i

ii
FAO IRRIGATION AND DRAINAGE PAPER 43
Water lifting
by
P.L. Fraenkel
Director
Intermediate Technology Power Limited Reading, UK

FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS

Rome, 1986
The designations employed and the presentation
of material in this publication do not imply the
expression of any opinion whatsoever on the part
of the Food and Agriculture Organization of the
United Nations concerning the legal status of any
country, territory, city or area or of its authorities,
or concerning the delimitation of its frontiers or
boundaries.
M-56
ISBN 92-5-102515-0
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical,
photocopying or otherwise, without the prior permission of the copyright owner.
Applications for such permission, with a statement of the purpose and extent of the
reproduction, should be addressed to the Director, Publications Division, Food and
Agriculture Organization of the United Nations Via delle Terme di Caracalla, 00100
Rome, Italy.
FAO 1986

iii

CONTENTS
Preface
Acknowledgements
1. INTRODUCTION
1.1 Scope and purpose of this paper
1.2 The increasing importance of irrigation
1.3 Irrigation and the "Energy Crisis"
1.4 Small-scale irrigation and development
1.5 The choice of water lifting technique
2. WATER LIFTING FOR IRRIGATION
2.1 General principles of water lifting
2.1.1 Definitions of Work, power, energy and efficiency
2.1.2 Efficiency of components: the importance of matching
2.1.3 Irrigation system losses
2.1.4 Flow through channels and pipes
2.1.5 Suction lift: the atmospheric limit
2.1.6 Drawdown and seasonal variations of water level
2.1.7 Review of a complete lift irrigation system
2.1.8 Practical power requirements
2.2 Outline of principles of small-scale irrigation
2.2.1 Irrigation water requirements
2.2.2 Nett irrigation requirement
2.2.3 Gross irrigation requirement
2.2.4 Pumping requirement

iv
3. REVIEW OF PUMPS AND WATER LIFTING DEVICES
3.1 Principles for moving or lifting water
3.2 Taxonomy of water lifts and pumps
3.3 Reciprocating and cyclic direct lift devices
3.3.1 Watering cans, buckets, scoops, bailers and the swing basket
3.3.2 Suspended scoop, gutters, dhones and the counterpoise- -lift or shadoof
3.3.3 Bucket hoists, windlasses, mohtes and water skips
3.4 Rotary direct lift devices
3.4.1 Bucket elevators, Persian wheels and norias
3.4.2 Improved Persian wheels, (zawaffa or jhallar)
3.4.3 Scoop-wheels; sakia, tympanum or tablia
3.5 Reciprocating displacement pumps
3.5.1 Piston or bucket pumps: basic principles
3.5.2 Double-acting piston pumps and plunger pumps
3.5.3 Pistons and valves
3.5.4 Reciprocating pumps and pipelines
3.5.5 Reciprocating borehole pumps
3.5.6 Hydraulically activated borehole pumps
3.5.7 Diaphragm pumps
3.5.8 Semi-rotary pumps
3.5.9 Gas displacement pumps
3.6 Rotary positive displacement pumps
3.6.1 Flexible vane pumps
3.6.2 Progressive cavity (Mono) pumps

v
3.6.3 Archimedean screw and open screw pumps
3.6.4 Coil and spiral pumps
3.6.5 Paddle wheels, treadmills and flashwheels
3.6.6 Water-ladders and Dragon Spine pumps
3.6.7 Chain and washer or Paternoster pumps
3.7 Reciprocating inertia (joggle) pumps
3.7.1 Flap valve pump
3.7.2 Resonant joggle pump
3.8 Rotodynamic pumps
3.8.1Rotodynamic pumps: basic principles
3.8.2 Volute, turbine and regenerative centrifugal pumps
3.8.3 Rotodynamic pump characteristics and impeller types
3.8.4 Axial flow (propeller) pumps
3.8.5 Mixed flow pumps
3.8.6 Centrifugal pumps
3.8.7 Multi-stage and borehole rotodynamic pumps
3.8.8 Self-priming rotodynamic pumps
3.8.9 Self-priming jet pumps
3.9 Air-lift pumps
3.10 Impulse (water hammer) devices
3.11 Gravity devices
3.11.1 Syphons
3.11.2 Qanats and foggara
3.12 Materials for water lifting devices

vi
3.13 Summary review of water lifting devices
4 POWER FOR PUMPING
4.1 Prime-movers as part of a pumping system
4.1.1 Importance of cost-effectiveness
4.1.2 Transmission systems
4.1.3 Fuels and energy storage
4.2 Human power
4.2.1 Human beings as power sources
4.2.2 Traditional water lifting devices
4.2.3 Handpumps
4.2.4 Handpump maintenance
4.3 Animal power
4.3.1 Power capabilities of various species
4.3.2 Food requirements
4.3.3 Coupling animals to water-lifting systems
4.4 Internal combustion engines
4.4.1 Different types of i.e. engine
4.4.2 Efficiency of engine powered pumping systems
4.5 External combustion engines
4.5.1 Steam engines
4.5.2 Stirling engines
4.6 Electrical power
4.6.1 Sources and types of electricity
4.6.2 AC mains power

vii
4.6.3 Electric motors
4.6.4 Electrical safety
4.7 Wind power
4.7.1 Background and State-of-the-Art
4.7.2 Principles of Wind Energy Conversion
4.7.3 The Wind Resource
4.7.4 Windpump Performance Estimation
4.8 Solar power
4.8.1 Background and State-of-the-Art
4.8.2 Principles of solar energy conversion
4.8.3 The solar energy resource
4.8.4 Performance estimation
4.9 Hydro power
4.9.1 Background and State-of-the-Art
4.9.2 Use of turbines for water lifting
4.9.3 The hydraulic ram pump (or hydram)
4.9.4 Water wheels and norias
4.9.5 Novel water powered devices
4.10 Biomass and coal (the non-petroleum fuels)
4.10.1 The availability and distribution of fuels
4.10.2 The use of solid fuels
4.10.3 The use of liquid biomass fuels
4.10.4 Gas from biomass: Biogas
5. THE CHOICE OF PUMPING SYSTEMS

viii
5.1 Financial and economic considerations
5.1.1 Criteria for cost comparison
5.1.2 Calculation of costs and benefits
5.1.3 Relative economics of different options
5.2 Practical considerations
5.2.1 Status or availability of the technology
5.2.2 Capital cost versus recurrent costs
5.2.3 Operational convenience
5.2.4 Skill requirements for installation, operation and maintenance
5.2.5 Durability, reliability and useful life
5.2.6 Potential for local manufacture
5.3 Conclusion
REFERENCES
Back Cover

ix

LIST OF TABLES
1
2
3A.
B.
C.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39

Irrigated areas of the world

Suggested maximum flow velocities, coefficients of roughness and side slopes for lined and
unlined ditches and flumes
Average conveyance efficiency
Average farm ditch efficiency
Average application efficiency
Average intake rates of water in mm/hr for different soils and corresponding stream size
Taxonomy of pumps and water lifts
Specifications of Chinese 'dragon spine' water lifts
Relative merits of materials for pumps
Review of pumps and water lifts
Cost attributes of prime movers
The calorific values of various staple foods
Power capability of human beings
Camparison of various water lifts in Bangladesh
Power and drawbar pull of various animals
Comparison of different petroleum-based engine fuels
Comparison of small I.C. engines
Power in the wind as a function of wind speed in units of power per unit area of wind stream
Variation of air density with altitude
Comparison between different rotor types
Calculation of windpump output using "binned" windspeed data combined with performance
data
Factors affecting windpump system efficiency
Efficiency of hydro-powered systems
Typical sizes and prices of Chinese turbine pumps
Comparison of irrigation costs in China
Some performance data of small turbine pumps
Hydram input capacity
Hydram performance
Power density in water currents as a function of water velocity
Land requirements in Brazil to produce grain for food or for fuel alcohol
Principal crop residues in developing countries
Typical cereal crop residues
Photosynthetic carbon production rates
Relative heat value of various fuels
Potential biomass values of selected crops
Biogas yield from various feedstocks
Quantities of excreta from various species
Principal operating parameters for farm biogas digesters
Sizing example to run an irrigation pump on biogas
Present value factors up to 25 years
Annualization factors up to 25 years

x
40 Analysis of unit water costs for four types of irrigation pumping systems
41 Cost and performance assumptions used for comparison of alternative pumping methods

xi

LIST OF FIGURES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

Typical pump installation

Key components of an irrigation system
The concept of an hydraulic gradient
Determination of head friction losses in straight pipes
Head loss nomogram calculated for rigid PVC pipes using Blasius formula
How pipeline and efficiency vary with flow
Effects of various physical conditions on the elevation of water surfaces in wells
Factors affecting system hydraulic efficiency
Energy flow through typical irrigation system
Hydraulic power requirements to lift water
Relationship between power, head and flow
Relationship between energy, head and daily output
Nomogram for calculating power needs for a given area, depth of irrigation and head
Rate of growth as a function of soil moisture content
Example of a crop coefficient curve for corn planted in mid-May at Cairo, Egypt
Typical curves showing relationship between head, flow, speed and efficiency
The scoop used as a simple hand tool
The swing basket in use
Scoop with a rope support
Dhone as used in Bangladesh
Counterpoise lift
Self-emptying mohte with inclined tow path
Persian wheel
Noria
Zawaffa type Persian wheel
Sakia or tympanum
The fathi is the optimum design of sakia
Basic principles of positive displacement pumps
Hand pump with single-acting, bucket piston
Piston pump for use in borehole
Crank operated piston pump
Different types of reciprocating displacement pumps
Vertical section through a borehole pump
Typical pump valve
Piston pump connected to a pipeline
Three methods for isolating reciprocating pumps from pipelines
Hydraulic shock absorber can serve as an alternative to an air chamber
Schematic cross-section through a borehole
The Vergnet hydraulic foot-pump
Cross-section of a diaphragm pump
Schematic drawing of the IRRI foot-operated diaphragm pump
Commercial portable double acting diaphragm pump
Detail of the New Alchemy Institute (USA) tyre pump

xii
44
45
46
47
48
49
50
51
52
53a.
53b.
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80a.
80b.
80c.
81.
82A.
82B.
83

