Cover Image

PAPERBACK
$125.25



View/Hide Left Panel
Click for next page ( 352


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 351
PHYSICAL TRANSPORT INVESTIGATIONS AT NEW BEDFORD, MASSACHUSETTS Allen M. Teeter U.S. Army Engineer Waterways Experiment Station ABSTRACT Migrations of sediment, sediment-associated contami- nant, and dissolved materials released by proposed dredging and disposal operations were predicted as part of an engi- neering feasibility study of dredging cleanup. Highly contaminated sediments blanket most of upper New Bedford Harbor (the Acushnet River estuary), and were found to be escaping from the upper harbor toward Buzzards Bay. Field measurements, laboratory tests, and computer models, each necessary to support the others, comprised the study. Field measurements characterized hydraulic conditions and trans- port mechanism for salt, sediment, and contaminants for a series of surveys. Suspended material was found to migrate from Buzzards Bay upstream in the estuary at concentrations generally below 10 ppm, and settle in the upper harbor at about 2,200 kg per tidal cycle. The flux of PCB-Aroclors was found to be seaward and averaged 1. 55 kg per tidal cycle . Laboratory tests for settling, depos ition, and era - sion of sediment material were carried out. The most mobile sediment fraction was found to make up 28 percent of the sediment. Models for hydrodynamics and sediment transport were applied to the upper New Bedford Harbor, and used to predict sediment and contaminant migration for dredging and disposal scenarios. Results indicated that the flux of sediment materials from the upper harbor would be 15 to 20 percent of the rate of sediment resuspension. INTRODUCTION An engineering feasibility study (EFS) of a possible Superfund dredging cleanup for upper New Bedford Harbor was conducted by the U.S. Army Corps of Engineers (COE). The COE's Waterways Experiment Station (WES) Hydraulics Laboratory evaluated hydraulic conditions and sediment migration as part of the WES Environmental Laboratory dredging and dis- posal EFS conducted for the U.S. Environmental Protection Agency (EPA), Region I, under the direction of COE's Missouri River Division, and in cooperation with its New England Division. Highly contaminated sedi- ments blanket most of the upper New Bedford Harbor (Acushnet River Estuary), threatening to spread to other harbor areas and adjoining 351

OCR for page 351
352 Buzzards Bay, and adversely impacting fisheries resources. The EFS is one component of EPA studies that will lead to a Superfund cleanup of the harbor and upper harbor. The Acushnet River estuary EFS had components addressing physical and chemical testing of sediments to determine appropriate dredging limits, acquisition of bathymetric and geotechnical information, and study of contaminant behavior under simulated field conditions. An EFS report series is in preparation. Averett (1988) gives an overview of the study. The WES Hydraulics Laboratory investigated hydraulic condi- tions and transport mechanisms for salt, sediments, and contaminants as part of the EFS (Teeter, 1988~. The approach of the study was to inte- grate prototype measurements, laboratory data, and model results to quantify present conditions and predict dredging and disposal effects. The objectives were to evaluate 1. contaminant and sediment migrations away from resuspension points, and out of the upper harbor, 2. the hydraulics of the present and dredged upper harbor, and 3. concentrations of sediment and contaminant in the upper harbor during dredging and disposal releases. Confined disposal facilities (CDF) are diked settling basins that can include treatment methods to enhance contaminant settling. Con- fined aquatic disposal (CAD) is a controlled operation using excavated cells or chambers in the estuary bed, which are capped after filling. Suspended sediments (total suspended material [TSM]) will be discharged with CDF effluent, and released during CAD filling. The rate of sedi- ment release, sediment characteristics, and ambient conditions will control the amount of sediment that will escape from the proximity of the dredge and disposal facilities, and from the upper harbor during dredging. This paper describes methods used to predict sediment and contaminant escape. FIELD DATA COLLECTION Figure 1 shows the layout and approximate dimensions of New Bedford Harbor, which is located on the north shore of Buzzards Bay and is the estuary of the Acushnet River. The mean tide range at New Bedford Har- bor is 1.1 m, and the spring range is 1.4 m. Surface-to-bottom sali- nity differences are generally less than 0.5 ppt. The Acushnet River has a mean annual freshwater discharge of about 0.85 cubic meters per second. The upper harbor is shallow, with an average depth of only about 1 m at mean low water. Survey dates, tides, freshwater flows, and winds were as follows:

