HomeMy WebLinkAboutPart 1 Guide on how to develop a small hydropower plant Final 2004 ESHA 2004
Guide on How to Develop a Small
Hydropower Plant
The present document is an updated version developed by the Thematic Network on Small
hydropower (TNSHP) of the Layman’s Guidebook on how to develop a small hydro site, by
Celso Penche1998.
This Guide has been translated by the TNSHP to German, French, and Swedish
European Small Hydropower Association - ESHA - esha@arcadis.be
Tel. +32-2-546.19.45 - Fax +32-2-546.19.47
ESHA is founding member of EREC, the European Renewable Energy Council
ESHA 2004
INDEX
Acknowledgements i
Executive Summary ii
Chapter 1. Introduction 1
Chapter 2. Fundamental of Hydraulic Engineering 12
Chapter 3. Evaluating Stream Flow 42
Chapter 4. Site Evaluation Methodologies 71
Chapter 5. Hydraulic Structures 91
Chapter 6. Electromechanical Equipment 152
Chapter 7. Environmental impact and its mitigation 199
Chapter 8. Economic analysis 236
Chapter 9. Administrative procedures 254
Glossary 290
European Small Hydropower Association - ESHA - esha@arcadis.be
Tel. +32-2-546.19.45 - Fax +32-2-546.19.47
ESHA is founding member of EREC, the European Renewable Energy Council
Guide on How to Develop a Small Hydropower Plant ESHA 2004
ACKNOWLEDGEMENTS
This Guide is an updated and adapted version of the publication “Layman’s Guidebook on How to
Develop a Small Hydro Site”, published by ESHA - the European Small Hydropower Association –
in 1998 in the frame of the European Commission DG-TREN (Directorate General for Transport
and Energy) ALTENER programme.
Although based on the original, this guide has been entirely updated and adapted due to significant
changes in the sector in the latest years as concern environmental and administrative aspects in
particular. The updated version is available in English, French, German and Swedish what has
added value to the already existing Spanish and Italian versions of the original publication.
The “Guide on how to develop a Small Hydro Site” has been carried out within the EC Project
“Thematic Network on Small Hydropower”, financed by the Fifth RD&D Framework Programme
(FP5). It has been updated and adapted by a Revision Committee under the coordination and
guidelines of ESHA. Members of the Revision Committee include the project partners Francis
Armand (ADEME), Anton Schleiss (EPFL-LCH), Erik Bollaert (EPFL-LCH), Pedro Manso
(EPFL-LCH), Jochen Bard (ISET), Jamie O’Nians (IT Power), Vincent Denis (MHyLab), Bernhard
Pelikan (ÖVFK), Jean-Pierre Corbet (SCPTH), Christer Söderberg (SERO), Jonas Rundqvist
(SERO) and Luigi Papetti (Studio Frosio). The network thanks Steve Cryer (BHA) for his input.
Special thanks to Celso Penche (ESHA), author of the Layman’s Guide, who has revised the
contents of the current Guide guaranteeing its consistency and accuracy.
ESHA 2004
i
Guide on How to Develop a Small Hydropower Plant ESHA 2004
EXECUTIVE SUMMARY
Developing a small hydropower site is not a simple task. There are many aspects which have to
be taken into consideration, covering many disciplines ranging from business, engineering,
financial, legal and administration. These will all be necessary at the different development
stages from, first choosing a site until the plant goes into operation.
The “Laymans Guide” guide brings together all of these aspects in a step-by-step approach, and
will serve as a useful tool for a potential developer of a small hydropower scheme.
This guide is divided into nine chapters and covers the basic concepts, meaning of definitions
and technological issues to be addressed.
Chapter 1 – Introduces basic concepts, such as the definition of small hydropower, types of
schemes, ways of exploiting the water resource available and gives a general overview of the
guide’s contents,
Chapters 2 through to 9 – describe the essential steps to be followed to evaluate a proposed
scheme before deciding whether to proceed to a detailed feasibility study.
The basic concepts considered in the guide are:
• Topography and geomorphology of the site.
• Evaluation of the water resource and its generating potential.
• Site selection and basic layout.
• Hydraulic turbines and generators and their control.
• Environmental impact assessment and mitigation measures.
• Economic evaluation of the project and financing potential.
• Institutional framework and administrative procedures to obtain the necessary consents
Reading this guide will inform the potential small hydropower developer and give a better
understanding of the different issues, phases and procedures that need be followed to develop
and run a small hydropower operation.
Bernhard Pelikan
President ESHA
ii
Guide on How to Develop a Small Hydropower Plant ESHA 2004
CHAPTER 1: INTRODUCTION
CONTENTS
1 INTRODUCTION ........................................................................................................................... 2
1.1 A free fuel resource potentially everlasting............................................................................. 2
1.2 Definition of small hydropower ............................................................................................... 3
1.3 Site configurations ................................................................................................................... 3
1.3.1 Run-of-river schemes ....................................................................................................... 3
1.3.2 Schemes with the powerhouse at the base of a dam ........................................................ 5
1.3.3 Schemes integrated within an irrigation canal ................................................................. 7
1.3.4 Schemes integrated in a water abstraction system ........................................................... 8
1.4 Planning a small hydropower scheme...................................................................................... 8
LIST OF FIGURES
Figure 1-1 High head scheme .................................................................................................................. 4
Figure 1-2 Low head scheme with penstock............................................................................................ 4
Figure 1-3 Low head scheme integrated in the dam ................................................................................ 5
Figure 1-4 Low head scheme using an existing dam ............................................................................... 6
Figure 1-5 Low head scheme – siphon intake ......................................................................................... 6
Figure 1-6 Integrated scheme using an irrigation canal ........................................................................... 7
Figure 1-7 Elongated spillway scheme using an irrigation canal ............................................................ 7
Figure 1-8 Scheme integrated in a water supply system.......................................................................... 8
1
Guide on How to Develop a Small Hydropower Plant ESHA 2004
1 INTRODUCTION i
1.1 A free fuel resource potentially everlasting.
Following the “Third Conference of the Parties to the United Nations Framework Convention on
Climate Change” held in Kyoto in December 1997, the European Union has recognized the urgent
need to tackle the climate change issue. It has also adopted a target to reduce greenhouse gas
emissions by 8 % by 2010 from 1990 levels, whereas for other industrialised countries the target is
5 %.
To facilitate the Member States achieving this objective, the Commission identified a series of
actions, focusing on reducing energy consumption and carbon emissions (CO2).
The development of energy from renewable resources is a very important step in the reduction of
CO2 emissions. Therefore the EU Council and Parliament has brought forward Directive
2001/77/EC for the promotion of electricity produced from renewable energy resources
Electricity production from hydropower has been, and still is today, the first renewable source used
to generate electricity. Nowadays hydropower electricity in the European
Union - both large and small scale – represents, according to the White Paper, 13% of the total
electricity generated, so reducing the CO2 emissions by more than 67 million tons a year. But
whereas the conventional hydro requires the flooding of large areas of land, with its consequential
environmental and social issues, the properly designed small hydro schemes are easily integrated
into local ecosystems.
In 2001, approximately 365 TWh of hydro energy was produced in the European Union from an
overall capacity of 118 GW. Small hydro plants accounted for 8.4% of installed capacity (9.9 GW)
and produced 39 TWh (about 11% of Hydropower generation). Given a more favorable regulatory
environment, the European Commission objective of 14000 MW by 2010 should be achievable and
that small hydro would be the second largest contributor behind windpower.
The large majority of small hydro plants are “run-of-river” schemes, meaning that they have no or
relatively small water storage capability. The turbine only produces power when the water is
available and provided by the river.
When the river flow falls below some predetermined value, the generation ceases. Some plants are
stand alone systems used in isolated sites, but in most cases in Europe, the electricity generated is
connected to the grid. Stand-alone, small, independent schemes may not always be able to supply
energy, unless their size is such that they can operate whatever the flow in the river is. In some
cases, this problem can be overcome by using any existing lakes or reservoir storage that exists
upstream, of the plant.
The connection to the grid has the advantage of easier control of the electrical system frequency of
the electricity, but has the disadvantage of being tripped off the system due to problems outside of
the plant operator’s control.
It is possible for grid connected systems to sell either all or some of their energy to supply
company. (Note: this may not necessarily be the grid operator). However, the price paid for this
2
Guide on How to Develop a Small Hydropower Plant ESHA 2004
energy is generally, in Europe particularly, fairly low. In recent years, supported by the RES-e
Directive an in some cases National Government legislation enhanced payments are available for
trading renewable energy states. This has helped small scale developments obtain a reasonable rate
of return on the investment. It has also led to an increase in small scale hydro schemes being
developed.
1.2 Definition of small hydropower
There is no consensus in EU member states on the definition of small hydropower: Some countries
like Portugal, Spain, Ireland and now, Greece and Belgium, accept 10 MW as the upper limit for
installed capacity. In Italy the limit is fixed at 3 MW (plants with larger installed power should sell
their electricity at lower prices) and in Sweden 1.5 MW. In France the limit has been recently
established at 12 MW, not as an explicit limit of SHP, but as the maximum value of installed
power for which the grid has the obligation to buy electricity from renewable energy sources. In
the UK 20MW is generally accepted as the threshold for small hydro. For the purposes of this text
any scheme with an installed capacity of 10 MW or less will be considered as small. This figure is
adopted by five member states, ESHA, the European Commission and UNIPEDE (International
Union of Producers and Distributors of Electricity).
1.3 Site configurations
The objective of a hydropower scheme is to convert the potential energy of a mass of water,
flowing in a stream with a certain fall to the turbine (termed the "head"), into electric energy at the
lower end of the scheme, where the powerhouse is located. The power output from the scheme is
proportional to the flow and to the head.
Schemes are generally classified according to the “Head”:-
• High head: 100-m and above
• Medium head: 30 - 100 m
• Low head: 2 - 30 m
These ranges are not rigid but are merely means of categorizing sites.
Schemes can also be defined as:-
• Run-of-river schemes
• Schemes with the powerhouse located at the base of a dam
• Schemes integrated on a canal or in a water supply pipe
1.3.1 Run-of-river schemes
Run-of-river schemes are where the turbine generates electricity as and when the water is available
and provided by the river. When the river dries up and the flow falls below some predetermined
amount or the minimum technical flow for the turbine, generation
ceases.
Medium and high head schemes use weirs to divert water to the intake, it is then conveyed to the
turbines via a pressure pipe or penstock. Penstocks are expensive and consequently this design is
usually uneconomic. An alternative (figure 1.1) is to convey the water by a low-slope canal,
running alongside the river to the pressure intake or forebay and then in a short penstock to the
turbines. If the topography and morphology of the terrain does not permit the easy layout of a canal
3
Guide on How to Develop a Small Hydropower Plant ESHA 2004
a low pressure pipe, can be an economical option. At the outlet of the turbines, the water is
discharged to the river via a tailrace.
Figure 1-1 High head scheme
Occasionally a small reservoir, storing enough water to operate only on peak hours, when prices
for electricity are higher, can be created by the weir, or a similarly sized pond can be built in the
forebay.
Figure 1-2 Low head scheme with penstock
Low head schemes are typically built in river valleys. Two technological options can be selected.
Either the water is diverted to a power intake with a short penstock (figure 1.2), as in the high head
schemes, or the head is created by a small dam, provided with sector gates and an integrated intake
(figure 1.3), powerhouse and fish ladder.
4
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 1-3 Low head scheme integrated in the dam
1.3.2 Schemes with the powerhouse at the base of a dam
A small hydropower scheme cannot afford a large reservoir to operate the plant when it is most
convenient, the cost of a relatively large dam and its hydraulic appurtenances would be too high to
make it economically viable. But if the reservoir has already been built for other purposes, such as
flood control, irrigation, water abstraction for a big city, recreation area, etc, - it may be possible to
generate electricity using the discharge compatible with its fundamental use or the ecological flow
of the reservoir. The main issue is how to link headwater and tail water by a waterway and how to
fit the turbine in this waterway. If the dam already has a bottom outlet, see figure 1.4, for a possible
solution.
5
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 1-4 Low head scheme using an existing dam
Provided the dam is not too high, a siphon intake can be installed. Integral siphon
intakes (figure 1.5) provide an elegant solution in schemes, generally, with heads up to 10 metres
and for units up to about 1000 kW, although there are examples of siphon intakes with an installed
power up to 11 MW (Sweden) and heads up to 30.5 meters (USA). The turbine can be located
either on top of the dam or on the downstream side. The unit can be delivered pre-packaged from
the works, and installed without major modifications to the dam.
Figure 1-5 Low head scheme – siphon intake
6
Guide on How to Develop a Small Hydropower Plant ESHA 2004
1.3.3 Schemes integrated within an irrigation canal
Two types of schemes can be designed to exploit irrigation canal:-
• The canal is enlarged to accommodate the intake, the power station, thetailrace and the lateral
bypass. Figure 1.6 shows a scheme of this kind, with a submerged powerhouse equipped with a
right angle drive Kaplan turbine. To safeguard the water supply for irrigation, the scheme
should include a lateral bypass, as in the figure, in case of shutdown of the turbine. This kind of
scheme must be designed at the same time as the canal, as additional works whilst the canal is
in full operation can be a very expensive option
Figure 1-6 Integrated scheme using an irrigation canal
• If the canal already exists, a scheme like the one shown in figure 1.7 is a suitable option. The
canal should be slightly enlarged to include the intake and the spillway. To reduce the width of
the intake to a minimum, an elongated spillway should be installed. From the intake, a penstock
running along the canal brings the water under pressure to the turbine. The water passes
through the turbine and is returned to the river via a short tailrace.
Figure 1-7 Elongated spillway scheme using an irrigation canal
Generally, migratory fish are not present in canals, fish passes are unnecessary.
7
Guide on How to Develop a Small Hydropower Plant ESHA 2004
1.3.4 Schemes integrated in a water abstraction system
The drinking water is supplied to a city by conveying the water from a headwater reservoir via a
pressure pipe. Usually in this type of installation, the dissipation of energy at the lower end of the
pipe at the entrance to the Water Treatment Plant is achieved through the use of special valves. The
fitting of a turbine at the end of the pipe, to convert this otherwise lost energy to electricity, is an
attractive option, provided that the water hammer phenomenon is avoided. Water hammer
overpressures are especially critical when the turbine is fitted on an old pressure pipe.
To ensure the water supply at all times, a system of bypass valves should be installed. In some
water supply systems the turbine discharges to an open-air pond. The control system maintains the
level of the pond. In case mechanical shutdown or turbine failure, the bypass valve system can also
maintain the level of the pond. Occasionally if the main bypass valve is out-of-operation and
overpressure occurs, an ancillary bypass valve is rapidly opened by a counterweight. All the
opening and closing of these valves must be slow enough to keep pressure variations within
acceptable limits.
The control system has to be more complex in those systems where the turbine outlet is subject to
the counter-pressure of the network, as is shown in figure 1.8.
Figure 1-8 Scheme integrated in a water supply system
1.4 Planning a small hydropower scheme
The definitive project or scheme comes as the result of a complex and iterative process, where
consideration is given to the environmental impact and different technological options. These are
then costed and an economic evaluation carried out.
8
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Although it is not easy to provide a detailed guide on how to evaluate a scheme, it is possible to
describe the fundamental steps to be followed, before deciding if one should proceed to a detailed
feasibility study or not. A list of the studies that should be undertaken:-
• Topography and geomorphology of the site.
• Evaluation of the water resource and its generating potential
• Site selection and basic layout
• Hydraulic turbines and generators and their control
• Environmental impact assessment and mitigation measures
• Economic evaluation of the project and financing potential
• Institutional framework and administrative procedures to attain the necessary consents
The water flowing along natural and man-made canals, conducted by low and high-pressure pipes,
spilling over weir crests and moving the turbines involves the application of fundamental
engineering principles in fluid mechanics. In Chapter 2 those principles are reviewed together with
shortcuts arising from the experience accumulated from centuries of hydraulic systems
construction.
To decide if a scheme will be viable it is necessary to begin by evaluating the water resource
existing at the site. The energy potential of the scheme is proportional to the product of the flow
and the head. Except for very low heads, the gross head can usually be considered as constant, but
the flow varies over the year. To select the most appropriate hydraulic equipment and estimate the
sites potential with calculations of the annual energy output, a flow-duration curve is most useful.
A single measurement of instantaneous flow in a stream has little value.
Measuring the gross head requires a topographical survey. The results obtained, by using a
surveyor's level and staff is accurate enough, but the recent advances in electronic surveying
equipment make the topographical surveying work much simpler and faster. To produce a flow-
duration curve on a gauged site is easier than producing a curve at an ungauged site. This requires
a deeper understanding of hydrology. In Chapter 3 various methods for measuring the quantity of
water flowing in a stream are analysed and hydrological models to calculate the flow regime at
ungauged sites are discussed.
Chapter 4 presents techniques such as orthophotography, RES, GIS, geomorphology, geotectonics,
etc - used nowadays for site evaluation. Some failures are also analysed and conclusions about how
they might have been avoided are explained.
In Chapter 5 the basic layouts are explained and the hydraulic structures, such as weirs, canals,
spillways, intakes and penstocks, studied in detail.
Chapter 6 deals with the electromechanical equipment used to convert the potential energy of the
mass of water to electricity. Turbines themselves are not studied in detail, but attention is focused
on turbine configurations, specifically for low head schemes, and on the process of turbine
selection, with emphasis on specific speed criteria. Since small hydro schemes are usually operated
unattended, the control systems, based on personal computers, are also reviewed.
9
Guide on How to Develop a Small Hydropower Plant ESHA 2004
An Environmental Impact Assessment may be required to obtain the necessary consents to build
the scheme and utilize the water available. Although several recent studies have shown that small
hydropower produce no emissions to atmosphere, nor do they produce toxic wastes, does not
contribute to climatic change, designers should implement all necessary measures to mitigate local
ecological impacts. Chapter 7 analyses those impacts and mitigating measures.
Chapter 8 reviews techniques for an economical evaluation of a scheme. Various methodologies of
economic analysis are described and illustrated with tables showing the cash flows generated by
the schemes.
In Chapter 9, the administrative procedures a developer will have to go through are presented.
Unfortunately the recent deregulation of much of the electricity industry in the EU has made it
difficult to establish a common procedure to follow.
A few years ago ESHA produced (December 1994) on behalf of the E.C. DGXVII, a report "Small
Hydropower. General Framework for Legislation and Authorisation Procedures in the European
Union", and though it is not current it still has many valid aspects. The report can be found in
www.esha.be, the ESHA web page.
Further important considerations for the developer to take into account are trading tariffs for green
and base energy and administrative procedures, for grid connection. These depend on the energy
policy and the institutional framework of each country. An overview has been provided in the
Appendix A of Chapter 9.
i By Celso Penche (ESHA), Francis Armand (ADEME), Vincent Dennis (MhyLab) and Christer Söderberg (SERO)
10
Guide on How to Develop a Small Hydropower Plant ESHA 2004
12
CHAPTER 2: FUNDAMENTALS OF HYDRAULIC ENGINEERING
CONTENTS
2. FUNDAMENTALS OF HYDRAULIC ENGINEERING ................................................ 13
2.1. Introduction ............................................................................................................................... 13
2.2. Water flow in pipes ................................................................................................................... 13
2.2.1. Head losses due to friction ................................................................................................ 16
2.2.2. Local head losses .............................................................................................................. 23
2.2.2 Transient flow ................................................................................................................... 28
2.3. Water flow in open channels..................................................................................................... 31
2.3.1. Classification of open channel flows ................................................................................ 31
2.3.2. Uniform flow in open channels......................................................................................... 32
2.3.3. Efficient cross-section in open channels ........................................................................... 33
2.3.4. Principles of energy in open channel flows ...................................................................... 33
Bibliography ............................................................................................................................................. 40
LIST OF FIGURES
Figure 2-1Velocity distribution for laminar and turbulent flow ............................................................... 14
Figure 2-2 Hydraulic gradient and energy gradient .................................................................................. 16
Figure 2-3 µ as a function of Reynolds number ....................................................................................... 20
Figure 2-4 Loss coefficients for trash racks.............................................................................................. 24
Figure 2-5 Kc and Kex values as a function of d/D .................................................................................. 25
Figure 2-6 Diffuser coefficients ................................................................................................................ 26
Figure 2-7 Entrance loss coefficients........................................................................................................ 27
Figure 2-8 Loss coefficients for flow in bends ......................................................................................... 27
Figure 2-9 Typical loss coefficients for flow through valves ................................................................... 28
Figure 2-10 Typical velocity distributions for open channel flow ........................................................... 31
Figure 2-11 Illustration of various types of varied flow ........................................................................... 32
Figure 2-12 Pressure distribution for channels with vertically curved bed .............................................. 34
Figure 2-13 Specific energy as a function of flow depth .......................................................................... 36
Figure 2-14 Moody’s Chart: Friction factors for pipe flow ...................................................................... 39
Figure 2-15 Illustration of pressure wave in pipes.................................................................................... 39
LIST OF TABLES
Table 2-1 Roughness height "e", for various commercial pipes ............................................................... 17
Table 2-2 Manning coefficient n for several commercial pipes ............................................................... 21
Table 2-3 Hazen-Williams coefficients .................................................................................................... 23
Table 2-4 Additional trash rack losses for non-perpendicular approach flows ........................................ 24
Table 2-5 Geometrical characteristics of different channel profiles......................................................... 36
Table 2-6 Empirical formulae used to estimate yc, in typical channel..................................................... 37
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2. FUNDAMENTALS OF HYDRAULIC ENGINEERING i
2.1. Introduction
Hydraulic engineering is based on the principles of fluid mechanics, although many empirical
relationships are applied to achieve practical engineering solutions. Until now there does not exist
and probably never will, a general methodology for the mathematical analysis of the movement of
fluids. Based on the experience accumulated, over many years of study and practice, there are
particular solutions to specific problems. Experience that goes back 2500 years, when a massive
irrigation system, that is still operative, was built in Sichuan, China, and to the many aqueducts built
during the period of the Roman Empire
In hydropower, hydraulic engineering is applied to:
• .Optimise the performance of waterways to reduce energy losses
• .Design spillways and structure for floods prevention
• .Design adequate energy dissipation works downstream of spillways
• .Control erosion and manage silt transportation
Control phenomena such as:
• Instability in waterways due to dynamic effects
• Air entrance into closed conduits
• Surges in long waterways
• Surge pressures in closed conduits
• Cavitation of structures and equipment
• Prevent reservoir sedimentation, intake obstruction and sediment related damage to the
hydraulic circuit and the equipment
In order to successfully develop small hydropower a thorough understanding of the principles of
hydraulics is required.
In this chapter, the fundamentals of hydraulic engineering are explained together with an explanation
of some of the phenomena mentioned above.
2.2. Water flow in pipes
A body of water will have a potential energy by virtue of its velocity and the vertical height through
which it drops, (as a difference in water levels is what drives the flow of water), which is known as
its “head”. This energy is its “Gravitational Potential Energy” which is product of mass, acceleration
due to the effects of gravity and head m.g.h and is generally expressed in Joules (J)
The energy head in the water flowing in a closed conduit of circular cross section, under a certain
pressure, is given by Bernoulli's equation:
g
VPhH2
2
11
11 ++=γ
(2.1)
Where:
13
Guide on How to Develop a Small Hydropower Plant ESHA 2004
H1 is the total energy head
h1 is the elevation above some specified datum plane,
P1 the pressure
γ the specific weight of water
V1 the velocity of the water, and
g the gravitational acceleration.
The total energy head at point 1 is then the algebraic sum of the potential energy h1, the pressure
energy P1/γ, and the kinetic energy V12/2g, commonly known as the “Velocity head”.
For an open channel, the same equation applies, but with the term P1/γ replaced by d1, the water
depth.
If water is allowed to flow very slowly in a long, straight, glass pipe of small bore into which a fine
stream of coloured water is introduced at the entrance to the pipe, the coloured water would appear
as a straight line all along the pipe. This effect is known as laminar flow. The water flows in lamina
(layers), like a series of thin walled concentric pipes. The outer virtual pipe adheres to the wall of the
real pipe, while each of the inner ones moves at a slightly higher speed, which reaches a maximum
value near the centre of the pipe. The velocity distribution has the form of a parabola and the average
velocity (figure 2.1) is 50% of the maximum centre line velocity.
Figure 2-1Velocity distribution for laminar and turbulent flow
If the flow rate is gradually increased, a point is reached when the lamina flow suddenly breaks up
and mixes with the surrounding water. The particles close to the wall mix up with the ones in the
midstream, moving at a higher speed, and slow them. At that moment the flow becomes turbulent,
and the velocity distribution curve is much flatter. Experiments carried out by Osborne Reynolds,
near the end of the 19th century, found that the transition from laminar flow to turbulent flow
depends, not only on the velocity, but also on the pipe diameter and on the viscosity of the fluid, and
14
Guide on How to Develop a Small Hydropower Plant ESHA 2004
is a ratio of the inertia force to the viscous force. This ratio, is known the Reynolds number and can
be expressed, in the case of a circular pipe, by the equation:-
ν
VDRe
⋅=
(2.2)
where:
D (m) is the pipe diameter
V is the average water velocity (m/s), and
ν is the kinematics viscosity of the fluid (m2/s).
From experimentation it has been found that for flows in circular pipes the critical Reynolds number
is about 2000. In fact this transition does not always happen at exactly Re=2000 but varies with the
conditions. Therefore there is more than a transition point, what exists is a transition range.
Example 2.1
A 60-mm diameter circular pipe carries water at 20oC. Calculate the largest flow-rate for
which the flow would be laminar.
The kinematics viscosity of water at 20oC is u = 1 x 10-6 m2/s.
Assuming a conservative value for Re = 2 000
V = 2 000 / (106x0.06) = 0.033 m/s
Q = AV = π /4x 0.062 x 0.033 = 3.73 x 10-4 m3/s = 0.373 l/s
Water loses energy as it flows through a pipe, fundamentally due to:
1. friction against the pipe wall
2. viscous dissipation as a consequence of the internal friction of flow
The friction against the pipe wall depends on the wall material roughness and the velocity gradient
nearest to the wall. Velocity gradient, as can be seen in figure 2.1, is higher in turbulent flow than in
laminar flow. Therefore, as the Reynolds number increases, the friction loss will also increase. At the
same time, at higher turbulences there is more intensive mixing of particles, and hence a higher
viscous dissipation. Consequently the energy losses in flow in the pipe increase with the Reynolds
number and with the wall pipe roughness. It can be verified that for water flowing between two
sections, a certain amount of the head of energy hf is lost:-
f
gg
hhPVhPV+++=++2
2
2
2
1
1
2
1
22 γγ (2.3)
Due firstly, to the friction of the water against the pipe wall, and secondly, to the internal friction of
the flow. In figure 2.2, HGL is the hydraulic gradient line and EGL the energy gradient line. If the
pipe cross-section is constant, V1 = V2 and both lines will be parallel. It is therefore necessary to
determine the value of hf ?
15
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2.2.1. Head losses due to friction
Darcy and Weisbach, applying the principle of conservation of mass to a certain volume of fluid in a
pipe, between two sections perpendicular to its axis - derived the following equation, valid for
incompressible and steady flows, through pipes:
Figure 2-2 Hydraulic gradient and energy gradient
g
V
D
Lfhf 2
2
⋅⎟⎠
⎞⎜⎝
⎛⋅=
(2.4)
where
f = friction factor, a dimensionless number
L = the length of the pipe in m
D = the pipe diameter in m
V = the average velocity in m/s, and
g = the gravitational acceleration (9.81 m/s2).
In a laminar flow f can be calculated directly by the equation:
eRDVf6464=⋅
⋅=ν
(2.5)
According to equation (2.5) the friction factor f in a laminar flow is independent of the wall
roughness and inversely proportional to the Reynolds number. The fact that, apparently, f decreases
when Re increases, does not mean that increasing the velocity decreases the friction losses.
Substituting f in equation (2.4) by its value in (2.5), gives:-
16
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2
2 32
2
64
Dg
VLv
g
V
D
L
DVhf ⋅
⋅⋅⋅=⋅⋅⋅
⋅=ν
(2.6)
This shows that the specific head loss, in laminar flow, is proportional to V and inversely
proportional to D2.
When the flow is practically turbulent (Re>2000), the friction factor become less dependent on the
Reynolds number and more dependent on the relative roughness height e/D, where "e" represents the
average roughness height of irregularities on the pipe wall and D the pipe diameter. Some values of
the roughness height "e” are provided in table 2.1.
Table 2-1 Roughness height "e", for various commercial pipes
Pipe material e (mm)
Polyethylene 0.003
Fiberglass with epoxy 0.003
Seamless commercial steel (new) 0.025
Seamless commercial steel (light rust) 0.250
Seamless commercial steel (galvanised) 0.150
Welded steel 0.600
Cast iron (enamel coated) 0.120
Asbestos cement 0.025
Wood stave 0.600
Concrete (steel forms, with smooth joints) 0.180
It is well known that, even in turbulent flows, immediately next to the wall pipe there exists, a very
thin layer of flow referred to as the laminar sub layer. When Re increases, the sub layer’s thickness
diminishes. Whenever the roughness height "e" is resolutely lower than the sub layer thickness the
pipe is considered hydraulically smooth.
