|
Items in cart [0] |
Questions? Call 888-624-8373 |
| Copyright © 2009. 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 178
COMPUTER S IMULATI ON OF DDT DI STRI BIrrI ON
IN PALO S VERDES SHELF SEDIMENTS
Bruce E. Logan and Robert G. Arnold
University of Arizona
and
Alex Steele
Los Angeles County Sanitation Districts
ABSTRACT
Prior to impos ition of effective source control meas-
ures in 1970, large quantities of DDT were discharged to the
Los Angeles County municipal sewer system and subsequently
to the Pacific Ocean.
among sediments of
of the DDT lies 10
Much of this material accumulated
the Palos Verdes shelf. While the bulk
to 40 cm below the sediment surface, its
fate may be affected by future wastewater treatment at the
Los Angeles County's 385-mgd, Joint Water Pollution Control
Plant (JUPCP).
To assess the potential impact of JWPCP secondary treat-
ment requirements on shelf sediment quality, processes that
may affect distribution of chemical tracers among those sedi-
ments (background sedimentation rate, effluent-related
solids contributions, sediment mixing, and diffusive trans-
port through pore waters) were incorporated in a mathemati-
cal model. Because physical mixing substantially affects sur-
ficial concentrations and the vertical distribution of DDT
at the 60-m depth contour, model projections of surface sed-
iment quality at the most heavily contaminated sites are sen-
sitive to projected effluent solids concentrations and thus
JWPCP treatment decisions. Dredging and capping do not rep-
resent physically or economically feasible remediation stra-
tegies at this site. The work described illustrates how
empirical models of contaminant fate can be used to (1) simp-
lify or avoid uncertainties associated with some mechanistic
models, and (2) serve as a basis for management decisions.
INTRODUCTION
Indicators of biological quality among benthos of the Palos Verdes
shelf are sensitive to chemical characteristics of surface sediments
(Word, 1978; Sherwood, 1976; Oshida and Wright, 1976; Cross, 1984~.
178
OCR for page 179
179
Species composition and primary feeding strategies among benthic inver-
tebrates are functions of both sediment organic content and local sur-
face concentrations ~ top 5 cm) of hazardous compounds including trace
metals, total DDT, and PCBs (Stull et al., 1986~. Despite uncertain-
ties caused by statistical correlations among individual chemical param-
eters, the importance of controlling surface concentrations of DDT in
sediments of the Palos Verdes shelf has been accepted; Prior to a ban
on its disposal within the Los Angeles County municipal sewage collec-
tion system in about 1970 (Norman Ackerman, Supervisor, Oceanographic
Monitoring Activities, LACSD, personal communication), DDT was held
responsible for endangerment of the California brown pelican (Keith et
al., 1970; Risebraugh et al., 1967, 1971~.
The Los Angeles County Sanitation Districts (LACSD) provides par-
tial secondary treatment (200-mgd secondary treatment capacity) for
approximately 360 mad (136,300 m3 d- ~ of domestic and industrial
wastewater at the Joint Water Pollution Control Plant (JWPCP) in Car-
son, California. Effluent, including some 150 metric tons of suspended
solids per day, is discharged via a system of ocean outfalls to waters
of the Palos Verdes shelf. During the last 15 years, JWPCP effluent
quality has shown marked improvements, in terms of solids emissions and
concentrations of specific contaminants including total DDT (Figure 2),
in response to improved treatment and solids handling at JWPCP and
implementation of source-control measures on tributary industries.
Much of the DDT that was discharged to the LACSD sewer system prior
to 1970 reached the coastal waters and sediments off the Palos Verdes
peninsula. Due to its persistence, there are residual links to environ-
mental quality and human health--muscle tissue concentrations of DDT
exceed FDA limitations in local populations of white croaker and other
bottom-feeding species (Gossett et al., 1982, 1983~. Surface sediment
DDT concentrations appear to drive fish tissue values, as opposed to
the DDT mass emission rate from the Whites Point outfall system (Young
et al., 1988~.
cat
cat
an
to
-
FIGURE 1 The Palos Verdes shelf
in relation to the Palos Verdes
Peninsula and JWPCP. Positions ~
and depth of the LACSD outfall O
system and sediment monitoring
stations are as indicated .
~1
0 1 2 ~ \
Miles
Depth in Meters
~ I
118 30 LONGITUDE 118 20
so
OCR for page 180
180
400 r 2.0
350 _
300 — 1.5 _
250 _
200 - 1.0 _
150 _
Coo - 0.5 _
FIGURE 2 Summary of
JWPCP mass emissions 50 _
records to total solids, _ _
DDT, and total flow. ° °-°
106
5
4
jo3
102
101 _ ~
_ ~
10° 1 1 1 1
1985 1965 1945 1925 1905
YEAR
I DDT 1la/ka
· SS Mton/yr
_~ FLOW mad
Sediment quality on the Palos Verdes shelf is a function of physi-
cal and chemical processes that are imperfectly understood. These
processes potentially include sedimentation, bioturbation by benthic
infauna, periodic resuspension by bottom currents or wave activity, and
chemical diffusion. Following implementation of effective source-
control measures for DDT in the early 1970s, the bulk of the sediment
DDT on the Palos Verdes shelf was buried under less contaminated sedi-
ment of both natural and sewage-related origin. Recent sediment pro-
files (Figure 3, for examples suggest that the bulk of sediment DDT
lies buried between 10 and 40 cm below the surface.