OCR for page 351
353 Survey Freshwater in- Tidal range at Wind direction, date flow m /see tide gauge 3, m speed, km/hr Mar 6 1.17 1.04 S. 24-32 Apr 24 1.50 1.65 NE, 8-12 then 32-48 SW, 16-24 Jun 5 0.25 1.04 Water temperature 4C 11C 17C Nine stations were sampled repeatedly over tidal cycles (Figure 1~. Current speed and direction, salinity, and TSM were sampled 0.6 m up from the bed and below the surface, and at mid-depth. Station 9 was sampled only at mid-depth. Bridge Flux of TSM The Coggeshall Street Bridge is a key point at which transport measurements and predictions were made. Bridge fluxes of TSM were estimated by integrating half-hour measurements of velocity (u) and TSM FIGURE 1 Sampling and gauging WOOD STREE7 locations for New Bedford Harbor. L UPPER HARBOR - TIDE GAGE & A UTOMA TIC AUTOMA TIC `~ f LUX RANGE SAMPLER . }` . Am,. ' 4 5 6 COGGESHALL STREETS-a I NSET AT COGGESHALL STREET BRIDGE \ ~ POPES ISLAND ODE GAGE =2 ~= ~~ ~ it. - FAI R HAVEN NEW BEDFORD-. `~\ 7~,,,: APPROXIMATE SCALES 500 0 500 1000 YD 50.0 0 500 1000 M LEGEND O BOAT STATION ~ Tl DE GAGE HURRICANE -x BA PRIER . .~ \ . ' , '. ~ \~\ 'N \\ BUZZA RDS BAY ".?~2 \\ -'\ \\ .- - ~ \\ \\ \\ W. .'--N CLAR K'S -\ Hi- POINT .} it, ~ TIDE GAGE =1

OCR for page 351
354 concentration over the tide-corrected, cross-sectional area, and integrating in time. Results are shown in Table 1. The net flux of TSM was always found to be upstream, although fluxes in either direction were at least twice the net flux values. Average flux of TSM into the upper harbor, corrected for tidal asym- metry, was about 2,200 kg per tidal cycle. The freshwater inflow added some additional sediment, on the order of a several hundred kg per tidal cycle. Shoaling resulting from the deposition of 2,500 kg per tidal cycle amounts to 3 mm per year when spread over the entire sur- face area of the upper harbor at a bulk wet density of 1.5 g/cm (775 dry-g/liter). The average sedimentation rate for the harbor has been estimated to be 7 mm per year (Summerhayes et al., 1977~. Estuarine TSM Flux Components Figure 2 shows a plot of the longitudinal distribution of depth- and tidal-averaged TSM concentration. TSM concentrations were lowest at the most seaward stations, and increase upstream. A turbidity maxi- mum occurred in the upper harbor. Differences in TSM concentration between spring- and reap-tide surveys were small, indicating that redis- persion of near-bed suspended material rather than erosion of bed sedi- ments contributed to tidal variations in TSM. The most important flux components for TSM were tidal pumping. TSM transport by steady vertical shear closely associated with transport by gravitational circulation, was small for New Bedford Harbor. Mechan- isms for upstream tidal pumping were evaluated. Maximum TSM resuspen- sion produced by the highest tidal currents, usually occurring near low water, were transported in the flood direction, producing upstream tid- al pumping. Water column redispersion time scales were much shorter than settling time scales, and produced phase differences between tidal velocities and TSM concentrations (tidal pumping). FIGURE 2 Longitudinal dis- 15 tribution of TSM concentration. ~ / O _ 1 2 3 5 7 8 9 STATION NUMBER