In a hydraulically smooth pipe flow, the friction factor f is not affected by the surface roughness of
the pipe, and for this case Von Karman, developed the following equation for the friction factor f:
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⋅=51.2log21
10
fR
f
e (2.7)
At high Reynolds numbers, the sub layer thickness becomes very small and the friction factor f
becomes independent of Re and depends only on the relative roughness height. In this case the pipe is
a hydraulically rough pipe, and Von Karman found that the friction factor f:
⎟⎠
⎞⎜⎝
⎛⋅=e
D
f 7.3log21
10 (2.8)
In between these two extreme cases, the pipe behaves neither completely smooth nor completely
rough, for this situation, Colebrook and White devised the following equation:
17
Guide on How to Develop a Small Hydropower Plant ESHA 2004
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛+⋅−=fR
De
f e
51.2
7.3
/log21
10 (2.9a)
Which can be expressed in terms of the average velocity U by:-
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
+−=
L
hgDD
De
L
hgDU
f
f
2
51.2
7.3log22 ν (2.9b)
Formulae 2.7 and 2.9 are difficult to solve by hand, prompting Moody to prepare his well-known
chart "Friction factors for pipe flow" (figure 2.15).
Looking to the chart it shows four different flow zones:
1. A laminar flow zone (shaded area in the diagram) where f is a linear function of R (equation 2.5)
2. A badly defined critical zone (shaded area)
3. A transition zone, starting with the smooth pipes (equation 2.7) and finishing in the dashed line
where, in between, f depends both of Re and e/D (equation 2.9a)
4. A developed turbulence zone where f depends exclusively of e/D (equation 2.8)
Example 2.2
Calculate, using the Moody chart, the friction loss in a 900-mm diameter welded steel pipe
along a length of 500 m, conveying a flow of 2.3 m3/s.
The average water velocity is 4Q /(π D2)= 1.886 m/s
From the table 2.1,
e = 0.6 mm and therefore e/D = 0.6/900 = 0.000617
ReNR =DV / u = (0.9 x 1.886)/ 1.31 = 1.3x106 (u = 1.31 10 -6)
In the Moody chart for e/D = 0.00062 and Re = 1.3*106 we find f=0.019
From equation (2.4):
mhf91.181.92
886.1
9.0
500019.0
2
=⋅⋅⋅=
In engineering practice the Colebrook-White formula (2.9) and the Moody diagram can be used to
solve the following typical problems with flows in closed pipes:
1. Given U (or Q), D and e, compute hf;
2. Given U (or Q), hf and e, compute D;
3. Given D, hf and e, compute U (or Q);
4. Given U (or Q), D, hf, compute e.
Problems in 3and 4 above can be solved directly by using formula (2.9b), whereas the remainder
problems require an iterative solution. The Moody’s diagram provides a direct solution for the 1st
and 4th problems. Alternatively, if you want to know what the maximum water velocity flowing in a
18
Guide on How to Develop a Small Hydropower Plant ESHA 2004
pipe of diameter D and length L, without surpassing a friction head loss hf you only need to use an
independent variable µ:
2
2
1
efR=μ
(2.10)
Substituting Re by its value in (2.2) and f by its value in (2.4) becomes:-
2
3
νμL
hgD f=
(2.11)
where all the parameters are known. Once µ is computed, f is derived from (2.10) and substituted in
(2.9) to attain:
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛+−=μμ2
51.2
7.3
/log2210
DeRe
(2.12)
An equation that makes it possible to plot the Re with respect to U for different values of e/D, is
shown in figure 2.3, a variation of the Moody Chart where Re can be estimated directly.
Example 2.3
Estimate the flow rate of water at 10oC that will cause a friction head loss of 2m per km in a
welded steel pipe, 1.5 m in diameter.
Substitute values in equation (2.12), with e/D=0.6/1500 = 4x104,
After computing U .
()10
26
3
1086.3
1031.11000
25.181.9 ⋅=⋅⋅
⋅⋅=−μ
6
10
4
10
10 1019.2
1086.32
51.2
7.3
104log1086.322 ⋅=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
⋅⋅+⋅⋅⋅−=
−
eR
=⋅⋅⋅=⋅=
−
5.1
1031.11019.2 66
D
RVe ν 1.913 m/s; Q=V⋅A=3.38 m3/s
Also based on the Colebrook-White equation there exists some other monographs, to compute the
friction head loss on a pipe, given a certain flow, a certain pipe diameter, with a certain roughness
coefficient such as the one shown in the next page and published by courtesy of Hydraulic Research,
Wallingford U.K.
19
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-3 µ as a function of Reynolds number
Empirical formulae
Over the years many empirical formulae, based on accumulated experience, have been developed.
They are, generally, not based on sound physics principles and even, occasionally, lack dimensional
coherence, but are intuitively based on the belief that the friction on a closed full pipe is:
1. Independent of the water pressure
2. Linearly proportional to its length
3. Inversely proportional to a certain power of its diameter
20
Guide on How to Develop a Small Hydropower Plant ESHA 2004
4. Proportional to a certain exponent of the water velocity
In turbulent flows it is influenced by the wall roughness
One of these formulae, widely used to estimate the flow in open channels, but also applicable to
closed pipes, is that developed by Manning (resp. Strickler):
Q =1
n ⋅A5 /3S1 /2
P 2/3 (2.13)
Where: n is the Manning roughness coefficient (s/m1/3, KStrickler=1/n)
P is the wetted perimeter (m)
A is cross-sectional area of the pipe (m2), and
S is the hydraulic gradient or head loss by linear meter (hf/L).
Applying the above formulae to a full closed circular cross section pipe:
333.5
2229.10
D
QnS⋅⋅= (2.14)
3162
223104
D
QnS
π
= (2.14a)
In Table 2.2 the Manning coefficient n for several commercial pipes is shown:
Table 2-2 Manning coefficient n for several commercial pipes
Kind of pipe n
Welded steel 0.012
Polyethylene (PE) 0.009
PVC 0.009
Asbestos cement 0.011
Ductile iron 0.015
Cast iron 0.014
Wood-stave (new) 0.012
Concrete (steel forms smooth finish) 0.014
In example 2.4 and more specifically in example 2.5 the results obtained by applying the Colebrook-
White equation and the Manning formulae can be compared.
Example 2.4
Using the parameters in example 2.2 compute the friction head loss applying the Manning
formulae
Accepting n=0.012 for welded steel pipe
00374.09.0
2.1012.029.10
333.5
22
=⋅⋅=L
h f
21
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Whereby for L=500 m, hf =1.87 m, slightly inferior to the value estimated with the Moody chart.
Example 2.5
Compute, using the Colebrook equation and the Manning formulae, the friction head loss on a
welded pipe 500 m long, of respectively 500 mm, 800 mm, 1 200 mm, and 1 500 mm diameter
respectively, under a 4 m/s average flow velocity.
D (mm) 500 800 1200 1500
Q (m*3*/s) 0.785 2.011 4.524 7.069
V (m/s) 4 4 4 4
L (m) 500 500 500 500
Applying Colebrook-White
e (mm) 0.6 0.6 0.6 0.6
hf (m) 17.23 9.53 5.73 4.35
Applying Manning
n 0.012 0.012 0.012 0.012
hf (m) 18.40 9.85 5.73 4.26
It can be observed that the solutions provided by the Manning formula do not differ much from those
offered by the Colebrook equation, except in the smaller diameters, where the head loss provided by
Manning is higher than that provided by Colebrook. In fact, both formulae agree for values of
e/D=9.17E-3 and provide results within a 5 % range for values of e/D between 9E-4 and 5E-2 in the
turbulent (rough) zone (Dubois, 1998). In this range of flows, the relation between the Darcy-
Weisbach and Manning’s coefficients is:
31
23424.2;2 D
ngfg
USD
f == (2.14b)
In North America for pipes larger than 5 cm diameter and flow velocities under 3 m/s the Hazen-
Williams formulae is typically used:
22
Guide on How to Develop a Small Hydropower Plant ESHA 2004
85.1
165.1
87.6 ⎟⎠
⎞⎜⎝
⎛⋅=C
V
D
Lhf (2.15)
Where V is the flow velocity (m/s), D the diameter (m), L the pipe length (m) and C the Hazen-
Williams coefficient such as shown in Table 2.3.
Table 2-3 Hazen-Williams coefficients
Pipe type C
Asbestos cement 140
Cast iron
New 130
10 years 107 - 113
20 years 89 - 100
30 years 75 - 90
Concrete
Cast on site - steel forms 140
Cast on site - wood forms 120
Centrifugal cast 135
Steel
Brush tar and asphalt 150
New uncoated 150
Riveted 110
Wood-stave (new) 120
Plastic pipes 135 - 140
2.2.2. Local head losses
In addition to friction losses, water flowing through a pipe systems experience head losses due to
geometric changes at entrances, bends, elbows, joints, racks, valves and at sudden contractions or
enlargements of the pipe section. This loss also depends on the velocity and is expressed by an
experimental coefficient K multiplied by the kinetic energy v2/2g.
2.2.1.1 Trash rack (or screen) losses
A screen is nearly always required at the entrance of both pressure pipes and intakes to avoid the
entrance of floating debris. The flow of water through the rack also gives rise to a head loss. Though
usually small, it can be calculated by a formula developed by Kirschmer:
Φ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⎟⎠
⎞⎜⎝
⎛=sing
V
b
tKtht 2
2
0
3/4
(2.16)
where the parameters are identified in figure 2.4.
23
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Ht
Φ
l
Figure 2-4 Loss coefficients for trash racks
For structural reasons, this formula is only valid if the length L of the bars is smaller than 5 times
their diameter. If the grill is not perpendicular but makes an angle β with the water flow (β will have
a maximum value of 900 for a grill located in the sidewall of a canal), there will be an additional
head loss. The result of equation 2.16 should be multiplied by a correction factor κ provided in the
table 2.4 (according to Mosonyi).
Table 2-4 Additional trash rack losses for non-perpendicular approach flows
t/b
β
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
0° 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
10° 1.06 1.07 1.08 1.09 1.10 1.11 1.12 1.14 1.50
20° 1.14 1.16 1.18 1.21 1.24 1.26 1.31 1.43 2.25
30° 1.25 1.28 1.31 1.35 1.44 1.50 1.64 1.90 3.60
40° 1.43 1.48 1.55 1.64 1.75 1.88 2.10 2.56 5.70
50° 1.75 1.85 1.96 2.10 2.30 2.60 3.00 3.80 …
60° 2.25 2.41 2.62 2.90 3.26 3.74 4.40 6.05 …
2.2.1.2 Loss of head by sudden contraction or expansion
When the pipe has a sudden contraction there is a loss of head due to the increase in velocity of the
water flow and to the large-scale turbulence generated by the change of geometry. The flow path is
so complex that, at least for the time being, it is impossible to provide a mathematical analysis of L
the phenomenon. The head loss is estimated by multiplying the kinetic energy in the smaller pipe
(section 2), by a coefficient Kc that varies with the ratio of contraction d/D:
24
Guide on How to Develop a Small Hydropower Plant ESHA 2004
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⋅=g
VKhcc2
2
2
(2.17)
For a ratio up to d/D = 0.76, Kc approximately follows the formula:-
k c =0,42 1 −d 2
D2
⎛
⎝ ⎜ ⎞
⎠ ⎟
(2.18)
The ratio, Kc is substituted by Kex, the coefficient used for a sudden expansion.
In sudden expansions, the loss of head can be derived from the momentum of flow and is given by:
()
g
V
D
d
g
V
A
A
g
V
V
V
g
VVhex 2121212
2
1
2
22
1
2
2
1
2
1
2
1
2
2
21 ⎟⎟⎠
⎞
⎜⎜⎝
⎛−=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛−=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛−=−=
(2.19)
where V1 is the water velocity in the smaller pipe.
Figure 2.5 is a graphic representation of the Kc and Kex values as a function o f d/D.
The head loss can be reduced by using a gradual pipe transition, known as a confuser for contraction,
or diffuser for expansion.
Figure 2-5 Kc and Kex values as a function of d/D
In the confuser the head loss varies with the confuser angle as it is shown in the table below where
Kc values are experimental:
25
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Angle Kc
300 0.02
450 0.04
600 0.07
In the diffuser the analysis of the phenomenon is more complex. Figure 2.6 shows the experimentally
found values of Kex for different diffuser angles. The head loss is given by:
g
VVKhexex2
2
2
2
1''−= (2.20)
A submerged pipe discharging in a reservoir is an extreme case of a sudden expansion, where V2,
given the size of the reservoir, compared with the pipe, can be considered as zero, and the
lossV12/2g.
On the other hand, the entrance from a reservoir to a pipe is an extreme case of a sudden contraction.
Figure 2.7 shows the value of the Ke coefficient that multiplies the kinetic energy V2/2g in the pipe.
Figure 2-6 Diffuser coefficients
26
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-7 Entrance loss coefficients
2.2.1.3 Loss of head in bends
In a bend, pipe flow experiences an increase of pressure along the outer wall and a decrease of
pressure along the inner wall. This pressure unbalance causes a secondary current such as shown in
the figure 2.10. Both movements together - the longitudinal flow and the secondary current -
produces a spiral flow that, at a length of around 100 diameters, is dissipated by viscous friction. The
head loss produced in these circumstances depends on the radius of the bend and on the diameter of
the pipe. Furthermore, in view of the secondary circulation, there is a secondary friction loss,
dependent of the relative roughness e/D.
Figure 2.8, taken from reference 3 gives the value of Kb for different values of the ratio R/D and
various relative roughness e/D. There is also a general agreement that, in seamless steel pipes, the
loss in bends with angles under 90o, is almost proportional to the bend angle. The problem is
extremely complex when successive bends are placed one after another, close enough to prevent the
flow from becoming stabilized at the end of the bend. Fortunately, this is hardly ever the case in a
small hydro scheme.
Figure 2-8 Loss coefficients for flow in bends
27
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2.2.1.4 Loss of head through valves
Valves or gates are used in small hydro schemes to isolate a component from the rest, so they are
either entirely closed or entirely open. Flow regulation is assigned to the distributor vanes or to the
needle valves of the turbine. The loss of head produced by water flowing through an open valve
depends of the type and manufacture of the valve. Figure 2.9 shows the value of Kv for different kind
of valves.
Kv=0.05 Kv=1.0Kv=0.2 Kv=0.6 Kv=0.05 Kv=1.0Kv=0.05 Kv=1.0Kv=0.2 Kv=0.6
Figure 2-9 Typical loss coefficients for flow through valves
2.2.2 Transient flow
In steady flows where the discharge is assumed to remain constant with time, the operating pressure
at any point along a penstock is equivalent to the head of water above that point. If a sudden change
of flow occurs, for instance when the plant operator, or the governor system, open or close the gates
too rapidly, the sudden change in the water velocity can cause dangerous high and low pressures.
This pressure wave is known as water hammer, or surge, and its effects can be dramatic. The
penstock can burst from overpressure or collapse if the pressures are reduced below atmospheric.
Although being transitionary the surge pressure induced by the “water hammer phenomenon” can be
of a magnitude several times greater than the static pressure due to the head. According to Newton's
second law of motion, the force developed in the penstock, by the sudden change in velocity, will be:
dt
dVmF=
(2.21)
If the velocity of the water column could be reduced to zero the resulting force would become
infinite. Fortunately this is not possible in practice; a mechanical valve requires some time for total
closure and the pipe walls are not perfectly rigid and the water column under large pressures is not
incompressible.
The following description, reproduced with the permission of the author, Allen R. Inversin from
Appendix F of his "Micro-Hydropower Sourcebook", is one of the best physical explanations of this
phenomenon. Figure 2.16, enclosed at the end of this chapter, illustrates how a velocity change,
caused by an instantaneous closure of a gate, or valve, at the end of a pipe creates a pressure wave
that travels the length of the pipe.
28
Guide on How to Develop a Small Hydropower Plant ESHA 2004
29
Initially, water flows at a velocity (Vo) as shown in (a). When the gate is closed, the water flowing
within the pipe has a tendency to continue flowing due to its momentum. Because this momentum is
physically stopped by the gate closing, it “piles up” behind it, the kinetic energy of the element of
water nearest the gate is converted to pressure energy, which slightly compresses the water and
expands the circumference of the pipe at this point (b). This action is repeated by the following
elements of water (c), and the wave front of increased pressure travels the length of the pipe until the
velocity of the water Vo is destroyed, the water is compressed, and the pipe is expanded over its
entire length (d). At this point, the water's kinetic energy has all been converted to strain energy
(under increased compression) and strain energy of the pipe (under increased tension).
Because the water in the reservoir remains under normal static pressure but the water in the pipe is
now under a higher pressure, the flow reverses and is forced back into the reservoir again with
velocity Vo (e). As the water under compression starts flowing back, the pressure in the pipe is
reduced to normal static pressure. A pressure “unloading” wave then travels down the pipe toward
the gate (f) until all the strain energy is converted back into kinetic energy (g). However, unlike case
(a), the water is now flowing in the opposite direction and because of its momentum the water again
tries to maintain this velocity. In so doing, it stretches the element of water nearest the gate, reducing
the pressure there and contracting the pipe (h). This happens with successive elements of water and a
negative pressure wave propagates back to the reservoir (i) until the entire pipe is under
compression and water under reduced pressure (j). This negative pressure wave would have the same
absolute magnitude as the initial positive pressure wave if it were assumed that friction losses do not
exist. The velocity then returns to zero but the lower pressure in the pipe compared to that in the
reservoir forces water to flow back into the pipe (k). The pressure surge travels back toward the gate
(e) until the entire cycle is complete and a second cycle commences (b). The velocity with which the
pressure front moves is a function of the speed of sound in water modified by the elastic
characteristics of the pipe material. In reality, the penstock pipe is usually inclined but the effect
remains the same, with the surge pressure at each point along the pipe adding to or subtracting from
the static pressure at that point. Also, the damping effect of friction within the pipe causes the kinetic
energy of the flow to dissipate gradually and the amplitude of the pressure oscillations to decrease
with time. Although some valves close almost instantaneously, closure usually takes at least several
seconds. Still, if the valve is closed before the initial pressure surge returns to the gate end of the
pipeline (g), the pressure peak will remain unchanged - all the kinetic energy contained in the water
near the gate will eventually be converted to strain energy and result in the same peak pressure as if
the gate were closed instantaneously. However, if the gate has been closed only partially, by the time
the initial pressure surge returns to the gate (g), not all the kinetic energy will have been converted to
strain energy and the pressure peak will be lower. If the gate then continues closing, the positive
pressure surge, which it would then create, will be reduced somewhat by the negative pressure (h)
surge which originated when the gate originally began closing. Consequently, if the gate opens or
closes in more time than that required for the pressure surge to travel to the reservoir and back to the
gate, peak surge pressures are reduced. This time is called the critical time, Tc, and is equal to:
Tc = 2L /c (2.22)
where c is the wave velocity. The wave velocity, or speed of sound, in water is approximately 1420
m/s. However, the wave velocity in a pipe - the speed with which the pressure surge travels along the
pipe - is a function of both the elastic characteristics of water and the pipe material. An expression
for the wave velocity is:
Guide on How to Develop a Small Hydropower Plant ESHA 2004
tE
Dk
/kc
⋅
⋅+
ρ=
1
(2.23)
where K = bulk modulus of water, 2.2x109 N/m2
ρ = density of water, 1 000 kg/m3
D = internal pipe diameter (m)
E = modulus of elasticity of pipe material (N/m2)
t = wall thickness (mm)
If the valve is already closed, when the pressure wave is on its way back (t<Tc) all the kinetic energy
of the water will be converted on an overpressure, and its value in meters of water column is:-
Vg
c
g
P Δ=Δ
ρ
(2.24)
where .V is the change of water velocity. In practical cases, .V can be assumed equal to the initial
flow velocity V0. However, if t is greater than Tc, then the pressure wave reaches the valve before the
valve is completely closed, and the overpressure will not develop fully, because the reflected
negative wave arriving at the valve will compensate for the pressure rise. In this case the maximum
overpressure may be calculated by the simplified Allievi formula, also known as the Michaud
formula:-
Vgt
L
g
P Δ=Δ 2
ρ (2.25)
where L = total pipe length (m)
ΔP/ρg = pressure difference between the initial static pressure P0/ρg and the maximum pressure
attained in the conduit (m column of water) t = closure time (s).
The total dynamic pressure experienced by the penstock will thus be:
P = Po + .P (2.26)
In chapter 5, several examples related to penstock design will clarify the above physics concepts. For
a more rigorous approach it would be necessary to take into consideration not only the elasticity of
the fluid and pipe material above, but also the hydraulic losses. The mathematical approach is rather
cumbersome and requires the use of computers. For interested readers Chaudry, Fox and Parmakian,
among others, give calculation methods, together with some worked examples.
30
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2.3. Water flow in open channels
In closed pipes the water fills the entire pipe, in an open canal there is always a free surface.
Normally, the free water surface is subject to the atmospheric pressure, commonly referred to as the
zero pressure reference, and usually considered as constant along the full length of the canal. In a
way this fact, by dropping the pressure term, facilitates the analysis, but at the same time introduces a
new dilemma. The depth of water changes with the flow conditions, and in unsteady flows its
estimation is a part of the problem. Any kind of canal, even a straight one, has a three-dimensional
distribution of velocities. A well-established principle in fluid mechanics is that any particle in
contact with a solid stationary border has a zero velocity. Figure 2.10 illustrates the iso-velocity lines
in channels of different profile. The mathematical approach is based on the theory of the boundary
layer; the engineering approach is to deal with the average velocity V.
2.3.1. Classification of open channel flows
A channel flow is considered steady when the depth at any section of the stretch does not change
with time, and unsteady if it changes with time. An open channel flow is said to be uniform if the
discharge and the water depth at every section of a channel length does not change with time.
Accordingly, it is said to be varied whenever the discharge and/or the water depth changes along its
length. Non uniform flow is a rare occurrence, and with uniform flow, steady uniform flow is
understood to occur. Steady variable flow is often stated as gradual or rapid.
Figure 2.11 represents different kinds of flows: steady uniform flow, steady gradually variable flow,
and steady rapidly variable flow. Unsteady flow occurs if either the flow depth, or the discharge,
over the length of the canal, changes as, for instance, in the case of upstream propagation of a small
perturbation wave due to closure or opening of a valve, or in the case of the discharge increase in a
collector channel.
Figure 2-10 Typical velocity distributions for open channel flow
31
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-11 Illustration of various types of varied flow
As with the analysis of fully closed pipe flows, channel flows also follow Bernoulli’s equation and
consequently formula (2.1) is valid. The amount of energy loss when water flows from section 1 to
section 2 is indicated by hL.
2.3.2. Uniform flow in open channels
By definition a flow is considered uniform when:-
1. The water depth, water area, and the velocity in every cross section of the channel are constant.
2. The energy gradient line, the free surface line and the bottom channel line are parallel to each
other.
Based on these concepts Chezy found that:-
V =CRi (2.27)
where:-
C = Chezy's resistance factor
Rh = Hydraulic radius of the channel cross-section
Se = Channel bottom line slope
Many attempts had been made to determine the value of C. Manning, using the results of his
own experiments and those of others, derived the following empirical relation:
6/11RnC= (2.28)
where n is the well-known Manning's roughness coefficient (see Chapter 5, Table 5.1). Substituting
C from (2.27) into (2.28) we have the Manning formula for uniform flows:
32
Guide on How to Develop a Small Hydropower Plant ESHA 2004
2/13/11 iRnV= (2.29)
or alternatively:-
2/13/21 iRAnQ⋅= (2.30)
The parameter ARh 2/3 has been defined as the section factor and is given, for various channel
sections, in table 2.5. The formula is entirely empirical and the n coefficient is not dimensionless, so
the formulae given here are only valid in S.I. units. Furthermore the formulae are only applicable to
channels with a flat bottom. The analysis of natural watercourses is more complex and the above
formulae can only be applied for first approximations.
2.3.3. Efficient cross-section in open channels
From (2.32) it may be deduced that for a channel with a certain cross-section area A and a given
slope S, the discharge increases by increasing the hydraulic radius. That means the hydraulic radius
is an efficiency index. As the hydraulic radius is the quotient of the area A and the wetted perimeter
P, the most efficient section will be the one with the minimum wetted perimeter. Among all cross-
sectional areas, the semicircle is the one, which has the minimum wetted perimeter for a given area.
Unfortunately such a channel, with a semicircular cross section is expensive to build and difficult to
maintain, and so is only used in small section channels built with prefabricated elements. Putting
aside the semicircular section, the most efficient trapezoidal section is a half hexagon. The most
commonly used channel section in small hydro schemes is the rectangular section, easy to build,
waterproof and maintain. In chapter 5 the selection of the channel section is considered from the
construction viewpoint, balancing efficiency, land excavation volumes, construction methods, etc.
2.3.4. Principles of energy in open channel flows
Uniform flows in open channels are mostly steady, and unsteady uniform flows are rather rare. If the
flow lines are parallel and we take the free surface of the water as the reference plane, the summation
of the elevation energy "h" and the pressure energy P/γ is constant and equal to the water depth. In
practice, most of the uniform flows and a large part of the steady varied flows can be considered
parallel to the bottom.
On a channel with a constant slope less than 6° (figure 2.12a), the pressure head at any submerged
point is equal to the vertical distance measured from the free surface to that point (depth of water).
The stress distribution is typically triangular. Nevertheless if the water is flowing over a convex path,
such as a spillway, the centrifugal flow acts in an opposite direction to the gravity, and the stress
distribution is distorted and looks like figure 2.12b. The pressure energy is given by the difference
between the depth and the centrifugal acceleration of the water mv2/r, being r the radius of curvature
of the convex path. If the path is concave, the acceleration force is added to the depth and the stress
distribution looks like in figure 2.12c. Consequently the resulting pressure head, for water flows
along a straight line, a convex path and a concave path is respectively:
33
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-12 Pressure distribution for channels with vertically curved bed
)();();(
22
crg
VyyPbrg
VyyPayP ⋅+=⋅−==γγγ (2.31)
where:
γ = the specific weight of water y = the depth measured from the free water surface to the point, y =
hCosα, h being the flow depth normal to the channel bottom
V = the water velocity at that point,
r = the radius of curvature of the curved flow path.
The specific energy in a channel section or energy head measured with respect to the bottom of the
channel at the section is:
g
VyE2
2
α+= (2.32)
where α is a coefficient that takes into account the actual velocity distribution in the particular
channel section, whose average velocity is V. The coefficient can vary from a minimum of 1.05 for a
very uniform distribution, to 1.20 for a highly uneven distribution. Nevertheless in a preliminary
approach a value of α = 1 can be used, a reasonable value when the slope is under 0.018 (α <1.03°).
Equation 2.32 then becomes:-
g
VyE2
2
+= (2.33)
A channel section with a water area A and a discharge Q, will have a specific energy:
2
2
2 gA
QyE+= (2.34)
Equation (2.34) shows that given a discharge Q, the specific energy at a given section, is a function
of the depth of the flow only. When the depth of flow y is plotted, for a certain discharge Q, against
the specific energy E, a specific energy curve, with two limiting boundaries, like the one represented
34
Guide on How to Develop a Small Hydropower Plant ESHA 2004
in figure 2.13 is obtained. The lower limit AC is asymptotic to the horizontal axis, and the upper AB
to the line E=y. The vertex point A on the specific energy curve represents the depth y at which the
discharge Q can be delivered through the section at a minimum energy. For every point over the axis
E, greater than A, there are two possible water depths. At the smaller depth the discharge is delivered
at a higher velocity and hence at a higher specific energy - a flow known as supercritical flow. At the
larger depth the discharge is delivered at a smaller velocity, but also with a higher specific energy - a
flow known as subcritical flow. In the critical state the specific energy is a minimum, and its value
can therefore be computed by equating the first derivative of the specific energy (equation 2.34) with
respect to "y" to zero.
dE
dy =−Q 2
gA3
dA
dy +1 =0 (2.35)
The differential water area near the free surface, δA/δy = T, where T is the top width of the channel
section (see figure 2.13). By definition:-
Y = A/t (2.36)
The parameter Y is known as the "hydraulic depth" of the section, and it plays a key role in studying
the flow of water in a channel.
Substituting in equation (2.35) δA/δy by T and A/T by Y one obtains:
1=gY
V (2.37 a)
Where:
gY
VFr= (2.37 b)
The quantity Fr is dimensionless and known as the Froude number. When Fr= 1, as in equation (2.37
a), the flow is in the critical state. The flow is in the supercritical state when Fr > 1 and in the
subcritical state when Fr < 1. In Figure 2.13, the AB line represents the supercritical flows and the
AC the subcritical ones. As shown in figure 2.13, a family of similar curves can be drawn for the
same section and different discharges Q. For higher discharges the curve moves to the right and for
lower discharges to the left.
In the critical state, y = yc (yc being the critical depth). It can be obtained from equation (2.37 a). For
a rectangular channel, the critical depth is given by:
35
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-13 Specific energy as a function of flow depth
3
2
3 2
2
g
q
gb
Qyc ==
(2.38)
where q=Q/b is the discharge per unit width of the channel. Table 2.5 shows the geometric
characteristics of different channel profiles and Table 2.6, taken from Straub (1982) presents the
empirical formulae used to estimate yc, in non-rectangular channel.