The contribution of effluent-related particulates to the overall
local sedimentation rate is not well established. The background sedi--
mentation rate (independent of outfall particulates) has been variously
estimated at 10 to 200 mgKcm~ Kyr~ (Hendricks, 1984), perhaps in
response to variations in local coastline stability, and the fraction
of solids discharged from the Mites Point outfall system that is
retained on the Palos Verdes shelf has been estimated at 0.01 to 1.0
(Myers, 1974; Hendricks, 1982~. There is no consensus regarding the
importance of periodic sediment resuspension as a determinant of
contaminant profiles.
Hendricks (1978, 1982, 1984, 1988) modeled the fate of chemical
tracers in Palos Verdes sediments by combining a solids deposition
model designed to yield the discharge- and current-dependent pattern of
particulate fluxes to the shelf with a sediment resuspension model that
produced time-dependent profiles of select contaminants in the Whites
Point sediments. Hendricks' work identified areas of theoretical
inadequacy and data gaps, which must be addressed before mixing of
marine sediments can be addressed mechanistically. At present, it is
impossible to address adequately the following areas of uncertainty:
1. potential interdependence of sediment resuspension parameters
(resuspension mass, resuspension frequency, and time between
OCR for page 181
181
50.0
40.0
30.0
3
o
CD
J
o
20.0
10.0
0.0
Station 8C 107
DATA 1 o6
— MODEL
r
10;'
~ On 104
103
~ 102
_ ~ 101
. 1 1 1 1 1
0 10 20 30 40 50
CORE DEPTH cm
Figure 3a
60 0
Station 8C
~ DATA
— · MODEL
~MIX/DIF
_
10° 1
it,
15 30 45 60 75 90
CORE DEPTH cm
Figure 3b
50.0— Station 6C Station 6C
10/
40.0
-
c~ 30 0
-
o
oh
g
20.0
10.0
o.e
· MODEL
· DATA 1 o6
105
I' 1 04
-
C]
~ 102
~ 1o1
) 1 1 1 1 1 1 10°
0 10 20 30 40 50 60
CORE DEPTH cm
Figure 3c
- ~ DATA
- · MODEL
- _ —· MIX/DIF
I ~ I . 1 1 1
0 15 30 45 60 75 90
CORE DEPTH cm
Figure ad
FIGURE 3 Calculated and observed concentrations of volatile solids or
DOT in Palos Verdes sediments among monitoring stations along the 61-m
depth contour. Volatile solids concentrations were calculated using
the basic sedimentation model without considering surface mixing or
diffusion. DDT profiles were calculated using the sedimentation/mixing
model. Model parameters used in calculations represented here are
summarized in Table 1. In all cases, the solid "data" line represents
1985 LACED monitoring data; the broken "model" line represents best-fit
results using the basic sedimentation model; and the broken "mix/dif"
illustrates the best-fit model calculations when mixing and diffusion
mechanisms are included.
OCR for page 182
182
50.0
40.0
Station 3C
10
—· DATA
_ 106
MODEL
_ 105
~, 30.0 _ ~ 104
O ~
5 20.0 _ ~ 103
o
10.0 <~~ 101
0.0 ~ I ~ ~ ' 10°
0 10 20 30 40 50 60
CORE DEPTH cm
Figure Be
50.0
Station OC
Station 3C
_
_ · DATA
_
_ · MODEL
— - ~ MIX/DIF
1 1 \1 1
0 1 5 30 45 60
CORE DEPTH cm
Figure of
107
Station OC
75 ~
· DATA — · MODEL
40.0 _ - · MODEL 1 o6 _ · MlX/DIf
~ 105 c
Ha) 30.0: == 104> h
5 ~ 102 ~
0.0 ~ 1 1 1 1 1 10°~ ~ ~ t~ 1 1 1 1
0 10 20 30 40 50 60 15 30 45 60 75 90
CORE DEPTH cm
Figure 39
FIGURE 3 Continued.
_~ DATA
CORE DEPTH cm
Figure ah
resuspension and redeposition) and
the ratio of effluent-
related to natural particulate material among local sediments,
2. biological determinants of sediment resuspension rates,
3. the dynamics of a sediment surface layer approximately 1 mm
thick, which may be subject to frequent short-duration resus-
pension events prior to stable incorporation into bulk sedi-
ments, and
4. aggregation processes among biologically active surface
sediments.
OCR for page 183
183
TABLE 1 Summary of Best-fit Parametric Values Resulting from Calibra-
tion of the Sediment Deposition/Mixing Model across the Palos Verdes
Shelf
Stationa Background sedi- ~ Effluent Local diffu- Depth of mix-
(depth) mentation rate solids reach- sion coeffi- ing zone (cm)
(m, dry wt ing grid tied
cm /yr) location (cm /see)
8C(61m) 400 0.5 10-8 6
6C(61m) 500 0.5 10-8 6
3C(61m) ' 500 0.2 10-8 4
oC(61m) 500 0.01 10-8 6
6A 200 0.075 10 8 4
NOTE:
aStations correspond to those indicated in Figure 1.
LACED and the U.S. Environmental Protection Agency (EPA) are now
forced to evaluate the merit of imposing an additional 80 percent reduc-
tion in suspended solids emissions by establishing full secondary treat-
ment requirements for JWPCP. It has been suggested that this action
could result in gradual reemergence of previously buried DDT and other
effluent-related contaminants for which DDT can serve as tracer. Resi-
dual uncertainties associated with mechanistic, state-of-the-art sedi-
ment models preclude their use for prediction of Palos Verdes surface
sediment quality characteristics as a function of the projected treat-
ment level at JWPCP.