OCR for page 351
355 a: U) X o 1 0 ~:s - - C) EN x: x m Cam P. ~ ~ ~ 3 x: U] of EN ~ of C) ~ 0 0 X to EN Hi: EN Cal o 0 x o ' t) of C) 0 om EN >m ~ 0 ~ m .c EN 0 > 0 CQ 0 0 U. 0 0 0 a' \0 0 0 1 1 1 coo coo coo moo coo coo - ~ - - no - ~ - ~ ~ ~ ~ ~ ~ ~ ~ lD ~ 0 1 1 _' 1 ~ ~ o 1 1 ~ ~ ~o ~n ~ o ~ 1 1 o ~ ~ 1 1 mo ~ mm m0 m~ ~o ~ ~ ~ ~ co ~ ~n ~ ~ ~ ~o ~ r~ mm om 0o ~ - - mo mm mmm - - o - 0 - ~ 0 ~ ~ ~ ~ to 0 ~ ~ ~ ~ ~ ~ ~ -e ~ ^ - o ooo 1 1 1 1 o o o 0= Q 0~ ,2 0= - 0 ~ - 0 ~ - 0 0 - ~ c~ ~ c 0 ~ ~ IO o X o X 0 E~ ~ O U] Z;~

OCR for page 351
356 Vertical mixing was found to be more intense during flood tidal phases and was less damped by density effects on flood tidal phases. PCB Fluxes Bridge fluxes of PCBs are shown in Table 1. Flow-proportioned ebb and flood PCB Aroclor concentrations were multiplied by the tidal volumes to obtain ebb and flood PCB fluxes. The difference between ebb and flood fluxes is the tidal net flux. Observed net fluxes were always seaward (negative) with a mean net flux of -1.25 kg per tidal cycle. Tidal biases were removed from the raw tidal net fluxes by summing net-f~ow fluxes (freshwater volume times mean concentration) and tidal pumping fluxes (the difference between ebb- and flood-mean concentra- tions times the mean tidal volume). Corrected flux values are also shown in Table 1, and were also seaward with a mean flux of -1.55 kg per tidal cycle. Floatable material samples at the bridge were low in PCBs. Accur- ate transport rates could not be estimated for floatable material, but were at least several orders of manitude less than that for suspension. PCBs attached to sediment particles at the surface of the bed could be exchanged into the overlying sediment suspension by a physical par- ticle exchange mechanism, and thus be mobilized for transport. Such an exchange could take place without a significant mass flux of sediment. A particle exchange theory, based on aggregation and disaggregation of cohesive particles at the sediment/bed interface was developed during this study. That analysis used laboratory data on another estuarine sediment. The results indicated that particle exchange could be an important transport mechanism. LABORATORY TESTING Depositional and erosional characteristics of fine-grained sedi- ments vary greatly, and are critical to the prediction of sediment and contaminant migration. Direct laboratory testing on sediments less than 74 Em from the study area was therefore undertaken. Deposition (D), or flux of sediment material to the bed, is the sum over a number of fractions of settling flux multiplied by deposition probability: n D ~ ~ PiWs Ci . ~ . 1~1 ~ (2) where Ws is the settling velocity, P is the probability that an aggre- gate reaching the bed will remain there, C is the concentration just above the bed, the subscript i indicates a sediment fraction, and n is the number of fractions. P varies linearly from O at a critical shear stress for deposition, ~cd, to 1 at zero bed shear stress, rb ~ 0