Example 2.6
In a trapezoidal section channel where b=6 m and z = 2, compute the critical depth flow for a
discharge of 17 m3/s.
From table 2.6 Ψ = α Q2/g = 29.46 for α=1
The solution is valid provided 0.1 < Q/b2 < 0.4; as q/b2 = 0.19 it is valid
mz
b
bzyc 86.081.0
27.0
25.175.0 =Ψ−⎟⎠
⎞⎜⎝
⎛Ψ=
The estimation of the critical depth, and the supercritical and subcritical ones, permits the profile of
the free surface to be determined, in cases such as, the sudden increase in the slope of a channel, the
free surface upstream from a gate and spillways, etc..
Table 2-5 Geometrical characteristics of different channel profiles
36
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Table 2-6 Empirical formulae used to estimate yc, in typical channel.
37
Guide on How to Develop a Small Hydropower Plant ESHA 2004
38
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 2-14 Moody’s Chart: Friction factors for pipe flow
Figure 2-15 Illustration of pressure wave in pipes
39
Guide on How to Develop a Small Hydropower Plant ESHA 2004
40
Bibliography
1. N.H.C.Hwang and Carlos Hita, "Fundamentals of Hydraulic Engineering Systems", Prentice
Hall Inc. Englewood Cliffs, New Jersey 1987
2. F.H. White, "Fluid Mechanics", MacGraw-Hill Inc. USA
3. A. Piqueras, "Evacuación de Broza" (in Castillan), ESHA Info nº 9 summer 1993
4. L. Allievi, The theory of waterhammer, Transactions ASME 1929
5. H. Chaudry. Applied Hydraulic Transients, Van Nostrand Reinhold Co. 1979
6. V.L. Streeter and E.B. Wylie, Hydraulic Transients, McGraw-Hill Book Co., New York 1967
7. J. Parmakian. Waterhammer analysis. Dower Publications, New York 1963
8. R.H. French, "Hidráulica de canales abiertos" (in Castillan), McGraw-Hill/Interamericana de
Mexico, 1988
9. V.T. Chow, Open Channel Hydraulics, McGraw-Hill Book Co., New York 1959
10. V.L. Streeter and E.B. Wylie, Fluid Mechanics, McGraw-Hill Book Co., New York 1975
11. A.C Quintela, « Hidráulica » (in Portuguese), Ed. Calouste Gulbenkian Foundation, 1981
12. J. Dubois, “Comportement hydraulique et modélisation des écoulements de surface" (in
French),
Communication LCH n° 8, EPFL, Lausanne 1998.
13. E. Mosonyi, “Water power development”, Tome I and II, Akadémiai Kiadó Budapest,
1987/1991
Other references on the topics of this subject :
W.King and E.F. Brater, Handbook of Hydraulics, McGraw-Hill Book Co., New York 1963
R. Silvester, Specific Energy and Force Equations in Open-Channel Flow, Water Power March 1961
i By Jonas. Rundqvist (SERO), Pedro Manso (EPFL) and Celso Penche (ESHA)
Guide on How to Develop a Small Hydropower Plant ESHA 2004
CHAPTER 3: EVALUATING STREAMFLOW
CONTENTS
3. EVALUATING STREAMFLOW .................................................................................................... 44
3.1. Introduction ............................................................................................................................... 44
3.2. Stream flow records .................................................................................................................. 45
3.3. Evaluating stream flows by discharge measurements .............................................................. 46
3.3.1. Velocity-area method ........................................................................................................ 46
3.3.2. Weir method...................................................................................................................... 53
3.3.3. Slope-area method ............................................................................................................ 54
3.4. Stream Flow Characteristics ..................................................................................................... 55
3.4.1. Hydrograph ....................................................................................................................... 55
3.4.2. Flow Duration Curves (FDC)........................................................................................... 55
3.4.3. Standardised FDC curves .................................................................................................. 56
3.4.4. FDCs for particular months or other periods .................................................................... 58
3.4.5. Water Pressure or “head”.................................................................................................. 58
3.5. Residual, reserved or compensation flow ................................................................................. 61
3.6. Estimation of plant capacity and energy output........................................................................ 61
3.6.1. How the head varies with the flow and its influence on the turbine capacity .................. 63
3.6.2. Peaking operation.............................................................................................................. 64
3.7. Firm energy ............................................................................................................................... 65
3.8. Floods........................................................................................................................................ 65
3.8.1. Flood Control Design........................................................................................................ 65
3.8.2. Statistical analysis of flood data ....................................................................................... 66
3.8.3. Hydrological modelling of the catchment area ................................................................. 68
Bibliography ............................................................................................................................................. 69
LIST OF FIGURES
Figure 3-1 Schematic layout of a hydro development .............................................................................. 44
Figure 3-2 Measuring the river stage, definitions ..................................................................................... 46
Figure 3-3 Rating curve ............................................................................................................................ 48
Figure 3-4 Measuring the cross-sectional area ......................................................................................... 49
Figure 3-5 Conductivity time curve .......................................................................................................... 52
Figure 3-6 Discharge measurements using weirs and notches ................................................................. 53
Figure 3-7 Example of hydrograph ........................................................................................................... 54
Figure 3-8 Example of a flow duration curve (FDC)............................................................................... 55
Figure 3-9 Example of FDC with logarithmic scale ................................................................................. 56
Figure 3-10 Example of standardised FDCs ............................................................................................. 57
Figure 3-11 Conveyance system (example 3.1)........................................................................................ 59
Figure 3-12 Residual flow ........................................................................................................................ 61
Figure 3-13 Example of turbine efficiency as a function of flow ............................................................. 63
Figure 3-14 Variation of net head vs. river flow ...................................................................................... 64
Figure 3-15 Components of hydrological model ...................................................................................... 68
42
Guide on How to Develop a Small Hydropower Plant ESHA 2004
LIST OF TABLES
Table 3-1 Typical values of Manning's n for watercourses ...................................................................... 54
Table 3-2 Minimum technical flow of turbines ........................................................................................ 62
Table 3-3 Typical design flood criteria..................................................................................................... 66
Table 3-4 Probability of occurrence ......................................................................................................... 66
LIST OF PHOTOS
Photo 3-1 Gauging station in a river ......................................................................................................... 47
Photo 3-2 Current meters .......................................................................................................................... 50
43
Guide on How to Develop a Small Hydropower Plant ESHA 2004
3. EVALUATING STREAMFLOW i
3.1. Introduction
All hydroelectric generation depends on falling water. This makes hydropower extremely site
dependent. First of all, a sufficient and dependable stream flow is required. Secondly, the
topographic conditions of the site must allow for the gradual descent of the river in a river stretch be
concentrated to one point giving sufficient head for power generation. This head can be created by
dams or by leading the water in parallel to the river in a waterway with low head losses compared to
the natural stream, or very often, by a combination of both.
Planning for the exploitation of a river stretch or a specific site is one of the more challenging tasks
that face a hydropower engineer, since there are an unlimited number of practical ways in which a
river or site can be exploited.
The hydropower engineer has to find the optimum solution for plant configuration, including dam
type, water conveyance system, installed generating capacity, location of various structures etc. The
success of the hydropower engineer depends on experience and an almost “artistic” talent, since a
strictly mathematical optimisation approach is impossible, due to the number of possibilities and site-
specific conditions.
When a site has been identified as topographically suitable for hydropower, the first task is to
investigate the availability of an adequate water supply. For an ungauged watercourse, where
observations of discharge over a long period are not available, it involves the science of hydrology,
the study of rainfall and stream flow, the measurement of drainage basins, catchment areas,
evapotranspiration and surface geology.
Figure 3-1 Schematic layout of a hydro development
Figure 3.1 illustrates how the water flowing from point A to point B, with elevations ZA and ZB, loses
potential energy corresponding to the drop in elevation. This loss of potential energy occurs
regardless of the path along the watercourse or via an open canal, penstock and turbine. The potential
energy lost can be converted to power lost according to the equation:
44
Guide on How to Develop a Small Hydropower Plant ESHA 2004
P = Q·Hg·γ
Where:
P is the power in kW lost by the water
Q is the flow in m3/s
Hg is the gross head in m, = ZA – ZB, and
γ is the specific weight of water, (9.81 kN/m3).
The water can follow the riverbed, losing power through friction and turbulence resulting in a
marginal rise in the temperature of the water. Or it can flow from A to B through an artificial
conveyance system with a turbine at its lower end. In this case the power will be used mainly for
running a turbine, and a smaller part of the power is lost in friction in the conveyance system. In the
latter case it is the power lost in pushing through the turbine that will be converted to mechanical
energy and then, by rotating the generator, to produce electricity.
The objective is to reduce construction costs while conserving the maximum amount of power
available to rotate the generator. To estimate the water potential one needs to know the variation of
the discharge throughout the year and how large the gross available head is. In the best
circumstances the hydrologic authorities would have installed a gauging station in the stretch of
stream under consideration, and stream flow time series data would have been gathered regularly
over several years.
Unfortunately, it is rather unusual for regular gauging to have been carried out in the stretch of river
where the development of a small hydro scheme is proposed. If, however, it does happen, then it will
suffice to make use of one of several approaches that can be used to estimate the long-term average
annual flow and the flow duration curve for the stretch in question (these approaches will be
explained later).
Whether or not regular gauging has taken place, the first step is to do some research, to ascertain if
there are stream flow records for the stretch of river in question. If not, then in other stretches of the
same river or a similar nearby river that permits the reconstitution of the time series for the referred
stretch of river.
3.2. Stream flow records
In Europe, stream flow records can be obtained from national hydrological institutes. These stream
flow records can be of several different types, each useful for the evaluation of the generating
potential of the considered site. These include:-
• Measured stream flow data for a number of gauged sites
• Stream flow characteristics for these sites such as mean flow and flow duration curves (both
expressed as actual flow and generalised as runoff per unit area of the catchment)
• Runoff maps, etc
There is a United Nations organisation, the “World Meteorological Organisation”, with a hydrologic
information service (INFOHYDRO) whose objective is to provide information regarding:
• National and international (governmental and non-governmental) organisations,
• Institutions and agencies dealing with hydrology;
• Hydrological and related activities of these bodies;
45
Guide on How to Develop a Small Hydropower Plant ESHA 2004
• Principal international river and lake basins of the world;
• Networks of hydrological observing stations of countries - numbers of stations and duration
of records;
• National hydrological data banks - status of collection, processing and archiving of data;
• International data banks related to hydrology and water resources.
Further information can be obtained at www.wmo.ch (At the date of printing, the INFOHYDRO
database was going through a major revision and was not available)
Figure 3-2 Measuring the river stage, definitions
3.3. Evaluating stream flows by discharge measurements
If appropriate stream flow time series cannot be found, the discharge should preferably be directly
measured for at least a year. A single measurement of instantaneous flow in a watercourse is of little
use. To measure the discharge several methods are available:
3.3.1. Velocity-area method
This is a conventional method for medium to large rivers, involving the measurement of the cross
sectional area of the river and the mean velocity of the water through it. It is a useful approach for
determining the stream flow with a minimum effort. An appropriate point must be selected on a
relatively straight, smoothly flowing portion of the river to be gauged (figure 3.2).
The river at this point should have a uniform width, with the area well defined and clean.
As discharge varies, the top water level (termed the stage of the river) rises and falls. The stage is
observed daily at the same time each day on a board - marked with metres and centimetres. In
modern gauging stations, instead of a board, that requires regular observations, any one of several
water-level measurement sensors is available which automatically register the stage. To calibrate the
stage observations or recordings, periodic discharge measurements from the lowest to the highest are
made over a time period of several months. Photo 3.1 shows a gauging station in a river.
46
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 3-1 Gauging station in a river
The correlation stage-discharge is called a rating curve (figure 3.3) and permits the estimation of the
river discharge by reading the river stage. To draw this curve, both the stage and the discharge must
be simultaneously read. It is strongly recommended that to begin measuring the low flows, one
should use the data to draw a curve that correlates the flows and the 'n' Manning coefficient. Later on
the river slope-area method (section 3.3.3) can be used to estimate the high flows, which are often
impossible to measure with the other methods.
When a rating curve has been graphically established, based on a number of readings, its
mathematical formulation can be readily derived, which facilitates interpretation of the stage
readings. The rating curve (figure 3.3) is represented by the function:-
Q = a (H+B)n
(3.1)
Where a and n = constants
H = river stage as measured or recorded
B = correction factor to get the actual level
To compute B (see figure 3.2) the data corresponding to two discharges should be noted, such as
Q1 = a (H1+B)n
Q2 = a (H2+B)n
47
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3-3 Rating curve
By introducing a third “factual” reading H3 vs. Q3, where Q3 (indexes in figure 3.3 are not
representative) is defined as the square root of the product of Q1 and Q2, and the corresponding H3 is
taken from the graphical representation of the rating curve Q3, this can be expressed as:-
()()()nnnBHaBHaBHaQQQ+⋅+=+=⋅=213213
consequently:
()()(BHBHBH+⋅+=+21
2
3 )
and therefore:
321
21
2
3
2 HHH
HHHB−+
−= (3.2)
There are ISO recommendations for the correct use of this technique.
Measuring the cross-sectional area
To compute the cross-sectional area of a natural watercourse it should be divided into a series of
trapezoids (figure 3.4). Measuring the trapezoid sides, by marked rules, illustrated in figure 3.4, the
cross-section would be given by:-
n
hhhbSna+++=.....2 (3.3)
48
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Measuring the velocity
Since the velocity both across the flow and vertically through it is not constant, it is necessary to
measure the water velocity at a number of points to obtain a mean value. There are several ways of
doing this, two of which are discussed below.
Figure 3-4 Measuring the cross-sectional area
By float
A floating object, which is largely submerged (for instance a wood plug or a partially filled bottle) is
located in the centre of the stream flow. The time t (seconds) elapsed to traverse a certain length L
(m) is recorded. The surface speed (m/s) would be the quotient of the length L and the time t. To
estimate the mean velocity, the above value must be multiplied by a correction factor that may vary
between 0.60 and 0.85 depending on the watercourse depth and their bottom and riverbank
roughness (0.65 is a well accepted value). The accuracy of this method is dependant on the range of
correction factor.
By mechanical current-meter
A current-meter is a fluid-velocity-measuring instrument. Current meters are classified in two types:-
Vertical axis rotor with cups: This type of instrument has a circle of small conical cups, disposed
horizontally which rotate about the suspension axis. (Photo 3.2 right photo) These current meters
operate in lower velocities than the horizontal axis rotor types, and have the advantage of bearings
being well protected from silty waters. The rotor can be repaired in the field.
Horizontal axis rotor with vanes (propeller): A small propeller rotates about a horizontal shaft, which
is kept parallel to the stream by tail fins. (Photo 3.2 left photo) The instrument is weighted to keep it
as directly as below the observer as possible. This rotor has the advantage of being less likely to
disturb the flow around the measuring point and also for being less likely to become entangled by
debris.
49
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 3-2 Current meters
Each revolution of the propeller is recorded electrically through a cable to the observer and the
number of revolutions is counted by the observer, or automatically by the instrument itself, over a
short period (say 1 or 2minutes). These observations are converted into water velocities from a
calibration curve for the instrument (modern instruments, with microprocessor technology will
compute this and display it almost immediately). By moving the meter vertically and horizontally to
a series of positions (whose coordinates in the cross-section are determined), a complete velocity
map of the cross-section can be drawn and the discharge through it calculated.
In the case of medium to large rivers, observations are made by lowering the meter from a bridge,
however, if the bridge is not single-span there will be divergence and convergence of the streamlines
caused by the piers, and this can cause considerable errors. In many instances, however, the gauging
site, which should be in as straight and uniform a reach of the river as possible, will have no bridge.
In such cases, particularly if it is deep and in flood, a cable to hold a stable boat must be provided,
together with a lighter measuring cable to determine horizontal position in the cross-section.
Since the drag on a boat, with at least two occupants and suspended current-meter, is considerable, a
securely fastened cable should be used. The presence of suitable large trees at a particular site often
necessitates its choice for this reason. Alternatively, for very large rivers, cableways are sometime
used to suspend the meter, directly from a cable car, the instrument in this latter case being
positioned by auxiliary cables from the riverbanks or from the cable car itself.
Depths should always be measured at the time of velocity observation since a profile can change
appreciably during flood discharges. Observers should also remember such elementary rules as to
observe the stage before and after the discharge measurement, and to observe the water slope by
accurate levelling to pegs at the water level as far upstream and downstream of the gauging site as is
practicable, up to (say) 500m in each direction.
As water velocities increase in high floods the weighted current meter will be increasingly swept
downstream on an inclined cable. The position of a meter in these circumstances can be found
reasonably accurately if the cable angle is measured. Ballast can be increased but only within limits.
50
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Rods can be used to suspend the meters but a rigid structure in the boat will then be required to
handle the rods, calling for a stable platform on a catamaran-type of craft. Rod vibration and bending
are common in deep rivers unless large diameter rods are employed, in which case the whole
apparatus is getting very heavy and unmanageable.
By electro-magnetic current-meter
An electro-magnetic (e/m) current-meter is an electrical induction-measurement instrument, with no
moving parts, mounted in a totally enclosed streamlined probe. The probe can be mounted on rods
and held at various depths or suspended on a cable.
The e/m meter has the advantages of being smaller and having a wider measurement range than the
propeller meters. It is particularly useful at very low velocities when propeller meters become erratic.
Its sensitivity and lower vulnerability to fouling from weeds and debris make it attractive for use in
heavily polluted or weedy streams.
Each unit is provided with a surface control box with a digital display and dry-cell batteries. A set of
stainless steel wading rods is also standard equipment. Latest models have built-in battery-charger
circuits.
It will be appreciated that since each river is unique, a careful assessment of its width, depth, likely
flood velocities, cable-support facilities, availability of bridges, boats, etc. needs to be made before a
discharge measurement programme can begin.
The discharge at the chosen measuring point is best obtained by plotting each velocity observation
on a cross section of the gauging site with an exaggerated vertical scale. Isovels (velocity profiles –
contours of equal velocity) are then drawn and the included areas measured by a planimeter.
Alternatively, the river may be subdivided vertically into sections and the mean velocity of each
section applied to its area. In this method the cross-sectional area of any one section, where
measurements, are taken should not exceed 10 per cent of the total cross-sectional area.
A check should always be made using the slope-area method of section 3.3.4 and a value obtained
for the Manning's number “n”. In this way knowledge of the n values of the river at various stages
will be built up and this may prove most valuable for subsequently extending the discharge-rating
curve.
To ensure uniformity in the techniques of current-meter gauging, ISO has published a number of
recommendations.
By dilution methods.
Dilution gauging is particularly suited to small turbulent streams where depths and flows are
inappropriate for current metering, and flow structures would be unnecessarily expensive. The
method involves the injection of a chemical into the stream and the sampling of the water some
distance downstream after complete mixing of the chemical in the water has occurred. The chemical
can either be added by constant-rate injection until the sampling downstream reveals a constant
concentration level, or administered in a single dose as quickly as possible, known as “gulp
injection”. In the latter case, samples over a period of time disclose the concentration-time
correlation. In both cases the concentration of chemical in the samples is used to compute the
dilution, and hence, the discharge of the stream can be obtained.
51
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3-5 Conductivity time curve
Analysis of the samples is by an automated colorimetric procedure that estimates the concentration
of very small amounts of chromium compound by comparing with a sample of the injection solution.
The equipment is expensive and somewhat specialised.
Recent developments have substituted the above procedures by a method developed by Littlewood7,
requiring only simple and relatively cheap equipment. The method depends on the electrical
conductivity of solutions of common salt (NaCl) in the stream water and is a version of the relative-
dilution gauging method developed by Aastad and Sognen.
The discharge is measured by gradually rleasing a known volume (V) of a strong salt solution (c1)
into the stream at a known rate (q), and measuring, at short intervals, the change in conductivity of
the water at the downstream end of the mixing length. It is then is possible to plot a conductivity-
time curve, along a time T as in figure 3.5. The average of the ordinates, of this curve, represent the
average of the difference in conductivity, between the salt solutions and the stream water, upstream
from the injection point. If a small volume, v, of the particular strong solution is added to a large
volume V* of the stream water, and the differences in conductivity Δc* are measured, the discharge
will be then given by the equation:
'
**
2 c
cxv
VxT
VQ Δ
Δ=
(3.5)
where:
v = volume of injection solution
T2 = duration of solute wave (s)
V = volume of the strong solution added to a larger
V* = volume of stream water
Δc* = change in conductivity (ohm-1) consequence of the dilution of v in V*
Δc* = ordinate's average curve conductivity-time
52
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3-6 Discharge measurements using weirs and notches
3.3.2. Weir method
If the watercourse being developed is reasonably small (say < 4 m3/s) then it may be possible to
build a temporary weir. This is a low wall or dam across the stream/river to be gauged with a notch
through which all the water may be channelled. Many studies have established accurate formulae for
the discharge through such notches. A simple linear measurement of the difference in level between
the upstream water surface and the bottom of the notch is sufficient to quantify the discharge.
However, it is important to measure the water surface level some distance back from the weir (at
least four times the depth over the base of the notch) and to keep the notch free of sediment and the
edge sharp and chamfered on the downstream side of the top of the weir/notch.
Several types of notches can be used:-
• Rectangular,
• V-notch
• Trapezoidal.
The V-notch is the most accurate at very low discharges but the rectangular or trapezoidal types are
capable of a much wider range of flows. The actual notches may be metal plates or planed hardwood
with sharp edges, built to the dimensions of figure 3.6.
Another fairly accurate method is to construct a “Flume”. A “Flume” is where a stream is channelled
through a particular geometrically shaped regular channel section for some distance before entering a
length of different cross-section, usually made so by side contraction or steps in the bed, generally in
the shape of a “Venturi”. These structures have the advantage over weirs in that they do not obstruct
the flow or “ponding” of the water upstream, they can also be very accurate and provide a permanent
gauging station.
53
Guide on How to Develop a Small Hydropower Plant ESHA 2004
To ensure uniformity in the techniques of current-meter gauging ISO has published various
recommendations. The catalogue with ISO recommendations can be obtained at:
http://www.iso.ch/iso/en/CatalogueListPage.CatalogueList?ICS1=17&ICS2=120&ICS3=20
Figure 3-7 Example of hydrograph
3.3.3. Slope-area method
This method depends on some of the hydraulic principles described in Chapter 2 and is useful for
high flows where other methods are impractical. It presupposes that it is practical to drive in pegs or
make other temporary elevation marks at the water-surface level (upstream and downstream of the
discharging site) at the time the flow measurements are taken. These marks can subsequently be used
to establish the water slope (S). Cross-sectional measurements are taken to establish the area (A) and
hydraulic radius of the section (R). Once these parameters are known the discharge is computed by
the Manning formula
n
SRAQ
2/13/2 ⋅⋅=
(3.6)
This method is sometimes criticised because of its dependence on the value of n. Since n for natural
streams is about 0.035, an error in n of 0.001 gives an error in discharge of 3 per cent. This error may
be partially reduced by plotting n against stage for all measured discharges, so that the choice of n
for high stages is not arbitrary but is taken from such a plot. If a high flood slope can be measured,
then this method may well be the best one for such flows. Typical values of Manning's n for
watercourses and common pipe materials are given Table 3.1
Table 3-1 Typical values of Manning's n for watercourses
Watercourses n
Natural stream channels flowing smoothly in clean conditions 0.030
Standard natural stream or river in stable conditions 0.035
River with shallows and meanders and noticeable aquatic growth 0.045
River or stream with rods and stones, shallows and weedy 0.060
54
Guide on How to Develop a Small Hydropower Plant ESHA 2004
3.4. Stream Flow Characteristics
A programme of stream gauging, at a particular site over a period of years, will provide a table of
discharges that, to be of any use, has to be organised into a usable form.
0
1
2
3
4
5
6
7
8
9
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
Duration [%]
Figure 3-8 Example of a flow duration curve (FDC)
3.4.1. Hydrograph
One way of doing this is to plot them sequentially in the form of a hydrograph, which shows
discharge against time, in chronological order (see figure 3.7).
3.4.2. Flow Duration Curves (FDC)
Another way of organising discharge data is by plotting a flow duration curve (FDC) An FDC shows
for a particular point on a river the proportion of time during which the discharge there equals or
exceeds certain values. It can be obtained from the hydrograph by organising the data by magnitude
instead of chronologically. If the individual daily flows for one year are organised in categories as
shown below:-
No of days % of the year
Flows of 8.0 m3/s and greater 41 11.23
Flows of 7.0 m3/s and greater 54 14.9
Flows of 6.5 m3/s and greater 61 16.8
Flows of 5.5 m3/s and greater 80 21.8
Flows of 5.0 m3/s and greater 90 24.66
Flows of 4.5 m3/s and greater 100 27.5
Flows of 3.0 m3/s and greater 142 39
Flows of 2.0 m3/s and greater 183 50
Flows of 1.5 m3/s and greater 215 58.9
55
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Flows of 1.0 m3/s and greater 256 70
Flows of 0.35 m3/s and greater 365 100
If the above figures are plotted then a graph like figure 3.8 will be obtained, which represents the
ordinates of figure 3.7 arranged in order of magnitude instead of chronologically.
0.1
1
10
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
Duration [%]Flow [m3/s]
Figure 3-9 Example of FDC with logarithmic scale
Most gauging stations (in the EU) are computerised, the easiest way to derive a FDC is to transpose
the digital data to a spreadsheet, sorting them in descending order, and then by hand or by using a
simple macro, classify the data as in the above table. Once done, the same spreadsheet, using its
graphic building capability will draw the curve FDC (like the one drawn in figure 3.8).
For many rivers the ratio of peak to minimum discharges may be two or more orders of magnitude
and FDCs for points on them are often more conveniently drawn with the (Q) ordinate to a
logarithmic scale, and a normal probability scale used for the frequency axis. On these, logarithmic
graphs, the discharges are distributed, such that the FDC plots as a straight line. Figure 3.9 represents
the graph in figure 3.8 with the vertical axis in logarithmic scale.
3.4.3. Standardised FDC curves
FDCs for different rivers can be compared by presenting them in a standardised way. The discharges
are divided firstly by the contributing catchment areas and secondly by weighted average annual
rainfall over the catchment areas. The resulting discharges, in m3/s or litres /s, per unit area, per unit
annual rainfall (typically m3/s/km2/m) can then be compared directly. Figure 3.10 shows twenty
FDCs corresponding to catchment areas of different geological composition, drawn to a double
logarithmic scale. A collection of regional flow-duration curves shows the effect of a basin’s
superficial geology on the shape of the curves. If the flow duration curves of different catchments are
standardised by the catchments mean flow, certain low flow statistics, such as Q95 can be used to
describe the whole flow duration curve.
56
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3-10 Example of standardised FDCs
Another method for standardising FDCs is to express Q in terms of Q/Qm, where Qm is the mean
flow. The use of such a non-dimensional ordinate allows all rivers, large and small, to be compared
on the same graph. If sufficient records are available from neighbouring rivers of similar
topographical character in a similar climate, these methods can be very useful in evaluating stream
flows at ungauged sites. If we know the FDC for another stretch of the same river, it will be possible
to extrapolate it using the ratio of areas of the respective catchment basins.
When there are no flow records for a particular location it is necessary to proceed from first
principles. Rainfall data is normally available from national agencies on an annual-average basis, but
often only on a fairly small scale. Therefore, attempts should always be made to find local records,
which will indicate seasonal variation. Failing that, a standard rain gauge should be installed in the
catchment area, as soon as the studies are being considered. Even one year's records will help in the
production of a synthesised FDC.
The first step is to estimate the mean annual flow Qm [also referred to average daily flow (ADF)].
In UK the mean flow is estimated using a rainfall catchment water balance methodology, the long
term average annual runoff of rainfall can be assumed to be equal to the difference between standard
average annual rainfall (SAAR) and actual evaporation (AE). Catchment values of SAAR and
potential evaporation are estimated from the rainfall and potential evaporation (PE) maps.
Actual evaporation is estimated from potential evaporation using a scaling factor r, where r increases
with SAAR and hence increasing water availability. For catchments with annual average rainfall in
excess of 850mm per year, it is assumed that actual evaporation is equal to potential. This
relationship between SAAR is given by:
r = 0.00061 x SAAR + 0.475 for SAAR < 850 mm
r = 1.0 for SAAR > 850 mm
Actual evaporation is calculated using: AE = r x PE
57
Guide on How to Develop a Small Hydropower Plant ESHA 2004
The average runoff depth (AARD in millimetres) over the catchment area (AREA in km2) is
converted to mean flow in m3/s by:
Qm = (AARD x AREA) / 31536
Although the mean annual flow gives an idea of a stream’s power potential, a firmer knowledge of
the stream’s flow regime, as obtained from a flow duration curve, is needed. The flow duration curve
depends on the type of soil on which the rain falls. If it is very permeable (sand) the infiltration
capacity will be high and the groundwater will be a large proportion of flow in the stream. If it is
impermeable (rock) the opposite will be the case. The catchments of high permeability and large
groundwater contributions will therefore tend to have more regular discharges with less fluctuating
flow than rocky catchments, where the variations will be great and will reflect the incidence of
rainfall to a much greater extent.