Here we apply an empirical approach to sediment quality modeling on
the Palos Verdes shelf in order to avoid scientific uncertainties that
have frustrated more mechanistic treatments. We have developed a ser-
ies of sediment models of increasing complexity in order to evaluate
the importance of specific processes, or classes of physical processes,
as determinants of sediment contaminant profiles. The background sedi-
mentation rate and fraction of discharged solids that reaches the sedi-
ment grid are treated as fitted parameters (selected to reproduce local
sediment profiles of volatile solids and DDT) as are sediment inter-
stitial dispersion coefficients and mixing parameters in higher order
models. Model results include surface sediment concentrations of DDT
as a function of projected treatment level (existing level, full
secondary, zero discharge, eta). at JWPCP.
METHODS
Deposition Model (Order 1)
The procedure for calculating sediment profiles of specific contam-
inants is described below and summarized in an appendix to this report.
OCR for page 184
184
The Palos Verdes shelf was divided into a two-dimensional (horizontal)
grid. The background sedimentation rate and fraction of the JWPCP
effluent solids deposited in each grid section were independently esti-
mated. Effluent-related solids were divided into volatile and nonvola-
tile fractions based on JWPCP monitoring records. The biodegradable
fraction of effluent solids was assumed to be completely oxidized in
the water column prior to deposition. All solids destroyed in this
~ -~ ~ ~ ~ ~ ~ ~ ~ - Residual
discharge-related solids were assumed to be refractory, as were solids
of natural origin that reached the local sediments.
Necessary measurements and parameter estimates included the record
of JWPCP solids emission rates dating to 1935 (Figure 2), the fraction
of discharged solids that are volatile, and the percentage of volatile
solids that are refractory in nature. With this information, the dry
mass of solids that reached the shelf sediments during each year of
record was calculated as a function of position on the shelf. The com-
All solids destroyed in this
fashion were assumed to come trom the volatile traction.
_
putational procedure also yielded an estimate of the volatile solids
fraction in the sediments for comparison with measured concentrations.
The total mass added to the sediment column during each year was
dependent upon an empirical relationship between sediment volatile
awl i Ac! =~ m^; c! - ll - = ^^r~t~=rit~ {Ar~r`~=Aiv)
~_~_~ ~~ ~~_~_. A- Or--. On the basis of the total mass
addition and the calculated (weighted average) sediment density, it was
possible to compute the thickness of an incremental layer of sediment
that accumulated in response to natural and effluent-related particle
deposition during each year. The position-dependent increment of mate-
rial predicted to have accumulated on the Palos Verdes shelf during the
lifetime of the Whites Point discharge (in the absence of resuspension
and mixing) was developed by integrating the calculated annual contribu-
tions. The computational procedure also yielded depth-dependent esti-
mates of volatile solids and moisture concentrations. These were com-
pared to profile data collected at several grid locations across the
shelf to determine goodness-of-fit. By varying input parameter values,
it was possible to select the most appropriate local background sedimen-
tation rate and fraction of effluent solids deposited within specific
grid boundaries.
In order to predict sediment profiles of specific effluent-related
contaminants, results of the foregoing procedure were combined with
measurements or estimates of annual mass emission rates (MERs) for the
contaminants of interest. ~ ~ ~ ~~~ ~ ~ ~ ---
In mode Lang ~VT, it was assumed that (ly DDT
was uniformly distributed (by mass) among effluent solids, and (2) dis-
charged DDT was entirely refractory in nature. Only total DDT was
modeled in this fashion. Sediment computations yielded depth-dependent
DDT concentrations in units of mg DDT per kg of dry solids. Annual
DDT mass emissions and other input data are summarized in Figure 2.
Fitted parameters in the initial modeling phase included only the
background sedimentation rate and the fraction of effluent solids that
is deposited in specific sectors of the sediment grid. Results of
LACED work using Hendrick's models were used to suggest appropriate
limits for parameter ranges. Grid sampling stations at which the mass
contribution of effluent solids is modest (in comparison to background
sedimentation), also provided an initial estimate of the background
sedimentation rate.
OCR for page 185
185
Sediment profiles of volatile solids, moisture content, and DDT con-
centration were developed in this manner for each sediment grid position
at which LACSD measured profile characteristics in 1985. Background
sedimentation rate and deposition of effluent-related solids were
adjusted to reproduce the volatile solids profile (depth and magnitude
of elevated volatile solids concentrations) to the extent possible.
Mixing Model (Order 2)
Perceived shortcomings in the level-one (deposition) model were
addressed by adding surface-mixing and DDT-dispersion components to the
computation of sediment profile characteristics. While the procedure
was limited to calculation of depth-dependent DDT concentrations, it
could be adapted to predict profile concentrations of any conservative
contaminant for which there is an adequate base of effluent data.
Computational procedures for the mixing model included those of the
deposition model mass balance, altered to account for surface mixing
and vertical diffusion of DDT throughout the sediment column. Diffu-
sion was incorporated by dividing the sediment profile into 2-cm com-
partments and permitting concentration-driven transport between adja-
cent compartments during specific finite time intervals. The 2 cm
represents both the depth interval for chemical determinations in LACSD
sediment cores and a practical compartment thickness for the diffusion
computation. A well-mixed zone, defined to consist of an arbitrary num-
ber of the uppermost 2-cm compartments, received the entire input of
deposited material during a given time interval; complete mixing was
assumed over the time necessary to add 2 cm to the overall profile
thickness. At that point, a new compartment was created at the top of
the sediment core, and a 2-cm compartment was pushed below the mixing
zone. Initial estimates of a molecular diffusion coefficient were
developed from the literature of mass transport through porous media
and later fitted using 1985 DDT profile data from Figure 1 monitoring
stations.