The Humphrey pump: a liquid piston internal combustion engine and pump combined
A flexible vane pump
The Permaprop tooth pump
Progressivity cavity or 'Mono' pump
An Archimedean screw
Cross-section through an open screw
Hydrostatic pressure pumps
Paddle-wheel or tread-wheel
Water ladder or Chinese 'dragon spine' pump
Chinese Liberation wheel chain and washer pump
A view of a hand-operation Liberation pump
Flap valve pump
Joggle pump
Early types of centrifugal pumps
The relationship between pressure and velocity through both a jet and diffuser
Centrifugal pumps
Typical rotodynamic pump characteristics
Axial flow (or propeller) pump
Surface mounted mixed flow pump
Submerged mixed flow pump
Portable axial flow pump (IRRI)
Typical surface mounted pedestal centrifugal pump
Surface mountd centrifugal pump installation
Below-surface (sump) centrifugal pump installation
Various types of centrifugal pump impellers
Effect of direction of curvature of vanes of centrifugal pump impellers
Combining centrifugal pumps in series or parallels
Multi-stage submersible electric borehole pumps
Schematic of complete electric submersible borehole pumping installation
Direct-coupled air-cooled diesel engine and pump installation with hand-operated
diaphragm pump for priming
Self-priming centrifugal pump
Schematic of a surface-suction jet pump
Borehole jet pump installation
Air lift pump (schematic)
Syphon arrangements
Cross-section through a qanat
Animal-powered Chinese Liberation pump mechanism uses steel components
Typical head and discharge capacities for different types of pumps and water-lifting devices
Discharge
Discharge
Linkages between energy resources and appropriate prime movers
How diminishing returns eventually defeat the benefits of seeking increased efficiency
beyond certain levels
The influence of efficiency on costs
Illustration of how correct speedmatching of a prime mover to a pump can be more

xiii

84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122

important than the efficiency of the prime mover

Two stage speed reduction transmission used in China to connect an electric motor to a
chain and washer pump
The number of people required to provide a specified quantity of water at different lifts
Rope and bag water lift from a dug well
Relative performance of the swing basket and the dhone
Nomograph for calculating hand pump discharge
Rotary drive hand pump
Power pump
The number of oxen required to provide a specified quantity of water at different lifts
Cross-section view of a mohte
Circular mohte utilizing two buckets with flap-valves in the bottom
A bullock-driven Persian wheel of the conventional chain and bucket type
Camel-driven Persian wheel showing over-head drive mechanism
Animal-powered Chinese Liberation pump
The back-axle from a car used as an animal-driven power transmission for an Archimedean
screw pump
Open flywheel low-speed single-cylinder diesel engine
Belt-driven 3 cylinder Lister diesel engine coupled to a centrifugal pump via multiple "V"
belts
Principle components of a small engine pumping system and their efficiencies
Effect of increasing the suction head on the output of a typical engine pump set
Schematic arrangement of a condensing steam engine
Rider-Ericsson hot air pumping engine
The four main types of electric motors
Direct coupled electric motor and centrifugal pump
Electric motor powered, belt-driven piston pump
Wooden indigenous windmill pump for pumping sea water into salt pans on the Island of
Sal, Cape Verde
All-steel 'American' farm wind pump
Gearbox from a typical back-geared 'American' farm windmill
Chinese chain windmill
Thai windpump
'Cretan' type of windmill used on an irrigation project in Southern Ethiopia
2 kw Dunlite wind electricity generator
55 kw Windamatic wind electricity generator
Typical farm windpump installation configurations
IT windpump, made in Kenya as the ' Kijito' and in Pakistan as the 'Tawana'
Savonius Rotor vertical-axis windpump in Ethiopia
Typical Troposkien shaped Darriens vertical-axis wind turbine
Turks and Caicos islands vertical-axis sail rotor
The power (A) and torque (B) of a wind rotor as a function of rotational speed for different
wind speeds
T he power coefficients (above) and the torque coefficients of various types of wind turbine
rotor plotted against tip-speed ratio
The trade-off between starting windspeed and output for differently loaded wind pumps

xiv
123 The operating characteristics of a windpump showing how the power output and matching
efficiency vary with windspeed
124 Typical windpump storm protection method in which rotor is yawed edge-on to the wind
125 Annual mean wind speeds
126 Typical presentation of long term wind data as monthly averages
127 Wind rose
128 Wind velocity-frequency histogram
129 Manufacturers' performance data for the Kenyan-made 'Kijito' windpump range based on the
IT windpump
130 Example of how to calculate the energy output of a windmill
131 Feasible options for solar-powered pumping systems
132 Losses in a typical solar thermodynamic pumping system
133 Construction of a silicon photovoltaic cell
134 Examples of solar pump configuration
135 Performance characteristics of silicon photovoltaic cells
136 Schematic arrangements of a photovoltaic solar pumping system
137 Losses in a typical solar photovoltaic pumping system
138 Solar energy availability compared with crop irrigation water demand
139 World map giving average annual Clearness Index for solar distribution
140 Flow measurement with rectangular weir
141 Method of stream gauging without the need to build a weir
142 Some of the main types of hydro turbine for low, medium and high heads
143 Single stage high lift turbine pump
144 Multi-stage high lift turbine pump
145 Turbine pump fitted with extension drive shaft
146 Typical turbine pump installation
147 Schematic diagram of hydram installation
148 Traditional European hydram design
149 South-east Asian type of hydram
150 Low-cost hydraulic ram using standard pipe fittings
151 Bamboo water wheels, Vietnam
152 Chinese type of small-scale Noria
153 Water wheel driven coil pump
154 Cut-away view showing general arrangement of a Plata pump installation
155 IT river current turbine pump
156 Routes for processing biomass fuels
157 2 kw Ricardo steam engine
158 The three main types of gasifier
159 Small producer gas irrigation pumping system
160 Fixed dome biogas digester
161 Biogas digester with floating gas holder and no water seal
162 Chinese two-wheel tractor running on biogas and being used to pump digester slurry on to
the field
163 Step-by-step procedure for a cost appraisal of a water pumping system
164 Human and animal power: output cost versus scale
165 Diesel and kerosene pumping sets

xv
166
167
168
169

Windpumps at various mean windspeeds

Solar pumps at various mean insolation levels
Hydrams
Mean values of the results of Figs. 164-168 plotted with linear scales on the same axes for
daily hydraulic requirements of 1 000 m3 m
170 Expected range of unit energy costs for three levels of demand

xvi

PREFACE
Since the first rudimentary agriculture started with man cultivating plants near his
dwelling, he has had to carry water to keep his crops alive. He developed simple
systems to raise water, like the shadoof, and later more sophisticated systems, such as
norias, Persian wheels and qanats. The power to move such devices came from man
himself, animals or the force of gravity. Over the years man learned to harness other
sources of energy.
This publication has been prepared to help planners and engineers consider the
potential and application of alternative sources of energy for prime movers. Research
and field application are continuous in many parts of the world, but the degree of
success varies. As new information becomes available, systems developed over recent
decades, and even long before that, may be modified or refined to take advantage of
new ideas and modern technology. It is an old subject, but innovations continue.
Consequently, this publication cannot be regarded as a comprehensive treatise on the
subject of alternatives for water lifting devices, nor can it be considered as the final state
of the art. It does, however, seek to provide enough technical background on promising
systems for possible application in the field, and as a base for further evolution.
This first edition has been prepared with a view to eliciting expert comment and further
contributions which would be considered for incorporation into and updated and possibly
more comprehensive publication. It is hoped that the document will serve as a practical
reference for the guidance not only of experts and counterparts in the field, but also for
government officers and others concerned with the planning, design and operation of
projects using these techniques in FAO's member countries.
Comments and suggestions for improvement of this Irrigation and Drainage Paper will
be welcome and should be addressed to:
Chief
Water Resources, Development and
Management Service
Land and Water Development Division
Food and Agriculture Organization
of the United Nations
Via delle Terme di Caracalla
00100 Rome, Italy

xvii

ACKNOWLEDGEMENTS
The Food and Agriculture Organization of the United Nations would like to express
sincere thanks to the technical author, Mr. Peter L. Fraenkel, and to the editor, Mr. Neil
Burton, Intermediate Technology Group, United Kingdom, for the preparation of the
publication.
The publishers would also like to thank the following especially for the use of material
from which illustrations have been drawn:
B. Y. S. Napel; Grundfos; Khan, H. R. and the Institute of Civil Engineers, London;
Lysen, E. and C W D, The Netherlands; National Aeronautical Laboratory, Bangalore,
India; Smithsonian Institute; and the UN Economic and Social Commission for Asia and
the Pacific.

1. INTRODUCTION
1.1

SCOPE AND PURPOSE OF THIS PAPER

This paper in effect replaces the excellent, but now long out of print booklet, entitled
"Water Lifting Devices for Irrigation", by Aldert Molenaar, published by FAO as long ago
as 1956 [1]. Since that time, little more than one generation ago, the human population
has almost doubled. In the same short period, over twice as much petroleum, our main
source of energy, has been consumed as in the whole of history prior to 1956. But there
has also been a much wider awareness of the constraints which must force changes in
technology.
The primary purpose of this paper is to provide a basis for comparing and choosing
between all present and (near) future options for lifting irrigation water on small and
medium sized . land-holdings (generally in the range 0.25 ha to say 25 ha). Small landholdings in this size range are most numerous in many of the developing countries, and
extension of the use of irrigation in this small farming sector could bring huge benefits in
increased food production and improved economic well-being. It is also hoped that this
paper will be useful to those seeking techniques for lifting water for purposes other than
irrigation.

1.2

THE INCREASING IMPORTANCE OF IRRIGATION

Water has always been a primary human need; probably the first consideration for any
community has always been the need for ready access to it. Irrigation water more
specifically can offer the following important benefits:
i.
ii.
iii.
iv.
v.

vi.

increases land area brought under cultivation

improves crop yield over rain-fed agriculture three or four-fold
allows greater cropping intensity
produces improved economic security for the farmer
reduced drought risk, which in turn allows:
use of high yield seeds
increased use of fertilizer, pesticides and mechanization
control of timing for delivery to market
control of timing for labour demand
allows introduction of more valuable crops

Feeding the rapidly growing human race is an increasingly vital problem. There is no
readily identifiable yield-increasing technology other than the improved seed-waterfertilizer approach. It is expected that in the next two decades about three quarters of all
the increases in the output of basic staples will have to come from yield increases, even
though during the past decade yield increases have only succeeded in supplying half the
increase in output [2]. This is because there is less and less fertile but as yet
uncultivated land available in the more densely populated regions. Irrigation of crops is a
primary route to bringing more land under cultivation and to increasing yields from
existing farm land. Irrigation will therefore be increasingly important in the future both to

2
increase the yield from already cultivated land and also to permit the cultivation of what
is today marginal or unusable land.
Table 1 indicates the irrigated regions of the world (adapted from [3]), and the principal
developing countries where irrigation is currently practised. The majority of the land
brought under irrigation since 1972 is mainly in countries where irrigation is already
generally practised. Not many countries have significant areas of irrigated land and the
two most populous countries, China and India, have about half of the entire world's
irrigated land area within their borders. These two large and crowded countries will have
to increase their irrigated land still further to improve their food production, while other
countries facing similar population pressures on the land will have to do tomorrow what
India and China do today.
Table 1 IRRIGATED AREAS OF THE WORLD (1972)
REGION
& principal irrigation
countries
1. SOUTH & S.E. ASIA

IRRIGATED AREA
million hectares
(Mha)

%of total

132

66

China

74

India

33

Pakistan

12

Indonesia

Taiwan

Thailand

2. NORTH AMERICA

17

3. EUROPE

13

4. MIDDLE EAST

11

Iraq

Iran

Turkey

5. USSR

10

6. AFRICA

Egypt

Sudan

7. CARIBBEAN & CENTRAL AMERICA

Mexico
8. SOUTH AMERICA

5
4
4.5

Argentina

1.2

Chile

1.3

9. AUSTRALASIA
WORLD TOTAL

2
2

1.4

201.9

100

1.3

IRRIGATION AND THE "ENERGY CRISIS"

Water and good land can often be found in juxtaposition, but it is the provision of the
necessary power for pumping which is so often the primary constraint. Human muscle
power or domestic animals have been used since antiquity, and still are being used in
many parts of the world, to lift and distribute water, but as will be explained later, these
techniques are often extremely costly in real terms due to the low productivity that is
achieved. Therefore, mechanized lift irrigation techniques are becoming increasingly
important to meet the enormous predictable future demand.
The area of irrigated land in the world has been estimated to have increased by about
70% in the period 1952 to 1972 [3] and much of this expansion will have been through
the increasing use of engine and mains-electrified pumps during that period of
decreasing fuel and electricity prices (in real terms). However, since then the price of
petroleum, and hence of electricity, has tended to rise, and this has reduced the margin
to be gained by farmers from irrigation, since food prices have generally been prevented
from rising in line with energy costs. Some governments attempt to mitigate this situation
by subsidizing oil and rural electricity for use in agriculture, but many of these
governments are the very ones that can least afford such a policy which exacerbates
balance of payments deficits by encouraging the use of oil.
Despite present short-term fluctuations in oil prices, conventional oil-based enginedriven power sources and mains electricity are expected to continue to increase in the
longer term. There are also major problems associated with maintenance of this kind of
machinery.
There is therefore a considerable incentive in most of the poorer developing countries to
discourage the use of oil, even though there is an equally strong incentive to encourage
the increase of agricultural production, which so often demands pumped irrigation. As a
result, there is an increasing need to find methods for energizing irrigation pumps that
are independent of imported oil or centralized electricity.