In UK, for instance, soils have been categorised into 29 discrete groups that represent different
physical properties and hydrological response. The classification system is referred to as the
Hydrology of Soil Types (HOST) classification. By measuring the areas of each of these categories,
within the catchment, as a proportion of the whole, the BFI (Base Flow Index) can be computed.
Knowing the BFI of the catchment, a standardised FDC can be selected from figure 3.11.
Multiplying the ordinates of the selected FDC by the catchment Qm, the particular flow duration
curve of the site is obtained.
In Spain, the distribution of the soils has been identified from the Soil Map of the European
Communities (CEC, 1985), which is based on the FAO/UNESCO Soil Classification of the World.
Nineteen soils are represented within the gauged catchments considered in the study.
There are actually many watershed models that permit calculation of the runoff for a certain
catchment basin taking into account the average daily rainfall, the potential evapotranspiration, the
soil composition, the basin slope and area, the stream length, and other parameters. All those
programs allow an analysis of the snowmelt and its contribution to the discharge, and also the
creation of flood inundation maps, flood depths maps and flood impact maps.
3.4.4. FDCs for particular months or other periods
It is always important to know when, during the year, water will be available for generation. This is
required when considering the economics of schemes in those networks where tariffs, paid by
utilities to independent producers, vary with the season of the year and time of day.
FDCs can be produced for particular periods of time as well as for particular years. Indeed, it is
standard practice to prepare FDCs for six "winter" months and six "summer"
months. This can be taken further, to obtain FDCs for individual months, if so desired. It is simply a
matter of extracting the flow records for a particular month from each year of record and treating
these data as the whole population. If sufficient flow records for this process do not exist, then the
rainfall record can be used.
3.4.5. Water Pressure or “head”
58
Guide on How to Develop a Small Hydropower Plant ESHA 2004
3.4.5.1.Evaluation of gross head
The gross head is the vertical distance that the water falls through in giving up its potential energy
(i.e. between the upper and lower water surface levels).
Field measurements of gross head are usually carried out using surveying techniques. The precision
required in the measurement will limit the methods that can be employed.
In the past, the best way to measure gross head was by levelling with a surveyor’s level and staff,
however this was a slow process. Accurate measurements were made by a tachometer or less
accurately by a clinometer or Abney level. Nowadays with digital theodolites, electronic digital and
laser levels and especially with the electronic total stations the job has been simplified.
The modern electronic digital levels provide an automatic display of height and distance within about
4 seconds with a height measurement accuracy of 0.4 mm, and the internal memory that can store
approximately 2,400 data points. Surveying by Global Positioning Systems (GSM) is now used
widely and a handheld GPS receiver is ideal for field positioning, and rough mapping.
3.4.5.2. Estimation of net head
Having established the gross head available, it is necessary to make allowances for the losses, from
trash racks, pipe friction, bends and valves. In addition to these losses, certain types of turbines need
to discharge their water to atmosphere, above the level of the tail water (the lower surface level).
Figure 3-11 Conveyance system (example 3.1)
The gross head minus the sum of all the losses equals the net head, which is available to drive the
turbine. Example 3.1 will help to clarify the situation:
Example 3.1
59
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3.13 shows the pipe layout in a small hydropower scheme. The nominal discharge is 3
m3/s and the gross head 85 m. The penstock is 1.5 m diameter in the first length and 1.2 m in
the second one. The radius of curvature of the bend is four times the diameter of the pipe. At
the entrance of the intake there is a trash rack inclined 600 with the horizontal. The rack is
made of stainless steel flat bars, 12 mm thick and the width between bars is 70 mm. Estimate
the total head loss?
According to experience the velocity at the entrance of the rack should be between 0.25 m/s and 1.0
m/s. The required trash rack area is estimated by the formula:-
αsin
11
01V
Q
bt
t
KS ⎟⎠
⎞⎜⎝
⎛
+=
where S is the area in m2, t the bar thickness (mm), b the distance between bars (mm), Q the
discharge (m3/s), V0 the water velocity at the entrance and K1 a coefficient which, if the trash rack
has an automatic cleaner, is equal to 0.80. Assuming V0 = 1 m/s, S=5.07 m2. For practical reasons a 6
m2 trash rack may be specified, corresponding to a V0 = 0.85 m/s, which is acceptable. The headloss
traversing the trash rack, as computed from the Kirschner equation
mhr007,081,92
8,0
70
124,2
24/3
=⋅⎟⎠
⎞⎜⎝
⎛=
The friction losses in the first penstock length are a function of the water velocity, 1.7 m/s. The
entrance to the pipe has a good design and coefficient Ke = 0.04 (see figure 2.11). The head loss in
the first length according to Manning’s equation is:
mhL
h
f
F 19,0;00177,0 ==
The headloss coefficient in the first bend is Kb = 0.085 (one half of the corresponding loss of a 901
bend); in the second Kb = 0.12 and in the third Kb = 0.14 The taper pipe, with an angle of 300, gives a
loss in the contraction hc = 0.02 m (for a ratio of diameters 0.8 and a water velocity in the smaller
pipe =2,65 m/s)
The friction headloss in the second length is computed in the same way as the first one, and
accordingly
mhL
h
f
F 10,1;0169,0 ==
The coefficient of headloss in the gate valve is Kv= 0.15. Therefore the headloss due to friction is
estimated to be
0.19 + 1,10 = 1.29 m
The additional headlosses will be as follows:-
• In the trash rack 0.007 m
• In the pipe entrance 0.04 x 0,147 0.059 m
• In the first bend 0.085 x 0.147 0.013 m
60
Guide on How to Develop a Small Hydropower Plant ESHA 2004
• In the second bend 0.12 x 0,359 0.043 m
• In the third bend 0.14 x 0,359 0.050 m
• In the confusor 0.02 x 0,359 0.007 m
• In the gate valve 0.15 x 0,359 0.054 m
Headlosses 0,233 m
The total head loss is equal to 1,29 m friction loss plus 0,23 m in local losses, giving a net head of
83.48 m. This represents a loss of power of 1,8% which is reasonable.
3.5. Residual, reserved or compensation flow
Uncontrolled abstraction of water from a watercourse (e.g. passing it through a turbine) even if it is
returned to the stream close to the intake, could lead to sections of the watercourse being left almost
dry with serious impacts on aquatic life.
Minimum
flow
Figure 3-12 Residual flow
To avoid this happening, permission to divert water through a hydro turbine or a licence to abstract
from a river or stream will almost always specify that a certain residual flow should remain. The
residual flow is sometimes called other names, depending on the country, or authority responsible,
e.g. "reserved flow", "prescribed flow" and "compensation flow" are terms commonly used.
This residual flow should be carefully evaluated since a flow that is too small would cause damage to
aquatic life in the stream. On the other hand an unnecessarily large flow effects the power production
and especially so in periods of low flow, thus reducing the benefits of the installation.
3.6. Estimation of plant capacity and energy output
The FDC provides a means of selecting the right design discharge, and by taking into account the
reserved flow and the minimum technical turbine flow, an estimate of the plant capacity and the
average annual energy output.
61
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 3.12 illustrates the FDC of the site it is intended to evaluate. The design flow has to be
identified through an optimisation process, studying a range of different flows, which normally gives
an optimum design flow significantly larger than the difference between the mean annual flow and
the reserved flow. Once the design flow is defined and the net head estimated, a suitable type of
turbine must be identified (refer chapter 6).
Figure 3.12 shows the useable region of the flow duration curve. Every selected turbine has a
minimum technical flow (with a lower discharge the turbine either cannot operate or has a very low
efficiency) and its efficiency is a function of the operating discharge.
The average annual energy production (E in kWh) is a function of:
E = fn (Qmedian, Hn, ηturbiner ηgearbox, ηtransformer, y, h)
Where:
Qmedian = flow in m3/s for incremental steps on the flow duration curve
Hn = specified net head
ηturbine = turbine efficiency, a function of Qmedian
ηgenerator = generator efficiency
ηgearbox = gearbox efficiency
ηtransformer = transformer efficiency
y = specific weight of the water (9.81 KN/m3)
h = number of hours for which the specified flow occurs.
The energy production can be calculated by dividing the useable area into vertical 5% incremental
strips starting from the origin. The final strip will intersect the FDC at Qmin or Qreserved which ever is
larger. For each strip Qmedian is calculated, the corresponding hturbine value is defined for the
corresponding efficiency curve, and the energy contribution of the strip is calculated using the
equation:
E = W x Qmedian x H x çturbine x çgenerator x çgearbox x çtransformer x Õ x h
Where: W = strip width = 0.05 for all strips except the last one that should be calculated
h = number of hours in a year
y = specific weight of the water (9.81 KN/m3)
The average annual energy production is then the sum of the energy contribution for each strip. The
capacity of each turbine (kW) will be given by the product of their design flow (m3/s), net head (m),
turbine efficiency (%), and specific weight of the water (kNm-3).
In Chapter 6, curves of turbine efficiency against flow for the commercial turbines are shown. Table
3.2 gives the minimum technical flow for different types of turbines as a percentage of the design
flow.
Table 3-2 Minimum technical flow of turbines
Turbine type Qmin (% of Qdesign)
Francis 50
Semi Kaplan 30
Kaplan 15
62
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Pelton 10
Turgo 20
Propeller 75
3.6.1. How the head varies with the flow and its influence on the turbine capacity
Depending on the river flow and the flow admitted to the turbines, the head can vary significantly.
The upstream water level may vary with flow. If the intake pond is controlled by an overflow weir
without any gates, the water level will rise with the flow. However, if the intake pond is controlled
by gates in order to operate at a specified reservoir level, the water level may remain constant even
during high flow periods. During low flow periods, the upstream water level may also be lower due
to draw down of the reservoir.
The head losses in the adduction system varies with the square of the admitted flow, and thus for
low flow seasons with low turbine flow the head loss in the adduction system can be substantially
reduced.
The downstream water level may vary with the flow. This depends on the water body into which
the water is discharged. If discharging directly into a headwater pond controlled by gates in a
downstream development, the water levels may remain almost constant even for higher flows. If the
water is discharged into a natural stream, the water levels again may vary considerably.
Figure 3-13 Example of turbine efficiency as a function of flow
In medium and high head schemes the head can be considered constant, because variations in the
upper or lower surface levels are small compared with the head. In low head schemes, when the flow
increases over the value of the rated flow of the water surface level, both in the intake and in the
tailrace, may increase but at different rates, so that the head can potentially increase or decrease.
If a turbine operates with a head H1 = ZUpstream - ZDownstream, other than the rated head Hd, the flow
admitted by the turbine will be:- note finished at 13/0305
d
d H
HQQ 1
1 ⋅= (3.7)
Headwater level is normally kept at spillway crest level when all the river discharge passes through
the turbines. When the river discharge exceeds maximum turbine discharge, excess flow will pass
over the spillway. The reservoir level corresponding to different spillway flows can easily be
63
Guide on How to Develop a Small Hydropower Plant ESHA 2004
calculated. In this case measuring the head on the spillway crest we have at the same time the level
of the intake water surface and the river discharge (including the discharge from the turbines).
The tailrace level is more difficult to estimate. The Hydrologic Engineering Centre (HEC) of the US
Army Corp of Engineers in Davis, California, has developed a computer program, HEC RAS, that
can be downloaded free of charge from INTERNET (http://www.usace.army.mil). Although freely
available and straightforward to use, the results as always depend on the quality of the input. Figure
3.14 shows an example of how the head varies with the flow in a real case and its influence on the
power delivered at different river discharges.
discharge m3/s
Figure 3-14 Variation of net head vs. river flow
3.6.2. Peaking operation
Electricity prices at peak hours can be substantially higher than in off-peak hours, hence the interest
in providing an extended forebay or pound, big enough to store the water necessary to operate, at
maximum during peak hours. To evaluate this volume:-
QR = river flow (m3/s)
QD = rated flow (m3/s)
QP = flow needed to operate in peak hours (m3/s)
QOP = flow needed to operate in off-peak hours (m3/s)
tP = daily peak hours
tOP = daily off-peak hours (24 - tP)
Qres = reserved flow (m3/s)
Qtmin = minimum technical flow of turbines (m3/s)
H = head (m)
The volume V will be given by:-
64
Guide on How to Develop a Small Hydropower Plant ESHA 2004
VR = 3.600·tP·(QP-(QR-Qres))
If the pound should be refilled in off-peak hours
tP (QP-(QR-Qres))δtOP (QR-Qres)
hence :-
Qp < top-tp/tp . (QR-Qres)
the flow available to operate in off-peak hours will be:-
min
)(24 Qt
QtQQQ
OP
PPresR
OP >−−=
3.7. Firm energy
Firm energy is defined as the power that can be delivered by a specific plant during a certain period
of the day with at least 90 –95% certainty. A run-of-river scheme has a low firm energy capacity. A
hydropower plant with storage does, however, have considerable capacity for firm energy.
If the hydropower scheme is to be connected to an electrical network that includes several types of
power and where the hydropower installations are geographically distributed, as is the case in
Europe, the firm power capacity of singular plants may, not be important.
If a small hydro scheme has been developed as the single supply to an isolated area, the firm energy
is extremely important. As failure to meet demand, could result in power shortages and blackouts.
3.8. Floods
The stream flow is the fuel of the plant, but stream flow in the form of floods is also a potential threat
to all structures built in rivers. Therefore the hydrological investigation must address not only water
availability for production, but also frequency and severity of floods so as to design flood protection
and control into the scheme. The design flood is not only characterized by its peak value of flow, but
a hydrograph flood flows should show the distribution of the flow over time.
3.8.1. Flood Control Design
It is important to distinguish between inflow floodwater and required spillway capacity, since
considerable routing effects take place in reservoirs.
For reservoirs with dams that are at risk from high floodwaters it is usual to consider two differing
criteria:
1. Maximum Inflow Design Flood - that the facilities should be able to accommodate the
floodwater, without unacceptable risk of dam failure or other serious damages to the
structures. This flood is normally defined as the PMF, (Probable Maximum Flood) or similar.
2. Normal Operation Design Flood - that the facilities should be able to accommodate
floodwater without exceeding normal conditions of operation. This flood is usually defined as
a flood with a specific return period.
65
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Whereas for medium and low hazard dams, the requirements often discard the routing effects of the
reservoir, and specify that the spillway capacity shall exceed the peak flow of a flood with a specific
return period, typically between 100 and 1000 years.
The requirements regarding the design flood are usually specified in national legislation or industry
guidelines, and distinguish between high, medium and low hazard structures. In table 3.3 below,
typical design flood requirements are given:
Table 3-3 Typical design flood criteria
Structure Design Flood
High Hazard Maximum Inflow Design Flood: PMF, Probable
Maximum Flood or similar.
Alternatively 10.000-year flood
Normal Operation Design Flood:
1000-year flood.
Medium Hazard 100- to 1000-year flood
Low Hazard Typically 100-year flood although in some
countries no formal requirements exist.
With a 100-year flood, an annual probability of occurrence is understood to be 1/100. In other words,
the Return Period is the inverse of the frequency. In the table below, the probability of occurrence
during different life spans for different event frequencies is shown.
Table 3-4 Probability of occurrence
Life Span
Frequency (Return Period)
10 years 50 years 100 years 200 years
0,01 (100) 9,6 % 39 % 63 % 87 %
0,001 (1 000) 1 % 5 % 9,5 % 18 %
0,0001 (10 000) 0,1 % 0,5 % 1% 2%
The economically optimal design flood return period for a specific dam, considering the marginal
cost of additional spillway capacity and the cost of failure, is usually higher than the 100-year flood
even for low hazard structures.
3.8.2. Statistical analysis of flood data
There are basically two ways of arriving at a design flood:
• Statistical analysis of stream flow records
• Hydrological modelling of the catchment area
Typically, statistical analysis is used for less important structures that would not cause dramatic
consequences to life and society in case of failure, whereas hydrological modelling is required for
important and potentially dangerous dams in case of failure.
The object of the hydrological modelling is to arrive at a Probable Maximum Flood, or similar, to be
used for dam and spillway design. Frequency analysis is a statistical method to calculate the
probability of an event based on a series of previous events.
66
Guide on How to Develop a Small Hydropower Plant ESHA 2004
The technique for estimating the return period of flows is straightforward and based on records of
annual maximum flows. For the evaluation, a probability distribution that fits to the phenomenon
must be chosen. Generally logPearson III is recommended for flood estimation since it allows for
non-symmetrical probability distributions around the mean value, which is often the case in
hydrology, however the lognormal distribution is still widely used.
The non-symmetrical distribution is expressed in a skew coefficient. LogPearson III and the
calculation of the skew coefficient are very sensitive for short data series. Therefore, it is
recommended to use a modified skew factor based not only on the actual data series, but also
includes general experience for the specific geographical region.
In the graphical method, the annual maximum floods are arranged in order of size and then plotted
on probability paper applicable for the desired distribution. Generally the ordinate represents the
value and the abscissa represents the probability. The data are expected to fit, as close as possible, to
a straight line. The graph can then be used for interpolation, extrapolation or comparison purposes.
In case of extrapolation, effects of errors are magnified and caution is recommended.
In the analytical method, the mean value, standard deviation as well as the skew coefficient (in case
of logPearson III) of the logarithmic value of the flow record is calculated. Based on the desired
frequency, a frequency factor is read from a diagram. The logarithms of floods corresponding to
certain frequencies are then calculated as the mean value plus the standard deviation multiplied with
the corresponding frequency factor. The logarithms are then converted to actual flow values.
Both methods are explained in more detail in standard hydrology textbooks.
As an illustrative example the 100-year flood is calculated using the analytical method for the
lognormal and logPearson III probability distribution based on the following time series of annual
maximum flows:
Flow (m3/s) 0 1 2 3 4 5 6 7 8 9
1970- 65 32 45 87 34 29 26 35 42 41
1980- 36 29 55 46 31 26 34 31 39 61
The steps are as follows:
1: Calculate the logarithmic value of the flow records
2: Calculate the mean of the logarithms
3: Calculate the standard deviation of the logarithms
(3b: Calculate the skew factor for LogPearson III)
4: Read the frequency factor for the desired probability (f = 0,01)
5: Calculate the logarithm of the 100-year flow
6: Convert the logarithm to a flow value:
Using the LogNormal distribution the 100-year annual maximum flow is estimated at 83 m3/s, and
for the Log Pearson III distribution almost 25 % higher, or 103 m3/s. Which value is the more
correct? This example illustrates that even though the methods are straightforward, a good
professional judgment is required as to applicability and choice of method.
67
Guide on How to Develop a Small Hydropower Plant ESHA 2004
3.8.3. Hydrological modelling of the catchment area
In order to arrive at a design flood using hydrological modelling, a design rainfall is introduced to a
hydrological model comprising various components. The design rainfall is combined with other
critical factors such as soil moisture content, snow melting, ground water magazine contents etc.
This task is best left to the experts.
Figure 3-15 Components of hydrological model
68
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Bibliography
1. José Llamas, “Hidrología General. Principios y Aplicaciones”. Servicio Editorial de la
Universidad del País Vasco, 1933.
2. ISO 1100-1: 1996 “Measurement of liquid flow in open channels. Part 1: Establishment and
operation of a gauging station”.
3. ISO/DIS 110-2 “Measurement of liquid flow in open channels – Part 2: Determination of the
stage-discharge relation” (revision of ISO 1100-2: 1982).
4. ISO 2537: 1988 “Liquid flow measurement in open channels – Rotating element
currentmeters”.
5. ISO 955-1: 1994 “Measurement of liquid flow in open channels – Tracer dilution methods for
the measurement of steady flow – Part 1: General”.
6. ISO 3846: 1989 “Liquid flow measurement in open channels by weirs and flumes –
Rectangular broad-crested weirs”.
7. ISO 3847: 1977: “Liquid flow measurement in open channels by weirs and flumes – End-
depth method for estimation of flow in rectangular channels with a free overfall”.
8. ISO 4359-1983 “Liquid flow measurement in open channels: Rectangular, trapezoidal and
Ushaped flumes”.
9. ISO 4360: 1984 “Liquid flow measurement in open channels by weirs and flumes –
Triangular profile weirs”.
10. ISO 4362: 1992 “Measurement of liquid flow in open channels – Trapezoidal profile”
i By Jonas Rundqvist (SERO), Bernhard Pelikan (ÖVFK), Vincent Denis (MHyLab) and Celso Penche (ESHA)
69
Guide on How to Develop a Small Hydropower Plant ESHA 2004
CHAPTER 4: SITE EVALUATION METHODOLOGIES
CONTENTS
4. SITE EVALUATION METHODOLOGIES .................................................................................... 72
4.1. Introduction ............................................................................................................................... 72
4.2. Cartography............................................................................................................................... 72
4.3. Geochemical Studies................................................................................................................. 73
4.3.1. Methodologies to be used ..................................................................................................... 73
4.3.2. Methodologies. The study of a practical case....................................................................... 75
4.3.2.1. The weir .................................................................................................................... 75
4.3.2.2. The open channel ...................................................................................................... 76
4.3.2.3. The channel in tunnel................................................................................................ 78
4.3.2.4. The powerhouse ........................................................................................................ 82
4.4. Learning from failures .............................................................................................................. 82
LIST OF FIGURES
Figure 4-1 Schematic representation of the site........................................................................................ 74
Figure 4-2 Weir location and structures of both slopes ............................................................................ 75
Figure 4-3 Geological section of the colluvial formation ......................................................................... 76
Figure 4-4 Geomorphologic scheme of the channel trace ........................................................................ 76
Figure 4-5 A schematic cut of the tunnel under the colluvium................................................................. 79
Figure 4-6 Concrete lining forming the final section of the canal ............................................................ 79
Figure 4-7 A thrust fault, present in the La Rienda tunnel ....................................................................... 81
Figure 4-8 results of the jet-grouting operation ........................................................................................ 82
Figure 4-9 Scheme of Ruahihi canal......................................................................................................... 83
Figure 4-10 Longitudinal scheme of La Marea plant ............................................................................... 86
Figure 4-11 Plan view f La Marea plant ................................................................................................... 86
LIST OF PHOTOS
Photo 4-1 General view of the right-side slope ........................................................................................ 77
Photo 4-2 Local instabilities generated during the excavation works ...................................................... 77
Photo 4-3 One of the existing sliding scarps before the beginning of the works..................................... 78
Photo 4-4 A view of the Cordiñanes colluvium, under which the tunnel runs ........................................ 78
Photo 4-5 View of tunnelling works ......................................................................................................... 80
Photo 4-6 View of the tunnel lining.......................................................................................................... 80
Photo 4-7 View of the tunnel lining.......................................................................................................... 81
Photo 4-8 The effects of failure ................................................................................................................ 84
Photo 4-9 La Marea basin ......................................................................................................................... 85
Photo 4-10 Weir undermined by seepage ................................................................................................. 87
Photo 4-11 Weir undermined by seepage ................................................................................................. 87
Photo 4-12 Channel destroyed by uplift pressure ..................................................................................... 88
71
Guide on How to Develop a Small Hydropower Plant ESHA 2004
4. SITE EVALUATION METHODOLOGIES i
4.1. Introduction
Adequate head and flow are necessary requirements for hydro generation. Consequently these
parameters are important factors in site selection.
Chapter 3 outlines the available methodologies for evaluating the flow that is available for power
production.
In this chapter, the methodologies needed, in order to evaluate the suitability of a site for hydropower
development, are presented.
The gross head may be rapidly estimated, either by field surveying or by using a GPS (Global
Positioning System) or by orthophotographic techniques. With the aid of the engineering hydraulic
principles outlined in Chapter 2 the net head can be determined.
Nevertheless, the selection of the most appropriate technical solution for the site will be the result of a
lengthy, iterative process, where the topography and the environmental issues for a particular site, are
most important. That is why a thorough knowledge of the principles is needed to avoid dangerous
failures in the operation of the plant.
Surveying technologies are undergoing a revolutionary change, and the use of the technologies
mentioned above may greatly assist in scheme design and reduce its cost.
4.2. Cartography
In industrialised countries, scaled maps are usually available. The E.U. territory has been or is being
digitised, and cartography at scale as large as 1:5 000 is already available. On the other hand, in the
developing countries, the developer will be fortunate if he can find maps at 1:25 000.
Aerial photographs of topography can be substituted for maps if they cannot be found at the required
scale. However aerial photographs are unlike maps in one important respect. A map has a uniform or
controlled variable scale, the latter being dependent on the choice of map projection. The aerial
photograph, on the other hand, does not have a constant or uniformly changing scale. Aside from lens
imperfections, which for all practical purposes can be considered negligible, two major factors are
responsible for variations in the scale of a photograph:-
1. The topographical relief - land, no matter how flat, is never horizontal – and…
2. The tilt of the optical axis of the camera.
Modern cameras are able to remove distortion, resulting from their axial tilt. Furthermore aerial
photographs can be viewed stereoscopically or in three dimensions. The stereoscopic effect enables the
geologist to identify rock types, determine geologic structures, and detect slope instability and the
engineer is able to gather data necessary for a dam, open channels and penstock construction.
Depending on the required accuracy, the digitised photographs can be geocoded (tied to a co-ordinate
system and map projection) and orthorectified. Distortion from the camera lens is removed by using
ground control points from maps, survey data or clients GPS vectors. This is a very cost-effective way
to orthorectify aerial photographs. Resolutions of 30 cm to one metre can be expected with digital
orthophotos. Both hard copy and digital orthophotos in diskettes, or CDROM can be produced.
72
Guide on How to Develop a Small Hydropower Plant ESHA 2004
With these maps it is possible to locate the intake, trace the open channel and penstock and locate the
powerhouse, with precision enough for the feasibility studies and even for the contractors to engage in
the bidding phase for construction.
With stereoscopic photographs geologic problems can often be spotted, especially those concerning
slope stability that can cause dangerous situations.
4.3. Geochemical Studies
Very often, the need to proceed with detailed geological studies of a site, are underestimated. In many
cases with regrettable consequences - seepage under the weir, open channel slides etc.
Fortunately in the E.U. member states and in many other countries all over the world, good geological
maps permit initial estimates, for the security of the dam foundations, the stability of the slopes and the
permeability of the terrain. However sometimes this information, should be complemented, with
fieldwork particularly, drilling and sampling.
Hydraulic structures should be founded on level foundations, with adequate side slopes and widths, not
subject to stability problems. There are a good number of slope stability computer programs, ranging
from a simple two-dimensional approach to the sophisticated three-dimensional full colour graphic
analysis. The catalogue of failures, especially in channel design is so large that a minimum
geomorphologic study of the terrain should be recommended in the first phase of the project. The
problem is especially acute in high mountain schemes, where the construction may be in a weathered
surface zone, affected by different geomorphologic features such as soil creep, solifluction, rotational
and planar soil slides and rock falls.
The weir and its corresponding reservoir can be affected by the instability of the superficial formations
that can be present within its zone of influence, but at the same time the pond itself can affect these
same formations. If the weir has to be founded on unconsolidated ground the variation of water level
can generate instability on the reservoir’s wetted slopes.
Along the open channel many geomorphologic features can adversely affect its selected line, which,
together with a steep slope of the terrain, may lead to potential instability. Colluvial formations, a
product of the surface mechanical weathering of the rock masses, and solifluction processes, are very
active in high mountain environments where the subsoil is seasonally or perennially wet – these are
some of the features that can compromise channel stability.
Drainage treatments, bench constructions and gunnite treatments, among many others, may be
recommended. At the end of the canal, the forebay acts as a mini-reservoir for the penstock.
Frequently, authorities require that all the water retaining embankment sections undergo stability
analysis regardless of their configuration. The layout of the penstock, usually placed on a steep slope,
poses problems both for its anchoring blocks and visual impact.
Deep in the valley, frequently built on an old river terrace, the powerhouse foundation poses problems
that many times only can be solved by using techniques as up today as the jet grouting (see 4.2.2.4).
4.3.1. Methodologies to be used
In geological science, there is a wide spectrum of geomorphologic techniques that can be used
including the most common ones:-
73
Guide on How to Develop a Small Hydropower Plant ESHA 2004
• Photogeology. As mentioned above photogrammetry - at scales from 1:10 000 to 1:5 000 –
allows the geologist to identify rock types, determine geologic structures, and detect slope
instability.
• Geomorphologic maps. The result of photogrammetric analysis complemented with the
results of the field survey must be combined on a Geomorphologic Map. This map is based on
a topographic one and is drawn at a scale between 1:10 000 and 1:5 000, duly classified using
simple symbols, should show all the surface formations affecting the proposed hydraulic
structures.
• Laboratory analysis. Traditional laboratory tests such as soil grading and classification, and
triaxial consolidation facilitate the surface formation classification. The results should be
included in the geomorphic map.
• Geophysical studies. A geophysical investigation either electrical or seismic (by refraction)
will contribute to a better knowledge of the superficial formation thickness, the location of the
landslide sections, the internal water circulation, and the volumetric importance of potentially
unstable formations.
