3.

HYDRAULIC DESIGN OF DESILTERS

3.1 BACKGROUND
Sediment transported in the flow, especially particles of hard materials such as
quartz, can be harmful to turbine components. The severity of damage to
equipment is a function of several variables, notably: sediment size, sediment
hardness, particle shape, sediment concentration and plant head.
The control of turbine wear problems due to silt erosion requires a comprehensive
design approach in which sediment properties, turbine mechanical and hydraulic
design, material selection and features to facilitate equipment maintenance are all
considered (Naidu, 2004). Accordingly the design parameters for desilter design
should be made in consultation with the mechanical designers and turbine
manufacturer.
Where the risk of damage is judged to be high a settling basin (or desilter) should
be constructed in the plant waterway to remove particles, greater than a selected
target size.
3.1.1 Need
The first design decision is to determine whether the sediment load in the river of
interest is sufficiently high to merit construction of a desilter. There is little
guidance available on this topic; however, the following limits are suggested by
Naidu (2004):
Table 2.2.3/1.0 Suggested Maximum Allowable Sediment
Concentration
versus Plant Head.
Parameter Low and Medium Head Turbines High Head Turbines
Head 150 m > 150 m
Maximum
allowable
sediment
concentration
200 ppm
150 ppm
3.1.2 Removal Size
There are also considerable divergences of opinion on the selection of design size
for sediment removal. Nozaki (1985) suggests a size range of between 0.3 mm to
0.6 mm for plant heads ranging from 100 m to 300 m. Indian practice is to design
for a particles size of 0.20 m regardless of head. Some authors suggest that
removal of particles smaller than 0.20 mm is not practical.
The adoption of 0.20 mm is the design (target) sediment size is recommended for
Indian SHP designs.
AHEC/MNRE/SHP Standards/ Civil Works - Guidelines For Hydraulic Design Of Small Hydro Plants /May 2011 37

3.1.3 Types of Desilters

There are two basic types of desilters:
Continuous flushing type
Intermittent flushing type
Guidelines for design of both types are given in this section.
3.2. DESIGN CONSIDERATIONS

3.2.1 Data Requirements (Small Hydro Plants)

It is recommended that a program of suspended sediment sampling be initiated
near the intake site from an early stage during site investigations to ensure that
sufficient data is available for design. The sampling program should extend
through the entire rainy season and should comprise at least two readings daily.
On glacier fed rivers where diurnal flow variations may exist, the schedule of
sampling should be adjusted to take this phenomenon into account and the
scheduled sampling times be adjusted to coincide with the hour of peak daily flow
with another sample taken about twelve hours later.
While it is often assumed that sediment load is directly related to flow, this is only
true on the average, in a statistical sense. In fact it is quite likely, that the peak
sediment event of a year may be associated with a unique upstream event such as
a major landslide into the river. Such events often account for a disproportionately
large proportion of the annual sediment flow. Therefore, it would also be
desirable to design the sediment measurement program to provide more detailed
information about such events, basically to increase the sampling frequency to
one sample per 1 or 2 hours at these times.
A five year long sediment collecting program would be ideal. Less than one
monsoon season of data is considered unsatisfactory.
Some authors suggest that the vertical variation of sediment concentration and
variations horizontally across the river be measured. However, on fast flowing
rivers inherent turbulence should ensure uniform mixing and sampling at one
representative point should be sufficient.
The data collected in a sediment sampling program should include:
Mean daily concentration of suspended sediment (average of two readings
twelve hours apart)
Water temperature
Flow (from a related flow gauging program)
The following additional information can then be derived from collected samples.

 A sediment rating curve (sediment concentration versus flow where

possible)
Particle size gradation curve on combined sample
Specific gravity of particles.
It is also recommended that a petrographic analysis be carried out to identify the
component minerals of the sediment mix. It is likewise recommended that
experiments be made on selected ranges of particles sizes to determine settling
velocities. A further discussion on the subject of sediment sampling is given in
Avery (1989)
The characteristics of the sediment on a given river as obtained from a data
collection program will assist in selection of appropriate design criteria.
3.2.2 Data Requirements (Mini Hydro Plants)
On mini hydro projects where resources and time may not be available to
undertake a comprehensive sampling program, selection of design parameters will
depend to a great extent on engineering judgment, supplemented by observations
on site and local information. The following regional formula by Garde and
Kothyari (1985) can be used to support engineering decision making.
Vs = 530.0 P0.6. Fe1.7.S0.25
0.19

P
DPd
Where
Vs = mean sediment load in (tonnes/km2/year)
s = average slope (m/m)
Dd = drainage density, as total length of streams divided by
catchment area (km/km2)
P = mean annual precipitation (cm)
Pmax = average precipitation for wettest month (cm)
Fe = ground cover factor, as below:
Fe = 1 [0.80 0.60 0.30 0.10 ]
AGFWA A A A
Ai
max 10 . 0

+ + +
AA = arable land area
AG = grass land area (all in km2)
AF = forested area
AW = waste land area (bare rock)
3.2.3 Design Criteria
The principle design criteria are:
1. The target size for removal (d): d = 0.20 mm is recommended
2. Flushing flow: QF = 0.2 QP is recommended
3. Total (design) flow: QT = QP + QF = 1.2 QP.

Where QP is plant flow capacity in (m3/s).