Sediment Quality Projection
Parameter estimates from the mixing model were used to project sedi-
ment quality characteristics as a function of the anticipated JWPCP sus-
pended solids MER. Treatment scenarios investigated and corresponding
solids emissions follow:
Treatment Level
No change (partial secon-
dary treatment)
Full secondary
Zero discharge
Reevaluation of treatment
resulting in 2x solids MER
Suspended solids MER (10 MT/yr)
(assumed constant from 1987 to 2005)
0.40
0.10
0.01
0.80
OCR for page 186
186
RESULTS
Deposition Model
A summary of fitted parameters corresponding to each of the core
sampling stations is provided as Table 1. A comparison of best-fit cal-
culations (deposition model) and measured (198S) DDT profiles at 61-m
monitoring stations is included within Figure 3b, 3d, 3f, and 3h. A
statistical measure of goodness-of-fit has not yet been incorporated
into the model structures, so model parameters were visually deter-
mined.
Order-one model calculations and sediment measurements of volatile
solids and DDT concentrations were within reasonable agreement at all
61-m monitoring stations along the Palos Verdes shelf. However, 1985
surface sediment concentrations of DDT were uniformly high relative to
model calculations, suggesting that addition of a mixing mechanism to
the deposition model would improve its performance substantially.
Mixing Model
A summary of best-fit parameter estimates (deposition/mixing model)
for stations along the 61-m contour is provided in Table 1. Estimated
background sedimentation rates are nearly uniform across the shelf
(400-500 mg/cm -yr) while the local contribution of effluent-related
solids varies by a factor of 50. Beyond the shelf, the background
sedimentation rate shrinks rapidly. In all cases, the local diffusion
coefficient for best fit was 10- /sec. The depth of the mixing zone
was 4-6 cm.
Calculated and measured profiles of volatile solids and DDT
concentrations for stations along the 61-m depth contour are provided
as Figure 3 (a-h). DDT profiles corresponding to the diffusion
(no-mixing) model are included to illustrate the importance of
surface-mixing to development of an effective sediment model.
Model Predictions
The calibrated sedimentation/mixing model was utilized to project
sediment profiles of DDT concentration along the 61-m depth contour.
Results are summarized in Figure 4 (a-d) and Table 2. Individual pro-
files correspond to JWPCP treatment alternatives ranging from zero dis-
charge to a doubling of the current JWPCP suspended solids, mass emis-
sion rate. The projected year-2005 surface concentration of DDT at
station do is independent of JWPCP treatment level. At the other
extreme, the model indicates that initiation of full secondary treat-
ment will produce year-2005 surface concentrations of DDT at station 6C
that are almost 75 percent higher than those that would result from no
additional treatment. In all cases, the predicted year-2005 surface
concentrations of DDT are much lower than current levels; there are
significant, treatment-dependent differences in those projections.
OCR for page 187
187
107
_
_
1o6 C
105
:l,
cl 1 03
1o2
1o1
10°
107
1o6 ~
105
t:n
At.
1 03
1 0Z .
10]
loo
1n7
1o6 ~
I;\ call 33 KNk/~\
~ 1 1 1 1` 1 10° ~ 1 ~ 1 ~ 1
0 15 30 45 60 75 90 0 15 30 45 60 75 90
CORE DEPTH cm CORE DEPTH cm
Figure 4a Figure 4t
1 _
10/ _
_ , ~ 1 g85
· 2005
~6 c
_
105
Ha\
en
1 04
103
1o2
1ol ~
F
. ) ~ 1 1 1 1 ~r 1 10°~ 1 1 1 1 1~\ 1
0 15 30 45 60 75 90 0 15 30 45 60 75 90
CORE DEPTH cm CORE DEPTH cm
Figure 4c Figure 4c
FIGURE 4 Year 2005 projections of sediment DDT concentrations at
monitoring station 6C (Figure 1) as a function of depth and treatment
level or solids removal efficiency at JWPCP. Corresponding solids MERs
are summarized in the text. 4a. No change scenario. Continuation of
partial secondary treatment at JWCPC. 4b. Full secondary treatment at
JWCPC. 4c. Zero discharge alternative. Hypothetical elimination of
ocean discharge via the Whites Point outfall system. 4d. Modification
of JWPCP treatment which results in a solids HER 2x the present (1987)
level.
Doubling the solids mass emission rate is predicted to lower the
year-2005, surface sediment concentration of DOT by a factor of four
relative to the secondary treatment alternative.
OCR for page 188
188
TABLE 2 Projecteda Year-2005, Surface Sediment Concentrations of
DDT along the 61-M Depth Contour of the Palos Verdes Shelf
Current No change 2x the
surface Zero Full (Partial current
concen- dis- secon- secondary) suspended
StationC tration charge dary treatment) solids MER
8C 11,500 4,000 3,500 2,100 800
6C 9,500 2,200 1,900 1,100 500
3C 3,500 550 500 360 240
dC 780 230 230 230 230
NOTES:
aProjections are based on the sedimentation/mixing model.
ball figures are in mg/dry kg of sediment.
CStation locations are indicated in Figure 1.
DISCUSSION
Sensitivity
In order to demonstrate model response to variation among fitted
parameters, results of representative sensitivity analyses are pre-
sented in Figure 5 (a-h). Sediment volatile solids concentrations res-
pond to variation in both background sedimentation rate and relative
size of the outfall-related solids contribution. DOT profile calcula-
tions were compared to measured values at station 6C to assess model
sensitivity to variation in the molecular diffusion coefficient and
depth of the mixing zone. For each of the four parameters tested,
appropriate profiles (volatile solids or DDT) resulting from both arti-
ficially high and low parameter estimates are presented.