• Structural geological analysis. Although not a proper geomorphologic technology can help to
solve problems in the catchment area and in those cases where hydraulic conduits must be
tunnels in rock massifs. The stability of the rock and seepage in the foundation of hydraulic
structures are problems that can be solved by this methodology, avoiding dramatic incidents
during the operation.
• Direct investigations - Borehole drilling. This is uncommon for small hydro scheme
development. However when the dam or weir has to be founded in unconsolidated strata, a
drilling programme, followed by laboratory tests on the samples extracted is essential. Some of
these recommended tests are:-
1. Permeability tests in boreholes, such as Lugeon or Low Pressure Test, to define the
water circulation in the foundation.
2. Laboratory tests to determine the compressive strength of the samples to define their
characteristics.
Complementing the above tests a geophysical refraction seismic essay to define the modulus of
dynamic deformation of the rock massif in depth can be recommended in the case of high dams.
Figure 4-1 Schematic representation of the site
74
Guide on How to Develop a Small Hydropower Plant ESHA 2004
4.3.2. Methodologies. The study of a practical case.
A short report on the geomorphologic techniques used in the Cordiñanes scheme, a high mountain
scheme located in the Central Massif of Picos de Europa (Leon, Spain) will help to demonstrate the
scope of the above-mentioned studies. Figure 4.1 is a schematic representation of the site, which
includes:-
• A gravity weir 11.5 meters high over foundations
• A reservoir with a storage capacity of 60 000 m3
• An open channel 2475 m long (776 m are in tunnel)
• A forebay at the end of the tunnel
• A 1.4 m diameter penstock, 650 m long with a 190 m drop
• A powerhouse
4.3.2.1.The weir
International regulations require that if there is a potential for direct shear failure or whenever sliding
is possible along joints or faults, rock foundations must be analysed for stability. When necessary,
additional rock excavation may be required.
Figure 4.2 shows the weir location and illustrates the entirely different structures of both slopes, the
left one, steeper, follows the nearly vertically bedded slate formation; the right one less steep is
associated to a colluvial formation.
Figure 4-2 Weir location and structures of both slopes
Figure 4.3 shows the geological complexity of the colluvial formation. The borehole drilling B-1
illustrates the existence of an alluvial terrace under the colluvial formation. Each formation behaves in
a different way to the requirements of the weir foundation.
75
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 4-3 Geological section of the colluvial formation
Figure 4-4 Geomorphologic scheme of the channel trace
4.3.2.2.The open channel
Figure 4.4 shows a geomorphologic scheme of the channel trace. Two large independent unstable
zones (b and c) can be seen in the right side of the river. Photographs 4.1 and 4.2 show a general view
of the right-side slope and the local instabilities generated during the excavation works, just as a detail
of one of these instabilities. Photograph 4.3 shows one of the existing sliding scarps before the
beginning of the works.
76
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 4-1 General view of the right-side slope
Photo 4-2 Local instabilities generated during the excavation works
The foundation of the channel should meet two requirements:-
• Must be stable. Channels are rigid structures and do not permit deformations.
• Should be permeable. Channels do not support thrusts or uplift pressures.
The geologic studies should aim to avoid settlements in the channel and to provide adequate drainage
to hinder the thrust and uplift stresses. The study should conclude with a recommendation to guarantee
the stability and suppress the uplift pressures
77
Guide on How to Develop a Small Hydropower Plant ESHA 2004
.
Photo 4-3 One of the existing sliding scarps before the beginning of the works
Photo 4-4 A view of the Cordiñanes colluvium, under which the tunnel runs
4.3.2.3.The channel in tunnel
The tunnel construction must comply with the following requirements:-
The excavation will be conditioned by the geologic formations that must traverse, either a rock
massif or a superficial formation.
The tunnel, being a hydraulic channel should be stable and watertight. Consequently the geologic
formations that exist in the massif to be traversed must be known in detail.
Photograph 4.4 shows a view of the Cordiñanes colluvium, under which the tunnel runs. Figure 4.5
shows a schematic cut of the tunnel under the colluvium and figure 4.6 illustrates the concrete lining
forming the final section of the canal.
78
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Figure 4-5 A schematic cut of the tunnel under the colluvium
Figure 4-6 Concrete lining forming the final section of the canal
The excavation works were extremely difficult due to the large variety and heterogeneity of the blocks,
which varied in size from simple stones to blocks of several cubic meters. The use of large explosive
charges was not permitted here. The use of tunnelling machines was not feasible. The excavation had
to proceed metre by metre, using small explosive charges to reduce the size of the blocks, which could
not be handled (Photograph 4.5).
The concrete lining was also difficult. Zone 2 in figure 4.6 was filled by injecting grout. In fact this
injection not only filled the empty space but also enclosed the supporting structure and reinforced the
loose terrain around the tunnel. This terrain is very permeable so to avoid lateral pressures or uplift
pressures a draining system was put in place.
The construction of tunnels through rocky massifs should take into account two important geologic
characteristics:-
79
Guide on How to Develop a Small Hydropower Plant ESHA 2004
• The lithologic variation, along its trace can decisively influence the construction method to be
used.
• The structural stability, of the massif along the trace. Even if the massif is lithologically
coherent the distribution of the potential discontinuities in stratification planes, joints, fissures -
will be far from homogeneous. Once again the knowledge of all those discontinuities must be
based on a detailed structural geological study.
As well as the relatively small discontinuities referred above, the designer should also deal with the
large tectonic discontinuities -large bending, faults, invert faults- that not only affect the work itself but
also the future operation of the canal.
Figure 4.7 shows a thrust fault, present in the La Rienda tunnel, second part of the tunnel of
Cordiñanes close to the forebay built right at the end of the tunnel. Due to the strains and deformations
supported in the past by this mass of rocks, the rocks originally found were completely altered. The
response to this excavation was of course very different from the response of the rest of the massif.
Only by knowing the presence of this fault beforehand could the tunnel be excavated without
unexpected incidents. As photographs 4.6 and 4.7 illustrate, the supporting structure during the tunnel
construction was very different in this area to the one used in the rest of the work.
Photo 4-5 View of tunnelling works
Photo 4-6 View of the tunnel lining
80
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 4-7 View of the tunnel lining
Figure 4-7 A thrust fault, present in the La Rienda tunnel
81
Guide on How to Develop a Small Hydropower Plant ESHA 2004
4.3.2.4.The powerhouse
If the powerhouse is founded on rock, the excavation work will remove the superficial weathered
layer, leaving a sound rock foundation. If the powerhouse is to be located on fluvial terraces near the
riverbanks that do not offer a good foundation then the ground must be reinforced.
The traditional cement grouting presents some difficulties and in many cases its results are not
satisfactory when the terrain is as heterogeneous and permeable as exists in fluvial terraces. A new
injection technique, jet grouting, can guarantee the terrain consolidation, replacing alluvial sediments
by an injected curtain. The technique, widely used by the DOE (Department of Energy of the U.S) to
cut the seepage in the underground storage reservoir for toxic wastes, is however very expensive at
present. Figure 4.8 illustrates the results of the jet-grouting operation that was performed to reinforce
the terrain supporting the powerhouse.
Figure 4-8 results of the jet-grouting operation
4.4. Learning from failures
Two well-known experts, Bryan Leyland of Australia and Freddy Isambert from France, presented at
the HIDROENERGIA95 Conference two independent papers dealing with the topic “Lessons from
failures”.
Mr Leyland quoting Sir Winston Churchill (the famous UK Prime Minister) – “he who ignores
history is doomed to repeat it” - claims that if one does not want to repeat the mistakes of others, the
reasons for their failures must studied and understood. According to Mr Isambert “case studies have
82
Guide on How to Develop a Small Hydropower Plant ESHA 2004
shown that a number of small hydro plants have failed because they were poorly designed, built or
operated”.
The authors presented, with the aid of graphics and photographs, several examples of schemes that
failed in the commissioning of the plant or later in the operation, and produces considerable loss of
money and dramatic delays.
Professor Mosony wrote in ESHA Info no. 15, “a fair and open discussion about failures is
indispensable in order to learn from failures and consequently to avoid their repetition”. Quoting
Marcus Tullius Ciceron (106-43 BC) “Every human being can make a mistake, but only the idiot
persists in repeating his mistake”. From the accounts of failures reported at HIDROENERGIA,
together with more than 50 others described in the ASCE publication “Lessons Learned from the
Design, Construction and Operation of Hydroelectric Facilities”, of which 28 of them concern schemes
of less than 10 MW capacity, examples have been selected for discussion below. They demonstrate the
importance of studying in depth, the stability of canals and the effects of uplift pressure on hydraulic
structures.
Figure 4-9 Scheme of Ruahihi canal
Ruahihi canal failure (New Zealand)
As shown in figure 4.9 the scheme had a 2000m canal laid along a side slope, leading to 750 m of
concrete and steel penstocks. The canal was excavated in soft ignimbrite (debris from a volcanic
explosion) and lined with a type of volcanic clay.
83
Guide on How to Develop a Small Hydropower Plant ESHA 2004
The brown ash dried and cracked during construction but due to its unusual characteristics, the cracks
did not seal when the canal was filled. So water leaked into the ignimbrite below. When these leaks
appeared perforated pipes were driven in to drain the bottom of the slope. This hid the problem and
also made it worse because the leaking water caused caverns to form in the fill.
Photo 4-8 The effects of failure
On the day after the scheme was officially opened, a large section of the canal suddenly collapsed.
Photograph 4.8 illustrates the magnitude of the catastrophe. Many options were examined and finally it
was decided that the only viable option was to replace the failed section of canal with 1100m of pipes.
This increased the length of the penstocks from 750 m to 1850 m and required that water hammer
pressures had to be reduced, because the original concrete pipes could only withstand a very limited
overpressure. It was necessary to modify the relief valves and the inlet valves so that there would only
be a 3% pressure rise under the worst conditions. A surge chamber was not an option because the
ground could not take the extra weight. Fortunately the turbine manufacturer was very cooperative and
had faith in the ability of his relief valves to limit the pressure rise to 3%, which they did. The
refurbishment was completed ahead of time and under budget.
The lessons learned were:
• The characteristics of volcanic materials are highly variable and often undesirable;
• When a canal leaks, be sure the problem is fully understood before repairs commence;
• When the alternative is to abandon a failed scheme, consider the seemingly impossible there
may not be a lot to lose!
84
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 4-9 La Marea basin
La Marea canal failure (Spain)
The La Marea scheme has a spiral Francis turbine of 1 100kW installed capacity a discharge of 1.3m3/s
and a 100m head. As shown in figure 4.11 the scheme includes a small weir for the water intake,
provided with a fish ladder. From the intake a rectangular canal built in reinforced concrete (3 x 2m
section) is followed by another 600m long canal in tunnel. At the outlet of the tunnel a reservoir was
built to store water for peak operation. The reservoir was built by compressing a mix of sand and clay,
and unfortunately proved to be insufficiently watertight. From the reservoir another canal, built with
prefabricated sections of concrete with thin steel plates between, brings the water to the forebay,
located 100m above the powerhouse.
The canal lies on a steep slope on strongly weathered sandstone. Heavy rain was pouring over the
canal both during its construction and during its commissioning. Immediately after opening the intake
gate, the reservoir was filled and the water began to seep into the terrain. The wetted sandstone could
not resist the shear stresses and a landslide broke the reservoir embankment (photograph 4.9), and
large masses of material reached the river, and through the river, to the coast.
The reservoir was replaced by a construction in reinforced concrete, which up to the present day has
served no useful purpose. Later on, the second section of the canal and the prefabricated reach, started
to leak. The terrain became saturated and, unable to resist the shear stresses, failed in a rotational slide.
About 200m of canal were replaced by a low-pressure welded steel pipe that up to now has been
performing adequately. The pipe runs under a storage pond, waterproofed by a thermo-welded plastic
sheet, and ends in the forebay.
The lessons learned were:-
• Weathered sandstone gives poor performance when resisting landslide, especially on slopes
with an angle over 35º to the horizontal.
85
Guide on How to Develop a Small Hydropower Plant ESHA 2004
• Hydraulic canals should be built to guarantee their watertightness; alternatively a drainage
system should be devised so the leakage does affect the terrain.
• The replacement of an open canal by a low pressure pipe on a steep slope may be a better
option, because it will be watertight and it will require only a few anchorage points
Figure 4-10 Longitudinal scheme of La Marea plant
Figure 4-11 Plan view f La Marea plant
Seepage under a weir (France)
This case concerns a small weir, which is the furthest structure upstream of a 600kW project
comprising of a buried culvert, a penstock and a powerhouse. The operating personnel had noticed
minor leakage at the downstream toe of the dam. The small reservoir was emptied, and a trench was
excavated so that the contact between the structure and the foundation could be examined. It was then
revealed that a conduit had formed between the upstream and the downstream faces of the weir (photo
4.11), which was actually founded on permeable deposits without a cut-off trench.
86
Guide on How to Develop a Small Hydropower Plant ESHA 2004
The weir in this condition would have eventually failed by undermining the foundation.
The key issues to learn from this case were the lack of a geomorphologic survey and inadequate
supervision of the design and construction of the weir.
Photo 4-10 Weir undermined by seepage
Photo 4-11 Weir undermined by seepage
The hydraulic canal in a low-head 2MW scheme
The hydraulic canal, 5m wide and 500m long ran along side a river. The river was known to
experience frequent flash floods. On one particular day, a flood occurred which was later calculated to
be a 100 year event. When the flood occurred, the turbines were stopped and all the gates closed. The
headrace channel had been almost emptied by leakage, and the channel was destroyed by uplift
pressure (photo 4.12). In this case the key technical issues were hydraulics, structural stability and
design.
87
Guide on How to Develop a Small Hydropower Plant ESHA 2004
Photo 4-12 Channel destroyed by uplift pressure
There are other cases that could be described to show the effects of poor judgement during either the
design or the construction phase. Such case studies show the number and diversity of parameters that
can cause failures. It is also unfortunately evident that design, construction and site supervision are
often carried out by companies, which may offer lower costs, but have little experience of hydraulic
works.
i By Luigi Papetti (Studio Frosio), Jonas Rundqvist (SERO) and Celso Penche (ESHA)
88
Guide on How to Develop a Small Hydro Site ESHA 2004
CHAPTER 5: HYDRAULIC STRUCTURES
CONTENTS
5 HYDRAULIC STRUCTURES .................................................................................................93
5.1 Introduction ........................................................................................................................93
5.2 Dams ..................................................................................................................................93
5.2.1 Embankment Dams ....................................................................................................95
5.2.2 Concrete Dams...........................................................................................................95
5.2.3 Other Dam types ........................................................................................................97
5.2.4 Loads and stability for concrete dams........................................................................98
5.2.5 Dam Safety.................................................................................................................99
5.3 Weirs and spillways .........................................................................................................100
5.3.1 Weirs ........................................................................................................................101
5.3.2 Gated Spillways .......................................................................................................103
5.3.3 Other spillways ........................................................................................................104
5.4 Energy dissipating structures ...........................................................................................109
5.5 Intake structures ...............................................................................................................109
5.5.1 General .....................................................................................................................109
5.5.2 Intake types ..............................................................................................................110
5.5.3 Head losses...............................................................................................................114
5.5.4 Trashracks ................................................................................................................115
5.5.5 Vorticity ...................................................................................................................119
5.6 Sediment traps ..................................................................................................................120
5.6.1 General .....................................................................................................................120
5.6.2 Efficiency of a sediment trap ...................................................................................121
5.6.3 Design ......................................................................................................................121
5.7 Gates and valves...............................................................................................................122
5.8 Open channels ..................................................................................................................126
5.8.1 Design and dimensioning .........................................................................................126
5.8.2 Excavation and stability ...........................................................................................130
5.9 Penstocks..........................................................................................................................134
5.10 Tailraces ...........................................................................................................................149
Bibliography.....................................................................................................................................150
LIST OF FIGURES
Figure 5.1: A zoned embankment dam with moraine core ................................................................95
Figure 5.2: Typical geometry for arch and cupola dams (single curvature arch dam to the left)......97
Figure 5.3: Masonry dam with vertical concrete upstream wall ........................................................98
Figure 5.4: Typical timber dams ........................................................................................................98
Figure 5.5: Loads on concrete dams ..................................................................................................99
Figure 5.6: Fixed and mobile spillway structures ............................................................................101
Figure 5.7: Discharge characteristics for weirs................................................................................102
Figure 5.8: Weir configurations .......................................................................................................103
Figure 5.9: Discharge characteristics for gated spillways................................................................104
Figure 5.10: Flashboards, articulated & embedded .........................................................................105
Figure 5.11: Inflatable weir..............................................................................................................106
91
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.12: Schematic layout of siphon spillway ...........................................................................108
Figure 5.13: Schematic view of morning glory (shaft) spillway .....................................................108
Figure 5.14: Labyrinth weir .............................................................................................................109
Figure 5.15: Secondary currents in river bends ...............................................................................111
Figure 5.16: Typical layout of lateral intake ....................................................................................112
Figure 5.17: Secondary current along the outer bend of a curved river...........................................113
Figure 5.18: “Tyrolean” intake ........................................................................................................114
Figure 5.19: French drop intake: a canal built in the streambed and covered by a trashrack ..........115
Figure 5.20: Trash boom layout .......................................................................................................116
Figure 5.21: Formula for computing head losses.............................................................................117
Figure 5.22: Oleo-Hydraulic cylinders............................................................................................118
Figure 5.23: Minimum degree of submergence ...............................................................................119
Figure 5.24: Sediment traps .............................................................................................................121
Figure 5.25: Wedge-shaped stopper.................................................................................................123
Figure 5.26: Butterfly valves ...........................................................................................................124
Figure 5.27: Globe and rotary valves ...............................................................................................124
Figure 5.28: Channel design ............................................................................................................131
Figure 5.29: Rectangular reinforced canal .......................................................................................131
Figure 5.30: Materials used for protection .......................................................................................132
Figure 5.31: Penstock.......................................................................................................................134
Figure 5.32: Penstock with concrete anchor blocks and expansion joints .......................................136
Figure 5.33: Manufactured steel pipe ..............................................................................................137
Figure 5.34: Energy loss ..................................................................................................................139
Figure 5.35: Friction and turbulence head losses.............................................................................141
Figure 5.36: Surge tower..................................................................................................................147
Figure 5.37: Surge height versus time .............................................................................................147
LIST OF TABLES
Table 5.1: Intake characteristics.......................................................................................................111
Table 5.2: Hydraulic parameters for common canal cross-sections ................................................127
Table 5.3: Optimum profile for different channel sections..............................................................128
Table 5.4: Different material’s characteristics .................................................................................137
LIST OF PHOTOS
Photo 5.1: Examples of gravity (RCC) and buttress dams ................................................................96
Photo 5.2: Example of an arch dam ...................................................................................................97
Photo 5.3: Failure of a small dam, the breach and the flooding downstream ..................................100
Photo 5. 4. Ogee weir .......................................................................................................................103
Photo 5.5: Articulated flashboard ....................................................................................................105
Photo 5.6: Flashboard controlled by inflatable rubber bladder ........................................................106
Photo 5.7: Hydroplus fusegates .......................................................................................................107
Photo 5.8: Drop intake .....................................................................................................................113
Photo 5.9: Prefabricated booms .......................................................................................................116
Photo 5.10: Telescopic hydraulic cylinders .....................................................................................118
Photo 5.11: wheel-and-axle mechanism ..........................................................................................123
Photo 5.12: Hydraulic Cylinder .......................................................................................................124
Photo 5.13: Large butterfly valve ....................................................................................................125
92
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.14: Butterfly valve hydraulically operated .........................................................................125
Photo 5.15: Tainter gate (left) and housing of its sector on a concrete pier ....................................126
Photo 5.15: Canal in the Cordinañes................................................................................................132
Photo 5.16: Lateral spillway ............................................................................................................133
Photo 5.17:Uplift..............................................................................................................................133
Photo 5.18: Flume ............................................................................................................................133
Photo 5.19: Water jet .......................................................................................................................148
93
Guide on How to Develop a Small Hydro Site ESHA 2004
5 HYDRAULIC STRUCTURES 1
5.1 Introduction
A hydropower development includes a number of structures, the design of which will be dependant
upon the type of scheme, local conditions, access to construction material and also local building
traditions in the country or region. The following structures are common in a hydro scheme:
• Diversion structure
o Dam
o Spillway
o Energy dissipation arrangement
o Fish pass
o Residual flow arrangements
• Water conveyance system
o Intake
o Canals
o Tunnels
o Penstocks
o Power house
Design aspects and common solutions for these structures are presented below.
5.2 Dams
Dams and weirs are primarily intended to divert the river flow into the water conveyance system
leading to the powerhouse. Dams also produce additional head and provide storage capacity. The
choice of dam type depends largely on local topographical and geotechnical conditions. For
instance if sound rock is not available within reasonable excavation depth, rigid structures such, as
concrete dams are difficult. Conversely, for narrow valleys, it can be difficult to find space for
separate spillways, and concrete dams can be the natural choice with their inherent possibilities to
integrate spillways etc in the dam body.
In the Nordic countries the ice age has left us with wide and open valleys and moraine material in
abundance. Not surprisingly the vast majority of dams are embankment dams with a central core of
moraine. South of the Alps natural clays suitable for dam core are not in abundance and the
topography in many locations favour concrete dams.
According to the ICOLD (International Committee of Large Dams), a dam is considered "small"
when its height, measured from its foundation level to the crest, does not exceed 15 m, the crest
length is less than 500 m and the stored water is less than 1 million cubic meters. These parameters
93
Guide on How to Develop a Small Hydro Site ESHA 2004
can be important, because of the complicated administrative procedures often associated with the
construction of large dams.
World wide, embankment dams are the more common partly due to the following characteristics,
which they possess:
• Can be adapted to a wide range of foundation conditions.
• Construction uses natural materials, which can often be found locally, limiting needs for
long transportation.
• The construction process can be continuous and highly mechanized.
• The design is extremely flexible in accommodating different fill materials.
Disadvantages with embankment dams are that they are sensitive to overtopping and leakage, and
erosion in the dam body and its foundation. There is a higher mortality rate among embankment
dams as compared to concrete dams.
Concrete dams on the other hand have drawbacks that correspond to the pros of the embankment
dams:
• Require certain conditions with respect to the foundations.
• Require processing of natural materials for aggregate at the site, hauling of large
quantities of cement and has a labour intensive and discontinuous construction process,
leading to large unit costs.
On the other hand concrete dams have several advantages:
• They are suitable for most ranges of topography that is for wide and narrow valleys,
provided that foundation conditions are right.
• They are not very sensitive to overtopping.
• A spillway can be placed at the crest, and if required over the entire length of the dam.
• Chambers or galleries for drainage, tubing and ancillary works can readily be housed
within the dam body.
• Powerhouses can be placed right at the toe of the dam.
The development of the Concrete Faced Rockfill Dam (CFRD) neutralizes many of the drawbacks
with core-type embankments. In particular, sensitivity to leakage and erosion is reduced, and
dependence of good core material is removed.
The development of the Roller Compacted Concrete Dams (RCC-dams) introduces a continuous,
highly mechanised construction process and low unit costs.
New large dams are almost always CFRD and RCC designs.
94
Guide on How to Develop a Small Hydro Site ESHA 2004
5.2.1 Embankment Dams
Homogeneous dams: These dams are used for low embankments (<4m) and often as secondary
dams. For dam safety reasons, some type of drainage is almost always provided.
Zoned embankment dams: These are used for dam heights from 4m and up. Constructions are
extremely sensitive to the engineering design and construction, and it is therefore vital to engage
highly skilled consultants and contractors require experienced site-supervision engineers. Critical
components of these dams are the core, the transition zones (filters) surrounding the core and
drainage capacity of the dam toe (see figure 5.1).
Embankments dams with membrane: The membranes can be of different types and be located
either at the upstream front of the embankment or vertically in the centre of the embankment.
Membranes can be made from concrete (as in the CFRD), asphalt (Norwegian type) or in the form
of a geomembrane on the upstream slope.
Figure 5.1: A zoned embankment dam with moraine core
Embankment dams are often categorised according to the main fill material, for example, rock-fill
dams, or earth-fill dams.
5.2.2 Concrete Dams
Generally, concrete dams are categorized according to how they function statically, and fall into one
of the following groups.
Gravity dams: These are dependent on their own mass for stability. Their cross-section is basically
triangular in order to provide adequate stability and stress distribution across the foundation plane.
The upper part is normally rectangular in order to provide adequate crest width for installation and
transportation.
Design issues include stability analysis (sliding and overturning), stress control, temperature control
during construction to avoid cracking, control of uplift pressures under the dam, etc. In photo 5.1 a
gravity dam constructed of RCC (left photo) is shown. Note the characteristic stepped downstream
slope.
95
Guide on How to Develop a Small Hydro Site ESHA 2004
Buttress dams: These dams consist of a continuous upstream face that is supported by buttresses at
regular intervals. The upstream face is normally divided into vertical sections by dilatation joints,
each section being supported by a buttress. Cross-sections are similar to those of gravitation dams.
In colder climates, the upstream face can be susceptible to freezing of the water contained in the
concrete, damaging the concrete. For this reason buttress dams in such locations are often covered
along the downstream contour of the buttresses in order to provide climate control. The right-hand
photo in photo 5.1 shows an example of a buttress dam. Note that the spillway is also a buttress type
structure.
Photo 5.1: Examples of gravity (RCC) and buttress dams
Arch and Cupola dams: These dams function structurally as horizontally laid out arches that
transfer the water pressure on the upstream face into the abutments rather than into the foundation.
Arch dams can be designed with a constant radius over the dam height, or with varying radii
(Cupola dams). Arch dams with a constant radius have a vertical and “straight” cross-section. These
dams will be subject to considerable vertical strain forces since the deformation of the dam will tend
to be greatest in the vertical centre of the dam. This requires that the dam be heavily reinforced to
avoid cracking with accompanying leakage.
The Cupola dam is designed to have only compression forces for all directions and at all sections.
This requires the radius of the curvature to vary over the dam height, which produces a curved
vertical cross-section.
The arch and cupola dams are structurally efficient and greatly reduce the required volume of
concrete. They require, however, a narrow valley topography and strong foundation rock in the
abutments. In photo 5.2 an example of an arch dam is shown, and in figure 5.2 the typical geometry
for single curvature arch dams versus double curvature cupola dams is displayed.
96
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.2: Example of an arch dam
Figure 5.2: Typical geometry for arch and cupola dams (single curvature arch dam to the left)
5.2.3 Other Dam types
Another type of concrete dam is the spillway dam, which can be gated or ungated. A gated dam
with large spillway openings compared to the dam height is often designed to function as a buttress
dam, whereas higher spillway dams with relatively small spillway openings normally are designed
to function as a gravity dam.
An ungated spillway dam is often referred to as a weir for lower dam heights. Weirs and spillways
are described in more detail below.
An old dam type still prevailing is the masonry dam. This dam was prevalent during the early days
of industrialization, utilizing the building techniques present at that time. The masonry structure
functioned as the load bearing structure and water tightness was provided by either vertical timber
sheeting on the upstream face or by filling impervious soils upstream of the masonry structure.
Figure 5.3 shows an example of a masonry dam, with an upstream wall. In many ways these dams
97
Guide on How to Develop a Small Hydro Site ESHA 2004
resemble the CFRD, a development in embankment dams, and they share a number of advantageous
characteristics.
Figure 5.3: Masonry dam with vertical concrete upstream wall
Timber dams: These dams can still be found although due to their limited durability they are
becoming increasingly scarce. These dams were constructed in two ways, as is shown in figure 5.4.
Figure 5.4: Typical timber dams
5.2.4 Loads and stability for concrete dams
In figure 5.5, the typical loads acting on concrete dams are shown. H denotes horizontal loads and V
vertical loads. The horizontal loads are: 1; Lateral water pressure, 2; pressure from soil or deposited
sediments, 3; Ice pressure, 4; Loads from floating objects and debris, 5; downstream water pressure,
6; dynamic acceleration from earthquakes, 7; incremental water pressure during earthquakes. The
vertical loads are: 1; self-weight of the dam, 2; weight of water on inclined upstream surface,
3;uplift pressure from pore water, 4; dynamic load from earthquakes. There is also a small vertical
load corresponding to the weight of water on the inclined downstream slope.
The understanding of uplift pressures and their importance for gravity dams has gradually
increased. The very existence of uplift pressures was not known until the beginning of the 20th
century. For the first gravity dams, made as masonry dams, uplift pressures were basically
98
Guide on How to Develop a Small Hydro Site ESHA 2004
eliminated due to the effective drainage provided by the porous structure of the masonry. As
masonry was replaced by concrete in new dams these dams were designed applying the same well-
proven dimensions used for masonry dams, which in many cases led to failure of the dams.
Modern concrete dams provide drainage in the form of drainage galleries, by drilling drainage holes
into the foundation rock. Using grouting curtains reduces foundation leakage. These measures can
be effective, but require maintenance. Concrete dams built as late as in the 1980’s regularly show
weaknesses due to the very optimistic assumptions regarding uplift pressures and the
ineffectiveness of individual counter measures.