AHEC/MNRE/SHP Standards/ Civil Works - Guidelines For Hydraulic Design Of Small Hydro Plants /May 2011 39

3.2.4 Siting
The following factors control site selection
1. A site along the water way of appropriate size and relatively level with respect
to cross section topography
2. A site high enough above river level to provide adequate head for flushing.
For preliminary layout a reference river level corresponding to the mean
annual flood and minimum flushing head of 1.50 m is recommended.
In principle a desilting tank can be located anywhere along the water conductor
system, upstream of the penstock intake. Sometimes it is convenient to locate the
desilting basin at the downstream end of the waterway system where the desilter
can also provide the functions of a forebay tank. However, a location as close to
the head works is normally preferred, site topography permitting.
3.3 Hydraulic Design
A desilter is made up of the following elements:
Inlet section
Settling tank
Outlet section
Flushing system
3.3.1 Inlet Section
The purpose of the inlet section is to reduce flow velocity from the relatively high
speed of the feeder canal to the low speed of the settling tank. For efficient
functioning of the settling tank the velocity should be as uniform as possible
without short circuits or localized high velocity areas. Where possible,
introducing flow into the settling section via a distribution weir or diffuser wall is
preferred. Alternatively, transition structure with walls diverging at a rate of 6:1 is
recommended. A design with vanes may also be considered, as shown in Figure.

Hydraulic losses in the inlet transition can be estimated as:

22

2
0.3
VV
g
h=
Where:
VF = velocity in feeder canal (m/s)
VT = velocity in settling basin (m/s)
3.3.2 Settling Section
The fundamental design objective is to remove all particles equal or greater than
the chosen target removal size (d). The methodology recommended follows the
approach given by Mosonyi:
a) Flow velocity in the tank should not entrain material that has settled out to the
bottom of the tank
Thus U a d
Where, U = velocity through tank (m/s)
d = target sediment size in (mm)
a = 0.36 for d > 1.0 mm
= 0.44 for 1.0 mm > d 0.10 mm
= 0.51 for d < 0.10 mm
b) Select fall velocity (wo) for d from Figure 2.2.3.2, assuming an appropriate
water temperature.
LFT

Vary the value of B to optimize the layout including: inlet section, settling basin
and outlet section. A length to width ratio

B
L of 8 to 10 is preferable, but the
minimum

B
L ratio should not be less than 4.0.
3.3.3 Outlet Section
The outlet section provides a transition between the settling tank and power canal.
A transition with walls converging at 2:1 will be satisfactory.
Hydraulic losses can be estimated as:

=
g
V
g
h VP T
L2 2
0.2
22

Where: VP = velocity in power canal (m/s)

VT = velocity in settling basin (m/s)
3.3.4 Flushing system Continuous Flushing Type
The recommended flushing system comprises a series of hoppers built into the
base of the settling tank with side slopes of 1:1 leading to a central outlet at the
bottom of the hopper. Flushing flow is withdrawn from the bottom of the hopper
and controlled by a manually operated valve, one for each row of hoppers. The
flushing system may be designed for either pressurized or non pressurized flow.
Where head is available the non pressurized flow design is to be preferred since
water passages can be made larger and therefore are easier to maintain. The usual
design procedure is to assume equal flow through each hopper. Figure 2.2.3.3
shows a typical design (at end of text).
3.3.5 Flushing system Intermittent Flushing Type
The same laws govern the design of intermittent flushing desilters, thus the main
basin dimensions can be obtained using the same procedures as outlined in Subsection
2.2.3/3.2. In place of hoppers used in a continuous flushing desilter a
sufficient storage volume must be provided. Determination of this volume should
be based on the incoming sediment load, trap efficiency and frequency of

flushing. It is recommended that this volume be computed from the mean

maximum monthly sediment load as measured or from comparable data from
another plant operating in similar conditions with respect to sediment and water
flows. In converting sediment flows in mass terms to volumes a relative density of
2.65 and a bulking factor of 1.25 are should be applied. Trap efficiency can be
calculated using Camps Sediment Removal Function as given in Figure 2.2.3.4.
The recommended flushing flow is 1.20QP and the flushing gates should be large
enough so as not the throttle this flow.
3.3.6 References:
Water Power Development
AHEC/MNRE/SHP Standards/ Civil Works - Guidelines For Hydraulic Design Of Small Hydro Plants /May 2011 43

Volume 2A: High Head Power Plants

(Pages: 18-26).
By E. Mosonyi
Akadmiai Kiad
Budapest, Hungary (1991)
Sediment Control at Intakes A Design Guide
Edited by P. Avery
BHRA Fluids Engineering Centre
Cranfield, England (1989)
Sediment Erosion from Indian Catchments
By R.J. Garde and U.C. Kothyari
Proc. of 2nd International Workshop on Alluvial River Problems,
Roorkee, India (1985)
Silt Erosion Problems in Hydro Power Stations and their Possible Solutions
By B.S.K. Naidu
Published by the National Power Training Institute,
Faridabad (Haryana), 2004
Estimation of Repair Cycle of Turbine due to Abrasion by Suspended Sand and
Determination of Desilting Basin Capacity
By Tsugo Nozaki
Electric Power Civil Engineering (Japan)
Volume 218 pp 143-152 January 1989.
(Original in Japanese)