Results indicate that the sedimentation/mixing model is reasonably
sensitive to selection of the background sedimentation rate, although a
50 percent increase in background sedimentation (Figure 5b) did not
substantially affect the quality of fit between calculation and meas-
urements. A 50 percent decrease in the assumed natural sedimentation
rate (Figure 5a) results in unrealistically high volatile solids con-
centration.
Volatile solids calculations are very sensitive to change in the
relative effluent solids contribution at station 6C. A 40 percent
decrease in discharge-related solids (Figure 5c) provides a marked
decrease in profile depth; an increase of similar magnitude (Figure Sd)
results in overestimation of the sediment organic content.
From the sensitivity of profile depth calculations to the local
flux of effluent-related solids, it is apparent that the Whites Point
OCR for page 189
50.0
40.0
-
-
cn
C, 30.0
J
o
an
IIJ
~ 20.0
o
10.0
50.OI
— DATA
_~ MODEL 40.0
-
~ 30.0
~ 111
-r ~ ~ 20.0
~ O
-0 - ~ 10.0
— DATA
· MODEL
\::;
0.0 ,, ~ ~ ~ ~ 1 1 0.0 I I I 1 1 ~
0 10 20 30 40 50 60 0 10 20 30 40 50 60
CORE DEPTH cm CORE DEPTH cm
Figure 5a Figure 5b
50.0 _ 50.0 _
- · DATA _~ DATA
40.0 · MODEL 40.0 _ - · MODEL
-
, 30.0 _
o
n
J
~ 20.0
o
10.0
0.0
-
O 30.0
oh
llJ
=
S
g 20.0
~ 10.0!
1 1 1 1
0 10 20 30 40
CORE DEPTH cm
Figure Sc
it'
I I 0.0 1 1 1 1 1 1
50 60 0 10 20 30 40 50 60
CORE DEPTH cm
Figure Ed
FIGURE 5 Analysis of parameter sensitivity. Comparison of model cal-
culations and 1985 sediment profiles of volatile solids or DOT concen-
tration at monitoring station 6C (Figure 1) in response to systematic
variation in model parameters. In all cases, the solid "data" line
represents 1985 LACED monitoring data; the broken "model" line repre-
sents results of the sedimentation (no mixing) model calculated using
the parametric values summarized below; and the broken "mix/dif" line
represents results of the sedimentation/mixing model corresponding to
the parameter estimates given.
OCR for page 190
190
10r ~
bolt
105
y
1 04
3
10
1o2
1o1
1oo
1n7 ~
+ DATA 107
_ MODEL
Hi_ MIX/DIF 1 o6
~ \L
f
105
r', 1 04
c, 10
- ~ DATA
· MODEL
— ~ MIX/DIF
:\\
101 _
1 1 1 1 1 1 10° 1 1 1
0 15 30 45 60 75 90 0 15 30 45 60 75 90
CORE DEPTH cm CORE DEPTH cm
Figure Se Figure 5f
102 ~
c
107 l=
_ — + DATA ~ ~ DATA
= _ MODEL · MODEL
1o6 = —~ MIX/DIF 1o6 · MIX/DIF
~ ~ 0~°~
10° _ ~ 10° I I I ~
0 15 30 45 60 75 90 0 15 30 45 60 75 90
CORE DEPTH cm CORE DEPTH crn
Figure 5g Figure 5h
FIGURE S. Continued
Summary of parameter estimates (by case):
Figure Background sedi-
desig- mentation rate
nation (mg/cm /yr)
% Effluent solids Diffusion Co- Mixing
depos iced in grid efficient depth
sector (cm /see) (cm)
5A 250 0.5 - -
SB 750 0.5
5C 500 0.3 - -
5D 500 0.7
5E 500 0.5 10- -
5F 500 0.S 10-7 -
5G 500 0.5 10 8 4 cm
5H S00 0.5 10-8 8 cm
NOTES:
The percent effluent solids deposited in the grid sector containing
station 6C was uncorrected for degradation of volatile solids.
Surface mixing was omitted from cases designed to test model
sensitivity to variation in the molecular diffusion coefficient.
OCR for page 191
191
discharge has contributed significantly to the accumulation of solids
on the bottom in the vicinity of the outfall. Based on the current
discharge rate and effluent quality, estimates of volatile and refrac-
tory solids fractions, etc., outfall-related solids now comprise about
9 percent of the total particulate flux to the bottom at station 6C.
From the record of JWPCP solids emissions, it is apparent that this
percentage was much higher prior to implementation of partial secondary
treatment. Effluent-related contributions at more distant stations are
much less significant.
Model simulations indicate that the effect of pore diffusion on She
distribution of DDT in Palos Verdes sediment is minimal. At D = 10-
cm /see, molecular diffusion smooths the calculated profile without
mitigating the steep (calculated) drop in DDT concentration near the
surface. If the diffusion coefficient is reduced by an order of
magnitude, its effect is negligible (Figure be). Increasing the
diffusion coefficient by a factor of 10 (Figure 5f) results in a
reasonable representation of near-surface DDT levels at station 6C but
broadens the DDT peak and extends elevated concentrations to depths
that are not supported by monitoring data. The estimated diffusion
coefficient (D ~ 10- cm /see) represents an upper limit since values
at and below that level are virtually indistinguishable.
DDT profiles are better reproduced by creating a well-mixed zone of
arbitrary dimensions among the surface sediments. The physical basis
of such a zone lies in either frequent sediment resuspension (unlikely
in light of the apparent depth of the mixing zone) or bioturbation.
