Morris & Fan. Reservoir Sedimentation Handbook
Подождите немного. Документ загружается.
SEDIMENT DEPOSITS IN RESERVOIRS 10.31
initial weights for clay, silt, and sand, respectively. As an example, in a continuously
submerged zone within the reservoir which traps 23 percent clay, 40 percent silt, and 37
percent sand, the solution is
W (416)(0.23)
(1120)(0.40)
(1550)(0.37)
1117kg/m
3
Because sediment composition can vary dramatically from one area of the reservoir to
another, and higher-elevation areas may be air-dried whereas deeper layers may be
continuously submerged, the reservoir should be divided into the appropriate zones for a
more accurate estimation of overall initial weight.
10.9.4 Sediment Compaction
It may be desired to know: (1) the bulk density of a sediment deposit at the end of a
specified consolidation period or (2) the average weight over a time interval. Lane and
Koelzer (1953) presented an empirical formula for the density-time relationship which takes
into account the grain size of the sediment and the method of operating the reservoir:
W
W
1
B
logt (10.13)
in which W = specific weight of a deposit with an age of t years; W
1
= initial weight,
usually taken to be the value after 1 year of consolidation; and B = constant. Both W
1
and
B are functions of sediment size and they are defined for four operational conditions: (1)
sediment submerged continuously or almost continuously, (2) normally moderate reservoir
drawdowns, (3) normally considerable reservoir drawdown, and (4) reservoir normally
empty. Parameter values are given in Table 10.4. For sediment containing more than one
TABLE 10.4 Coefficient Values for Consolidation Calculation
B, kg/m
3
(lb/ft
3
)
Operational condition Sand Silt Clay
Continuously submerged 0 91 (5.7) 256 (16)
Periodic drawdown 0 29 (1.8) 135 (8.4)
Normally empty reservoir 0 0 0
size class, the weighted value of the coefficient B should he computed. For the previous
example of continuously submerged sediment with 23 percent clay, 40 percent silt, and 37
percent sand:
B (0.37)(0)
(0.40)(91)
(0.23)(256)
95.3kg/m
3
In computing the average compaction over a period of time, each year's sediment deposits
will have a different compaction time. The average density of all sediment deposited
during t years of consolidation may be computed using the equation presented by Miller
(1953):
W
t
W
1
0.4343B
t
t 1
(lnt)1
(10.14)
Where W
t
= average specific weight after t years of consolidation, W
1
= initial specific
weight, and B = constant as given in table 10.4. For the previous example, with W
1
=
1117 kg/m
3
and B = 95.3, the average specific weight of the deposit after 100 years is
computed by
SEDIMENT DEPOSITS IN RESERVOIRS 10.32
W
100
1117 (0.4343)(95.3)[(100/99)(4.61)
1]
1268kg/m
3
This equation incorporates the assumption that sediment accumulates at a constant rate every
year. The error introduced by this assumption is negligible over a period of many years,
but over shorter periods in environments with highly irregular sediment inputs, year-by-
year computations can be made with an electronic spreadsheet and Eq. (10.14)
10.10 EMPIRICAL PREDICTION OF DEPOSITION PATTERNS
Sediments are deposited in reservoirs at all elevations, causing the stage-capacity curve to
shift. Empirical methods have been developed to distribute sediment deposits within a
reservoir as a function of depth, thereby projecting the shift in the stage-storage curve. These
methods are much quicker and easier to use than mathematical modeling and also require less
data. When sediment survey data are available for an existing reservoir, the observed
deposition pattern can be used to select the proper empirical relationship to compute the
future shift in the stage-area and stage-capacity relationships. As a limitation, empirical
methods do not identify the specific locations in a reservoir which will be affected by
sediment; they predict only the change in the stage-area and stage-capacity curves. A
significant shift in the operating regime, such as implementation of sediment management,
will affect the deposition pattern. Empirical methods cannot be user to simulate these effects,
and the evaluation of management alternatives requires numerical modeling.
10.10.1 Deposits in Flood Control Pool
Reservoir levels are normally held below the flood control level, resulting in small
amounts of sediment accumulation in the portion of the impoundment between the
normal and maximum (flood) pool levels. However, large amounts of sediment can be
delivered during floods, and if the pool level is held above the normal pool for a
significant period of time a substantial amount of sediment may he deposited in this zone.