1.4

SMALL-SCALE IRRIGATION AND DEVELOPMENT

Intensive irrigation of small-holdings is likely to become increasingly important and

widely used during the next few decades, particularly in the developing countries. This is
because the majority of land-holdings, particularly in Asia and Africa are quite small,
under 2 ha [4]. Even in South America, where the maximum percentage of farmed land
consists of very large land-holdings, the most numerous type of land-holding is under 5
ha.
Studies have shown that small land-holdings are often more productive, in terms of yield
per hectare, than larger units. An. Indian farm management study [5], indicated that
small family run land-holdings are consistently more productive than larger units,
although they are more demanding in terms of labour inputs. A similar survey in Brazil
[5], also showed better land utilization on small land-holdings; however this was
achieved by applying between 5 and 22 times as much labour per hectare compared
with large farms.

4
Small land-holdings also generally achieve better energy ratios than large ones; i.e. the
ratio of energy available in the crop produced, to the energy required to produce it.
Energy ratios for tropical subsistence and semi-subsistence agriculture are in the range
10 to 60 (i.e. the food product has 10 to 60 times as much energy calorific value as the
energy input to grow it) [6]. Mechanised large scale commercial agriculture, which
usually, but not necessarily produces a better financial return, generally has energy
ratios in the range from about 4 to less than 1. Therefore, in a situation where
commercial fuels will get both scarcer and more expensive, there is more scope for
increasing food production through improving the productivity of small labour-intensive
land-holdings which have the potential capability to produce most food from a given
investment in land and energy.
Small-scale irrigation has been shown to offer positive results in alleviating poverty. For
example, the introduction of irrigation can double the labour requirements per hectare of
land [5], and raise the incomes thereby not only of the farmers but also of landless
labourers. The same reference gives examples from actual surveys of the average
percentage increase in income for farmers who practised irrigation compared with those
who did not; examples of increases obtained were 469% in Cameroon, 75% in South
Korea, 90% in Malaysia, and 98% in Uttar Pradesh, India. In the Malaysian case, the
increased income for landless labourers resulting from the introduction of irrigation
averaged 127%.
Finally, there is probably more scope for significantly increasing yields in the small farm
sector through irrigation than with large farms. For example, the average rice yield in the
poorer South and South East Asian countries is typically 2 t/ha, while in Japan, with
sophisticated small-scale irrigation and land management, 6 t/ha is commonly achieved
[7]. The Asian Development Bank has reported that a doubling of rice production per
hectare should be possible in the region within 15 years [7]. Obviously irrigation is not
the only factor necessary to achieve such improvements, but it is perhaps one of the
primary needs.

1.5

THE CHOICE OF WATER LIFTING TECHNIQUE

There are many different types of human and animal powered water lift, some of which
are better than others for different purposes. While the power source or prime-mover so
often attracts most interest, the correct selection of water conveyance and field
distribution system can often have a greater influence on the effectiveness (technically
and economically) of any irrigation system than differences between pumping power
sources. In fact the use of a well-optimized and efficient water distribution system is vital
when considering certain renewable energy systems where the cost is closely related to
the power rating, and therefore a minimum power system needs to be selected.
Before looking for radical new water lifting techniques, there is also much scope for
improving traditional and conventional pumping and water distribution methods; for
example, petroleum-fuelled engines are commonly badly matched to both the pump and
the piping system used for water distribution, which can waste a considerable proportion
of fuel used.
The wide range of options for providing power for pumping water include some
traditional technologies, such as windmills, and some entirely new technologies owing

5
their origins to very recent developments, such as solar photovoltaic powered pumps.
There are also technologies which have been widely and successfully used in just one
area but which remain unknown and unused elsewhere with similar physical conditions;
an example is the hydro-powered turbine pump, which has been used in tens of
thousands solely in China. There are also some interesting new (and some not so new)
options which are currently being experimented with, some of which may become
available for general use in the near future; for example, steam pumps, Stirling engine
pumps, and gasifiers for running internal combustion engines. All of these can produce
pumping power from agricultural residues or other biomass resources, perhaps in future
even from fuel crops, and may become more important as oil becomes scarcer and
more expensive.

2. WATER LIFTING FOR IRRIGATION

2.1
2.1.1

GENERAL PRINCIPLES OF WATER LIFTING

Definitions of Work, Power, Energy and Efficiency

Energy is required, by definition, to do work; the rate at which it is used is defined as

power (see Annex I for detailed definitions of units and their relationships). A specific
amount of work can be done quickly using a lot of power, or slowly using less power, but
in the end the identical amount of energy is required (ignoring "side-effects" like
efficiencies).
The cost of pumping or lifting water, whether in cash or kind, is closely related to the rate
at which power is used (i.e. the energy requirement in a given period). Since there is
often confusion on the meaning of the words "power" and "energy", it is worth also
mentioning that the energy requirement consists of the product of power and time; for
example, a power of say, 5kW expended over a period of say, 6h (hours), represents an
energy consumption of 30kWh (kilowatt-hours). The watt (W), and kilowatt (kW) are the
recommended international units of power, but units such as horsepower (hp) and footpounds per second (ft.lb/s) are still in use in some places. The joule (J) is the
internationally recommended unit of energy; however it is not well known and is a very
small unit, being equivalent to only 1 Ws (watt-second). For practical purposes it is
common to use MJ (megajoules or millions of joules), or in the world outside scientific
laboratories, kWh (kilowatt-hours). lkWh (which is one kilowatt for one hour or about the
power of two horses being worked quite hard for one hour) is equal to 3.6MJ. Fuels of
various kinds have their potency measured in energy terms; for example petroleum fuels
such as kerosene or diesel oil have a gross energy value of about 36MJ/litre, which is
almost exactly 10kWh/litre. Engines can only make effective use of a fraction of this
energy, but the power of an engine will even so be related to the rate at which fuel (or
energy) is consumed.
The hydraulic power required to lift or pump water is a function of both the apparent
vertical height lifted and the flow rate at which water is lifted.

In other words, power needs are related pro rata to the head (height water is lifted) and
the flow rate. In reality, the actual pumping head imposed on a pump, or "gross working
head", will be somewhat greater than the actual vertical distance, or "static head", water
has to be raised. Fig. 1 indicates a typical pump installation, and it can be seen that the
gross pumping head, (which determines the actual power need), consists of the sum of
the friction head, the velocity head and the actual static head (or lift) on both the suction
side of the pump (in the case of a pump that sucks water) and on the delivery side.

35

3. REVIEW OF PUMPS AND WATER LIFTNG

TECHNIQUES
3.1

PRINCIPLES FOR LIFTING AND MOVING WATER

Water may be moved by the application of any one (or any combination) of six different
mechanical principles, which are largely independent, i.e. by:

i.
ii.
iii.
iv.
v.

3.2

direct lift
this involves physically lifting water in a container
displacement
this involves utilizing the fact that water is (effectively)incompressible and can
therefore be "pushed" or displaced
creating a velocity head
when water is propelled to a high speed, the momentum can be used either to
create a flow or to create a pressure
using the buoyancy of a gas
air (or other gas) bubbled through water will lift a proportion of the water
gravity
water flows downward under the influence of gravity

TAXONOMY OF WATER LIFTS AND PUMPS

Families of pumps and lifting/propelling devices may be classified according to which of

the above principles they depend on. Table 5 is an attempt to classify pumps under the
categories given above. It will be seen that most categories sub-divide into the further
classifications "reciprocating/cyclic" and "rotary". The first of these relates to devices that
are cycled through a water-lifting operation (for example a bucket on a rope is lowered
into the water, dipped to make it fill, lifted, emptied and then the cycle is repeated); in
such cases the water output is usually intermittent, or at best pulsating rather than
continuous. Rotary devices were generally developed to allow a greater throughput of
water, and they also are easier to couple to engines or other types of mechanical drive.
Therefore, by definition, a rotary pump will generally operate without any reversal or
cessation of flow, although in some cases the output may appear in spurts or pulsations.
Before considering the differences between the diverse options available for lifting water,
it is worth briefly noting the factors they all have in common. Virtually all water lifting
devices can best be characterized for practical purposes by measuring their output at
different heads and speeds. Normally the performance of a pump is presented on a
graph of head versus flow (an H-Q graph, as in Fig. 16 ) and in most cases curves can
be defined for the relationship between H and Q at different speeds of operation.
Invariably there is a certain head, flow and speed of operation that represents the
optimum efficiency of the device, i.e. where the output is maximized in relation to the
power input. Some devices and pumps are more sensitive to variations in these factors
than others; i.e. some only function well close to a certain design condition of speed,
flow and head, while others can tolerate a wide range of operating conditions with little
loss of efficiency. For example, the centrifugal pump characteristic given in Fig. 16

36
shows an optimum efficiency exceeding 80% is only possible for speeds of about 2000
rpm.

Fig. 16 Typical curves showing relationship between head, flow, speed and efficiency
(example given for a centrifugal pump)
The rest of this Section describes in some detail each of the devices given in the
Taxonomy of Pumps and Water Lifting Devices of Table 5.

3.3
3.3.1

RECIPROCATING AND CYCLIC DIRECT LIFT DEVICES

Watering cans, Buckets, Scoops, Bailers and the Swing-Basket

These are all variations on the theme of the bucket, are hand-held and must be the
earliest artificial methods for lifting and carrying water. The watering can is effectively a
bucket with a built-in sprinkler and represents an efficient, but labour intensive method
for irrigating very small land-holdings. Artisan-made watering cans are quite widely used
in Thailand. Scoops, bailers (Fig. 17) and the swing-basket (Fig. 18) represent methods
of speeding up the process of filling, lifting and emptying a bucket; the latter also uses
two people rather than one and thereby increases the mass of water that can be
scooped in each swing. These are more fully described in Section 4.2 in the context of
using human muscle power, since they have evolved in such a way as to fit the human
prime-mover but to be unsuitable for any kind of mechanization. They are rather
inefficient as water is lifted over lm and allowed to fall back to 0.3-0.5m, which is the
approximate operating head for devices of this kind.
Table 5 TAXONOMY OF PUMPS AND WATER LIFTS
Category and Name

Construction Head Power Output Efficiency Cost Suction

range range
Lift?

Status
for

37
(M)

(W)

Irrigation

I DIRECT LIFT
DEVICES
Reciprocating .'cyclic
Watering car.