The estimated depth of such a well-mixed or biologically active zone at
station 6C is 6 cm. The primary effect of increasing or decreasing the
dimension of the mixing zone is evidenced in the estimated surface DDT
concentration, which increases nearly an order of magnitude in response
to variation in mixing-zone depth between 4 and 8 cm (Figure 5g and
5h).
Sedimentation model (no mixing) results were included in Figure
5(e-h) to illustrate further the importance of mixing to accurate
representation of surface sediment DDT concentrations. In Figure Se
and Sf, surface mixing effects were eliminated in order to isolate the
effect of order-of-magnitude-scale variation in diffusivity.
Insight and Limitations
Although the sedimentation/mixing model (1) is capable of repro-
ducing sediment profiles of volatile solids and DDT concentrations, and
(2) exhibits a satisfactory degree of sensitivity relative to selection
of parameter estimates, it provides an incomplete mechanical framework
within which to identify the primary determinants of DDT distribution.
Nevertheless, we feel that the following mechanistic observations
relative to model behavior are justified:
1. Physical rather than chemical processes will determine the
long-term fate of DDT among sediments of the Palos Verdes
OCR for page 192
192
shelf. Chemical factors of practical importance are that (1)
DDT will not chemically or biochemically degrade within the
foreseeable future and (2) surface affinity or hydrophobic
effects ensure that DDT does not behave as a solute.
. The primary physical determinants of DDT distribution along the
61-m depth contour are discrete particle sedimentation and sedi-
ment mixing due to bioturbation. No attempt was made to ration-
alize either process--natural and outfall-related sedimentation
rates were developed via parameter fitting using sediment pro-
files of volatile solids; bioturbation was simulated by creating
a well-mixed zone of arbitrary depth among the surface
sediments.
3. Natural sedimentation rapidly decreases with distance beyond the
break of the Palos Verdes shelf.
4. There is little spatial variation in the background sedimenta-
tion rate along the 61-m contour suggesting that (1) local
variation within the shoreline erosion rate is not a primary
determinant of sediment accumulation at that depth and (2)
effluent-related solids are not responsible for appreciable
flocculation of slow-settling natural particulates.
I. The sedimentation/mixing model could not adequately reproduce
the sediment profiles of volatile solids and DDT at the ED
(Figure 1, 30-m depth) sampling location (data not shown). That
site is characterized by unusually high surface concentrations
of volatile solids and DDT. It is possible that in shallower
waters processes not considered here, such as wave-driven
sediment resuspension, are responsible for re-exposure of buried
materials, including relatively high-level DDT concentrations.
Extensions
Results suggest that there is a great deal to gain from an approach
to sediment modeling in which component physical, chemical, and biolog-
ical processes are added sequentially to the overall model. Such an
approach is especially useful when potentially important unit processes
are complex--when existing data do not support a mechanistic treatment.
The deposition of effluent-related solids on the Palos Verdes shelf
has been successfully modeled by several investigators (Hendricks,
1978, 1982, 1984; Koh, 1982~. Comparison of LACED sediment core
measurements with quality characteristics predicted on the basis of
effluent quality data and sedimentation pattern predictions alone
were used to assess the importance of processes (sediment resuspension,
biological mixing) that are less amenable to mathematical modeling.
The importance of surface-mixing processes as determinants of the ver-
tical distribution of sediment DDT is clear from the exercise. Sedi-
ment mixing was then treated empirically to reproduce sediment pro-
files. From the latter analysis, it is apparent that periodic sediment
resuspension is an important determinant of the vertical distribution
of sediment contaminants at the shelf 3)-m depth contour. A similar
approach might be applied to sediment modeling in any situation in
OCR for page 193
193
which sediment quality data are more reliable or more easily obtained
than the physical data necessary for construction of mechanistic
models.
In summary, rational treatment of the fate of sediment contaminants
awaits development of an adequate base of physical oceanographic data.
Shortcomings in this area, even among the most heavily studied coastal
sediments in the country, were outlined in the text. Absent reliable
theory or supporting data, empirical approaches based in sediment chem-
istry can identify determinants of sediment quality and contaminant
distribution. In cases where current-driven sediment resuspension can
be neglected, such models permit prediction of sediment quality changes
following perturbations in appropriate forcing functions. Improvements
in existing theory and a greater commitment to collection of physical
data are necessary to support rational model--when sediment resuspen-
sion is an important determinant of sediment quality.
In light of Marine Board instructions to consider engineering reme-
dies for contaminated sediments, it is worthwhile considering the magni-
tude of problems imposed by contaminants in place on the Palos Verdes
shelf and the likelihood of finding engineered solutions to these prob-
lems. On the bases of specific chemical and biological indicators of
sediment quality, Palos Verdes sediments are among the most impacted in
the nation. Dr. Zarba's work (this volume) on the severity of national
sediment contamination has produced suggested contaminant levels for
categorizing the severity of DOT and metals contamination in
sediments. A significant portion of the shelf sediments fall among
those most heavily contaminated in the nation using either metals or
DDT criteria. Dr. Robertson's mussel tissue measurements of metals and
pesticides (also in this volume), which were not designed to identify
contaminant "hot spots" but to suggest portions of the country likely
to contain a high level of general sediment contamination, found that
animals exposed to Palos Verdes contaminants contained the highest
levels of total DOT and cadmium encountered in his survey. Despite the
apparent severity of chemical effects, economic impacts and human
health effects of shelf contamination are modest or unproven.
Nevertheless, these sediments would be granted high priority for -
remedial activities should remediation measures prove economically
feasible and cost effective. They will not.