Figure 5.5: Loads on concrete dams
Concrete dams are designed for:
• Stability against rotation and overturning
• Stability against translation and sliding
• Over-stress and material failure
5.2.5 Dam Safety
Dams have been identified as “the single man-made structures capable of causing most deaths”.
Hazards with dam failure have largely been associated with large dams and reservoirs, but
depending on localization and circumstances even smaller and medium sized dams and reservoirs
can be potentially dangerous, and considering their large number they do pose a significant threat to
health and environment. In Sweden, for example, the only fatality as a result of dam failure was
caused by the failure of a dam less than 4m high. Photo 5.3 shows two photos of the failure of a
“small” dam. The left photo shows the breach and the right photo show the damage downstream.
In order to identify potentially hazardous dams most countries now employ a classification system
for dams, requiring dam owners to classify their dams. The hazard level is described and identified
subjectively using terms such as low, significant and high (USACE 1975).
Dam safety can be improved by installation of monitoring systems, performing reviews and
undertaking dam inspections on a regular basis.
99
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.3: Failure of a small dam, the breach and the flooding downstream
5.3 Weirs and spillways
A dam failure can have severe effects downstream of the dam. During the lifetime of a dam
different flow conditions will be experienced and a dam must be able to safely accommodate high
floods that can exceed normal flow conditions in the river by orders of magnitude. For this reason
carefully designed overflow passages are incorporated in dams or weirs as part of the structure.
These passages are known as spillways. Due to the high velocities of the spilling water, some form
of energy dissipation is usually provided at the base of the spillway.
The large majority of small hydro schemes are of the run-of-river type, where electricity is
generated from discharges larger than the minimum required to operate the turbine. In these
schemes a low diversion structure is built on the streambed to divert the required flow whilst the
rest of the water continues to flow over it. Such a structure is commonly known as a weir, whose
role is not to store the water but to increase the level of the water surface so the flow can enter into
the intake.
Weirs and spillways can be subdivided into fixed and mobile structures (Figure 5.6). Smaller fixed
structures are generally referred to as weirs, whereas larger structures are often referred to as
spillways. Spillways are often divided into ungated and gated spillways, corresponding to fixed and
mobile structures, the ungated spillway in fact being a large-scale weir.
Fixed storage structures, such as weirs and ungated spillways have the advantage of security,
simplicity, easy maintenance, and are cost effective. However, they cannot regulate the water level
and thus both the water level and energy production fluctuates as a function of discharge.
Mobile storage structures such as gated spillways can regulate the water level such that it stays
more or less constant for most incoming flow conditions. Depending on gate configuration and
discharge capacity they may also be able to flush accumulated sediment downstream. These
structures are generally more expensive than fixed structures, for both construction and
maintenance, and their functioning is more complicated.
100
Guide on How to Develop a Small Hydro Site ESHA 2004
Bassin amortisseurCoursier
Déversoir Niveau
variable
Alluvionnement
Protection cont
érosion locale
Joint
Ecran d'étanchéité
Marche positive
Parafouilles
Vanne segment
Pilier
Niveau variable
Seuil / barrages fixes
Seuil / barrages mobiles Mobile structure
Fixed structure
water
level
Weir crest
Weir face
Joint
Impermeablescreen
A lluvial
deposition
Energy dissipation basin
Erosion
protection
Step
water level Radial gate
Pile
Bassin amortisseurCoursier
Déversoir Niveau
variable
Alluvionnement
Protection cont
érosion locale
Joint
Ecran d'étanchéité
Marche positive
Parafouilles
Vanne segment
Pilier
Niveau variable
Seuil / barrages fixes
Seuil / barrages mobiles Mobile structure
Fixed structure
water
level
Weir crest
Weir face
Joint
Impermeablescreen
A lluvial
deposition
Energy dissipation basin
Erosion
protection
Step
water level Radial gate
Pile
Figure 5.6: Fixed and mobile spillway structures
5.3.1 Weirs
Weirs can be constructed perpendicular, angular or lateral compared to the river axis. Most often
the weir crest is rectilinear and perpendicular to the river axis. For relatively low downstream water
levels, the weir controls the flow and defines the relationship between the upstream water level and
the discharge. As a function of the type of weir, different discharge relationships are obtained as
indicated in Figure 5.7.
The sharp-crested weir is easy to construct and relatively cost-effective. Its discharge is defined by
means of a coefficient Cd. Special attention has to be paid to the shape of the downstream face of
the upper part of the weir in order to obtain sufficient aeration between the lower nappe (sheet of
water that flows over the weir) of the jet and the structure. If the lower nappe of the jet sticks to the
structure, vibrations may be transferred from the flow to the structure.
The broad-crested weir is often applied for temporary structures or for structures of secondary
importance, such as in case of temporary flow diversion. Its design is simple and cost-effective. The
hydraulic conditions are far from optimal, expressed by a low discharge coefficient and the
presence of under-pressures along the weir crest and downstream face. The discharge depends on
the form of the structure.
101
Guide on How to Develop a Small Hydro Site ESHA 2004
Sharp-crested
weir
Broad-crested
weir
Ogee
weir
Type Design Discharge
relationship
H
w
H
ΔH
w
Characteristics
θ Le
h0H
w
Q = b·Cd·H3/2·√2g
Cd =0.42
Q = b·CdD·H3/2·√2g
CdD =0.494
(for H = HD)
Q = b·ce·Cd·H3/2·√2g
Cd,mean =0.42 Simple design,
underpressures
on crest
Cost effective
Highest discharge
Costly design
Simple design
Cost effective
ce = 1-2sinθ
9(1+ξe )4
ξe=H-w
Le
Sharp-crested
weir
Broad-crested
weir
Ogee
weir
Type Design Discharge
relationship
H
w
H
w
H
ΔH
w
H
ΔH
w
Characteristics
θ Le
h0H
w
θ Le
h0H
w
Q = b·Cd·H3/2·√2g
Cd =0.42
Q = b·CdD·H3/2·√2g
CdD =0.494
(for H = HD)
Q = b·ce·Cd·H3/2·√2g
Cd,mean =0.42 Simple design,
underpressures
on crest
Cost effective
Highest discharge
Costly design
Simple design
Cost effective
ce = 1-2sinθ
9(1+ξe )4ce = 1-2sinθ
9(1+ξe )4
ξe=H-w
Le
Figure 5.7: Discharge characteristics for weirs
The ogee weir is hydraulically the most ideal solution giving the highest discharge coefficient. Its
curved shape is defined by the jet trajectory that would appear for the design discharge HD. For
lower or higher discharges, over-or under-pressures will appear along the downstream face. For
discharges much higher than the design discharge, these under-pressures may lead to cavitation and
damage to the downstream concrete face. Recent work suggests fortunately that separation will not
occur until H > 3HD. The U.S. Waterways Experimental Station has provided a set of profiles that
have been found to agree with actual prototype measurements. The exact relationship between the
discharge coefficient and the ratio H/HD can be found in Sinniger & Hager (1989).
For downstream water levels that are equal to or higher than the spillway crest level, the spillway
becomes progressively submerged and its corresponding discharge decreases. Furthermore, in
presence of piles, the governing discharge will depend on the shape and dimensions of the piles. All
these aspects influence the functioning of a spillway and for a detailed and correct design; the
reader is referred to classical works in this field, such as Sinniger & Hager (1989).
102
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5. 4. Ogee weir
Figure 5.8: Weir configurations
5.3.2 Gated Spillways
The installation of mobile elements on dams or weirs allows control of the flow conditions without
changing the water level. This is performed by means of gates, which are designed such that, when
the gate is fully open (and the structure functions as if it where fixed) the discharge has to pass the
structure without noticeable water level increase upstream. Gate operation needs permanent
maintenance and an external energy source. As a result, there is a risk that the gate remains blocked
during floods.
The most used types of gates are presented in Figure 5.9. Depending on the type of gate, the
possible gate movements are rotating, sliding or turning. The discharge through the gates depends
not only on the type of gate and the relative gate opening and gate lip angle, but also on the shape of
the supporting weir.
103
Guide on How to Develop a Small Hydro Site ESHA 2004
Flat gate
Sector or
radial gate
Type Design Discharge
relationshi
Q = a · b · C d · √ 2 gh 1
Position
rectangula
canal
rectangula
canal
ogeeweir
ogeeweir
H = cteh 1
a Cca
1 2
h2
Ho
H D zHH
e
91
D
e
2 3
D
e
D
0
D
23
0
D
g
H
z
6
1
H
z
H
H
H
H
Q
Q ⎟⎟⎠
⎞⎜⎜⎝
⎛+⋅ ⎥ ⎥
⎦
⎤
⎢⎢
⎣
⎡
⎟ ⎟ ⎠
⎞ ⎜ ⎜ ⎝
⎛ − − =
QD = b · C dD · H 3/2 · √ 2 g
C dD = 0.494
⎥⎥
⎦
⎤
⎢ ⎢
⎣
⎡
⎟⎟⎠
⎞⎜⎜⎝
⎛⋅ + ⋅ ⎥ ⎥
⎦
⎤
⎢⎢
⎣
⎡
⎟ ⎟ ⎠
⎞ ⎜ ⎜ ⎝
⎛ ⋅ −=
85.1
D
l
D
l 2 3
D
l
H
x
2
1
H
z
H
x
g
21G
Q = H D · b · G · C dg · √ 2 gHe
12.0
D
dg H
H
271 90.0C ⎟⎟⎠
⎞⎜ ⎜ ⎝
⎛ ⋅ ⎟ ⎠
⎞ ⎜ ⎝
⎛
°
β − ⋅ =
α − γ = β
⎥ ⎦
⎤ ⎢ ⎣
⎡
⎟ ⎟
⎠
⎞
⎜ ⎜
⎝
⎛ δ − ⋅ − = b1 h 2
a exp C C
2
1 0 d d
⎥ ⎦
⎤ ⎢ ⎣
⎡ ⋅ + ⋅ =
δ⋅−
9
e 5 4 96 . 0 C
76.0
0d
Q = a · b · C d · √ 2 gh 1
⎥ ⎦
⎤ ⎢ ⎣
⎡
⎟ ⎟ ⎠
⎞
⎜ ⎜ ⎝
⎛ δ − ⋅ − = b1 h 2
a exp CC
2
1 0 d d
⎥ ⎦
⎤ ⎢ ⎣
⎡ ⋅ + ⋅ =
δ⋅−
9
e 5 4 98 . 0 C
76.0
0d°=δ 90
h 1
a
Hl
H H D Xl
Zl
α
γ
G
δ
Flat gate
Sector or
radial gate
Type Design Discharge
relationshi
Q = a · b · C d · √ 2 gh 1
Position
rectangula
canal
rectangula
canal
ogeeweir
ogeeweir
H = cteh 1
a Cca
1 2
h2
Ho
H D zHH
e
91
D
e
2 3
D
e
D
0
D
23
0
D
g
H
z
6
1
H
z
H
H
H
H
Q
Q ⎟⎟⎠
⎞⎜⎜⎝
⎛+⋅ ⎥ ⎥
⎦
⎤
⎢⎢
⎣
⎡
⎟ ⎟ ⎠
⎞ ⎜ ⎜ ⎝
⎛ − − =
QD = b · C dD · H 3/2 · √ 2 g
C dD = 0.494
⎥⎥
⎦
⎤
⎢ ⎢
⎣
⎡
⎟⎟⎠
⎞⎜⎜⎝
⎛⋅ + ⋅ ⎥ ⎥
⎦
⎤
⎢⎢
⎣
⎡
⎟ ⎟ ⎠
⎞ ⎜ ⎜ ⎝
⎛ ⋅ −=
85.1
D
l
D
l 2 3
D
l
H
x
2
1
H
z
H
x
g
21G
Q = H D · b · G · C dg · √ 2 gHe
12.0
D
dg H
H
271 90.0C ⎟⎟⎠
⎞⎜ ⎜ ⎝
⎛ ⋅ ⎟ ⎠
⎞ ⎜ ⎝
⎛
°
β − ⋅ =
α − γ = β
⎥ ⎦
⎤ ⎢ ⎣
⎡
⎟ ⎟
⎠
⎞
⎜ ⎜
⎝
⎛ δ − ⋅ − = b1 h 2
a exp C C
2
1 0 d d
⎥ ⎦
⎤ ⎢ ⎣
⎡ ⋅ + ⋅ =
δ⋅−
9
e 5 4 96 . 0 C
76.0
0d
Q = a · b · C d · √ 2 gh 1
⎥ ⎦
⎤ ⎢ ⎣
⎡
⎟ ⎟ ⎠
⎞
⎜ ⎜ ⎝
⎛ δ − ⋅ − = b1 h 2
a exp CC
2
1 0 d d
⎥ ⎦
⎤ ⎢ ⎣
⎡ ⋅ + ⋅ =
δ⋅−
9
e 5 4 98 . 0 C
76.0
0d°=δ 90
h 1
a
Hl
H H D Xl
Zl
α
γ
G
δ
Figure 5.9: Discharge characteristics for gated spillways
More detailed design accounts also for the shape of the gate lip. Furthermore, the above-presented
discharges are only valid for un-submerged flow conditions. Similar to the fixed structures, when
the downstream water level becomes equal to or higher than the crest level, the mobile structure
becomes progressively submerged and the corresponding discharge decreases. For more
information, the reader is encouraged to consult classical textbooks on this subject.
5.3.3 Other spillways
Flashboards
To raise the water level slightly behind the weir to ensure adequate depth of water at the intake,
without endangering the flooding of the upstream terrain, flashboards may be installed on the crest
of the weir (Figure 5.10). The flashboards are commonly made of wood and supported by steel pins
embedded in steel sockets (pipes cut down to size) in the spillway crest. The flashboards have to be
removed by hand during flood flows so that high water levels do not flood the upstream terrain, an
operation that in such circumstances is very difficult. The articulated flashboard is somewhat easier
to remove.
104
Guide on How to Develop a Small Hydro Site ESHA 2004
flashboard with embedded
supports
articulated flashboard with
strut
wood block pipe
wood board
weir crest
strut
Figure 5.10: Flashboards, articulated & embedded
Photo 5.5: Articulated flashboard
Inflatable weirs
Another method, capable of remote control, is the inflatable weir, which employs a reinforced
rubber bladder instead of concrete, steel or wood flashboards. This offers an alternative to more
conventional methods of weir construction, with the inherent advantages of low inital cost, simple
operation and minimal maintenance.
In effect, inflatable weirs are flexible gates in the form of a reinforced, sheet-rubber bladder inflated
by air or water, anchored to a concrete foundation (Figure 5.11) by anchor bolts embedded into the
foundation. Like any other gate, the inflatable weir needs a mechanism by which it is opened and
closed. The weir is raised when filled with water or air under pressure. An air compressor or a water
pump is connected, via a pipe, to the rubber bladder. When the bladder is filled the gate is raised;
when it is deflated the weir lies flat on its foundation, in a fully opened position. The system
becomes economic when the width of the weir is large in relation to the height.
When the management and operational safety of the system is rather critical, the use of inflatable
weirs can give substantial advantages over conventional systems. An electronic sensor monitors the
upstream water level and the inner pressure of the bladder. A microprocessor maintains a constant
level in the intake entrance by making small changes in the inner pressure of the bladder. To avoid
flooding land, a similar device can regulate the inflatable weir regulated to correspond to a pre-set
upstream water level.
Inflatable gate control systems can be designed to fully deflate the bladder automatically in rivers
prone to sudden water flow surges. On a typical weir, two meters high and thirty meters wide, this
can be done in less than thirty minutes. Photo 5.6 illustrates a new type of inflatable weir - patented
by Obermeyer Hydro - where the sheet rubber incorporates a steel panel that behaves as a
flashboard, which is quickly and easily moved in the event of sudden floods. By controlling the
pressure in the rubber blade the steel panels may be more or less inclined, varying the level of the
water surface. The system incorporates an additional advantage: the rubber blade is always
protected against boulders carried during flood flows (buoyancy causes heavy boulders to loose a
portion of their weight in water, making it easier for the flood flow to carry them downstream). A
105
Guide on How to Develop a Small Hydro Site ESHA 2004
synthetic rubber flap anchored to one of the panels closes the free space between panels or between
panel and the buttress.
h
Control shaft
Connection conduitfixationConcrete weir
Min. water pressure .
1.2 to 1.3 ρ*h*g
Δh ≅0.2 to 0.3 h
rubber
bladder
h
Control shaft
Connection conduitfixationConcrete weir
Min. water pressure .
1.2 to 1.3 ρ*h*g
Δh ≅0.2 to 0.3 h
rubber
bladder
Figure 5.11: Inflatable weir
Photo 5.6: Flashboard controlled by inflatable rubber bladder
Fusegates
In large installations, but also sometimes in small ones, it is advisable to place fusegates, such as
those supplied by Hydroplus 2 . In the event of a major flood, when the water reaches a pre-set level,
one or more of the fusegates (basically hinged structures) will tilt to increase the section of the
spillway (Photo 5.7).
106
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.7: Hydroplus fusegates
Siphon spillways
Alternatively where space available for the spillway is limited, a siphon spillway or a shaft spillway
may be used. Both solutions help to keep the upstream water level within narrow limits. A siphon
spillway is basically a curved enclosed duct (Figure 5.12). When the water level rises above the
elbow of the siphon, water begins to flow down the conduit just as in an overflow, but it is when it
rises further that the siphon is primed and increases the discharge considerably. Usually siphons are
primed when the water level reaches or passes the level of the crown, but there are designs where
priming occurs when the upstream level has risen only to about one third of the throat height.
If badly designed, the siphon process can become unstable. At the beginning the siphon discharges
in a gravity mode, but when the siphon is primed the discharge suddenly increases. Consequently
the reservoir level drops, the siphon is de-primed and the discharge is reduced. The level of the
reservoir increases anew until the siphon primes again, and the cycle of events is repeated
indefinitely, causing severe surges and stoppages. Multiple siphons with differential crest heights or
aerated siphons can be the solution to this problem. When the siphon is primed the flow through a
siphon spillway is governed, as in penstocks, by Bernoulli's equation. Assuming that the velocity of
water in the conduit is the same at the inlet and outlet, the head loss may be calculated from the
formulae in Chapter 2, paragraph 2.2.1.
If the pressure at the crown of the siphon drops below the vapour pressure, the water vaporises
forming a large number of small vapour cavities, which entrained in the flow, condense again into
liquid in a zone of higher pressure. This phenomenon is known as cavitation and it can be extremely
damaging. To avoid it, the distance between the crown of the siphon and the maximum level at the
reservoir, depending on height above sea level and prevailing barometric pressure, should normally
not exceed 5 m. Further details on this kind of spillway can be found in the literature3 .
107
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.12: Schematic layout of siphon spillway
Shaft (or Morning glory) spillways
Shaft spillways are rarely used in small scale-hydro. As illustrated in Figure 5.13, a shaft spillway
incorporates a funnel-shaped inlet to increase the length of the crest, a flared transition which
conforms to the shape of the nappe as in the overflow spillway though it is sometimes stepped to
ensure aeration, a vertical shaft and an outlet tunnel that sometimes has a slight positive slope to
ensure that at the end it never flows full. The US Bureau of Reclamation reports (USBR) 6 and 7
describe the design principles for these spillways.
Figure 5.13: Schematic view of morning glory (shaft) spillway
Labyrinth weir
In some small hydropower schemes (e.g. small schemes in an irrigation canal) there is not enough
space to locate a conventional spillway. In these cases, U shaped or labyrinth weirs (Figure 5.14)
should help to obtain a higher discharge in the available length.
108
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.14: Labyrinth weir
5.4 Energy dissipating structures
The discharge from the aforementioned fixed or mobile structures is usually supercritical at the
outlet. The corresponding high flow velocities and turbulence may produce severe erosion at the toe
of the structure, especially if the riverbed is not erosion resistant, such as for example in the case of
silt, clay, loose sand, gravel or even fractured rock.
To avoid such damage, several structural solutions may be applied, some of them being very costly.
The most often used solutions are:
• Stilling basin
• Baffled apron drop
• Plunge pool
• Chute cascades
Most of these structures dissipate the flow energy by the formation of a hydraulic jump, which
dissipates a lot of energy over a relatively short distance. The design and construction of energy
dissipating structures is quite complex and vast and the reader is encouraged to contact specialized
engineers. More detailed information can be found for example in Vischer & Hager (1995).
In RCC-dams the stepped chute downstream of the spillway has proven effective in reducing flow
velocities and reducing the dimensions of the subsequent stilling basin.
5.5 Intake structures
5.5.1 General
A water intake must be able to divert the required amount of water into a power canal or into a
penstock without producing a negative impact on the local environment and with the minimum
possible head losses. Also, a major challenge consists of handling debris and sediment transport.
The intake serves as a transition between a stream that can vary from a trickle to a raging torrent,
109
Guide on How to Develop a Small Hydro Site ESHA 2004
and a controlled flow of water both in quality and quantity. Its design, based on geological,
hydraulic, structural and economic considerations, requires special care to avoid unnecessary
maintenance and operational problems that cannot be easily remedied and would have to be
tolerated for the life of the project.
A water intake designer should take three criteria into consideration:
• Hydraulic and structural criteria common to all kind of intakes
• Operational criteria (e.g. percentage of diverted flow, trash handling, sediment exclusion,
etc.) that vary from intake to intake
• Environmental criteria characteristics of each project (eg requiring fish diversion systems,
fish passes, etc).
The location of the intake depends on a number of factors, such as submergence, geotechnical
conditions, environmental considerations (especially those related to fish life) sediment exclusion
and ice formation, where necessary. The orientation of the intake entrance to the flow is a crucial
factor in minimising debris accumulation on the trashrack, a source of possible future maintenance
problems. The best disposition of the intake is with the screen at right angles to the spillway so, that
during flood seasons, the flow pushes the debris over its crest. The intake should not be located in
an area of still water, far from the spillway, because the eddy currents common in such waters will
accumulate trash at the entrance.
The intake should be equipped with a trashrack to minimise the amount of debris and sediment
carried by the incoming water; a settling basin where the flow velocity is reduced, to remove all
particles over 0.2 mm; a sluicing system to flush the deposited silt, sand, gravel and pebbles with a
minimum of water loss; and a spillway to divert the excess water.
5.5.2 Intake types
The first thing for the designer to do is to decide what kind of intake the scheme needs. These can
be classified according to the following criteria:
• Power intake: The intake supplies water directly to the turbine via a penstock. These intakes
are often encountered in lakes and reservoirs and transfer the water as pressurized flow.
• Conveyance intake: The intake supplies water to other waterways (power canal, flume,
tunnel, etc.) that usually end in a power intake (Figure 1-1 Chapter 1). These are most
frequently encountered along rivers and waterways and generally transfer the water as free
surface flow.
Conveyance intakes along rivers can be classified into lateral, frontal and drop intakes. The main
characteristics of these three types are summarized in Table 5.1.
110
Guide on How to Develop a Small Hydro Site ESHA 2004
Table 5.1: Intake characteristics
River slope River width B Plan view of river Sediment transport
in outer river
bend 0.001% < J < 10%All widths Curved path is optimal
Strong bedload, small
suspended transport (Qeq <
Qcr)
with gravel
deposition
canal
0.01% < J < 10%B < 50 m Possible rectilinear path if
countermeasures
Strong bedload with
continuous flushing, strong
suspended load
Frontal
intake
with gravel
deposition
tunnel
0.01% < J < 10%B < 50 m, (B < 500 m for
economical dams/weirs)
Rectilinear is optimal, curved
path is possible if
countermeasures
Strong bedload with
continuous flushing, very
strong suspended load
Drop intake J > 10% favorably, possible
already at 2.5%
B < 50 m, (B < 500 m is
possible for dams/weirs over
part of river width)
Rectilinear Strong bedload (only large
grain sizes)
Lateral
intake
The lateral intake functions by using a river bend or by using a gravel deposition channel. The
former is presented in Figure 5.15. This intake favourably applies the presence of a strong
secondary current along the outer bend of the curved river. This secondary current prevents bedload
from entering the intake. The installed discharge Qep has to be smaller than 50 % of the critical river
discharge Qcr, where the latter is defined as the discharge for which the bedload transport starts.
Outer bank
Section a - a
Inner bank
c Surface flow →Lateral acceleration
d Bottom flow with sediment transport (centripetal force)r
vb
2
c =
c
dg
bc
a
a
Material deposition
Bank erosion
c
d
c
d
Outer bank
Section a - a
Inner bank
c Surface flow →Lateral acceleration
d Bottom flow with sediment transport (centripetal force)r
vb
2
c =r
vb
2
c =
c
dg
bc
a
a
Material deposition
Bank erosion
c
d
c
d
Figure 5.15: Secondary currents in river bends
The latter type of lateral intake uses a gravel deposition canal in front of the intake in order to
prevent both bed and suspended load from entering the intake. Hence, there is no discharge
restriction. The channel makes use of a gravel weir of minimum 1-1.5 m, as indicated in Figure
5.13. Furthermore, its slope should be at least 2%, preferably 5%. The channel bottom has to be
111
Guide on How to Develop a Small Hydro Site ESHA 2004
protected against abrasion (using high quality concrete, stones, etc.). A partially submerged wall
(0.8-1.0 m submersion) is installed in order to prevent debris from entering the intake.
The main elements of the lateral intake structure are presented in Figure 5.16: a mobile weir/dam,
gravel deposition channel and intake with trashrack.
Cross section a - a : Intake
step
gravel weir
flush gate
submerged wall
≈0.8 - 1.0 m
1.0 - 1.5 m
Cross section b - b : Weir /dam Cross section c - c : Gravel weir
→ 5 %
flush gateflushing
channel
protection against abrasion
Plan view
gates
flush gategravel weir
b o t t o m f l o w
surfac
e flow
a
a
b
b
c
c
g r a v e ld e p o s it io n
c a n a lmobile weir
step
trashrack
energy dissipation
bassin
trashrack
Cross section a - a : Intake
step
gravel weir
flush gate
submerged wall
≈0.8 - 1.0 m
1.0 - 1.5 m
Cross section b - b : Weir /dam Cross section c - c : Gravel weir
→ 5 %
flush gateflushing
channel
protection against abrasion
Plan view
gates
flush gategravel weir
b o t t o m f l o w
surfac
e flow
a
a
b
b
c
c
g r a v e ld e p o s it io n
c a n a lmobile weir
step
trashrack
energy dissipation
bassin
trashrack
Figure 5.16: Typical layout of lateral intake
The frontal intake is always equipped with a gravel deposition tunnel and is well adapted for
rectilinear river reaches. The deposition tunnel has to be flushed in a continuous manner and the
maximum river width is 50 m. A major advantage of this type of intake is its ability to handle large
quantities of both bed and suspended load. However, this needs continuous flushing and thus large
losses of water. The frontal intake is largely applied in regions with very large bed and suspended
loads, such as for example in India and Pakistan. In Europe, its application is largely restricted.
The drop intake is generally used in steep sloped rivers, such as torrents, and for rectilinear reaches.
The "French" drop intake (Figure 5.17) is essentially a canal built in the streambed, stretching
across it and covered by a trashrack with a slope greater than the streambed slope. The trashrack
bars are oriented in the direction of the streamflow. Photo 5.8 shows a drop intake installed in a
mountain stream in Asturias (Spain).
112
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.17: Secondary current along the outer bend of a curved river.
Photo 5.8: Drop intake
The Coanda type screen is an advanced concept of the drop intake, incorporating the "Coanda
effect", well known in the ore separation industry, to separate fish and debris from clean water.
Essentially it consists of a weir with a downward sloping profiled surface of stainless steel wire
screen mesh on the downstream side and a flow collection channel below the mesh - as in the drop
intake. The mesh wires are held horizontal -unlike the drop intake- and are of triangular section to
provide an expanding water passage. Water drops through the mesh with debris and fish carried off
the base of the screen. The screen is capable of removing 90% of the solids as small as 0.5 mm, so a
silt basin and sediment ejection system can be omitted. The intake is patented by AQUA SHEAR
and distributed by DULAS 11 in Europe.
In the Alps, a drop intake has been developed that is particularly adapted to very steep torrents in
high mountainous regions with difficult access, called the “Tyrolean” intake (Figure 5.18).
113
Guide on How to Develop a Small Hydro Site ESHA 2004
Flush gate
Longitudinal viewTyroleanintake
Intake
Flush gate
Longitudinal viewTyroleanintake
Intake
trashrack
Lateral viewTyrolean intake
trashrack
Lateral viewTyrolean intake
Figure 5.18: “Tyrolean” intake
Power intakes are mostly used on lakes and reservoirs. The water is transferred under pressure and
the problems associated with these kinds of intakes are different than for conveyance intakes. For
example, sediments are much less able to enter the intake, although they may pose a problem by
deposition in the lake itself. On the other hand, pressurized intakes with low pressure heads contain
the risk of vortex formation at their entrance and thus the formation of air pockets inside the
downstream conduit. This is discussed later on.