>3

Scoops and bailers

>1

**

Swing basket

>1

**

Pivoting gutters
and "dhones"

1-1.5

**

**

**

Counterpoise '.lit or
"Shadoof"

1-4

**

**

**

Rope & bucket and

windlass

5-50

Self-emptying
bucket or "mohte"

3-8

**

***

**

Reciprocating
bucket hoist

100****
500

****

***

****

Continuous bucket
pump:

5-50

**

**

***

**

Persian when', or
"tablia"

3-10

**

***

***

**

Improved Persian
wheel "zawaffa"

3.15

***

****

****

***

Scoop wheels or
"sakia"

>2

**

****

****

****

Waterwheels or
"noria"

>5

**

**

**

2-200 ***

***

*****

****

100500

***

**

****

*****

5-10

**

***

****

***

10100

**

**

*****

****

Semi-rotary pumps

5-10

**

**

**

Gas or vapour
displacement

5-50 ****

****

***

***

or x

Rotary/continuous

II DISPLACEMENT
PUMP
Reciprocating/cyclic
Piston/bucket
pumps
Plunger pumps
Diaphragm pumps
"Petropump"

Rotary/continuous

2& 3

38
Gear and lobe
pumps

10-20

**

***

Flexible vane
pumps

10-20

**

***

***

****

Progressive cavity
(Mono)

10***
100

***

****

****

Archimedean screw

>2m

****

***

***

Open screw pumps

>6m **** *****

****

*****

Coil and spiral

pumps

>6m

**

**

***

***

Flash-wheels &
treadmills

2& 3

>2 m

**

****

**

**

Water-ladders
"Dragon spines"

>2m

**

***

***

***

2&3

320m

***

***

****

****

>3m

***

***

310m

**

**

Inertia and "joggle"

pumps

2& 3

2-4

**

****

**

Flap valve pump

1&2

2-4

**

Resonating joggle
pump

2-10

**

****

****

***

Rebound inertia

2-60

**

****

***

Propeller (axial
flow) pumps

5-3

**** *****

****

****

Mixed flow pumps

2-10 **** *****

****

****

Centrifugal (volute)
pumps

3-20+ ***** *****

****

***

Centrifugal
(turbine) pumps

3-20+ ***** *****

****

****

10-30 ***

***

***

****

2-20

**

**

***

Chain (or rope) and

washer
Peristaltic pump
Porous rope

**

III VELOCITY
PUMPS
Reciprocating/cyclic

Rotary/continuous

Centrifugal
(regenerative)
pump;
Jet pumps (water,
air or stream)
IV BUOYANCY

***

39
PUMPS
Air lift

5-50

10100

** ***

**

****

V IMPULSE PUMPS
Hydraulic ram

**

**

***

***

*****

**

**

*****

VI GRAVITY
DEVICES
Syphons
Qanats or foggara

1, 2 & 3

1.(110)

Construction: 1 Basic 2 Traditional

Very low
3 Industrial
Low-medium
Medium

** Medium***** Yes Yes

* high
**
Possible
High
***** No x
*
?
**
Unlikely
*
x

Fig. 17 The scoop used as a simple hand tool

40
Fig. 18 The swingbasket in use (after T. Schioler [24])

Fig. 19 Scoop with a rope support

3.3.2 Suspended Scoop, Gutters, Phones (or Doons) and the Counterpoise-lift
or Shadoof
The next: stage of technical advance is to support the mass of water being lifted by
mounting the scoop or bucket on a suspended pivoted lever to produce a swinging
scoop (Fig. 19) or a see-sawing gutter or "dhone" (Fig. 20) which also operate through
low lifts (0.5-lir.) at relatively high speed. The water container can be balanced with a
weight; Fig. 21 shows a counterpoise lift, alias water-crane or "shadoof".
If the terrain permits, such as on a sloping river bank, several "shadoofs" can be used to
lift water in stages through a greater height than is possible with one.

41

Fig. 20 Dhone as used in Bangladesh

Fig. 21 Counterpoise lift

42
3.3.3

Bucket hoists, Windlasses, Mohtes and Water Skips

To increase the lift it becomes necessary to introduce a rope to pull the container of
water from the source to a level where it can be tipped into the conveyance channel.
There is therefore a family of devices for pulling up a container of water on a rope. The
simplest form for this is a rope and bucket, which in an improved form has a simple
windlass, i.e. a hand operated winch, to increase the leverage and hence the size of
bucket that can be lifted.
The output of such systems is generally too snail for irrigation, (they tend to be used
mainly for domestic or livestock water supply duties), but by powering the device with
animals, usually oxen, sufficient water can be lifted to irrigate even through heads of 510m. This encouraged the evolution of the "self-emptying bucket", known in India as a
"mohte", (Figs. 22 and 92). These commonly have a bucket made of leather or rubber,
with a hole in its bottom which is held closed by a flap which is pulled tight by a second
rope harnessed to the animals. The number in use today is still in the order of a million
or more, so this device remains of considerable importance in some regions. Mohtes are
discussed in more detail under Section 4.3 dealing with animal power as a prime mover.

Fig. 22 Self-emptying mohte with inclined tow path

43

3.4

ROTARY DIRECT LIFT DEVICES

It generally improves both efficiency and hence productivity if the water lifting element
can move on a steady circular path rather than being cycled or reciprocated. The reason
for this is that the energy input to any water lifting device is usually continuous, so that if
the output is intermittent, unless energy can be stored during the parts of the cycle when
no water is lifted, it is lost. Therefore reciprocating/cyclic devices tend to be less efficient
than rotary devices; this is not a firm rule however, as some reciprocating devices
include means to store energy through the non-productive part of the cycle while some
rotary devices are less efficient for other reasons.

3.4.1

Bucket elevators, Persian wheels and Norias

An. obvious improvement to the simple rope and bucket is to fit numerous small buckets
around the periphery of an endless belt to form a continuous bucket elevator. The
original version of this, which is ancient in origin but still widely used, was known as a
"Persian wheel" (Figs. 23, 94 and 95); the earliest forms consisted of earthenware pots
roped in a chain which is hung over a drive wheel. The water powered "noria" (Figs. 24,
151 and 152), a water wheel with pots, buckets or hollow bamboo containers set around
its rim, is similar in principle except the containers are physically attached to the drive
wheel circumference rather than to an endless belt suspended from it.
The flow with any of these devices is a function of the volume of each bucket and the
speed at which the buckets pass across the top of the wheel and tip their contents into a
trough set inside the wheel to catch the output from the buckets. Therefore, for a given
power source and speed of operation roughly the same number of containers are
needed regardless of head. In other words, a higher head Persian wheel requires the
buckets to be proportionately more spaced out; double the head and you more or less
need to double the spacing.
The Persian wheel has been, and still is, widely used particularly in the north of the
Indian sub-continent and is discussed in more detail under Section 4.3 on Animal Power,
while the noria was widely used in China, S.E. Asia and to some extent in the Middle
East and being normally water powered is discussed in more detail in Section 4.9. Both
devices are tending to be replaced by more modern mechanical water lifting techniques
as they are old-fashioned and low in output. It should be noted that the term "Persian
wheel" is sometimes used to describe other types of animal powered rotary pumps.
Although Persian wheels and norias are mechanically quite efficient, the main source of
loss from these types of device is that some water is spilled from the buckets and also
there is a certain amount of friction drag caused when the buckets scoop up water,
which again reduces efficiency. Also, the Persian wheel is obliged to lift the water at
least 1 m (or more) higher than necessary before discharging it into a trough, which can
significantly increase the pumping head, particularly in the case of low lifts. The
traditional wooden Persian wheels also inevitably need to be quite large in diameter to
accommodate a large enough collection trough to catch most of the water spilling from
the pots; this in turn requires a large well diameter which increases the cost.

44

Fig. 23 Persian wheel

45

Fig. 24 Noria
Some performance figures for animal powered Persian wheels are given in Molenaar;
[1 ].
Height lifted
9m
6m
3m
1.5m

Discharge
8-10m3/h
10-12m3/h
15-17m3/h
20-22m3/h

Depending on the assumption used for the power of the animals in the above examples,
the implication is that the efficiency is in the order of 50% at medium lifts such as 6m and
perhaps marginally better at higher lifts but worse at lower lifts.
The water-powered noria uses the same principle as the Persian wheel and therefore
also needs to be of larger diameter than the pumping head, which either limits it to very
low lift pumping or requires very large, cumbersome and expensive construction. Small,
low-lift norias, used in Thailand and China, are very inexpensive while much larger
norias are used in Vietnam and Syria. Some of the largest in Syria exceed 10m in
diameter, but in relation to their size they tend to be unproductive compared with more

46
modern pumping systems. A fuller description of Vietnamese and other norias is given in
Section 4.9

3.4.2

Improved Persian wheel, (Zawaffa or Jhallar)

Traditional wooden Persian wheels were fitted with earthenware water containers, but a
variety of all-metal improved Persian wheels have been built, some as commercial
products, in China, India, Pakistan and Egypt. Metal Persian wheels can be made
smaller in diameter, which reduces the extra height the water needs to be lifted before it
is tipped out of the containers, and also reduces the well diameter that is necessary.
A modified version of the Persian wheel used in Syria and also in Egypt (where it is
called the zawaffa or jhallar) includes internal buckets within the drive wheel which catch
the water and direct it through holes in the side plate near the hub into a collection
trough; Fig. 25. This reduces both splashing and spillage losses and the extra height
above the collection channel at which the water is tipped. Roberts & Singh [13] gave
figures for a modernised metal Persian wheel, of 153m3/h lifted through 0.75m. This
implies that efficiencies as high as 75% are possible with modernised devices, which are
rather good.

3.4.3

Scoop-wheels; Sakia, Tympanum or Tablia

The scoop-wheel (sakia in Egypt where it originated) has some factors in common with
the noria. Although widely used in Egypt it has failed to become popular anywhere else.
It is however an efficient and effective device; Fig. 26.

47

Fig. 25 Zawaffa type Persian wheel

(side wall shown partially removed)

48

Fig. 26 Sakia or Tympanum (electrically powered in this case)

It consists of a large hollow wheel with scoops around its periphery, which discharges
water at or near its hub rather than from its top. The diameters for sakias range from
about 2-5m; since water discharges at their hub level, the rule of thumb used in Egypt is
that a sakia will lift water through a head of half its diameter less 0.7 m, to allow for the
depth of submergence of the rim in order to scoop up water effectively. Therefore sakias
of diameters from 2-5m will lift water from 0.3-1.8m respectively.
Sakias are now normally made from galvanized sheet steel. Second-hand vehicle roller
bearings are commonly used to support the substantial weight of a sakia and its water
contents. Most sakias are animal powered, but they are increasingly being driven by
either mains electric motors or small engines, via suitable reduction gearing. The normal
operating speed is 2-4 rpm for animal-driven sakias, and 8-15 rpm for motorised or
engine-powered units.
Various different spiral shapes have evolved for the internal baffle plates in the sakia,
and the Hydraulic Research and Experimental Station (HRES) in Egypt tested various
models to try to determine the optimum design. The best shaped designs, such as in
Fig. 27, were measured as being as much as 50% better than the worst. Since some
300 000 sakias are in use in the Nile valley and Delta, optimization of the design could
yield substantial aggregated benefits. An important feature of the the three most
successful sakia variants tested is that the outer compartments divided by the internal
baffle plates discharge first into individual collection chambers which in turn discharge
through holes surrounding the hub instead of having a common discharge orifice as on
the more traditional designs. This prevents water running back into the compartment
adjacent to it.