OCR for page 351
357 The functional form 1 ~ rb/rcd where rb < Ad is used for P.A sus- pension of uniform material in a steady, uniform flow will either deposit completely or remain entirely suspended depending on whether the bed shear stress is below or above ~cd, according to Eq. 2. The objective of the deposition testing was to determine ~cd, and the magnitude of the product P Ws for each sediment fraction identified. The mode of resuspension (used synonymously with erosion) considered important to potential contaminant migration at New Bedford Harbor is particle erosion. At rb above a critical value, particles individually dislodged from the sediment bed as interaggregate bonds are broken. Particle resuspension (E) is related to the shear stress in excess of a critical value, and to an erosion rate constant (M), thus: E = M ( b 1) , rb > arc tic }, where arc is the critical erosion shear stress (Ariathurai et al., 1977~. Observed erosion does not follow Eq. 3 indefinitely. Sus- pension concentrations above experimental eroding beds often reach equilibrium values that depend on the bed shear stress. Equilibrium suspensions form as erosion rates decrease with time to zero, while the flow remains constant. Eqilibrium suspensions have been related to vertical inhomogeneity in the bed (either particle characteristics or bed density) or to armoring by selective erosion at the bed surface. See Figure 3 for the configuration of the sediment water tunnel. This testing device was developed for this study to safely test contami- nated sediments. It was a closed-conduit sediment water tunnel, open to the air only at a small expansion chamber. The water tunnel had a uniform cross-section area, which changed from rectangular in the hori- zontal, deposi tion/resuspension sections to circular in the vertical settling and pumping sections. The water tunnel was calibrated so that propeller speed could be related to average velocity and bed shear stress. Three sediment fractions were identified, and designated 1, 2, 3. The 'developed from the analysis of the data were 0.42, 0.33, and 0.043 N/m for fractions 1, 2, and 3, respectively. Considering all results, the approximate composition of the sieved composite sample was 30, 30, and 40 percent for fractions 1, 2, and 3, respectively. Ws's were about 2, 1, and O.006 mm/see for fractions 1, 2, and 3, respectively. Median Ws values are shown In Figure 4. For the most erodible fraction, Sac was found to be 0.06 N/m . The most easily eroded fraction is the same fraction identified as the slowest to deposit (fraction 3~. Only about an additional 15 percent of the total bed material, or half of fraction 2, eroded between 0.06 and 0.9 N/m2, and the remainder of the material had Acts greater than 0.6 N/m . (3) MATHEMATICAL MODELING Near-field Plume and CAD Models Suspended-sediment plume calculations were performed to evaluate

OCR for page 351
358 3o - \ ,~ am x em x 213.4cm DUCT `~ SAMPL ING TUBE `` ~ AND 3 SAMPLING \5o. ''by PORTS ~ VARIABLE SPEED \\ hi\ At/ DRI VE MO TOR \\ ~~< ~ F W \~' W: t ~ ) ME ER ~ 'I ~ D 20.3cm ID ~\ ~ ~ it\ ~ ~ ACCESSPOflT \\ ; nF.5rF~nn`lG TURf \: NOTE NOT TO SCALE CONSTRUCTED OF 1.25cm THICK CLEAR ACRYLIC. _ 20.3ctn ID ASCENDING TUBE - DRI VE SEA F T ~ AND PROPEL L ENS \ / ~ TRANSI T/ON ~ EL BOW FIGURE 3 Isometric view of sediment water tunnel. the escape of sediments and contaminants from proposed dredging and dis- posal site outfalls in upper New Bedford Harbor. Near-field analyses of the escape of sediments from a CAD during the filling phase were also performed. The near-field plume and CAD models assumes an infinitely small, vertically well-mixed, suspended sediment source. Model suspended sed- iment plumes are advected away from the source in the X direction, spread or diffused in the Y direction, and allowed to settle. A diffu- sion velocity formulation was used to introduce a length scale-depen- dence into model plume spreading, similar to those observed in field experiments. The required data for the plume model were Qs, H. U. and Ws. A site depth of 1 m was assumed, the average depth of the upper harbor. The remaining two variables, U and Us, were assigned distributions. Plume calculations were made for a test matrix of 16 conditions formed by four values of current speed, and four values of settling velocity. Plume predictions for dredging in upper New Bedford Harbor indicate that on average about 35 and 29 percent of the material released at the dredge head will escape from 50 and 100 m of the site, respectively. The remainder settled within this radius. Sediment that escaped did so at the highest current speed, and had the lowest set- tling rate. However, escape totals were highest for the moderate