The areal extent of severely contaminated sediments (by Dr. Zarba's
criteria and others) is on the order of 10 km (alongshore) by 4 km
(cross - shore ~ . Contaminant removal would demand dredging to an average
depth of about 25 cm over that region. Thus th7 total volume of sedi-
ments to be removed would be on the order of 10 m . or a
cube roughly 2.5 football fields in each dimension. Could these be
removed, they would present a remarkable disposal problem. The feasi-
bility of dredging from 30 to 150 m depth is itself uncertain. Dr.
Herbich ~ this volume ~ indicated that modern dredge capabilities encom-
pass removal of 2, 000 m /hr~ from a depth of 20 m. Even if this
capacity could be extended to the depths of the contaminated sediments
it would take seven months of continuous operation of the best
equipment known to remove this material.
OCR for page 194
194
To cap the same area wit) clean material to a depth of even 0.6 m
would require a minimum 2x10 m of clean sediment, and the environmental
problems associated with sediment resuspension during such an activity
might be exceptionally severe. Engineered remediation strategies of a
nature applicable in shallower waters in which contamination is more
contained are inappropriate in this setting.
SUMMARY AND CONCLUS IONS
The primary physical determinants of the distribution of refrac-
tile, nondiffusing chemicals among sediments of the Palos Verdes shelf
61-m depth contour include sedimentation of natural and outfall-related
particles and bioturbation of surface sediments. Processes of secon-
dary or perhaps negligible importance to the long-term fate of sediment
DDT at that depth include sediment resuspension and diffusion through
sediment pores. In shallower water, sediment resuspension may be an
important factor.
The background or natural sedimentation rate along the 61-m depth
contour is approximately 500 dry mg/cm lyre Flocculation of
natural particles due to the discharge of effluent-related solids is
not apparent. The background sedimentation rate decreases rapidly
beyond the shelf break.
No matter what treatment scenario is selected for the JWPCP, the
flux of natural particles to the bottom will gradually diminish the
surface sediment concentration of DDT along the 61-m depth contour with
attendant benefits for fish tissue concentrations of DDT and sediment
infaunal ecology. However, the rate at which DDT will decrease at the
surface is sensitive to the solids HER from the Whites Point outfall
system. Implementation of full secondary treatment at JUPCP results in
nearly a 7S percent increase in the predicted, year-2005 surface sedi-
ment concentration of DDT at station 6C relative to a partial secondary
treatment (or no-change) scenario.
ACKNOWLEDGMENTS
Funding for this research was provided by Los Angeles County Sanita-
tion Districts. We thank Irwin Haydock and Jan Stull for general assis-
tance and advice. The manuscript was prepared by Mrs. Sharon Solomon
of the University of Arizona Civil Engineering staff. Portions of this
work were presented at the Sediment Dynamics Workshop, October 19-21,
California State Polytechnic University, Pomona, California.
REFERENCES
Cross, J. N. 1984. Tumors in fish collected on the Palos Verdes
shelf . Southern California Coastal Water Research Proj ect Biennial
Report, 1983-1 984, W. Bascom, ed. Long Beach, California: SCOURS.
OCR for page 195
195
Gossett, R. W., H. W. Puffer, R. H. Arthur, J. F. Alfafara, and D. R.
Young. 1982. Levels of trace organic compounds in sportfish from
southern California. Southern California Coastal Water Research
Project Biennial Report 1981-1982, W. Bascom, ed. Long Beach,
California: SCCWRP.
Gossett, R. U., H. W. Puffer, R. H. Arthur, and D. R. Young. 1983.
DDT, PUB and benzo~a~pyrene levels in white croaker (Genyonemus
lineatus) from Southern California. Mar. Poll. Bull. 14~29:60-65.
Hendricks, T. J. 1988. Development of methods for estimating the
changes in marine sediments as a result of the discharge of sewered
municipal wastewaters through submarine outfalls : Part II , resuspen-
s~on processes, draft copy. Final report for U.S. EPA.
Hendricks, T. J. 1984. Predicting sediment quality around outfalls.
Southern California Coastal Water Research Project Biennial Report
1983-1984' W. Bascom, ed. Long Beach, California: SCCWRP.
Hendricks, T. J. 1982. An advanced sediment quality model. Southern
California Coastal Water Research Project Biennial Report
1981-1983, U. Bascom, ed. Long Beach, California: SCCURP.
Hendricks, T. J. 1978. Forecasting changes in sediments near outfalls.
Southern California Coastal Water Research Proj ect Annual Report
1978, W. Bascom, ed. Long Beach, California: SCCWRP .
Keith , J . O ., L. A. Woods , and E . G . Hunt . 1970 . Reproductive failure
in brown pelicans on the Pacific coast. Trans. North Am. Wildl,
Nat . Res our . Conf . pp . 56 - 6 ~ .
Koh, R. C. V. 1982 . Initial sedimentation of waste particulates dis -
charged from ocean outfalls. Environ. Sci. Technol. 16:757-763.
Myers, E. P. 1974. The concentration and isotopic composition of carbon
in marine sediments affected by a sewage discharge. Ph.D. Thesis.
California Insti tute of Technology, Pasadena, California .
Oshida, P. S. and J. L. Wright. 1976. Acute responses of marine inverte-
brates to chromium. Southern California Coastal Water Research
Proj ect Annual Report 1976 . E1 Segundo, California: SCCWRP .
Risebraugh, R. W., O. B. Mengel, D. J. Martin and H. S. Olcott. 1967.
DOT residues in Pacific seabirds: A persistent insecticide in
marine food chains . Nature 216: 589 - 591.
Risebraugh, R. W., F. C. Sibley, and M. N. Kirven. 1971. Reproductive
failure of the brown pelican on Anacapa Island in 1969. Amer. Birds
25: 8 - 9.