Data from 11 reservoirs in the Great Plains area of the United States are plotted in Fig 10.18,
which shows the percentage of total sediment inflow trapped within the flow control pool as
a function of the flood pool index, computed as
Depth of Pool
( )
(percent of time in flood pool) In
dex =
Dept
h Below Flood Pool
The
graphed relationship is only a rough guideline and reservoirs having different
geometry (e.g., mountain reservoirs) may depart from this relationship.
10.10.2 Area-increment and Empirical Area Reduction Methods
Empirical methods to distribute sediment below the normal pool elevation developed by the
U.S. Bureau of Reclamation have been widely used. They are based on a four-step
process:
SEDIMENT DEPOSITS IN RESERVOIRS 10.33
FIGURE 10.18 Percent of sediment trapped within flood control pool in 11 reservoirs in the U.S.
Great Plains region (Strand and Pemberton. 1987).
1. Det
ermine the amount of sediment to be distributed. In an existing reservoir, the
volume of sediment might be estimated by assuming that the observed historical
rate of volume loss will continue into the future. At a new site, volume loss
would be estimated from the predicted inflowing load, sediment release
efficiency, and deposit specific weight.
2. On the basis of the site characteristics, select the appropriate empirical curve for
sediment distribution. In an existing reservoir with surveyed sediment deposits,
the historical deposition pattern is used as the basis for selecting the empirical
curve for distributing future sediment deposition.
3. Determine the height of sediment accumulation at the dam, termed the new
zero-capacity elevation.
4. Use the selected empirical curve to distribute sediment as a function of depth
above the new zero-capacity elevation. These values are then subtracted from the
original stage-area and stage-capacity curves to produce the adjusted curves.
This entire procedure is repeated for each sediment volume to be analyzed. These
computations are readily adapted to solution by electronic spreadsheet.
The first empirical method developed was the area-increment method, which uses
the assumption that an equal volume of sediment will be deposited within each depth
increment in the reservoir. From resurveys of many reservoirs, the Bureau found that the
deposition pattern varied from one site to another in a somewhat predictable fashion.
Reservoir geometry, operation, and sediment grain size all affect sediment distribution
SEDIMENT DEPOSITS IN RESERVOIRS 10.34
FIGURE 10.19 Type curves for the empirical area-reduction method for estimating storage loss
due to sediment deposition (Strand and Pemberton, 1987).
wi
thin the impoundment, and four different empirical type curves were developed on the
basis of these characteristics. Use of the appropriate type curve can produce a more
realistic sediment distribution than the area increment method. The four type curves
developed from reservoir resurvey data are shown in Fig. 10.19 along with the curve
corresponding to the area-increment method. Notice that the area-increment curve is very
similar to the Type II curve. Both empirical methods were described by Borland and
Miller (1958). The empirical area reduction method was revised by Lara (1962), and
the example given here using the Theodore Roosevelt Dam located on the Salt River upstream of
Phoenix, Arizona, is based on Strand and Pemberton (1987).
Data for the example computations at Theodore Roosevelt Dam are given in Table 10.5.
The dam was constructed in 1902 with an original capacity of 1888 Mm
3
to the top of the
conservation pool, located 71.3 m above the original streambed elevation at the dam. The
original stage-area and stage-capacity curves are given in columns 1, 2, and 3 of Table 10.5.
The sediment survey conducted in 1981 determined that 239 Mm
3
of sediment had
accumulated during the 72.4 years since closure. In the example problem the distribution for
100 years of sediment accumulation will be calculated.
SEDIMENT DEPOSITS IN RESERVOIRS 10.35
New zero-capacity elevation.
Source: After Strand and Pemberton (1987).
TABLE 10.5 Empirical Area Reduction Method, Sample Computations, Theodore Roosevelt Dam
SEDIMENT DEPOSITS IN RESERVOIRS 10.36
Either the area-increment method or the empirical area reduction method can be solved
by performing the following steps. The difference between the methods is the type curve
selected to distribute sediment.