OCR for page 351
1 .O ~ 0.1 - C: o LD z - z C) 0.01 0.00 J-8 I ' Arks- -l - I-r I I I I L _ ~ J-8 COMPOSlTEm y my. COtPOSlTEl~ WS = 0.000113 C-4/3, MM/SEC y~COMPOSI TE ~ I I I I I,l I I I ~ I I To 100 no logo CONCENTRATION, MG/9 FIGURE 4 Median settling velocity/concentration relationship for settling test phases and field samples at grid cell J-8. current speeds, and for the lowest settling rat" speeds had the greatest frequency of occurrences. example plot of plume concentration contours. Average CAD release ratios were calculated for _ __ test matrix in U. Ws, and H. Four depositional classes and current speed ranges were identical to those used in the plume calculations. Results indicated that only the finest or slowest settling fraction escaped from the CAD, and that the escape of this fraction was almost complete. Moderate current Figure 5 shows an a three-dimens tonal Estuarine Numerical Modeling Computer codes RMA-2V and RMA-4 of the TABS-2 numerical modeling system (Thomas and McAnally, 1985) were used to schematically model ver- tically averaged hydrodynamics and sediment transport' respectively. To properly describe boundary conditions, the model domain was extended downstream to the hurricane barrier. A numerical mesh of 219 elements

OCR for page 351
25 15 n in, (n C) _5 -tO 360 NEW BEDFORD DREDGE PLUME ISOPLETHS AT 40 20, 10, 5, 2.5, t.25 mg/Q -15 -20 -25 0 ~ 0 20 30 I I 1 1 . I 1 1 40 50 60 X-DISTANCE, M 70 80 90 100 FIGURE 5 Example plume concentration isopleths for Qs ~ 5 g/sec, IJ 0.03 m/see and Ws ~ 0.01 mm/sec. was developed to cover the study area for use by both RMA- 2V and RMA- 4 . The upper harbor portion of the mesh is shown in Figure 6 . A mean-tide sequence was applied to the seaward boundary of the model . At the upper end, velocities corresponding to a constant fresh- water inflow of 0.85 m /see were specified. lathe numerical hydrodynamics model was verified to field data by adjusting friction coefficients and turbulent exchange coefficients. Hydrodynamic computations were performed by "spinning up" the model from a steady, flat, water surface condition. Results from the hydro- dynamics model were used to construct an 8 - tidal - cycle sequence for nput to the sediment transport model ~ Sediment transport modeling was performed to estimate escape prob- abilities from the upper harbor for various sediment materials which might be resuspended as a result of dredging. Transport of resuspended material was modeled as a steady mass loading at specified points. Only sediments released from the mass-load~ng point were included in computations . Three mass- loading locations were employed to represent various locations where sediment releases might occur along the axis of the estuary. Five sediment fractions were modeled to characterize the range of sediments that might be actually released. The effective sedi- ment depos ition coeff icient used was fix ~ WSP H The five depositional fractions tested covered the range 0.10 C ox 25.6, where c' has the units of 1/day. An example contour plots of the concentration field with ~ ~ O

OCR for page 351
361 FIGURE 6 New Bedford upper Harbor mesh showing CAD exclusion zone (hashed) for 5.5-ft spring tide. / // rr 66. \ ~ ~- _~ 5 4 U7 ~S~ ~ COGGESHALL STREET BRIDGE ~2N 1 1/ 17~\ for flood and ebb tide phases are shown in Figure 7. Maximum concentra- tions were about 3 mg/liter for a release rate of 15 g/sec. However, the estuarine model over-estimates spreading near the source, and near-field predictions should be applied here. Concentrations were proportional to release rates. Sediment transport results were used to calculate the escape proba- bilities of resuspended sediments Pleased in the upper harbor. Average transport rates under the Coggeshall Street Bridge were computed for flood and ebb tidal phases after the model had reached dynamic equil- ibrium. The difference between ebb and flood transport rates, normal- ized by the mass loading rate, represents the escape probability. Escape probabilities were calculated for each depositional sediment fraction at three source locations in the upper harbor, and with three variations of geometry representing dredging changes. A plot of escape probability versus a-infinity for three source locations is shown in Figure 8 for the existing estuary geometry. Source locations are shown in Figure 6.

