sediment layers may exist over the consolidating and the permanent beds, with a loosely defined sediment-water interface and the presence of benthos (Figure 39. It is clear that comprehensive field monitoring in such an envi- ronment presents tremendous difficulties. Ideally, we would like to measure the vertical profiles of all the following parameters within the bottom boundary layer and the surficial sediments: mean velocity, turbulence, salinity, temperature, suspended sediment concentration, density, size distribution, settling velocity, benthos, pore pressure, and sediment composition. Even in the absence of any suspended sediments, it is difficult to measure the mean flow and turbulence in the vicinity of the bed where significant vertical variation in flow quantities exists. Field meas- urement using two-axis submersible LDV failed to capture the wave boun- dary layer even at 7 cm above the bed (Agrawal et al., 1988~. In spite

OCR for page 166
172 of the fine-scale flow structure, researchers typically rely on mean flow measured at 1 m above the bottom to calculate a turbulent bottom stress using the law of the wall (e.g., Grant et al., 1984~. This method may give erroneous results in the presence of large roughness features, strong waves, and significant stratification. While measurement of turbidity and optical scattering at selected locations can be obtained with existing instruments, the accurate determination of suspended sediment concentration and particle size distribution is still unresolved and remains the topic of many researchers. Due to the difficulties of field measurements, erosion and depo- sition processes have been studied in laboratory using rotating annul) (Figure 4~. A critical review of sediment studies using rotating annu- li was recently given by Sheng (1988b). Deficiencies in such labora- tory studies include the following: 1. Scaling problem--the small sizes of the rotating annul) (d ~ 1 m) makes it very difficult to generate fully rough turbulent flow, which is often encountered in the field. 2. Secondary flow problem--significant secondary flow (Figure 5) in rotating annul) (with rotating top and/or bottom) alters bottom friction and sediment dispersion patterns. 3. Turbulence damping problem--significant vertical gradient of sus- pended sediment concentration near the bottom can damp turbu- lence and reduce the amount of sediment erosion (Sheng and Vil- laret, 1988~. 4. Bioturbation problem--feeding activities of benthos can affect the sediment erosion and deposition, but the extent varies sig- nificantly with time and location (McCall and Trevesz, 1982~. Existing laboratory-based models of sediment erosion and deposition generally neglected the above problems and, although capable of repro- ducing the particular laboatory experiments, may produce large errors when extrapolated to a new field application. L r_ ,1 d OCR for page 166
~ l / ~ ~ t I - '<15~7 ~ it t:///~: r = rj \ >: ~J r = rO / ~ _~_L j I ~ I i'J , /: Fixed 73 l A /i d / , ~ it/ r- r (l)~ r = rO _ ._ - / FIGURE 5 Left: flow patterns and particle trajectory in a rotating annulus with a rotating top. Right: flow patterns in the azimuthal and radial planes of a rotating annulus with a rotating top and bottom. The laboratory-based erosion models are empirical, site-specific, and can vary by several orders of magnitude, as is shown in Figure 6 (Lavelle et al., 1983) which compares 10 erosion models compiled in the form of E = a|~|0 where r is the dimensionless bottom stress. Curve is determined from Puget Sound field data by Lavelle et al., curves 3 and 4 are determined from laboratory experiments of Lake Erie sediments (Sheng and Lick, 1979), while curves 9 and 10 are determined from labor- atory experiments of San Francisco mud (Partheniades, 1965~. Recently, comprehensive mathematical models have been used to quan- tify some of the deficiencies of the laboratory sediment experiments. For example, Sheng (1988a) used an integral boundary layer formula to calculate the secondary flow within a rotating annulus (Sheng and Lick, 1979) and found the radial flow to be 20 to 50 percent of the azimuthal flow. In addition, the law of the wall was modified to include the effect of a radial pressure gradient to allow the calculation of the vertical profile of radial and azimuthal velocities. Sheng and Vil- laret (1988) developed a simplified second-order closure model to inves- tigate the effect of sediment concentration gradient on the flow. They found that erosion models developed without considering such effect can cause very significant error in the prediction of suspended sediment concentration. In some cases, erosion may stop because the turbulent stress becomes significantly reduced by the concentration gradient. The deposition is treated as a separate process in Sheng and Lick (1979) and Sheng (1986a), which derived a rigorous formula for deposi- tion velocity. Thus, erosion and deposition may be allowed to occur at the same time. Krone (1962) considered the erosion and deposition to- gether, however, and defined a critical stress for net erosion