1. Determine sediment inflow. The sediment volume to be distributed within the
reservoir must be determined. At the Theodore Roosevelt Dam, the sediment survey
data were extrapolated to determine an average volume loss of 330 Mm
3
per 100
years.
2. Select design curve. Plot the original depth-capacity relationship on log-log paper
and calculate the slope m of the fitted line, which is the reciprocal of the slope of the
depth versus capacity plot (Fig. 10.20). Use the resulting slope m to classify the
reservoir shape using the table below.
m Reserv
oir shape Type
Lake I 3.5-4.5
Floodplain-foothill II 2.5-3.5
Hill and gorge III 1.5-2.5
Gorge IV 1.0-1.5
The type curves in Fig. 10.19 reflect the tendency for sediment in lake-type reservoirs to
accumulate in shallower water and in gorge-type reservoirs to accumulate in deeper water.
When the slope m does not plot as a straight line, use the shape type corresponding
to the predominate overall slope, or the slope in the area of the reservoir where most of
FIGURE 10.20 Reservoir depth-capacity relationship as used in the empirical area-
reduction method (Strand and Pemberton, 1987).
SEDIMENT DEPOSITS IN RESERVOIRS 10.37
the sediment will deposit. Strand and Pemberton state that the reservoir does not change
type with continued sediment deposition, unless reservoir operation changes. Thus, the stage-
capacity plot should be based on the original reservoir bathymetry, not the bathymetry
following sediment accumulation.
Classify the reservoir operation as stable pool, moderate drawdown, considerable drawdown,
or normally empty. Giving equal weight to reservoir shape and reservoir operation,
determine the weighted reservoir type from the table below.
Operational
class
Shape
class
Weighted
class
R
eservoir Operation
Sediment submerged (continuous high pool level) I I
II
III
I
I or II
II
Moderate drawdown II I
II
III
I or II
II
II or III
Considerable drawdown III I
II
III
II
II or III
III
Normally empty IV All IV
Where a choice of two types is given, or in borderline situations, select the type according
to whether reservoir shape or reservoir operation is felt to be more influential at the site.
For the Theodore Roosevelt Dam, a Type II classification was chosen. Sediment grain
size may also be considered using the following guideline:
Predominant Grain size Type
Sand or coarser I
Silt II
Clay III
In most river basins the grain size distribution has been found to be the least important factor
influencing sediment distribution, and grain size should be used as an auxiliary variable to
select the weighted type curve only in those cases when there is a choice between two type
numbers.
For an existing reservoir having a sediment survey, plot the survey results on Fig.
10.19 to determine which empirical curve should be used. By plotting the data from
Theodore Roosevelt Dam, it was confirmed that the Type II empirical curve should be used
at that site.
3. Compute new zero-capacity elevation at dam. Use the original elevation-area
and capacity curves to compute the value of the dimensionless function F at several
different pool elevations in the deeper part of the reservoir:
F
S
V
h
HA
h
(10.15)
SEDIMENT DEPOSITS IN RESERVOIRS 10.38
where S = total sediment deposition (330 Mm
3
), V
h
= reservoir capacity (m
3
) from
column 3 at each elevation h from column 1, H = original depth of reservoir below
normal pool (71.3 m), A
h
= reservoir area (m
2
, from column 2) at a given elevation h. Note
that hectares must be converted to square meters before computing. Also compute
decimal values for the relative depth p as
p
h
h
min
H
(10.16)
where h
min
= original bottom elevation (579.73 m). Plot the resulting F and p values
(columns 4 and 5) on the type curves presented in Fig. 10.21. The intersection of the
FIGURE 10.21 Type curves for determining the new zero depth at the dam based on the
dimensionless F function. The values for each of the type curves are shown in both graphical
and tabular form (Strand and Pemberton, 1987).
plotted F values with the type curve selected for the reservoir defines the p
0
value for the
new zero-capacity elevation at the dam. The example intersection occurs at p
0
= 0.284.
SEDIMENT DEPOSITS IN RESERVOIRS 10.39
The new zero-capacity elevation is given by h
0
= (p
0
H + h
min
) = (0.284)(71.3) +
579.73 = 600.0m. The zero-capacity elevation is included as a row in Table 10.5, and the
corresponding area is computed from the original area-elevation curve as A
0
= 518.5 ha.