OCR for page 351
362 I SOLINES 1 .1 2 .4 3 .7 4 1. 5 1.3 6 1.6 7 1.9 8 2.2 9 2.5 10 2.8 I ~1 ~ ~~ ~ ~~ ~ As) ~ = FIGURE 7 Sediment concentration field, ebb tide. SUMMARY AND CONCLUSIONS Assessments of sediment and contaminant migration out of the upper New Bedford Harbor for proposed dredging and disposal were made from information and analyses developed by field, laboratory, and various model studies. Upper New Bedford Harbor is a sheltered area with low current speeds, typical of many areas where contaminated sediments reside. Paradoxical tidal fluxes were found for suspended sediments and sediment-associated contaminants, implying that depositional sites do not retain all particle-associated contaminants. Average PCB flux from the upper estuary was seaward about 1.55 kg Aroclor per tidal cycle. However, the Acushnet estuary was found to be depositional, and a reasonably efficient sediment trap. TSM was imported from the coastal areas of Buzzards Bay, pumped upstream by tidal action, and formed turbidity maximums in the upper harbor. The sediment fraction slowest to settle and deposit comprised 28 percent of the composite sediment and represents by far the greatest potential for sediment and contaminant migration. Other fractions will

OCR for page 351
363 100 80 60 co m O 40 LL Q An lo ~ ~s 0 0.1 LEGEND \3 ~ \ 1) LOADING AT NODE 66 2) LOADING AT NODE 54 3) LOADING AT NODE 19 a, day~1 FIGURE 8 Escape probabilities for sediments released at three points along the upper harbor (see Figure 6 for locations). not be highly mobile in the upper harbor. Experimentally determined erosion thresholds were used with numer- ical hydrodynamics results to identify areas where CAD cells should not be sited. Based on these results, the area of the channel upstream from the bridge were not recommended for CAD sitting. See Figure 6. Results from the dredge plume model indicated that an average, weighted by occurrence frequencies, of about 33 percent of the resus- pended material will escape from a 100-m radius of the dredging site. Results from the CAD cell model indicated that all of the fine fraction sediment expelled with slurry pore water will escape from CAD cells. Modeling results indicated that the CAD may cause the greatest release of sediments and contaminant materials. The escape probability for bulk sediments resuspended at the point of dredging will average 27 percent but will depend on the highly variable bed sediment composi- tion. General information on dredge resuspension rates and suspended sediment releases during CAD filling is scarce. Pilot dredging and disposal, planned for the fall of 1988, will provide direct measure- ments for this site. The methods employed in this study could be applied to similar sites. However, studies of high-current systems will have to give greater attention to the difficult experimental and theoretical aspects of erosion processes. ACKNOWLEDGMENTS This work was sponsored by EPA Region I, and monitored by Mr. Frank

OCR for page 351
364 Ciavettieri. The EPA has not reviewed this paper, and the views expressed herein are those of the author and not necessarily those of the EPA. Permission to publish was granted by the Office, Chief of Engineers. REFERENCES Ariathuria, R., R. C. MacArthur, and R. B. Krone. 1977. Mathematical Model of Estuarine Sediment Transport. Technical Report D-77-12. Vicksburg, Miss.: U.S. Army Engineer Waterways Experiment Station. Averett, D. E. 1988. Study Overview. Report 1 of 12, New Bedford Superfund Project: Acushnet River Estuary Engineering Feasibility Study Series. Technical Report EL-88-15. Vicksburg, Miss.: U.S. Army Engineer Waterways Experiment Station. In preparation. Summerhayes, C. P., et al. 1977. Fine-Grained Sediment and Industrial Waste Distribution and Dispersal in New Bedford Harbor and Western Buzzards Bay, MA. WHOI-76-115. Woods Hole, Mass.: Woods Hole Oceanographic Institution. Teeter, A. M. 1988. Sediment and Contaminant Hydraulic Transport Inves- tigations. Report 2 of 12, New Bedford Superfund Project: Acushnet River Estuary Engineering Feasibility Study Series, Technical Report EL-88-15. Vicksburg, Miss.: U.S. Army Engineer Waterways Experiment Station. In preparation. Thomas, W. A. and W. H. McAnally. 1985. User's Manual for the General- ized Computer Program System: Open-Channel Flow and Sedimentation: TABS-2. Instr. Rpt HL-85-1. U. S. Army Engineer Waterways Experi- ment Station, Vicksburg, Miss.