49

Fig. 27 The Fathi is the optimum design of Sakia

The types of sakia with separate discharge points for each compartment are
distinguished by the generic name "tablia". A further advantage of the tablia type of
device is that the water discharges a few centimeters above the centre shaft and
therefore increases the useful head in relation to the diameter; especially with smaller
machines. Typically a 3m tablia will lift water 1.5m compared with 0.90m for a centredischarge sakia.
Another important conclusion from the tests by HRES was that for wheels operated in
the 2-15rpm range, 6-8 compartments provide the optimum discharge. According to
Molenaar [1], the following performance might be expected from traditional sakia
designs:
diameter of
sakia
5m
4m
3m
2m

head
lifted
1.8m
1.3m
0.9m
0.3m

output
36m3/h
51m3/h
75m3/h
114m3/h

Comparison of the above outputs with those from a traditional persian wheel indicate
that the sakia is somewhat more efficient, although of course it cannot lift water as high
as is possible with a persian wheel.

Fig. 1 Typical pump installation

The friction head consists of a resistance to flow caused by viscosity of the water,
turbulence in the pump or pipes, etc. It can be a considerable source of inefficiency in
badly implementedwater distribution systems, as it is a function which is highly sensitive
to flow rate, and particularly to pipe diameter, etc. This is discussed in more detail in
Section 2.1.4.
The velocity head is the apparent resistance to flow caused by accelerating the water
from rest to a given velocity through the system; any object or material with mass resists
any attempt to change its state of motion so that a force is needed to accelerate it from
rest to its travelling velocity. This force is "felt" by the pump or lifting device as extra
resistance or head. Obviously, the higher the velocity at which water is propelled through
the system, the greater the acceleration required and the greater the velocity head. The
velocity head is proportional to the square of the velocity of the water. Therefore, if the
water is pumped out of the system as a jet, with high velocity (such as is needed for
sprinkler irrigation systems), then the velocity head can represent a sizeable proportion
of the power need and hence of the running costs. But in most cases where water
emerges from a pipe at low velocity, the velocity head is relatively small.

2.1.2

Efficiency of Components; the Importance of Matching

The general principle that:

applies to any water lifting technique, whether it is a centrifugal pump, or a rope with a
bucket on it. The actual power and energy needs are always greater then the hydraulic
energy need, because losses inevitably occur when producing and transmitting power or
energy due to friction. The smaller the friction losses, the higher the quality of a system.
The quality of a system in terms of minimizing losses is defined as its "efficiency":

using energy values in the equation gives the longer-term efficiency, while power values
could be used to define the instantaneous efficiency.
A truly frictionless pumping system would in theory be 100% efficient; i.e. all the energy
applied to it could reappear as hydraulic energy output. However, in the real world there
are always friction losses associated with every mechanical and hydraulic process. Each
component of a pumping system has an efficiency (or by implication, an energy loss)
associated with it; the system efficiency or total efficiency is the product of multiplying
together the efficiencies of all the components. For example, a small electrically driven
centrifugal pump consists of an electric motor, (efficiency typically 85%), a mechanical
transmission (efficiency if direct drive of say 98%), the pump itself (optimum efficiency
say 70%) and the suction and delivery pipe system (say 80% efficient). The overall
system efficiency will be the product of all these component efficiencies. In other words,
the hydraulic power output, measured as (static head) x (flow) (since pipe losses have
been considered as a pipe system efficiency) will in this case be 47%, derived as
follows:

The efficiency of a component is generally not constant. There is usually an operating

condition under which the efficiency is maximized or the losses are minimized as a
fraction of the energy throughput; for example a centrifugal pump always has a certain
speed at a given flow rate and head at which its efficiency is a maximum. Similarly, a
person or draft animal also has a natural speed of operation at which the losses are
minimized and pumping is easiest in relation to output.
Therefore, to obtain a pumping system which has a high overall efficiency depends very
much on combining a chain of components, such as a prime-mover, transmission, pump
and pipes, so that at the planned operating flow rate and static head, the components
are all operating close to their optimum efficiencies - i.e. they are "well matched". A most
important point to consider is that it is common for irrigation systems to perform badly
even when all the components considered individually are potentially efficient, simply
because one or more of them sometimes are forced to operate well away from their
optimum condition for a particular application due to being wrongly matched or sized in
relation to the rest of the system.

2.1.3

Irrigation System Losses

The complete irrigation system consists not only of a water source and water lifting
mechanism and its prime-mover and energy supply, but then there must also be a water
conveyance system to carry the water directly to the field or plots in a controlled manner
according to the crop water requirements. There may also be a field distribution system
to spread the water efficiently within each field. In some cases there could be a water
storage tank to allow finite quantities of water to be supplied by gravity without running
the water lifting mechanism. Fig. 2 indicates the key components of any irrigation
system, and also shows some examples of common options that fulfil the requirements
and which may be used in a variety of combinations.

Fig. 2 Key components of an irrigation system

Most of the irrigation system components influence the hydraulic power requirements.
For example, if pipes are used for distribution, even if they transfer water horizontally,
pipe friction will create an additional resistance "felt" at the pump, which in effect will
require extra power to overcome it. If open channels are used, extra power is still
needed because although the water will flow freely by gravity down the channel, the
input end of the channel needs to be high enough above the field to provide the
necessary slope or hydraulic gradient to cause the water to flow at a sufficient rate. So
the outlet from the pump to the channel needs to be slightly higher than the field level,
thus requiring an increased static head and therefore an increased power demand.
For the same reason the secondary or field distribution system will also create an
additional pumping head, either because of pipe friction, or if sprinklers are used then
extra pressure is needed to propel the jets of water. Even open channels or furrows
imply extra static head because of the need to allow for water to flow downhill.
The power needed is the product of head and flow, and any losses that cause water to
fail to reach the plants also represent a reduction in effective flow from the system. Such
losses therefore add to the power demand and represent a further source of inefficiency.
Typical water losses are due to leakage from the conveyance system before reaching
the field, evaporation and percolation into the soil away from crop roots.
Therefore, in common with the prime-mover and the water lifting device an entire
irrigation system can be sub-divided into stages, each of which has a (variable)
efficiency and a discrete need for power, either through adding to the actual pumping
head or through decreasing the effective flow rate due to losses of water (or both).
Most components have an optimum efficiency. In the case of passive items like pipes or
distribution systems this might be redefined as "cost-effectiveness" rather than
mechanical efficiency. All components need to be chosen so as to be optimized close to
the planned operating condition of the system if the most economical and efficient
system is to be derived. The concept of "cost-effectiveness" is an important one in this
connection, since most irrigation systems are a compromise or trade-off between the
conflicting requirements of minimizing the capital cost of the system and minimizing the
running costs. This point may be illustrated by a comparison between earth channels
and aluminium irrigation pipes as a conveyance; the channels are usually cheap to build
but require regular maintenance, offer more resistance to flow and, depending on the
soil conditions, are prone to lose water by both percolation and evaporation. The pipe is

10
expensive, but usually needs little or no maintenance and involves little or no loss of
water.
Because purchase costs are obvious and running costs (and what causes them) are less
clear, there is a tendency for small farmers to err on the side of minimizing capital costs.
They also do this as they so often lack finance to invest in a better system. This
frequently results in poorer irrigation system efficiencies and reduced returns then may
be possible with a more capital-intensive but better optimized system.

2.1.4

Flow Through Channels and Pipes

The proper design of water conveyance systems is complex, and numerous text-books
deal with this topic in detail. It is therefore only proposed here to provide an outline of the
basic principles so far as they are important to the correct choice and selection of water
lifting system. Useful references on this subject are [3] and [8].
i.Channels
When water is at rest, the water level will always be horizontal; however, if water flows
down an open channel or canal, the water level will slope downwards in the direction of
flow. This slope is called the "hydraulic gradient"; the greater the frictional resistance to
flow the steeper it will be. Hydraulic gradient is usually measured as the ratio of the
vertical drop per given length of channel; eg. lm per 100m is expressed as 1/100 or 0.01.
The rate of flow (Q) that will flow down a channel depends on the cross sectional area of
flow (A) and the mean velocity (v). The relationship between these factors is:

For example, if the cross sectional area is 0.5m2, and the mean velocity is lm/s, then the
rate of flow will be:

Table 2 SUGGESTED MAXIMUM FLOW VELOCITIES, COEFFICIENTS OF

ROUGHNESS AND SIDE SLOPES, FOR LINED AND UNLINED DITCHES AND
FLUMES
Type of surface

Maximum flow velocities

Coefficients of
roughness
(n)

Side slopes or
shape

Metres
per second

Feet
per second

Sand

0.3-0.7

1.0-2.5

0.030-0.040

3:1

Sandy loam

0.5-0.7

1.7-2.5

0.030-0.035

2:1 to 2 1/2:1

Clay loam

0.6-0.9

2.0-3.0

0.030

11/2:l to 2:1

Clays

0.9-1.5

3.0-5.0

0.025-0.030

1:1 to 11/2:1

UNLINED DITCHES

11
Gravel

0.9-1.5

3.0-5.0

0.030-0.035

1:1 to 11:1

Rock

1.2-1.8

4.0-6.0

0.030-0.040

1/4:1 to 1:1

Cast-in-place

1.5-2.5

5.0-7.5

0.014

1:1 to 1 1/2 :1

Precast

1.5-2.0

5.0-7.0

0.018-0.022

11/2:1

Bricks

1.2-1.8

4.0-6.0

0.018-0.022

11/2:1

Concrete

1.2-1.8

4.0-6.0

0.015

1:1 to 11/2:1

Exposed membrane

0.9-1.5

3.0-5.0

0.015

11/2:1 to 1:1

Buried membrane

0.7-1.0

2.5-3.5

0.025-0.030

2:1

0.6-0.9

2.0-3.0

0.025-0.030

21/2:11

1.5-2.0

5.0-7.0

0.0125

Smooth

1.5-2.0

5.0-7.0

0.015

Corrugated

1.2-1.8

4.0-6.0

0.021

Wood

0.9-1.5

3.0-5.0

0.014

LINED DITCHES
Concrete

Asphalt

Plastic
Buried membrane
FLUMES
Concrete
Metal

The mean velocity (v) of water in a channel can be determined with reasonable accuracy
for typical irrigation channels by the Chezy Formula:

where C is the Chezy coefficient which is dependent on the roughness of the surface of
the channel (n), its hydraulic radius (r), (which is the area of cross-section of submerged
channel divided by its wetted submerged perimeter), and the hydraulic gradient (s) of the
channel (measured in unit fall per unit length of channel).
The Chezy coefficient is found from Manning's Formula:

in this formula; K = 1 if metric units are used, or K = 1.486 if feet are used; r is the
previously defined hydraulic radius and n is the Manning's Coefficient of Roughness
appropriate to the material used to construct the channel, examples of which are given in
Table 2. This table is also of interest in that it indicates the recommended side slopes
and maximum flow velocities for a selection of commonly used types of channels,
ranging from earth ditches to concrete, metal or wooden flumes. Combining the above

12
equations gives an expression for the quantity of water that will flow down a channel
under gravity as follows:

where Q will be in m3/s, if A is in m2, r is in metres, and K is 1.