5.5.3 Head losses
For small hydro plants, head losses can be of huge importance to the feasibility of the project and
should thus be minimized as much as possible. Accounting for the following issues can do this:
• Approach walls to the trashrack designed to minimise flow separation and head losses
• Piers to support mechanical equipment including trashracks, and service gates
• Guide vanes to distribute flow uniformly
• Vortex suppression devices
• Appropriate trashrack design
The velocity profile decisively influences the trashrack efficiency. The velocity along the intake
may vary from 0.8 - 1 m/s through the trashrack to 3 - 5 m/s in the penstock. A good profile will
achieve a uniform acceleration of the flow, minimising head losses. A sudden acceleration or
deceleration of the flow generates additional turbulence with flow separation and increases the head
losses. Unfortunately a constant acceleration with low head losses requires a complex and lengthy
intake, which is expensive. A trade-off between cost and efficiency should be achieved. The
maximum acceptable velocity dictates the penstock diameter; the need for a reasonable velocity of
the flow approaching the trashrack dictates the dimensions of the rectangular section.
114
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.19: French drop intake: a canal built in the streambed and covered by a trashrack
The research department of "Energy, Mines and Resources" of Canada commissioned a study of
entrance loss coefficients for small, low-head intake structures to establish guidelines for selecting
optimum intake geometry. The results showed that economic benefits increase with progressively
smoother intake geometry having multiplane roof transition planes prepared from flat form work. In
addition, it was found that cost savings from shorter and more compact intakes were significantly
higher than the corresponding disadvantages from increased head losses.
Analyses of cost/benefits suggests that the best design is that of a compact intake with a sloping
roof and converging walls (Figure 5.19, alternative 2 in the study), whilst the length of the intake is
unlikely to be the major factor contributing to the overall loss coefficient. The K coefficient of this
transition profile was 0.19. The head loss (m) in the intake is given by:
ΔH = 0.19 V2/2g (5.1)
where V is the velocity in the penstock (m/s). Head losses due to the trashrack depend on spacing
and shape of the bars, orientation of the trashrack compared to the flow and eventual obstruction
due to debris and are discussed in more detail below.
5.5.4 Trashracks
One of the major functions of the intake is to minimise the amount of debris and sediment carried
by the incoming water, so trashracks are placed at the entrance to the intake to prevent the ingress
of floating debris and large stones. A trashrack is made up of one or more panels, fabricated from a
series of evenly spaced parallel metal bars. If the watercourse, in the flood season, entrains large
debris, it is convenient to install, in front of the ordinary grill, a special one, with removable and
widely spaced bars (from 100 mm to 300 mm between bars) to reduce the work of the automatic
trashrack cleaning equipment.
115
Guide on How to Develop a Small Hydro Site ESHA 2004
Trashracks are fabricated with stainless steel or plastic bars. Since the plastic bars can be made in
airfoil sections, less turbulence and lower head losses result. The bar spacing varies from a clear
width of 12 mm for small high head Pelton turbines to a maximum of 150 mm for large propeller
turbines. The trashrack should have a net area (the total area less the bars frontal area) so that the
water velocity does not exceed 0.75 m/s on small intakes, or 1.5 m/s on larger intakes, to avoid
attracting floating debris to the trashrack. Trashracks can be either be bolted to the support frame
with stainless steel bolts or slid into vertical slots, to be removed and replaced by stoplogs when
closure for maintenance or repair is needed. In large trashracks it must be assumed that the grill may
be clogged and the supporting structure must be designed to resist the total water pressure exerted
over the whole area without excessive deformation.
Photo 5.9: Prefabricated booms
Figure 5.20: Trash boom layout
When the river entrains heavy debris, floating booms may be located ahead of the trashracks. The
simplest boom consists of a series of floating pieces of timber connected end to end with cables or
chains. However modern booms are built with prefabricated sections of steel and plastic (Photo 5.9)
supported by steel cables. Their location is critical, because their inward bowed configuration does
not lend itself to a self-cleaning action during flood flows. Figure 5.20 (reproduced from reference
11) shows a rather complex trash boom layout designed for a dual-purpose: preventing boats
passing over the spillway and protecting the adjacent intake. A section of the boom is hinged at one
end of the fixed section so that winches can handle the other end to let the trash pass downstream
over the spillway, when large quantities are passing.
The trashrack is designed so the approach velocity (V0) remains between 0.60 m/s and 1.50 m/s.
The maximum possible spacing between the bars is generally specified by the turbine
manufacturers. Typical values are 20-30 mm for Pelton turbines, 40-50 mm for Francis turbines and
80-100 mm for Kaplan turbines.
116
Guide on How to Develop a Small Hydro Site ESHA 2004
angle of flow:
slope:
spacing: b
opening: a
Thickness of bars: d
α
δ
L : height of barsFlow velocity v 0
ΔH
bars
Horizontal support
α V0
angle of flow:
slope:
spacing: b
opening: a
Thickness of bars: d
α
δ
L : height of barsFlow velocity v 0
ΔH
bars
Horizontal support
α V0dd
bb aaa
δ
L
bar
support
1.00.760.76 0.430.370.300.74gβ
Shape factor of bars:gβ
3/4
0
g
1b
a
3
7
:5.0b
aand5d
Lfor
A
A,d
L
Head loss factor:
⎟⎠
⎞⎜⎝
⎛−=ξ
>≈
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛=ξ
ξ
f
Rack with manual cleaning:2c1.5
Rack with mechanical cleaning:1.3c1.1
Non-obstructed rack:1c
Trashrack coefficient:c
<<
<<
=
()κ⋅δ⋅⋅ξ⋅β=ζ
⋅⋅ζ=Δ
sinc
g2
vH
gg
20g
αd/a
1.00.760.76 0.430.370.300.74gβ1.00.760.76 0.430.370.300.74gβ
Shape factor of bars:gβ
3/4
0
g
1b
a
3
7
:5.0b
aand5d
Lfor
A
A,d
L
Head loss factor:
⎟⎠
⎞⎜⎝
⎛−=ξ
>≈
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛=ξ
ξ
f
Rack with manual cleaning:2c1.5
Rack with mechanical cleaning:1.3c1.1
Non-obstructed rack:1c
Trashrack coefficient:c
<<
<<
=
()κ⋅δ⋅⋅ξ⋅β=ζ
⋅⋅ζ=Δ
sinc
g2
vH
gg
20g
()κ⋅δ⋅⋅ξ⋅β=ζ
⋅⋅ζ=Δ
sinc
g2
vH
gg
20g
()κ⋅δ⋅⋅ξ⋅β=ζ
⋅⋅ζ=Δ
sinc
g2
vH
gg
20g
αd/a
Figure 5.21: Formula for computing head losses
As can be seen, the head loss coefficient depends on several factors, such as for example the way of
cleaning of the rack. The presented equations (Figure 5.21) are strictly only valid for rectangular
bars, but experience has proven that they can also be used for other bar shapes. Cleaning of
trashracks is very important to reduce possible head losses through the system. Manual cleaning is
very difficult, especially during floods. Therefore, mechanical cleaning is recommended.
Another formula for computing head losses in clean trashracks is the Kirscmer formula, detailed in
Chapter 2, section 2.2.2.1. This formula is only valid when the flow approaches the screen at right
angles.
117
Guide on How to Develop a Small Hydro Site ESHA 2004
Trashrack
Drain
Oil cylinder
Travelling carriage
Rack
Figure 5.22: Oleo-Hydraulic cylinders
Photo 5.10: Telescopic hydraulic cylinders
The trashrack should be removable for repair and maintenance and provided with facilities to clean
it. To facilitate the hand cleaning of the trashrack it should be inclined at an angle 300o from the
horizontal although steeper angles are often used. Trashracks can be cleaned by hand up to 4 meters
depth. A horizontal platform above high-water level should be provided to facilitate the operation.
On unattended plants operated by remote control, mechanical rakers are used. The mechanical raker
can be designed to be operated either on a timed basis or on a head differential basis. The latter uses
a sensor to detect the drop in head across the trashrack. An accumulation of trash on the trashrack
creates an increased differential head across the trashrack. The raker begins when a predetermined
differential head is reached.
The raker in Figure 5.22 is operated through oleo-hydraulic cylinders. The secondary cylinder
pushes out or retracts the raker, which rides on a hinged arm. The raker pushes out in its way down
to the bottom of the screen and then retracts to travel up along the screen. The raker itself is a series
of prongs protruding from a polyamide block that moves along the spaces between the bars. The
trash is conveyed to the top to be dumped on a conduit or on to a conveyor. If dumped into a
conduit a small water pump delivers enough water to wash the trash along the canal. The problem
of trash disposal must be solved case by case, bearing in mind that a trash raker can remove large
amount of debris.
When the trashrack is very long the trash raker described above is assembled on a carriage that can
move on rails along the intake. Automatic control can be programmed to pass along the supporting
structures without human aid. Using telescopic hydraulic cylinders the raker can reach down to 10
m deep, which combined with the almost limitless horizontal movement, makes it possible to clean
large surface screens (Photo 5.10).
118
Guide on How to Develop a Small Hydro Site ESHA 2004
5.5.5 Vorticity
A well-designed intake should not only minimise head losses but also preclude vorticity. Vorticity
can appear for low-head pressurized intakes (power intakes) and should be avoided because it
interferes with the good performance of turbines - especially bulb and pit turbines. Vortices may
effectively:
• Produce non-uniform flow conditions
• Introduce air into the flow, with unfavourable results on the turbines: vibration, cavitation,
unbalanced loads, etc.
• Increase head losses and decrease efficiency
• Draw trash into the intake
The criteria to avoid vorticity are not well defined, and there is not a single formula that adequately
takes into consideration the possible factors affecting it. According to the ASCE Committee on
Hydropower Intakes, disturbances, which introduce non-uniform velocity, can initiate vortices.
These include:
• Asymmetrical approach conditions
• Inadequate submergence
• Flow separation and eddy formation
• Approach velocities greater than 0.65 m/sec
• Abrupt changes in flow direction
Lack of sufficient submergence and asymmetrical approach seem to be the most common causes of
vortex formation. An asymmetric approach is more prone to vortex formation than a symmetrical
one. When the inlet to the penstock is deep enough and the flow is undisturbed, vortex formation is
unlikely.
Empirical formulas exist that express the minimum degree of submergence of the intake in order to
avoid severe vortex formation. Nevertheless, no theory actually exists that fully accounts for all
relevant parameters. The minimum degree of submergence is defined as shown in Figure 5.23.
D
h t
DD
h tht
Figure 5.23: Minimum degree of submergence
119
Guide on How to Develop a Small Hydro Site ESHA 2004
The submersion is defined as ht. The following formulas express the minimum values for ht:
KNAUSS ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛
⋅⋅+⋅≥Dg
V3.21Dht (5.2)
NAGARKAR ()54.050.0
t DV4.4h ⋅⋅≥ (5.3)
ROHAN (5.4) 76.048.0
t DV474.1h ⋅⋅≥
GORDON DVcht⋅⋅≥ (5.5)
with c = 0.7245 for asymmetric approach conditions
c = 0.5434 for symmetric approach conditions
It is important to highlight that V is the velocity inside the downstream conduit in m/s and D is the
hydraulic diameter of the downstream conduit in m.
Beside a minimum submersion, constructive measures might help to prevent vortex formation. For
example, asymmetric flow conditions may be prevented by means of vertical walls, piles, screens,
floating rafts or by appropriate design of the entrance shape.
5.6 Sediment traps
5.6.1 General
Conveyance intakes are designed on rivers in order to eliminate possible floating debris and bedload
transport. However, they cannot prevent the entrance of suspended sediment transport. For this, a
sediment trap is projected downstream of an intake. The main objective of such a trap is to avoid
sedimentation of downstream structures (canals, shafts, etc.) as well as to limit the possible damage
of sediments on the hydro mechanical equipment.
A sediment trap is based on the principle of diminishing the flow velocities and turbulence. This
results in a decantation of suspended sediments in the trap. This diminishing is obtained by an
enlargement of the canal, controlled by a downstream weir as shown in Figure 5.24.
120
Guide on How to Develop a Small Hydro Site ESHA 2004
Sediment deposit
decantation
Flushing channel
Flushing element
decantation
weir
channel
sediment deposit flushing
vT
vD
Q
L
h
Sediment deposit
decantation
Flushing channel
Flushing element
decantation
weir
channel
sediment deposit flushing
vT
vD
Q
L
h
Figure 5.24: Sediment traps
A sediment sluicing system that minimises the sluicing time and the wasted water can be used 4 .
5.6.2 Efficiency of a sediment trap
The efficiency of the sediment trap is defined by the grain diameter that deposits in the trap. The
choice of efficiency depends on the type of hydro mechanical equipment and on the gross head
difference of the power plant. For a Francis turbine, the abrasive power of sediment grains is
expressed as a function of the velocity of the grains and the gross head of the plant as follows:
3Es
e VRP⋅ρ−ρ⋅∀⋅μ= (5.6)
in which μ is a friction coefficient between the turbine blades and the grains, ∀ is the volume of the
grains, ρs and ρE are the densities of grains and water, R is the radius of the blades and V is the
grain velocity. The volume of the grains is directly related to the efficiency of the trap.
Reparation intervals of Francis turbines are around 6-7 years for a sediment trap efficiency of
0.2 mm, 3-4 years for an efficiency of 0.3 mm and 1-2 years for an efficiency of only 0.5 mm. It is
obvious that the cost of a sediment trap increases with its efficiency. Hence, an optimum efficiency
may be found as a function of the construction costs, the energy losses, the reparation costs of the
turbines and the exploitation costs. Experience has shown that the most economical solution is
around 0.2 mm efficiency for severe conditions (significant gross head, quartz particles) and around
0.3 mm for normal conditions.
5.6.3 Design
The necessary length of a sediment trap is defined by the equipped discharge of the intake and by
the chosen efficiency of the trap (grain diameter that still deposits inside the trap). The length has to
be such that all grains have the time to deposit before leaving the trap. This happens when the
deposition time tD equals the transfer time tt. The former is defined as h/vD and the latter as L/vT
(see Figure 5.24). Hence, the minimum length required to deposit a grain of diameter dD is given:
Bv
QL
D ⋅≥ (5.7)
121
Guide on How to Develop a Small Hydro Site ESHA 2004
The width B has to stay smaller than 1/8 times the length L and also smaller than twice the flow
depth h. The deposition velocity vD is defined by the Newton or Prandtl formula for spherical
particles and under ideal conditions, i.e. pure water, no turbulence and no wall effects. It depends on
the form drag of the particle, which on its turn depends on the Reynolds number. For real situations,
no formula exists and experiments should be carried out. For practice, the empirical formula of
Zanke is often used as a first-hand approach in still water flow conditions:
(1d1057.11d9
100v 32
D −⋅⋅+⋅=) (5.8)
in which vD is expressed in mm/s and the grain diameter d in mm. This expression is strictly valid
for T = 20° and a grain-to-water density ratio of 2.65.
For turbulent flow conditions, the deposition velocity decreases and the following expression
becomes more appropriate:
0vvvT0DD≥⋅α−= (5.9)
in which vD0 is the deposition velocity in still water and α a reduction factor (in [1/m1/2]) expressed
as a function of the trap water depth h (m):
h
132.0=α (5.10)
Finally, for appropriate design, the critical transfer velocity of the trap has to be defined. This
critical velocity defines the limit between the suspension regime and the deposition regime. If the
velocity is too high, deposited sediments risk to be entrained again by the flow. For a Manning-
Strickler roughness value of K = 60 m1/3/s (K = 1/n, average value for concrete) and for a grain-to-
water density ratio of 2.65 the following formula is valid:
dR13v61
hcr ⋅⋅= (5.11)
Typical values for vcr are 0.2-0.3 m/s.
Further information regarding design and construction details can be found for example in Bouvard
(1984).
5.7 Gates and valves
In every small hydropower scheme some components, for one reason or another (maintenance or
repair to avoid the runaway speed on a shutdown turbine, etc) need to be able to be temporarily
isolated. Some of the gates and valves suited to the intakes for small hydro systems include the
following:
• Stoplogs made up of horizontally placed timbers
• Sliding gates of cast iron, steel, plastic or timber
• Flap gates with or without counterweights
• Globe, rotary, sleeve-type, butterfly and sphere valves
122
Guide on How to Develop a Small Hydro Site ESHA 2004
Almost without exception the power intake will incorporate some type of control gate or valve as a
guard system located upstream of the turbine and which can be closed to allow the dewatering of
the water conduit. This gate must be designed so it can be closed against the maximum turbine flow
in case of power failure, and it should be able to be opened partially, under maximum head, to allow
the conduit to be filled.
For low pressure the simplest type of gate is a stoplog; timbers placed horizontally and supported at
each end in grooves. Stoplogs cannot control the flow and are used only to stop it. If flow must be
stopped completely, such as when a repair is needed downstream, the use of two parallel sets of
stoplogs is recommended. They should be separated by about 15 cm, so that clay can be packed in
between. Gates and valves control the flow through power conduits. Gates of the sliding type are
generally used to control the flow through open canals or other low-pressure applications. This is
the type of flow control used on conveyance intake structures where, if necessary, the flow can be
stopped completely to allow dewatering of the conduit. Cast iron sliding-type gates are those mostly
used for openings of less than two square meters. For bigger openings fabricated steel sliding gates
are cheaper and more flexible. Gates of the sliding type are seldom used in penstocks because they
take too long to close. The stopper slides between two guides inside the gate.
Photo 5.11: wheel-and-axle mechanism
Figure 5.25: Wedge-shaped stopper
When used in a high-pressure conduit the water pressure that forces the stopper against its seat
makes the valve difficult to operate. This difficulty is overcome with a wedge-shaped stopper
(Figure 5.25), so that the seal is broken over the whole face as soon as it rises even a small distance.
To provide a good seal around a sliding gate different kinds of rubber seals are used. They can be
made of natural rubber, styrene-butadiene or chloroprene compounds. The seal path is located
adjacent to the roller path.
Using a wheel-and-axle mechanism (Photo 5.11), a hydraulic cylinder (Photo 5.12) or an electric
actuator on a screw thread can raise small sliding gates controlling the flow.
123
Guide on How to Develop a Small Hydro Site ESHA 2004
In butterfly valves a lens shaped disk mounted on a shaft turns to close the gap (Figure 5.26). Under
pressure each side of the disk is submitted to the same pressure, so the valve is easy to manoeuvre
and closes rapidly. Butterfly valves are used as the guard valves for turbines and as regulating
valves. Is easy to understand that when used for regulation their efficiency is rather low because the
shaped disk remains in the flow and causes turbulence.
Photo 5.12: Hydraulic Cylinder
Figure 5.26: Butterfly valves
Figure 5.27: Globe and rotary valves
Butterfly valves are simple, rugged and uncomplicated and can be operated manually or
hydraulically. Photo 5.13 shows a large butterfly valve being assembled in a powerhouse and Photo
5.14 shows a butterfly valve, hydraulically operated, with an ancillary opening system and a
counterweight, at the entrance to a small Francis turbine.
124
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.13: Large butterfly valve
Photo 5.14: Butterfly valve hydraulically operated
Globe and rotary valves (Figure 5.27) have lower head losses than the slide and butterfly gate
valves and are also commonly used in spite of their higher price.
The radial gates, conceptually different, are a method of forming a moveable overflow crest and
allow a close control of headwater and tailwater. Photo 5.15 shows a Tainter gate at the left, ready
to be installed, and the housing of the sector on a concrete pier at the right. The radial gate is
operated by raising or lowering to allow water to pass beneath the gate plate. The curved plate that
125
Guide on How to Develop a Small Hydro Site ESHA 2004
forms the upstream face is concentric with the trunnions of the gate. The trunnions are anchored in
the piers and carry the full hydrostatic load. Because the hydrostatic load passes through the
trunnions, the lifting force required by the hoisting mechanism is minimised. The head losses in
gates and valves are relatively high, especially when they are operated as regulating devices. For
further details refer to Chapter 2, Section 2.2.4 and the enclosed bibliography.
Photo 5.15: Tainter gate (left) and housing of its sector on a concrete pier
5.8 Open channels
5.8.1 Design and dimensioning
The flow conveyed by a canal is a function of its cross-sectional profile, its slope, and its roughness.
Natural channels are normally very irregular in shape, and their surface roughness changes with
distance and time. The application of hydraulic theory to natural channels is more complex than for
artificial channels where the cross-section is regular in shape and the surface roughness of the
construction materials - earth, concrete, steel or wood - is well documented, so that the application
of hydraulic theories yields reasonably accurate results.
Table 2.4, Chapter 2, illustrates the fundamental geometric properties of different channel sections.
In small hydropower schemes the flow in the channels is in general in the rough turbulent zone and
the Manning equation can be applied:
3/2
2/13/52/13/2
Pn
SA
n
SRAQ ⋅
⋅=⋅⋅= (5.12)
where n is Manning's coefficient, which in the case of artificial lined channels may be estimated
with reasonable accuracy, and S is the hydraulic gradient, which normally is the bed slope.
Alternatively:
126
Guide on How to Develop a Small Hydro Site ESHA 2004
2
3/2
2
3/5
3/2
⎟⎠
⎞⎜⎝
⎛⋅=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⋅⋅=AR
nQ
A
PnQS (5.13)
The above equation applies when metric or SI units are used. To use Imperial or English units the
equation must be modified to:
3/2
2/13/549.1
Pn
SAQ⋅
⋅= (5.14)
where Q is in ft3/s; A in ft2 and P in ft. n has the same value as before.
The above equation shows that for the same cross-sectional area A and channel slope S, the channel
with a larger hydraulic radius R, delivers a larger discharge. That means that for a given cross-
sectional area, the section with the least wetted perimeter is the most efficient hydraulically.
Semicircular sections are consequently the most efficient. A semicircular section however, unless
built with prefabricated materials, is expensive to build and difficult to maintain. The most efficient
trapezoidal section is the half hexagon, whose side slope is 1 vertical to 0.577 horizontal. Strictly,
this is only true if the water level reaches the level of the top of the bank. Actual dimensions have to
include a certain freeboard (vertical distance between the designed water surface and the top of the
channel bank) to prevent water level fluctuations overspilling the banks. Minimum freeboard for
lined canals is about 10 cm, and for unlined canals this should be about one third of the designed
water depth with a minimum of fifteen centimetres. One way to prevent overflow of the canal is to
provide spillways at appropriate intervals; any excess water is conveyed, via the spillway, to an
existing streambed or to a gully.
Table 5.2: Hydraulic parameters for common canal cross-sections
Type of Channel Manning's n
Excavated earth channels
Clean 0.022
Gravelly 0.025
Weedy 0.030
Stony, cobbles (or natural streams) 0.035
Artificially lined channels
Brass 0.011
Steel, smooth 0.012
Steel, painted 0.014
Steel, riveted 0.015
Cast iron 0.013
Concrete, well-finished 0.012
Concrete, unfinished 0.014
Planed wood 0.012
Clay tile 0.014
Brickwork 0.015
Asphalt 0.016
Corrugated metal 0.022
Rubble masonry 0.025
127
Guide on How to Develop a Small Hydro Site ESHA 2004
It should be noted that the best hydraulic section does not necessarily have the lowest excavation
cost. If the canal is unlined, the maximum side slope is set by the slope at which the material will
permanently stand under water. Clay slopes may stand at 1 vertical to 3/4 horizontal, whereas sandy
soils must have flatter slopes (1 to 2).
Table 5.3 defines for the most common canal sections the optimum profile as a function of the
water depth y, together with the parameters identifying the profile.
Table 5.3: Optimum profile for different channel sections
Channel Area Wetted Hydraulic Top Water
section perimeter radius width depth
A P R T d
Trapezoid:
half hexagon 1.73 y2 3.46 y 0.500 y 2.31 y 0.750y
Rectangle :
half square 2 y2 4 y 0.500 y 2 y y
Triangle:
half square y2 2.83 y 0.354 y 2 y 0.500y
Semicircle 0.5πy2 π y 0.500 y 2 y 0.250πy
Example 5.1
Assuming a flow depth of 1 m, a channel base width of 1.5 m and side slopes of 2 vertical to 1
horizontal, a bed slope of 0.001 and a Manning's coefficient of 0.015, determine the discharge (Q),
the mean velocity (V).
According to Table 2.4 for b=1.5, x=1/2 and y=1
A=(1.5+0.5x1)x1=2m2; mxP736.35.0125.1 2 =++=
Applying 5.6) for A=2 and P=3.736
smxxQ/78.2001.0736.3
2
015.0
1 3
3/2
3/5
==
V=Q/A=2.78/2=1.39 m/s
Example 5.2
Determine the slope knowing the discharge and the canal dimensions. Assuming a canal paved with
smooth cement surface (n=0.011), a channel base of 2 m, side slopes with inclination 1v:2h and a
uniform water depth of 1.2 m, determine the bed slope for a discharge of 17.5 m3/s.
Applying the formulae of table 2.4:
128
Guide on How to Develop a Small Hydro Site ESHA 2004
002.0717.028.5
011.05.17 2
3/2 =⎟⎠
⎞⎜⎝
⎛
⋅
⋅=S
When the canal section, the slope and discharge are known and the depth "d" is required, equation
5.6 - nor any other - does not provide a direct answer, so iterative calculations must be used.
Example 5.3
A trapezoidal open channel has a bottom width of 3 m and side slopes with inclination 1.5:1. The
channel is lined with unfinished concrete. The channel is laid on a slope of 0.0016 and the discharge
is 21 m3/s. Calculate the depth.
According to 5.6 the section factor:
A=(b+zy)y = (3 + 1.5y)y P=b+2y(1+z2)0.5 = 3+3.6y
Compute the factor section for different values of y, up to find one approaching closely 6.825:
For y = 1.5 m A=7.875, R=0.937, AR2/3=7.539
For y = 1.4 m A=7.140, R=0.887. AR2/3=6.593
For y = 1.43 m A=7.357, R=0.902, AR2/3=6.869
According to the above results the normal depth is slightly under 1.43. Using the software program
FlowPro, mentioned in Chapter 2 it would be instantaneously calculated, as shown in the enclosed
captured screen: a depth of 1.425, with A=2.868, P=8.139, R=0.900 and a section factor 6.826
Summarising, the design of fabricated channels is a simple process requiring the following steps:
• Estimate the coefficient n from table 5.2
• Compute the form factor AR2/3=nQ/S1/2 with the known parameters in second term
• If optimum section is required apply values in table 5.3. Otherwise use values in table 2.4
• Check if the velocity is high enough to form deposit or aquatic flora
• Check the Froude number NF to determine if it is a subcritical or a supercritical flow
• Define the required freeboard
129
Guide on How to Develop a Small Hydro Site ESHA 2004
Example 5.4
Design a trapezoidal channel for an 11 m3/s discharge. The channel will be lined with well-finished
concrete and the slope 0.001.
Step 1. Manning n = 0.012
Step 2. Compute form factor
174.4001.0
11012.03/2 ===x
S
nQAR
Step 3. Not intended to find the optimum section.
Step 4. Assuming a bottom width of 6 m and side slopes with inclination 2:1 compute the depth d
by iteration as in example 5.3.
d = 0.87 m A = 6.734 m2
Step 5. Compute the velocity
V = 11/6.734 = 1.63 m/s OK
Step 6. Total channel height. The tables of the US Bureau of Reclamation (USA) recommend a
freeboard of 0.37 m. The FlowPro software would provide all these results.
5.8.2 Excavation and stability
In conventional hydropower schemes and in some of the small ones, especially those located in
wide valleys where the channels must transport large discharges, the channels are designed in the
manner shown in Figure 5.28. According to this profile, the excavated ground is used to build the
embankments, not only up to the designed height but to provide the freeboard, the extra height
necessary to account for the height increase produced by a sudden gate closing, waves or the excess
130
Guide on How to Develop a Small Hydro Site ESHA 2004
arising in the canal itself under heavy storms. These embankment channels although easy to
construct are difficult to maintain, due to wall erosion and aquatic plant growth.
The stability of the walls is defined by the eventual sliding of the material. This sliding can be
enhanced by rapid water level changes in the canal. The velocity of water in unlined canals should
be kept above a minimum value to prevent sedimentation and aquatic plant growth, but below a
maximum value to prevent erosion.
If the canal is unlined, the maximum velocity to prevent erosion is dependent on the mean grain
diameter of the bank material dm:
61
h
31
m Rd7.5V ⋅⋅≤ (5.15)
Where Rh stands for the hydraulic radius of the canal. For grain diameters of 1 mm and hydraulic
radius of 1 to 3 m, critical velocities of 0.6-0.7 m/s are obtained. For grain diameters of 10 mm, the
critical velocities are between 1.2 and 1.5 m/s for the same hydraulic radius. The above equation
can be used for grain diameters larger than 0.1 mm.
For cohesive soils, the critical velocities are between 0.4 and 1.5 m/s. Concrete-lined canals may
have clear water velocities up to 10 m/s without danger. Even if the water contains sand, gravel or
stones, velocities up to 4 m/s are acceptable.
On the other hand, to keep silt in suspension after the intake, the flow velocity should be at least
0.3-0.5 m/s. To prevent aquatic plant growth, the minimum velocities are 0.5-0.75 m/s and the
minimum water depths are 1.5 to 2.0 m.
Figure 5.28: Channel design
Figure 5.29: Rectangular reinforced canal
An appropriate lining provides bank protection. Possible materials to be used for protection are
vegetation, rock blocks with or without mortar, bituminous material, or concrete. Some examples
are presented in Figure 5.30.