OCR for page 166
174 5 -6- _ 11J 0 7- cr o 8 FIGURE 6 Various models of sediment erosion. SOURCE: Lavelle et al., 1983. / o/ it; ,10 O 1 2 10910T (dYneS/Cm2) (rce) and a critical stress for net deposition (red) Although Tcewas assumed to be greater than ~cd, both must be prescribed empirically for each study. Teeter (1988) found the deposition of New Bedford sediments varied by a factor of 256. CONCLUSIONS AND RECOMMENDATIONS A brief review of our understanding of the dominant processes of sediment dispersion has been given here. Presently there are suffi- cient understanding and predictive capabilities of the advection and turbulent mixing processes in marine environments. However, our under- standing of the processes of flocculation, settling, erosion, deposi- tion, and bed evolution is rather limited. Existing models of these processes are primarily based on data from laboratory experiments, which often contain limiting simplifying assumptions, and hence are generally site-specific and contain large uncertainties for general application. Extrapolation of these empirical models to new field application may lead to large errors. It is thus extremely important to develop mechanistic process models (flocculation model, erosion and deposition model, and bed model) using comprehensive field data. These models can then be combined with the circulation model, wave model, and bottom boundary layer model to produce an overall sediment dispersion model (Figure 7), which may then be used for field validation and mass- balance predictions. Unless the process models have been validated by field data, the overall sediment dispersion model cannot be expected to produce reliable "prediction." Research should be carried out in the following two areas:

OCR for page 166
175 ~ . CIRCULATION MODEL d , . -, 1 _ . WAVE MODEL LAB and FIELD STUDIES I ADYECTION | . MIXING BBL _ BOTTOM MODEL TURBULENCE . L BOTTOM CURRENT 1 1 BOTTOM ORBITAL CURRENT MODEL ~ | EVOLUTION 1 BOTTOM l . _ ~ EROSION and DEPOSITION MODEL . FLOCCULATION MODEL 1 FIGURE 7 Framework of a sediment dispersion model. SEDIMENT DISPERSION MODEL FLOCCULATION I arm _ SETTLING 1 . Re - examination and improvement of exis tiny laboratory -based models of sediment dispersion processes: comprehensive flow and sediment data should be collected in rotating annul) and syn- thesized to allow the determination of less empirical models. A second-order closure model of turbulent transport can be used to simulate the complete flow field withing the rotating annul). 2. A comprehensive field program at a contaminated marine site to allow monitoring of the extent of contamination and to allow development of field-based models of sediment dispersion pro- cesses: due to the availability of more advanced instrumentation and modeling technique, this study should yield results which are much more useful than previously possible under the DMRP program. Due to the large uncertainties contained in the various parameters (e.g., size distribution, settling velocity, and bottom sediment distri button, etc.) appearing in a sediment dispersion model, it is important to perform an uncertainty analysis for any sediment mass balance study. Rather than treating the various model parameters as adj ustable tuning parameters to achieve a single "best fit" with limited field data, it is more reasonable to attempt to predict the mean value as well as the variance ~ or uncertainty) of the sediment/contaminant concentration distribution in marine environments. -