Sherwood, M. 1976 . Fin erosi on disease induced in the laboratory.
Southern California Coastal Water Research Project Annual Report
1976. E1 Segundo, California: SCC~RP.
Stull, J. K., C. I. Haydock, R. W. Smith, and D. E. Montagne. 1986.
Long-term changes In the benthic community on the coastal shelf of
Palos Verdes, Southern California. Mar. Biol. 91:539-551.
Word, J. Q. 1978. The ~nfaunal trophtc index. Southern California
Coastal Water Research Project Annual Report 1978, U. Bascom, ed.
Long Beach, California: SCCWRP.
Young, D. R., R. W. Gossett, and T. C. Heesen. 1988. Persistence of
chlorinated hydrocarbon contamination in a California marine eco-
system. In Urban Wastes in Coastal Marine Environments, Vol. 5,
Oceanic Processes in Marine Pollution Series, D. Wolfe and T P.
O'Connor, eds.
OCR for page 196
196
APPENDIX
MODELING APPROACH FOR SEDIMENTATION AND
SEDIMENTATION/MIXING MODELS
Computational procedures are summarized below:
1. The uncorrected flux of outfall-related solids to specific sediment
locations, Suncorr (mg/cm -yr), is estimated as follows
fraction of effluent
Suncorr solids which reach (solids MER) ~ 1
the designated area LareaJ
- [solids fractions SS x 105 MT x 103kg x 106 mg
yr MT kg
x 1
X {I l X I l
0.25 ~ 1 km2 103m J ~ lOOcm J
Suncorr [ g ~ ~ (solids fractions) (SS) x 40,000
cm2-yr
Notes: 1. The fraction of effluent solids reaching specific grid
sectors or "solids fraction" was initially estimated from
the modeling efforts of LACED personnel using Hendricks
model for sediment deposition on the Palos Verdes shelf.
Values were subsequently fitted to reproduce volatile
solids profiles in sediment cores.
2. Grid-sector dimensions are 0.25 km x 1.0 km.
3. SS values, or suspended solids mass emission rates at
JWPCP are taken from Figure 3.
2. Biological degradation of solids reduces the volatile solids
content of discharged solids by the fraction fVSd in the water
column prior to deposition. Thus, the total solids of sewage
origin which reach a core site, Ssr (mg/cm2-yr) is:
Ssr Suncorr (1 - fVS x fad)
where fVS is the ratio of volatile solids to total solids in JWPCP
effluent (estimated at 70 percent from LACED records ~ .
3. The flux of volatile solids of sewage origin to the sediment site
is given by:
VSsr (SUnCOrr) (fVS) (1 - fVSd)
OCR for page 197
197
4. The total flux of solids to a site is the sum of the sewage-
related solids and the background (or natural) sediment flux:
Sflux Ssr + Snat
Note: Snag was originally estimatesd from the depth of elevated
DOT concentrations at the TIC monitoring site. Values
were subsequently fitted to reproduce sediment volatile
solids profiles.
The volatile solids content of natural particulate material
(fVSnat) was estimated at 7 percent (Hendricks, 1984~. The
contribution of background sedimentation to the local flux of
volatile solids is equal to Snat times f~Snat. That is:
Reseat = Snat X f\JSnat
6. The average volatile solids content (VS in percent) of solids which
reach the sediments is therefore:
VS ~ VSsr + VSnat X 100
Sflux
7. The water content (M, in percent) and the sediment wet density (p in
g/cm3) were estimated using the following empirical relationships:
M ~ 1.15 VS ~ 26.238
p ~ -0.01673 VS + 1.7938
Note: Relationships shown were determined using LACSD sediment
data and a linear regression algorithm.
8. The depth of sediment (DEPtot in cm) added during time intervals
of 1 year is given by:
DEPtot c Sflux . ~0~3
{1 - 0.01M) p
9. The diffusion of DDT within sediments was calculated using a one-
dimensional forward explicit finite difference equation:
ct ~x' ~ C (x) -
DAt tC`x+ly ~ C(x-l)
Ax2
- 2 C(x)]
OCR for page 198
198
where C is the DDT concentration in the sediment [pg/kg],
ct is the new concentration after a time interval At ~ 1
day, Ax ~ 2cm is the node spacing, with 40 nodes per
sediment profile. No loss of DDT from sediments was assumed.
10. Mixing was incorporated by equally distributing DDT within the
designated number of 2 cm sediment compartments over which mixin
would occur. New sediment was incorporated into the mising mode
only after 2 cm of new material had accumulated.
NOTATION
Influx
VSnat
AS
M
p
DEPtot
C
D
-
~x
At
~1
solids flux (uncorrected from outfall (mg/cm2-yr)
Mass emission rate (metric tons/yr)
fraction of volatile solids that are biologically degraded
fraction of total solids that are volatile
volatile solids flux of sewage origin to sediment (mg/cm2-yr)
solids flux of sewage origin to sediment (mg/cm2-yr)
solids flux of natural origin to sediment (mg/cm2-yr)
fraction of volatile solids of natural (non-sewage) origin in
sedimenting material
sum of sewage and natural (non-sewage) sediment solids
(mg/cm2-yr)
volatile solids flux of natural origin to sediment (mg/cm2-yr)
percent of sediment solids that is volatile
moisture, or percent water content of sediment
sediment wet density (g/cm3)
depth of sediment accumulation each year (cm)
concentration of DDT in sediment (pg/kg)
diffusivity (cm2/s)
distance between nodes in finite difference model (cm)
time integral in f inite difference model ~ d)
Representative terms from entire chapter:
palos verdes