To obtain a greater flow rate, either the channel needs to be large in cross section (and
hence expensive in terms of materials, construction costs and land utilization) or it needs
to have a greater slope. Therefore irrigation channel design always introduces the
classic problem of determining the best trade-off between capital cost or first cost (i.e.
construction cost) and running cost in terms of the extra energy requirement if flow is
obtained by increasing the hydraulic gradient rather than the cross sectional area. The
nature of the terrain also comes into consideration, as channels normally need to follow
the natural slope of the ground if extensive regrading or supporting structures are to be
avoided.
Obviously in reality, the design of a system is complicated by bends, junctions, changes
in section, slope or surface, etc. The reader wishing to study this topic in greater detail
should refer to an appropriate text book on this subject.
A further point to be considered with channels is the likely loss of water between the
point of entry to the channel and the point of discharge caused by seepage through the
channel walls and also by evaporation from the open surface. Any such losses need to
be made up by extra inputs of water, which in turn require extra pumping power (and
energy) in proportion. Seepage losses are of course most significant where the channel
is unlined or has fissures which can lose water, while evaporation only becomes a
problem for small and medium scale irrigation schemes with channels having a large
surface area to depth ratio and low flow rates, particularly under hot and dry conditions;
the greatest losses of this kind occur generally within the field distribution system rather
than in conveying water to the field. The main factors effecting the seepage rate from a
channel or canal are:
i.
ii.
iii.

soil characteristics
depth of water in the channel in relation to the wetted area and the depth of the
groundwater
sediment in water in relation to flow velocity and length of time channel has been
in use

This latter point is important [9], as any channel will leak much more when it has been
allowed to dry out and then refill. Seepage decreases steadily through the season due to
sediment filling the pores and cracks in the soil. Therefore, it is desirable to avoid letting
channels dry out completely to reduce water losses when irrigating on a cyclic basis.
Typical conveyance efficiencies for channels range at best from about 90% (or more)
with a heavy clay surface or a lined channel in continuous use on small to medium land
holdings down to 60-80% in the same situation, but with intermittent use of the channel.

13
In less favourable conditions, such as on a sandy or loamy soil, also with intermittent
use, the conveyance efficiency may typically be 50-60% or less; (i.e. almost half the
water entering the channel failing to arrive at the other end).
Methods for calculating conveyance losses have been derived and are discussed in
detail in specialist references (such as [9]). For example, an approach used by the
Irrigation Department in Egypt [9] uses an empirical formula attributed to Molesworth and
Yennidumia:

where S will be the conveyance loss in m3/s per length L

ii.Pipes
A pipe can operate like a channel with a roof on it; i.e. it can be unpressurized, often with
water not filling it. The advantage of a pipe, however, is that it need not follow the
hydraulic gradient like a channel, since water cannot overflow from it if it dips below the
natural level. In other words, although pipes are more expensive than channels in
relation to their carrying capacity, they generally do not require accurate levelling and
grading and are therefore more cheaply and simply installed. They are of course
essential to convey water to a higher level or across uneven terrain. As with a channel, a
pipe also is subject to a hydraulic gradient which also necessarily becomes steeper if the
flow is increased; in other words a higher head or higher pressure is needed to
overcome the increased resistance to a higher flow. This can be clarified by imagining a
pipeline with vertical tappings of it (as in Fig. 3). When no flow takes place due to the
outlet valve being closed, the water pressure along the pipe will be uniform and the
levels in the vertical tappings will correspond to the head of the supply reservoir. If the
valve is opened so that water starts to flow, then a hydraulic gradient will be introduced
as indicated in the second diagram and the levels in the vertical tappings will relate to
the hydraulic gradient, in becoming progressively lower further along the pipe. The same
applies if a pump is used to push water along a pipe as in the lowest diagram in the
figure. Here the pump needs to overcome a resistance equal to the static head of the
reservoir indicated in the two upper diagrams, which is the pipe friction head. In low lift
applications, as indicated, the pipe friction head can in some cases be as large or larger
then the static head (which in the example is all suction head since the pump is mounted
at the same level as the discharge). The power demand, and hence the energy costs will
generally be directly related to total head for a given flow rate, so that in the example,
friction losses in the pipe could be responsible for about half the energy costs.
Those wishing to undertake scientifically rigorous analysis should consult a specialized
hydraulics text book, (eg. [8] or [9]) but an

14
approximate value of the head loss through a pipe can be gained using the empirical
equation [8],[10]:

Fig. 3The concept of an 'hydraulic gradient'

15

The head loss due to friction is expressed as an "hydraulic gradient", i.e. head per length
of pipe (m per m or ft per ft).
Note: use K = 10 with metric units, (L and D in metres and Q in cubic metres per
second), and K = 4.3 with L and D in feet and Q in cubic feet per second. Values of C
are typically 1.0 for steel, 1.5 for concrete, 0.8 for plastics.
An easy way to estimate pipe friction is to use charts, such as Fig. 4. Reference to this
figure indicates that a flow, for example, of 6 litres/second (95 US gall/minute) through a
pipe of 80mm (3" nominal bore) diameter results in a loss of head per 100m of pipe of
just over 2m., As an alternative method, Fig. 5 gives a nomogram (from reference [10])
for obtaining the head loss, given in this case as m/km, for rigid PVC pipe. These results
must be modified, depending on the type of pipe, by multiplying the result obtained from
the chart by the roughness coefficient of the pipe relative to the material for which the
chart of nomogram was derived; for example, if Fig. 4 is to be used for PVC pipe, the
result must be multiplied by the factor 0.8 (as indicated at the foot of the figure) because
PVC is smoother than iron and typically therefore imposes only 80% as much friction
head.
Account must also be taken of the effects of changes of cross section, bends, valves or
junctions, which all tend to create turbulence which in effect raises the effective friction
head. Ageing of pipes due to growth of either organic matter or corrosion, or both, also
increases the friction head per unit length because it increases the frictional resistance
and it also decreases the available cross section of flow. This is a complex subject and
various formulae are given in text books to allow this effect to be estimated when
calculating head losses in pipes.
The head loss due to friction in a pipeline is approximately related to the mean velocity
and hence the flow rate squared; i.e.:

therefore, the total head felt by a pump will be approximately the sum of the static head,
the friction head and (if the water emerges from the outlet with significant velocity) the
velocity head:

16
i.e. (total head) = (static head) + (friction head) + (velocity head)

17

Fig. 4 Determination of head friction losses in straight pipes

18

Fig. 5 Head loss nomogram calculated for rigid PVC pipes using Blasius formula
Since the velocity of flow is proportional to the flow rate (Q), the above equation can be
re-written:
where

Fig. 6 illustrates the relationship between the total head and the flow rate for a pumped
pipeline, and the pipeline efficiency which can be expressed in energy terms as:

Fig. 6 How pipeline and efficiency vary with flow

2.1.5

Suction Lift: the Atmospheric Limit

Certain types of pump are capable of sucking water from a source; i.e. the pump can be
located above the water level and will literally pull water up by creating a vacuum in the
suction pipe. Drawing water by suction depends on the difference between the
atmospheric pressure on the free surface of the water and the reduced pressure in the
suction pipe developed by the pump. The greater the difference in pressure, the higher

19
the water will rise in the pipe. However, the maximum pressure difference that can be
created is between sea level atmospheric pressure on the free surface and a pure
vacuum, which theoretically will cause a difference of level of water of 10.4m (or 34ft).
However, before a drop in pressure even approaching a pure vacuum can be produced,
the water will start gassing due to release of air held in solution (just like soda water
gasses when released from a pressurized container); if the pressure is reduced further,
the water can boil at ambient temperature. As soon as this happens, the pump loses its
prime and the discharge will cease (due to loss of prime) or at least be severely reduced.
In addition, boiling and gassing within the pump (known as cavitation) can cause
damage if allowed to continue for any length of time.
The suction lifts that can be achieved in practice are therefore much less than 10.4m.
For example, centrifugal pumps, which are prone to cavitation due to the high speed of
the water through the impeller, are generally limited to a suction lift of around 4.5m (15ft)
even at sea level with a short suction pipe. Reciprocating pumps generally impose lower
velocities on the water and can therefore pull a higher suction lift, but again, for practical
applications, this should never normally exceed about 6.5m (21ft) even under cool sea
level conditions with a short suction pipe.
At higher altitudes, or if the water is warmer than normal, the suction lift will be reduced
further. For example, at an altitude of 3 000m (10 000ft) above sea level, due to reduced
atmospheric pressure, the practical suction lift will be reduced by about 3m compared
with sea level, (and proportionately for intermediate altitudes, so that 1 500m above sea
level will reduce suction lift by about 1.5m). Higher water temperatures also cause a
reduction in practical suction head; for example, if the water is at say 30C, (or 86F) the
reduction in suction head compared with water at a more normal 20C will be about 7%.
Extending the length of the suction pipe also reduces the suction head that is
permissible, because pipe friction adds to the suction required; this effect depends on
the pipe diameter, but typically a suction pipe of say 80m length will only function
satisfactorily on half the above suction head.

2.1.6

Draw-down and Seasonal Variations of Water Level

Groundwater and river water levels vary, both seasonally and in some cases due to the
rate of pumping. Such changes in head can significantly influence the power
requirements, and hence the running costs. However, changes in head can also
influence the efficiency with which the system works, and thereby can compound any
extra running costs caused by a head increase. More serious problems can arise,
resulting in total system failure, if for example a surface mounted suction pump is in use,
and the supply water level falls sufficiently to make the suction lift exceed the practical
suction lift limits discussed in the previous section.
Fig. 7 illustrates various effects on the water level of a well in a confined aquifer. The
figure shows that there is a natural ground water level (the water table), which often rises
either side of a river or pond since ground water must flow slightly downhill into the open
water area. The water table tends to develop a greater slope in impermeable soils (due
to higher resistance to flow and greater capillary effects), and is fairly level in porous soil
or sand.

20
If a well is bored to below the water table and water is extracted, the level in the well
tends to drop until the inflow of water flowing "downhill" from the surrounding water table
balances the rate at which water is being extracted. This forms a "cone of depression" of
the water table surrounding the well. The greater the rate of extraction, the greater the
drop in level. The actual drop in level in a given well depends on a number of factors,
including soil permeability and type, and the wetted surface area of well below the water
table (the greater the internal surface of the well the greater the inflow rate that is
possible). Extra inflow can be gained either by increasing the well diameter (in the case
of a hand-dug well) or by deepening it (the best possibility being with a bore-hole).