OCR for page 166
176 ACKNOWLEDGMENTS Support the from U.S. Environmental Protection Agency under Coopera- tive Agreement AERL-87-01, with Dr. Steve C. McCutcheon as the Scienti- fic Officer, is acknowledged. REFERENCES Agrawal, Y. C., D. G. Aubrey, and F. Dias. 1988. Field Observations of the coastal bottom boundary layer under surface gravity waves. Proc. Conf. App. Laser Anemometry to Fluid Dynamics, Lisbon, July 11-14, 1988. Byrne, R. J., A. Y. Kuo, R. L. Mann, J. M. Brubaker, E. P. Ruzecki, P. V. Hyer, R. J. Diaz, and J.H. Posenau. 1987. Newport Island: An Evaluation of Potential Impacts on Marine Resources of the Lower James River and Hampton Roads. Special Report in Applied Marine Science and Ocean Engineering No. 283. Gloucester Point, Va.: Virginia Institute of Marine Science, College of William and Mary. Dyer, K. R. 1986. Coastal and Estuarine Sediment Dynamics. New York: John Wiley & Sons. 342 pp. Grant, W. D., A. J. Williams, and S. Glenn. 1984. Bottom stress esti- mates and their prediction on the northern California continental shelf during CODE-1. J. Phys. Oceanogr. 14:506. Krone, R. B. 1962. Flume Studies in the Transport of Sediment in Estuar- ine Shoaling Processes. Hydraulics Engineering Laboratory Report. Berkeley: University of California. Lavelle, J. W., H. O. Mofjeld, and E. T. Baker. 1983. An In situ Ero- sion Rate for a Fine-Grained Marine Sediment. NOAA/ERL PMEL Con- tribution Number 654. Washington, D.C.: National Oceanic and Atmos- pheric Administration. McCall, P. L. and M. Trevesz. 1982. Effects of benthos on physical pro- perties of freshwater sediments. In Animal-Sediment Relations, P. L. McCall and M. J. Trevesz, eds. New York: Plenum Press. Pp. 105-176. Nihoul, J. C. J. and B. Jamart. 1987. Three-dimensional Models of Marine and Estuarine Dynamics. London: Elsevier. Partheniades, E. 1965. Erosion and deposition of cohesive soils. J. Hyd. Div. ASCE, 91(HY1):105-138. Sheng, Y. P. 1988a. Curvilinear-grid model for estuarine and coastal hydrodynamics. Proceedings of the 21st International Conference on Coastal Engineering, Spain, June, 1988. New York: ASCE (in press). Sheng, Y. P. 1988b. Consideration of flow in rotating annul) for sediment erosion and deposition studies. To be published in J. Coastal Research. In Press. Sheng, Y. P. 1987. Numerical modeling of estuarine hydrodynamics and dispersion of cohesive sediments. In Sedimendation Control to Reduce Maintenance Dredging of Navigational Facilities in Estuar- ies. Washington, D.C.: Marine Board, National Research Council. Pp. 94-117.

OCR for page 166
177 . Sheng, Y. P. 1986a. Modeling bottom boundary layers and cohesive sedi ment dynamics . In Estuarine Cohesive Sediment Dynamics. New York: Springer-Verlag. Pp. 360-400. Sheng, Y. P. 1986b. Finite-Difference Models for Hydrodynamics of Lakes and Shallow Seas: Physics-Based Modeling of Lakes, Reser voirs, and Impoundments. New York: American Society of Civil Engineers. Pp. 146-228. Sheng, Y. P. and W. J. Lick. 1979. The transport and resuspension of sediments in a shallow lake. J. Geophysical Research 84:713-727. Sheng, Y. P. and C. Villaret. 1988. Second-order closure modeling of sediment-laden turbulent boundary layers. Paper presented at American Geophysical Union Chapman Conference on Sediment Transport Processes in Estuaries, Bahia Blanca, Argentina, June 13-17, 1988. To be published in J. Geophysical Research. Sheng, Y. P. 1983. Modeling of Three-dimensional Coastal Currents and Sediment Dispersion, Vol. 1, Model Development and Application. Technical Report CERC-83-2. Vicksburg, Miss: U.S. Army Engineer Waterways Experiment Station. Teeter, A. 1988. Case Study: Physical transport investigations at New Bedford, Mass. Paper presented at the Contaminated Marine Sediments Symposium, Marine Board, National Research Council. Tampa, May 31-June 2, 1988.