The area-increment curve is often also used to compute the intersection point because
it will always intersect the F curve and offers a good method to check the new zero-capacity
elevation.
4.Distribute sediment. The specified volume of trapped sediment is distributed
within the reservoir according to the selected type curve. Compute the values for relative
sediment area a (Col. #6) at each relative depth p using the appropriate equation:
Type I: a = 5.047p
1.85
(1-p)
0.36
(10.17)
Type II: a = 2.487p
0.57
(1-p)
0.41
(10.18)
Type III: a = 16.967p
1.15
(1-p)
2.32
(10.19)
Type IV: a = 1.486p
-0.25
(1-p)
1.34
(10.20)
Also compute the relative sediment area a at the new zero elevation. (For the example use
Type II, resulting in a = 1.059 when p = 0.284.) Compute the area correction factor as
A
0
/a = 549.1/1.059 = 518.5 ha.
Compute the area at each pool elevation occupied by sediment (column 7) by multiplying
the area correction factor by the relative sediment area (column 6) at each level above the
new zero-capacity elevation. In the fully sedimented part of the reservoir extending from the
new zero-capacity elevation down to the original bottom, the sedimented pool area equals
the original pool area.
Compute the sediment volume for each stage increment above the new zero-capacity
elevation using the end area method (column 8). For example, at elevation 603.50 the
computation is (549.1 + 583.8)(10
4
)/2(603.50-600.0) = 19.83 × 10
6
m
3
. From the zero-
capacity elevation to the reservoir bottom, the sediment volume equals the original reservoir
capacity, since this zone is entirely sedimented. The cumulative volume of deposited
sediment (column 9) is computed by summing the values in column 8. The total sediment
volume should match the predetermined sediment volume from step 1 (V
s
) within about 1
percent.
Compute the revised area and capacity curves (columns 9 and 10) by subtracting the
sediment area (column 7) and cumulative sediment volume (column 9) from the original
area and capacity values (columns 2 and 3).
The original and adjusted curves for Theodore Roosevelt Dam for the example and
other points in time are presented in Fig. 10.22.
10.11 SAMPLING SEDIMENT DEPOSITS
Sediment sampling equipment is required to obtain undisturbed samples of reservoir
deposits for determining parameters such as bulk density, grain size distribution, and
chemical characteristics. The selection of sediment sampling technique depends on the
parameter to be measured, sediment grain size and consolidation (which affects core
penetration), water depth, and sediment thickness. Sediment sampling can be performed from a
boat, raft, or ice cover. Aquatic sampling techniques are reviewed by Mudroch and
MacKnight (1991).
SEDIMENT DEPOSITS IN RESERVOIRS 10.40
FIGURE 10.22 The original and adjusted stage-capacity and stage-area curves for
Theodore Roosevelt Dam. Also, the curve from the 1981 reservoir survey is compared to
the relationship computed from the Type II empirical distribution, indicating good overall
agreement (Strand and Pemberton, 1987).
10.11.1 Planning the Sampling Program
In design of the sampling program, several considerations in addition to cost are
important. The sediment volume required depends on the tests to be performed. For example,
137
cesium counts tend to require large samples, indicating the use of cores on the order of 8
cm in diameter. In sampling for contaminants associated with fine sediment, it may not be
necessary to use coring equipment capable of collecting sands because the contaminants of
interest are not associated with sands. Deeper sand layers and deeper cores may produce a
very different picture of deposit composition compared to surface sampling with a Ponar
dredge, for example, which will sample only the most recently deposited material. In
sampling for any type of contaminants required by regulatory agencies, the most important
preparatory task will be the development of a sampling plan specifying sampling locations,
depths of penetration, sample collection and handling protocols, and laboratory methods
agreeable to all parties involved. This may be the most difficult part of the sampling
program.
10.11.2 Sampling for Chemical Contaminants
In sampling sediments for contaminants, including a variety of organic chemicals and
heavy metals, special sample collection and preservation protocols are required to ensure
the integrity of samples to be analyzed for trace contaminants. For specialized sampling
consult the pertinent regulatory agency (e.g., U.S. Environmental Protection Agency) for