Fig. 7 Effects of various physical conditions on the elevation of water surfaces in wells
Draw-down usually will increase in proportion to extraction rate. A danger therefore if
large and powerful pumps are used on small wells or boreholes is to draw the water
down to the pump intake level, at which stage the pump goes on "snore" (to use a
commonly used descriptive term). In other words, it draws a mixture of air and water
which in many cases causes it to lose its prime and cease to deliver. As with cavitation,
a "snoring" pump can soon be damaged. But not only the pump is at risk; excessive
extraction rates on boreholes can damage the internal surface below the water table and
cause voids to be formed which then leads to eventual collapse of the bore. Even when

21
a fully lined and screened borehole is used, excessive extraction rates can pull a lot of
silt and other fine material out with the water and block the screen and the natural voids
in the surrounding sub-soil, thereby increasing the draw-down further and putting an
increasing strain on the lowermost part of the bore. Alternatively, with certain sub-soils,
the screen slots can be eroded by particles suspended in the water, when the extraction
rate is too high, allowing larger particles to enter the bore and eventually the possible
collapse of the screen.
Neighbouring wells or boreholes can influence each other if they are close enough for
their respective cones of depression to overlap, as indicated in Fig. 7 . Similarly, the
level of rivers and lakes will often vary seasonally, particularly in most tropical countries
having distinct monsoon type seasons with most rain in just a few months of the year.
The water table level will also be influenced by seasonal rainfall, particularly in proximity
to rivers or lakes with varying levels, (as indicated in Fig. 7).
Therefore, when using boreholes, the pump intake is best located safely below the
lowest likely water level, allowing for seasonal changes and draw-down, but above the
screen in order to avoid producing high water velocities at the screen.
When specifying a mechanized pumping system, it is therefore most important to be
certain of the minimum and maximum levels if a surface water source is to be used, or
when using a well or borehole, the draw-down to be expected at the proposed extraction
rate. A pumping test is necessary to determine the draw-down in wells and boreholes;
this is normally done by extracting water with a portable engine-pump, and measuring
the drop in level at various pumping rates after the level has stablized. In many
countries, boreholes are normally pumped as a matter of routine to test their draw-down
and the information from the pumping test is commonly logged and filed in the official
records and can be referred to later by potential users.

2.1.7

Review of a Complete Lift Irrigation System

The factors that impose a power load on a pump or water lifting device are clearly more
complicated than simply multiplying the static head between the water source and the
field by the flow rate. The load consists mainly of various resistances to flow which when
added together comprise the gross pumping head, but it also is increased by the need to
pump extra water to make up for losses between the water source and the crop.
Fig. 8 summarizes these in a general way, so that the advantages and disadvantages of
different systems discussed later in this paper can be seen in the context of their general
efficiency. The table indicates the various heads and losses which are superimposed
until the water reaches the field; actual field losses are discussed in more detail in
Section 2.2 which follows.
The system hydraulic efficiency can be defined as the ratio of hydraulic energy to raise
the water delivered to the field through the static head, to the hydraulic energy actually
needed for the amount of water drawn by the pump:

22

Where Estat is the hydraulic energy output, and Egross is hydraulic energy actually applied.
Finally, Fig. 9 indicates the energy flow through typical complete irrigation water lifting
and distribution systems and shows the various losses.

Fig. 8 Factors affecting system hydraulic efficiency

23

Fig. 9 Energy flow through typical irrigation system (showing percentage of original
energy flow that is transmitted from each component to the next).

24

2.1.8

Practical Power Requirements

Calculating the power requirement for water lifting is fundamental to determining the type
and size of equipment that should be used, so it is worth detailing the principles for
calculating it. In general the maximum power required will simply be:

where the mass flow is measured in kg/s of water. 1kg of water is equal to 1 litre in
volume, so it is numerically equal to the flow in litres per second; g is the acceleration
due to gravity of 9.81m/s2 (or 32.2 ft/s2). Therefore, for example, 5 litre/sec through 10m
with a system having an overall efficiency of 10% requires:

The daily energy requirement will similarly be:

eg. for 60m3/day lifted through 6m with an average efficiency of 5%

Note: 60m3 = 60 000 litres which in turn has a mass of 60 000kg (= 60 tonne). Also,
since 1kWh = 3.6MJ, we can express the above result in kWh simply by dividing by 3.6:
so

It follows from these relationships that a simple formula can be derived for converting an
hydraulic energy requirement into kWh, as follows:

If the above calculation relates to a gasoline engine pump irrigation system, as it might
with the figures chosen for the example, then we know that as the energy input is
19.6kWh/day and as gasoline typically has an energy content of 32MJ/litre or
8.9kWh/litre, this system will typically require an input of 2.2 litre of gasoline per day.

25
Fig. 10 illustrates the hydraulic power requirement to lift water at a range of pumping
rates appropriate to the small to medium sized land-holdings this publication relates to.
These figures are the hydraulic output power and need to be divided by the pumping
system efficiency to arrive at the input power requirement. For example, if a pump of
50% efficiency is used, then a shaft power of twice the hydraulic power requirement is
needed; (pump efficiencies are discussed in more detail in Chapter 3). The small table
on Fig. 10 indicates the typical hydraulic power output of various prime movers when
working with a 50% efficient water lifting device; i.e. it shows about half the "shaft power"
capability. The ranges as indicated are meant to show "typical" applications; obviously
there are exceptions.

Fig. 10 Hydraulic power requirements to lift water

26

Fig. 11 Relationship between power, head and flow

These power curves, which are hyperbolas, make it difficult to show the entire power
range of possible interest in connection with land-holdings from less than 1ha to 25ha,
even though they cover the flow, head and power range of most general interest. Fig. 11
is a log-log graph of head versus flow, which straightens out the power curves and
allows easier estimation of the hydraulic power requirement for flows up to 100 l/s and
hydraulic powers of up to 16kW.
Fig. 12 is perhaps more generally useful, being a similar log-log graph, but of daily
hydraulic energy requirement to deliver different volumes of water through a range of
heads of up to 32m. The area of land that can be covered, as an example to 8mm depth,
using a given hydraulic energy output over the range of heads is also given.

27

Fig. 12 Relationship between energy, head and daily output (areas that can be irrigated
to a depth of 8mm are shown in parentheses)
Finally, Fig. 13 is a nomograph which allows the entire procedure of calculating power
needs for a given irrigation requirement to be reduced to ruling a few lines so as to arrive
at an answer. The following example of the procedure is indicated and helps to illustrate
the process; starting with the area to be irrigated (in the example 3ha is used), rule a line
vertically upwards until it intersects the diagonal. This point of intersection gives the
required depth of irrigation; 8mm is used in the example but field and distribution losses
are not accounted for in this nomograph, so the irrigation demand used must be the
gross and not the nett requirement. Rule horizontally from the point of intersection,
across the vertical axis (which indicates the daily water requirement in cubic metres per
day - 240 in the example) until the line intersects the diagonal relating to the pumping
head; 10 m head is used in the example. Dropping a vertical line from the point of
intersection gives the hydraulic energy requirement (6.5kWh (hyd)/day). This is
converted to a shaft energy requirement by continuing the line downwards to the
diagonal which corresponds with the expected pumping efficiency; 50% efficiency is
assumed for the example (the actual figure depends on the type of pumping system) and
this gives a shaft power requirement of 13kWh/day when a line is ruled horizontally
through the shaft power axis. The final decision is the time per day which is to be spent
pumping the required quantity of water; 5h is used as the example. Hence, ruling a line
vertically from the point of intersection to the average power axis (which coincides with
the starting axis), shows that a mean power requirement (shaft power) of about 2.6kW is
necessary for the duty chosen in the example. It should be noted that this is mean shaft
power; a significantly higher peak power or rated power may be necessary to achieve
this mean power for the number of hours necessary.
This nomograph readily allows the reader to explore the implications of varying these
parameters; in the example it is perhaps interesting to explore the implications of

28
completing the pumping in say 3h rather than 5h and it is clear that the mean power
requirement then goes up to about 4.25kW.
In some cases it may be useful to work backwards around the nomograph to see what a
power unit of a certain size is capable of doing in terms of areas and depths of irrigation.
The nomograph has been drawn to cover the range from 0 - 10ha, which makes it
difficult to see clearly what the answers are for very small land-holdings of under lha.
However, the nomograph also works if you divide the area scale by 10, in which case it
is also necessary to divide the answer in terms of power needed by 10. In the example,
if we were interested in 0.3ha instead of 3ha, and if the same assumptions are used on
depth of irrigation, pumping head, pump efficiency and hours per day for pumping, the
result will be 0.26kW (or 260W) instead of 2.6kW as indicated. Obviously the daily water
requirement from the top axis will also need to be divided by ten, and in the example will
be 24m3/day. Similarly, it is possible to scale the nomograph up by a factor of ten to look
at the requirements for 10 to 100 ha in exactly the same way. Note that in most real
cases, if the scale is changed, factors like the pump efficiency ought to be changed too.
An efficiency of 50% used in the example is a poorish efficiency for a pump large
enough to deliver 240m3/day, but it is rather a high efficiency for a pump capable of only
one tenth of this daily discharge.

29

Fig. 13 Nomogram for calculating power needs for a given area, depth of irrigation and
head

30

2.2

OUTLINE OF PRINCIPLES OF SMALL SCALE IRRIGATION

2.2.1

Irrigation Water Requirements

The quantity of water needed to irrigate a given land area depends on numerous factors,
the most important being:
-

nature of crop
crop growth cycle
climatic conditions
type and condition of soil
topography
conveyance efficiency
field application efficiency
water quality
effectiveness of water management

Few of these factors remain constant, so that the quantity of water required will vary
from day to day, and particularly from one season to the next. The selection of a small
scale irrigation system needs to take all of the above factors into account.
The crop takes its water from moisture held in the soil in the root zone. The soil therefore
effectively acts as a water storage for the plants, and the soil moisture needs
replenishing before the moisture level falls to what is known as the "Permanent Wilting
Point" where irreversible damage to the crop can occur. The maximum capacity of the
soil for water is when the soil is "saturated", although certain crops do not tolerate waterlogged soil and in any case this can be a wasteful use of water. In all cases there is an
optimum soil moisture level at which plant growth is maximized (see Fig. 12 ). The art of
efficient irrigation is to try to keep the moisture level in the soil as close to the optimum
as possible.
References such as [3], [8], [10] and [11] give a more detailed treatment of this subject.

2.2.2

Nett Irrigation Requirement

The estimation of irrigation water requirements starts with the water needs of the crop.
First the "Reference Crop Evapotranspiration" ET is determined; this is a standardized
rate of evapotranspiration (related to a reference crop of tall green grass completely
shading the ground and not short of water) which provides a base-line and which
depends on climatic factors including pan evaporation data and windspeed. A full
description on the determination of ETo is presented in reference [11]. Because ETo
depends on climatic factors, it varies from month to month, often by a factor of 2 or
more. The evapotranspiration of a particular crop (ETcrop ) will of course be different from
that of the reference crop, and this is determined from the relationship:

31
Kc is a "crop coefficient" which depends on the type of crop, its stage of growth, the
growing season and the prevailing climatic conditions. It can vary typically from around
0.3 during initial growth to around 1.0 (or a bit over 1.0) during the mid-season maximum
rate of growth period; Fig. 13 shows an example. Therefore the actual value of the crop
water requirement, ETcrop usually varies considerably through the growing season.
The actual nett irrigation requirement at any time is the crop evapotranspiration demand,
minus any contributions from rainfall, groundwater or stored moisture in the soil. Since
not all rainfall will reach the plant roots, because a proportion will be lost through run-off,
deep percolation and evaporation, the rainfall is factored to arrive at a figure for
"effective rainfall". Also, some crops require water for soil preparation, particularly for
example, rice, and this need has to be allowed for in addition to the nett irrigation
requirement.
To give an idea of what these translate into in terms of actual water requirements an
approximate "typical" nett irrigation requirement under tropical conditions with a
reasonably efficient irrigation system and good water management is 4 000m3/ha per
crop, but under less favourable conditions as much as 13 000m3/ha per crop can be
needed. This is equal to 400-1300mm of water per crop respectively. Since typical
growing cycles are in the range of 100-150 days in the tropics, the average daily
requirement will therefore be in the 30-130m3/ha range (3-13mm/day). Because the
water demand varies through the growing season, the peak requirement can be more
than double the average, implying that a nett peak output of 50-200m3/ha will generally
be required (which gives an indication of the capacity of pumping system needed for a
given area of field).
Inadequate applications of irrigation water will not generally kill a crop, but are more
likely to result in reduced yield [11]. Conversely, excessive applications of water can also
be counterproductive apart from being a waste of water and pumping energy. Accurate
application is therefore of importance mainly to maximise crop yields and to get the best
efficiency from an irrigation system.