131
Guide on How to Develop a Small Hydro Site ESHA 2004
Bituminous
material
Drainage
layer
Geotextile
or filter
Rock blocks
protection against scour
1 - 1.5 m minimum
mortar
Rock blocs
concrete (with armoring,
200 à 300 kg/m3)
ev. impermeable screen
Construction
joints filled with
bituminous
material Drainage
layer
Bituminous
material
Drainage
layer
Geotextile
or filter
Rock blocks
protection against scour
1 - 1.5 m minimum
mortar
Rock blocs
concrete (with armoring,
200 à 300 kg/m3)
ev. impermeable screen
Construction
joints filled with
bituminous
material Drainage
layer
Figure 5.30: Materials used for protection
In high mountain schemes the canal is usually built from reinforced concrete, so that environmental
legislation may require it to be covered and revegetated. Figure 5.29 shows the schematic section of
a rectangular reinforced concrete canal in the Cordinañes scheme, referred to in Chapter 4 and
Photo 5.15 shows the same canal not yet covered with the concrete slab that would serve as a basis
for new ground and new vegetation. Sometimes, to ensure that no seepage will occur, the canal is
lined with geotextile sheets, to prevent landslides consequent to the wetting of clayey material.
As is shown in the following examples, once the canal profile has been selected it is easy to
compute its maximum discharge.
Photo 5.15: Canal in the Cordinañes
132
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.16: Lateral spillway
To ensure that the channel never overflows, endangering the slope stability, and in addition to
provide a generous freeboard, a lateral spillway (as in Photo 5.16) should be provided.
Before definitely deciding on the channel route, a geologist should carefully study the
geomorphology of the terrain. Photo 5.17 shows clearly how uplift can easily ruin a power channel
(6 m wide and 500 m long, in a 2 MW scheme). On one particular day, a flood occurred which was
later calculated to be a 100 year event. At the time the flood occurred, the headrace channel had
been empty, and uplift pressures destroyed the channel. Consideration should also be taken of the
type of accidents detailed in Chapter 4, section 4.4.
Photo 5.17:Uplift
Photo 5.18: Flume
133
Guide on How to Develop a Small Hydro Site ESHA 2004
Circumventing obstacles
Along the alignment of a canal obstacles may be encountered, and to bypass them it will be
necessary to go over, around or under them.
The crossing of a stream or a ravine requires the provision of a flume, a kind of prolongation of the
canal, with the same slope, supported on concrete or steel piles or spanning as a bridge. Steel pipes
are often the best solution, because a pipe may be used as the chord of a truss, fabricated in the
field. The only potential problem is the difficulty of removing sediment deposited when the canal is
full of still water. Photo 5.18 shows a flume of this type in China.
Inverted siphons can also solve the problem. An inverted siphon consists of an inlet and an outlet
structure connected by a pipe. The diameter calculation follows the same rules as for penstocks,
which are analysed later.
5.9 Penstocks
Arrangement and material selection for penstocks
Conveying water from the intake to the powerhouse (this is the purpose of a penstock) may not
appear a difficult task. However deciding the most economical arrangement for a penstock is not so
simple. Penstocks can be installed over or under the ground, depending on factors such as the nature
of the ground itself, the penstock material, the ambient temperatures and the environmental
requirements.
A flexible and small diameter PVC penstock for instance, can be laid on the ground, following its
outline with sand and gravel surrounding the pipe to provide good insulation. Small pipes installed
in this way do not need anchor blocks and expansion joints.
Larger penstocks are usually buried, as long as there is only a minimum of rock excavation
required. Buried penstocks must be carefully painted and wrapped to protect the exterior from
corrosion, but provided the protective coating is not damaged when installed, further maintenance
should be minimal. From the environmental point of view the solution is optimal because the
ground can be returned to its original condition, and the penstock does not constitute a barrier to the
movement of wildlife.
Figure 5.31: Penstock
134
Guide on How to Develop a Small Hydro Site ESHA 2004
A penstock installed above ground can be designed with or without expansion joints. Variations in
temperature are especially important if the turbine does not function continuously, or when the
penstock is dewatered for repair, resulting in thermal expansion or contraction. Usually the
penstock is built in straight or nearly straight lines, with concrete anchor blocks at each bend and
with an expansion joint between each set of anchors (Figure 5.31). The anchor blocks must resist
the thrust of the penstock plus the frictional forces caused by its expansion and contraction, so when
possible they should be founded on rock. If, due to the nature of the ground, the anchor blocks
require large volumes of concrete, thus becoming rather expensive, an alternative solution is to
eliminate every second anchor block and all the expansion joints, leaving the bends free to move
slightly. In this case it is desirable to lay the straight sections of the penstock in steel saddles, made
to fit the contour of the pipe and generally covering 120 degrees of the invert (Figure 5.32). The
saddles can be made from steel plates and shapes, with graphite asbestos sheet packing placed
between saddle and pipe to reduce friction forces. The movement can be accommodated with
expansion joints, or by designing the pipe layout with bends free to move.
If a pipeline system using spigot and socket joints with O-ring gaskets is chosen, then expansion
and contraction is accommodated in the joints.
Today there is a wide choice of materials for penstocks. For the larger heads and diameters,
fabricated welded steel is probably the best option. Nevertheless spiral machine-welded steel pipes
should be considered, due to their lower price, if they are available in the required sizes. For high
heads, steel or ductile iron pipes are preferred, but at medium and low heads steel becomes less
competitive, because the internal and external corrosion protection layers do not decrease with the
wall thickness and because there is a minimum wall thickness for the pipe.
For smaller diameters, there is a choice between: manufactured steel pipe, supplied with spigot and
socket joints and rubber "O" gaskets, which eliminates field welding, or with welded-on flanges,
bolted on site (Figure 5.33); plain spun or pre-stressed concrete; ductile iron spigot and socket pipes
with gaskets; cement-asbestos; glass-reinforced plastic (GRP); and PVC or polyethylene (PE)
plastic pipes. Plastic pipe PE14 is a very attractive solution for medium heads (a PVC pipe of 0.4 m
diameter can be used up to a maximum head of 200 meters) because it is often cheaper, lighter and
more easily handled than steel and does not need protection against corrosion. PVC15 pipes are
easy to install because of the spigot and socket joints provided with "O" ring gaskets. PVC pipes are
usually installed underground with a minimum cover of one metre. Due to their low resistance to
UV radiation they cannot be used on the surface unless painted, coated or wrapped. The minimum
radius of curvature of a PVC pipe is relatively large (100 times the pipe diameter)– and its
coefficient of thermal expansion is five times higher than that for steel. They are also rather brittle
and unsuited to rocky ground.
Pipes of PE16 – (high molecular weight polyethylene) can be laid on top of the ground and can
accommodate bends of 20-40 times the pipe diameter (for sharper bends, special factory fittings are
required). PE pipe floats on water and can be dragged by cable in long sections but must be joined
in the field by fusion welding, requiring a special machine. PE pipes can withstand pipeline freeze-
up without damage, may be not available in sizes over 300 mm diameter.
135
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.32: Penstock with concrete anchor blocks and expansion joints
Concrete penstocks, both pre-stressed with high tensile wires or steel reinforced, featuring an
interior steel jacket to prevent leaks, and furnished with rubber gasket spigot and socket joints
constitute another solution. Unfortunately their heavy weight makes transportation and handling
costly, but they are not affected by corrosion.
In developing countries, pressure creosoted wood-stave, steel-banded pipe is an alternative that can
be used in diameters up to 5.5 metres and heads of up to 50 metres (which may be increased up to
120 meters for a diameter of 1.5 metres). The advantages include flexibility to conform to ground
settlement, ease of laying on the ground with almost no grade preparation, no requirement for
expansion joints and no necessity for concrete supports or corrosion protection. Wood-stave pipe is
assembled from individual staves and steel bands or hoops that allow it to be easily transported even
over difficult terrain. Disadvantages include leakage, particularly in the filling operations, the need
to keep the pipe full of water when repairing the turbine, and considerable maintenance such as
spray coating with tar every five years. Table 5.3 shows the main properties of the above material.
Some of these properties are not always typical, particularly the values of the Hazen Williams
coefficient which depends on the surface condition of the pipe.
136
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.33: Manufactured steel pipe
Table 5.4: Different material’s characteristics
Material Young’s modulus of
elasticity E(N/m2)E9
Coefficient of linear
expansion a (m/m 0c)E6
Ultimate tensile
strength (N/m2)E6
n
Welded Steel 206 12 400 0.012
Polyethylene 0.55 140 5 0.009
Polyvinyl Chloride (PVC)2.75 54 13 0.009
Asbestos Cement n/a 8.1 n/a 0.011
Cast iron 78.5 10 140 0.014
Ductile iron 16.7 11 340 0.013
Hydraulic design and structural requirements
A penstock is characterised by materials, diameter, wall thickness and type of joint:
• the material is selected according to the ground conditions, accessibility, weight, jointing
system and cost,
• the diameter is selected to reduce frictional losses within the penstock to an acceptable level,
137
Guide on How to Develop a Small Hydro Site ESHA 2004
• the wall thickness is selected to resist the maximum internal hydraulic pressure, including
transient surge pressure that will occur.
Penstock diameter
The diameter is selected as the result of a trade-off between penstock cost and power losses. The
power available from the flow Q and head H is given by the equation:
P=QHγη
where Q is the discharge in m3/s, H the net head in m, γ the specific weight of water in kN/m3 and η
the overall efficiency.
The net head equals the gross head minus the sum of all losses, including the friction and turbulence
losses in the penstock, that are approximately proportional to the square of the velocity of the water
in the pipe. To convey a certain flow, a small diameter penstock will need a higher water velocity
than a larger diameter penstock, and therefore the losses will be greater. Selecting a diameter as
small as possible will minimise the penstock cost but the energy losses will be larger and vice versa.
Chapter 2 details the friction loss calculations, putting special emphasis on the graphic
representation of the Colebrook equations (the Moody diagram and the Wallingford charts) and on
the Manning's formula. In this chapter the above principles are used and some examples will
facilitate their application in real cases.
A simple criterion for diameter selection is to limit the head loss to a certain percentage. Loss in
power of 4% is usually acceptable. A more rigorous approach is to select several possible
diameters, computing power and annual energy. The present value of this energy loss over the life
of the plant is calculated and plotted for each diameter (Figure 5.34). On the other side the cost of
the pipe for each diameter is also calculated and plotted. Both curves are added graphically and the
optimum diameter would be that closest to the theoretical optimum.
Actually the main head loss in a pressure pipe are friction losses. The head losses due to turbulence
passing through the trashrack, in the entrance to the pipe, in bends, expansions, contractions and
valves are minor losses. Consequently a first approach will suffice to compute the friction losses,
using for example the Manning equation:
333.5
22
3.10 D
Qn
L
h f = (5.16)
Examining the above equation, it can be seen that dividing the diameter by two would lead to the
losses being multiplied by 40.
From this it follows that:
1875.0223.10
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛⋅⋅=
fh
LQnD (5.17)
If we limit hf at 4H/100, D can be computed knowing Q, n and L, by the equation:
138
Guide on How to Develop a Small Hydro Site ESHA 2004
1875.022
69.2 ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛=H
LQnD (5.18)
Figure 5.34: Energy loss
Example 5.5
A scheme has a gross head of 85 m, a discharge of 3 m3/s, and a 173 m long penstock in welded
steel. Calculate the diameter so the power losses due to friction do not surpass 4%.
According to equation (5.18):
mxxD88.085
173012.0369.2
1875,022
=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛=
We select a 1m steel welded pipe and compute all the losses in the next example
Example 5.6
Compute the friction and turbulence head losses in a scheme as that illustrated in Figure 5.35. The
rated discharge is 3 m3/s and the gross head 85 m. The steel welded penstock diameter 1.0 m. The
radius of curvature of the bends is four times the diameter. At the entrance of the power intake there
is a trashrack with a total surface of 6 m2, inclined 600 to the horizontal. The bars are 12-mm thick
stainless steel bars, and the distance between bars is 70 mm.
The flow velocity approaching the screen is: (with K1=1)
smxxxV/7.0866.0
1
6
1
70
127030 =+=
139
Guide on How to Develop a Small Hydro Site ESHA 2004
The head loss through the trashrack is given by the Kilchner formula:
mxxxxhf0049.0866.081.92
7,0
70
124.2
23/4
=⎟⎠
⎞⎜⎝
⎛=
The head loss at the inlet of the penstock is given in Figure 2.11, Chapter 2: K=0.08. The velocity in
the penstock is 3.82 m/s, so the head loss at the inlet:
he= 0.08 x 3.822/(2 x 9.81) = 0.06 m
The gross head at the beginning of the penstock is therefore
85-0.005-0.06=84.935 m
The friction loss in the penstock, according Manning equation (2.15) is:
mxxxhf30.21730.1
3012.03.10
333.5
22
==
The Kb coefficient for the first bend is 0.05. The coefficient for the second bend Kb=0.085 and for
the third bend Kb=0.12. The head losses in the three bends amount to:
(0.05 + 0.085 + 0.12) x 3.822/(2 x 9.81) = 0.19 m.
The head loss in the gate valve 0.15 x 3.822/(2 x 9.81) = 0.11 m
Summarising: head loss in trashrack plus pipe inlet: 0.065
head loss in three bends and valve : 0.30 m
head loss by friction in the penstock: 2.30 m
Total head loss: 2.665 m equivalent to 3.14% of the gross power.
140
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.35: Friction and turbulence head losses
Wall thickness
The wall thickness required depends on the pipe material, its ultimate tensile strength (and yield),
the pipe diameter and the operating pressure. In steady flows (discharge is assumed to remain
constant with time) the operating pressure at any point along a penstock is equivalent to the head of
water above that point. The wall thickness in this case is computed by the equation:
f
DPeσ2
1 ⋅= (5.19)
where
e = Wall thickness in mm
P1= Hydrostatic pressure in kN/mm2
D = Internal pipe diameter in mm
fσ = Allowable tensile strength in kN/mm2
In steel pipes the above equation is modified by:
s
ff
ek
DPe +⋅
⋅=σ2
1
where
es= extra thickness to allow for corrosion
141
Guide on How to Develop a Small Hydro Site ESHA 2004
kf= weld efficiency
kf = 1 for seamless pipes
kf = 0.9 for x-ray inspected welds
kf = 1.0 for x-ray inspected welds and stress relieved
fσ = allowable tensile stress (1400 kN/mm2)
The pipe should be rigid enough to be handled without danger of deformation in the field. ASME
recommends a minimum thickness in mm equivalent to 2.5 times the diameter in metres plus 1.2
mm. Other organisations recommend as minimum thickness tmin=(D+508)/400, where all
dimensions are in mm.
In high head schemes it can be convenient to use penstock of uniform diameter, but with different
thickness as a function of the hydrostatic pressures.
A certain area of the penstock can remain under the Energy Gradient Line and collapse by sub-
atmospheric pressure. The collapsing depression will be given by:
3
882500 ⎟⎠
⎞⎜⎝
⎛=D
exPc (5.20)
where e and D are respectively the wall thickness and diameter of the pipe in mm.
This negative pressure can be avoided by installing an aeration pipe with a diameter in cm given by:
cP
Qd47.7= (5.21)
provided Pc ≤ 0.49 kgN/mm2 ; otherwise d=8.94 Q .
Sudden changes of flow can occur when the plant operator or the governing system opens or closes
the gates rapidly. Occasionally the flow may even be stopped suddenly due to full load rejection, or
simply because an obstruction becomes lodged in the nozzle of a Pelton turbine jet. A sudden
change of flow rate in a penstock may involve a great mass of water moving inside the penstock.
The pressure wave which occurs with a sudden change in the water's velocity is known as water
hammer; and although transitory, can cause dangerously high and low pressures whose effects can
be dramatic: the penstock can burst from overpressure or collapse if the pressures are reduced
below ambient. The surge pressures induced by the water hammer phenomenon can be of a
magnitude several times greater than the static pressure due to the head, and must be considered in
calculating the wall thickness of the penstock.
Detailed information on the water hammer phenomenon can be found in texts on hydraulics, and
information is given in Chapter 2, section 2.2.3. Some examples will show the application of the
recommended formulae.
As explained in Chapter 2, the pressure wave speed c (m/s) depends on the elasticity of the water
and pipe material according to the formula:
142
Guide on How to Develop a Small Hydro Site ESHA 2004
ρ⎟⎠
⎞⎜⎝
⎛+
=
−
Et
kD
kc
1
10 3
(5.22)
where
k = bulk modulus of water 2.1x109 N/m2
E = modulus of elasticity of pipe material (N/m2)
t = wall thickness (mm)
The time taken for the pressure wave to reach the valve on its return, after sudden closure is known
as the critical time:
T= 2L/c (5.23)
For instantaneous closure (the pressure wave reaches the valve after its closure) the increase in
pressure, in metres of water column, due to the pressure wave is:
gcPvΔ= (5.24)
where is the velocity change. vΔ
Examples 6.4 and 6.5 shows that surge pressures in steel pipes are more than three times greater
than in PVC, due to the greater stiffness of the steel.
Example 5.7
Calculate the pressure wave velocity, for instant closure, in a steel penstock 400mm diameter and
4mm-wall thickness.
Applying the above equations gives:
smc/1024
4101.2
400101.21
101.2
11
9
6
=
××
××+
×=
b) The same for a PVC pipe 400 mm diameter and 14 mm wall thickness.
smc/305
141075.2
400101.21
101.2
9
9
6
=
××
××+
×=
143
Guide on How to Develop a Small Hydro Site ESHA 2004
Example 5.8
What is the surge pressure, in the case of instant valve closure, in the two penstocks of example 5.7,
if the initial flow velocity is 1.6 m/s?
a) steel penstock:
mxPs4178.9
41024 ==
b) PVC penstock:
mxPs1238.9
4305
==
As the example 5.8 shows, the surge pressure in the steel pipe is three times higher than in the PVC
pipe, due to the greater rigidity of the steel.
If the change in velocity occurs in more than ten times the critical time T, little or no overpressure
will be generated and the phenomenon may be ignored. In between, if T>2L/c, Ps will not develop
fully, because the reflected negative wave arriving at the valve will compensate for the pressure
rise. In these cases the Allievi formula may compute the maximum overpressure:
⎟⎟
⎠
⎞
⎜⎜
⎝
⎛+±=Δ NNNPPo42
2
(5.25)
where P0 is the hydrostatic pressure due to the head and:
2
0
0 ⎟⎟
⎠
⎞
⎜⎜
⎝
⎛=tgP
LVN (5.26)
where:
V0 = water velocity in m/s
L = total penstock length (m)
P0 = gross hydrostatic pressure (m)
t = closing time (s)
The total pressure experienced by the penstock is P = P0 + ΔP
The next example illustrates the application of the Allievi formula, when the closure time is at least
twice but less than 10 times the critical time.
144
Guide on How to Develop a Small Hydro Site ESHA 2004
Example 5.9
Calculate the wall thickness in the penstock analysed in example 5.6 if the valve closure time is 3
seconds.
Summarising the data,
Gross head: 84.935 m
Rated discharge: 3 m3/s
Internal pipe diameter 1.0 m
Total pipe length: 173 m
Estimating in a first approach at 5 mm wall thickness to compute the wave speed c:
sm
x
xc /7.836
5101.2
1000101.21
101.2
11
9
6
=
×
×+
×=
The closure time is bigger than the critical one (0.41 s) but smaller than 10 times its value, so the
Allievi formula can be applied.
The water velocity in the pipe is:
smx
xV /82.30.1
34
2 ==π
N would be computed for a gross head in the pipe of 84.935 m
070.03935.8481.9
17382.3 2
=⎟⎠
⎞⎜⎝
⎛=xx
xN
and therefore
mmP58.19;65.254
07.007.02
07.0935.84
2
−+=⎟⎟
⎠
⎞
⎜⎜
⎝
⎛+±=Δ
The total pressure would be 84.935+25.65 = 110.585 tf/m2 = 11.06 kN/mm2.
It requires a wall thickness:
mmx
xe 95.4114002
100006.11 =+=
That agrees with the initial estimation and covers the specification for handling the pipes in the field
(tmin=2.5x1+1.2=3.7 mm)
145
Guide on How to Develop a Small Hydro Site ESHA 2004
To compute the air vent pipe diameter:
2
3
/11.01000
5882500 mmkNPc=⎟⎠
⎞⎜⎝
⎛=
And the diameter:
cmd46.2211.0
347.7 ==
The waterhammer problem becomes acute in long pipes, when the open channel is substituted by a
pressure pipe all along the race. For a rigorous approach it is necessary to take into consideration
not only the elasticity of fluid and pipe material, as above, but also the hydraulic losses and the
closure time of the valve. The mathematical approach is cumbersome and requires the use of a
computer program. For interested readers, Chaudry 19, Rich 20, and Streeter and Wylie 21 give
some calculation methods together with a certain number of worked examples.
To determine the minimum pipe thickness required at any point along the penstock two water
hammer hypotheses should be taken into consideration: normal water hammer and emergency water
hammer. Normal water hammer occurs when the turbine shuts down under governor control. Under
these conditions, the overpressure in the penstock can reach 25% of the gross head, in the case of
Pelton turbines, and from 25% to 50% in the case of reaction turbines (depending on the governor
time constants). The turbine manufacturer's advice should be taken into consideration. Emergency
water hammer, caused for example by an obstruction in the needle valve of a Pelton turbine, or a
malfunction of the turbine control system, must be calculated according to the aforementioned
equation.
In steel penstocks, the compounded stresses (static plus transitory) are a function both of the
ultimate tensile and yield strength. In the case of normal water hammer, the combined stress should
be under 60% of the yield strength and 38% of the ultimate tensile strength. In the case of
emergency water hammer, the combined stresses should be under 96% of the yield strength and
61% of the ultimate tensile strength.
Commercial pipes are often rated according to the maximum working pressure under which they are
designed to operate. The pressure rating of a pipe already includes a safety factor, and sometimes
may include an allowance for surge pressures. Safety factors and surge pressure allowances depend
on the standards being used.
146
Guide on How to Develop a Small Hydro Site ESHA 2004
Figure 5.36: Surge tower
If the scheme is liable to surge pressure waves a device to reduce its effects must be considered.
The simplest device is the surge tower, a sort of large tube, connected at its base to the penstock and
open to the atmosphere. The fundamental action of a surge tower is to reduce the length of the
column of water by placing a free water surface closer to the turbine (Figure 5.36). Some authors
consider that the surge tower is unnecessary if the pipe length is inferior to 5 times the gross head. It
is also convenient to take into account the water acceleration constant th in the pipe:
gH
LVth
⋅= (5.27)
where
L = length of penstock (m),
V = flow velocity (m/s) and
H = net head (m).
Figure 5.37: Surge height versus time
147
Guide on How to Develop a Small Hydro Site ESHA 2004
Photo 5.19: Water jet
If th is inferior to 3 seconds the surge tower is unnecessary but if it surpasses 6 seconds, either a
surge tower or another correcting device must be installed to avoid strong oscillations in the turbine
controller.
With the valve open and a steady flow in the penstock, the level of the water in the tower will
correspond to the pressure in the penstock - equivalent to the net head. When by a sudden closure of
the valve the pressure in the penstock rises abruptly, the water in the penstock tends to flow into the
tower, raising the level of the water above the level in the intake. The level in the tower then begins
to fall as the water flows from the tower into the penstock, until a minimum level is reached. The
flow then reverses and the level in the tower rise again and so on. Figure 5.37 shows a graph
plotting the surge height versus time. The maximum height corresponds to the overpressure in the
penstock due to the waterhammer. The throttling introduced by a restricted orifice will reduce the
surge amplitude by 20 to 30 per cent. The time th plays an important role in the design of the turbine
regulation system. In a badly designed system, the governor and the tower surge can interact,
generating speed regulation problems too severe for the governor to cope with.
In instances, when the closure time of the turbine valves must be rapid, a relief valve placed in
parallel with the turbine, such that it opens as the turbine wicket gates close, can be convenient.
This has the effect of slowing down the flow changes in the penstock 5 . Photo 5.19 shows the water
jet ejecting from the open valve.
Saddles, supporting blocks and expansion joints
The saddles are designed to support the weight of the penstock full of water, but not to resist
significant longitudinal forces. The vertical component of the weight to be supported, in kN, has a
value of:
F1=(Wp+Ww)⋅L⋅cosΦ (5.28)
148
Guide on How to Develop a Small Hydro Site ESHA 2004
where
Wp = weight of pipe per metre (kN/m)
Ww = weight of water per metre of pipe (kN/m)
L = length of pipe between mid points of each span (m)
Φ = angle of pipe with horizontal
The design of support rings is based on the elastic theory of thin cylindrical shells. The pipe shell is
subject to beam and hoop stresses, and the loads are transmitted to the support ring by shear. If
penstocks are continuously supported at a number of points, the bending moment at any point of
penstock may be calculated assuming that it is a continuous beam, and using the corresponding
equation. The rings are welded to the pipe shell with two full length fillet welds and are tied
together with diaphragm plates
The span between supports L is determined by the value of the maximum permissible deflection
L/65000. Therefore the maximum length between supports is given by the equation:
()3
440147.061.182 P
DDL−+⋅= (5.29)
where D = internal diameter (m) and P = unit weight of the pipe full of water (kg/m).
5.10 Tailraces
After passing through the turbine the water returns to the river trough a short canal called a tailrace.
Impulse turbines can have relatively high exit velocities, so the tailrace should be designed to
ensure that the powerhouse would not be undermined. Protection with rock riprap or concrete
aprons should be provided between the powerhouse and the stream. The design should also ensure
that during relatively high flows the water in the tailrace does not rise so far that it interferes with
the turbine runner. With a reaction turbine the level of the water in the tailrace influences the
operation of the turbine and more specifically the onset of cavitation. This level also determines the
available net head and in low head systems may have a decisive influence on the economic results.
149
Guide on How to Develop a Small Hydro Site ESHA 2004
BIBLIOGRAPHY
1. http://www.obermeyhydro.com
2. H.C. Huang and C.E. Hita, “Hydraulic Engineering Systems”, Prentice Hall Inc.,
Englewood Cliffs, New Jersey 1987.
3. British Hydrodynamic Research Association, “Proceedings of the Symposium on the Design
and Operation of Siphon Spillways”, London 1975.
4. Allen R. Inversin, “Micro-Hydropower Sourcebook”, NRECA International Foundation,
Washington, D.C.
5. USBR, “Design of Small Canal Structure”, Denver Colorado, 1978a.
6. USBR, “Hydraulic Design of Spillways and Energy Dissipaters”, Washington DC, 1964.
7. T. Moore, “TLC for small hydro: good design means fewer headaches”, HydroReview,
April 1988.
8. T.P. Tung y otros, “Evaluation of Alternative Intake Configuration for Small Hydro”, Actas
de HIDROENERGIA 93. Munich.
9. ASCE, Committee on Intakes, “Guidelines for the Design of Intakes for Hydroelectric
Plants”, 1995.
10. G. Munet y J.M. Compas, “PCH de recuperation d’energie au barrage de “Le Pouzin””,
Actas de HIDROENERGIA 93, Munich.
11. G. Schmausser & G. Hartl, “Rubber seals for steel hydraulic gates”, Water Power & Dam
Construction September 1998.
12. ISO 161-1-1996 “Thermoplastic pipes for conveyance of fluids – Nominal outside diameters
and nominal pressures – Part 1: Metric series.”
13. ISO 3606-1976 “Unplasticized polyvinyl chloride (PVC) pipes. Tolerances on outside
diameters and wall thickness.”
14. ISO 3607-1977 “Polyethylene (PE) pipes. Tolerance on outside diameters and wall
thickness.”
15. ISO 3609-1977 “Polypropylene (PP) pipes. Tolerances on outside diameters and wall
thickness.”
16. ISO 4065-1996 “Thermoplastic pipes – Universal wall thickness table.”
17. H. Chaudry, “Applied Hydraulic Transients”, Van Nostrand Reinhold Company, 1979.
18. J. Parmakian, “Waterhammer Analyses”, Dover Publications, Inc, New York, 1963.
150
Guide on How to Develop a Small Hydro Site ESHA 2004
19. Electrobras (Centrais Eléctricas Brasileiras S.A.) “Manual de Minicentrais Hidrelétricas.”
20. M. Bouvard, “Mobile barrages and intakes on sediment transporting rivers” IAHR
Monograph, AA Balkema, 1984.
21. Sinniger & Hager, “Constructions Hydrauliques”, PPUR, Lausanne, 1989.
1 By Erik Bollaert (LCH-EPFL), Jonas Rundqvist (SERO) and Celso Penche (ESHA)
2 J.L. Brennac. “Les Hauses Hydroplus”, ESHA Info n° 9 Estate 1993
3 USBR “Design of Small Dams” - 3rd ed., Denver, Colorado, 1987.
4 One of these, the SSSS (Serpent Sediment Sluicing System) has been described in detail in the issue 9 -spring/summer
1993- of ESHA Info
5 In the ESHA NEWS issue of spring 1991 there is a description of such a valve.
151