2.2.3

Gross Irrigation Requirement

The output from the water lifting device has to be increased to allow for conveyance and
field losses; this amount is the gross irrigation requirement. Typical conveyance and field
distribution system efficiencies are given in Table 3 [11] [12], from which it can be seen
that conveyance efficiencies fall into the range 65-90% (depending on the type of
system), while "farm ditch efficiency" or field application efficiency will typically be 5590%. Therefore, the overall irrigation system efficiency, after the discharge from the
water lifting device, will be the product of these two; typically 30-80%. This implies a
gross irrigation water requirement at best about 25% greater than the nett requirement
for the crop, and at worst 300% or more.
The previous "typical peak nett irrigation" figures of 50-200m3/day per hectare imply
"peak gross irrigation" requirements of 60-600m3/day; a wide variation due to
compounding so many variable parameters. Clearly there is often much scope for
conservation of pumping energy by improving the water distribution efficiency;
investment in a better conveyance and field distribution system will frequently pay back
faster than investment in improved pumping capacity and will achieve the same result.

32
Certainly costly pumping systems should generally only be considered in conjunction
with efficient conveyance and field distribution techniques. The only real justification for
extravagant water losses is where pumping costs are low and water distribution
equipment is expensive.
Table 3 A AVERAGE CONVEYANCE EFFICIENCY
Irrigation Method

Method of Water Delivery

Irrigated
Area (ha)

Basin for rice cultivation

Continuous supply with no substantial change

in flow

90

3,0005,000

88

< 1,000

70

Surface irrigation (Basin, Rotational supply based on

Border and Furrow
predetermined schedule with effective
management
Rotational supply based on
predetermined schedule with
less effective management

Efficiency
(%)

> 10,000

Rotational supply based on

advance request

< 1,000

65

> 10,000

B AVERAGE FARM DITCH EFFICIENCY

Irrigation Method Method of
Delivery

Soil Type and Ditch

Condition

Block Size (ha)

Efficiency (%)

Basin for rice

Continuous

Unlined: Clay to heavy clay

Lined or piped

up to 3

90

Surface

Rotation or

Unlined: Clay to heavy clay

<20

80

Irrigation

Intermittent

Lined or piped

>20

90

Rotation or

Unlined: Silt clay

<20

60-70

Intermittent

Lined or piped

>20

80

Rotation or

Unlined: Sand, loam

<20

55

Intermittent

Lined or piped

>20

65

C AVERAGE APPLICATION EFFICIENCY

Irrigation Method Method of Delivery
Basin

Continuous

Soil type
Clay

Depth of
Efficiency (%)
Application (mm)
>60

40-50

Heavy clay
Furrow

Intermittent

Light soil

>60

60

Border

Intermittent

Light soil

>60

60

Basin

Intermittent

All soil

>60

60

33
Sprinkler

Intermittent

Sand, loam

<60

70

A summary of the procedure so far outlined to arrive at the gross crop irrigation water
requirement is given in Fig. 14.

Fig. 14 Rate of crop growth as a function of soil moisture content

2.2.4

Pumping Requirement

In order to specify a water lifting system the following basic information is needed:
i.
ii.

the average water demand through the growing season

the peak daily water demand (which generally will occur when the crop
coefficient and rate of plant growth are at their peak)

Having determined the daily application required by the plants, a further consideration is
the "intake rate" as different soil types absorb water at different rates, (see Table 4). Too
rapid a rate of application on some soils can cause flooding and possible loss of water
through run-off. This constraint determines the maximum flow rate that can usefully be
absorbed by the field distribution system. For example, some silty clay soils can only
take about 7 l/sec per hectare, but in contrast sandy soils do not impose a serious
constraint as they can often usefully absorb over 100 l/s per hectare. Obviously lower

34
rates than the maximum are acceptable, although the application efficiency is likely to be
best at a reasonably high rate in most cases, and farmers obviously will prefer not to
take longer than necessary to complete the job.
Taking account of the above constraint on flow rate, it is then possible to calculate how
many hours per day the field will require irrigating, for example by using the nomogram
given in Fig. 15
Table 4 AVERAGE INTAKE RATES OF WATER IN mm/hr FOR DIFFERENT SOILS
AND CORRESPONDING STREAM SIZE 1/sec/ha
Soil Texture

Intake Rate mm/hr

Stream size q 1/sec/ha

Sand

50 (25 to 250)

140

Sandy loam

25 (15 to 75)

70

Loam

12.5 (8 to 20)

35

Clay loam

8 (2.5 to 5)

Silty clay

2.5 (0.03 to 5)

5 (1 to 15)

14

Clay

Fig. 15 Example of a crop coefficient curve for corn planted in mid-May at Cairo, Egypt;
e.g. initial stage is 8.4mm/day with irrigation

50

3.5

RECIPROCATING DISPLACEMENT PUMPS

Water is for most practical purposes incompressible. Consequently, if a close fitting

piston is drawn through a pipe full of water (Fig. 28 A), it will displace water along the
pipe. Similarly, raising a piston in a submerged pipe will draw water up behind it to fill the
vacuum which would otherwise occur (Fig. 28 B); this applies of course only up to a
certain limit of the height water can be pulled by a vacuum, as discussed earlier in
Section 2.1.5. In the first case water is displaced by the piston, but in the second case,
the piston serves to create a vacuum and the water is actually displaced by atmospheric
pressure pressing on its external surface, as indicated in the figure. So water car. be
displaced either by "pushing" or by "pulling", but it can also be "displaced" by a solid
object being pushed into water so that the level around it rises when there is no where
else for the water to go, as indicated in Fig. 28 C.

Fig. 28 Basic principles of positive displacement pumps

The displacement principle can be applied either through reciprocating/cyclic
mechanisms, or continuously via rotary devices.The following sections deal first with
reciprocating displacement pumps and later with rotary displacement pumps.

51
3.5.1

Piston or Bucket Pumps: Basic Principles

The most common and well-known form of displacement pump is the piston or "bucket"
pump, a common example of which is illustrated in Fig. 29. These work by applying both
the principles shown in Fig. 28 A and B; i.e., in the example of Fig. 29, water is sucked
into the cylinder through a check valve on the up-stroke, and the piston valve is held
closed by the weight of water above it (as in Fig. 28 B); simultaneously, the water above
the piston is propelled out of the pump as in Fig. 28 A. On the down-stroke, the lower
check valve is held closed by both its weight and water pressure, while the similar valve
in the piston is forced open as the trapped water is displaced through the piston ready
for the next up-stroke.

Fig. 29 Hand pump with single-acting, bucket piston (piston valve shown open as on the
down-stroke, and foot valve or inlet valve is closed)
Fig. 30 shows a typical traditional design of brass-lined cylinder borehole pump with a
metal foot valve and a metal piston valve; the piston has two leather cup-washer seals
(indicated on the diagram). The outer casing and end fittings are normally cast iron in a
pump of this kind.
There are various basic relationships between the output or discharge rate (Q), piston
diameter (d), stroke or length of piston travel (s), number of strokes per minute (n), and

52
the volumetric efficiency, which is the percentage of the swept volume that is actually
pumped per stroke (nVol):

53

54
Fig. 30 Piston pump for use in borehole

(multiplying by 60 gives Q in cubic metres per hour)

To use this result, if n is known in strokes/minute (or rpm), volumetric efficiency should
be expressed as a decimal fraction (eg. 0.9), with d and s in metres. To convert the
answer to litres per second simply divide the above answer by 3.6, (1 000 litres/m3 - 3
600 seconds per hour).
Volumetric efficiency may be expressed as a decimal fraction and is sometimes called
"Coefficient of Discharge". Another commonly used and related term is "Slippage" (X);
this is the difference between the swept volume per stroke and the output per stroke;
i.e.:

Slippage arises partly because the valves take time to close, so they are often still open
when the piston starts its upward travel, and also because of back leakage past piston or
valve seats. Slippage is therefore normally less than unity, typically 0.1 or 0.2; it tends to
be worse with shorter stroke pumps and with higher heads. With high flow rates at low
heads, the moving water in the pipe can sometimes keep moving upwards with both
valves remaining open for part of the down stroke so that discharge continues for part of
the down stroke as well as on the up stroke. In such situations the "slippage" will be less
than zero (known as "negative slip") i.e. the pump passes more water per stroke then its
actual swept volume. In extreme cases with high speed pumping at low heads, the
slippage can be in the region of -0.5, giving an equivalent volumetric efficiency in the
region of 150%.
The force (F) required to lift the piston, will be the weight of the piston and pump rods
(Wp), plus the weight of the column of water having a cross section equal to the piston
area and a height equal to the head (H). There is also a dynamic load which is the force
needed to accelerate these masses. If the acceleration is small, we can ignore the
dynamic forces, but in many cases the dynamic forces can be large; dynamic loads due
to accelerating water are discussed later in Section 3.5.4. In principle, the dynamic force,
to be added to the static force, will be the summed product of the mass and acceleration
of all moving components (i.e. water, plus piston, plus pump rod). In situations such as
boreholes, where the pump rods are submerged within the rising main, their weight will
be partially offset by an upward buoyancy force due to displacing water (see Fig. 38),
which also should come into the equation. However this is not a text book on pump
theory, so if we ignore possible "second order effects" such as buoyancy of pump rods
and the dynamic forces, for simplicity, we arrive at a figure for the primary "static" force
necessary to initiate movement of the piston;

if Wp is in Newtons (or kilograms X 9.81), A is in m2, p the density of water is 1,000 kg/m3,
g is 9.81 m/s2 and H is in metres, then F will be in Newtons.

55
If the pump rod is connected to a lever, as in a hand pump (Fig. 29), then the downward
force required to lift the pump rod will be reduced by the ratio of the leverage, however,
the distance the hand of the operator will have to move, compared with the stroke, will
be proportionately increased.
The pump rod can also be connected to a flywheel via a crank (as in Figs. 31 and 89);
this is the coventional way of mechanizing a reciprocating piston pump. The torque (or
rotational couple) needed to make the crank or flywheel turn will vary depending on the
position of the crank. When the piston is at the bottom of its travel (bottom dead centre
or b.d.c), marked as "a" on the figure, the torque will be zero as the pump rod pull is
acting at right angles to the direction of movement of the crank and simply hangs on the
crank; as it rotates to the horizontal position marked "b", the torque will increase
sinusoidally to a maximum value of Fs/2 (force F times the leverage which is s/2); the
resisting force will then decrease sinusoidally to zero at top dead centre (t.d.c.) marked
"c". Beyond t.d.c. the weight of the pump rod and piston will actually help to pull the
crank around and while the piston is moving down the water imposes no significant force
on it other than friction. If, for convenience, we assume the weight of the piston and
pump rod is more or less cancelled out by friction and dynamic effects, the torque is
effectively zero for the half cycle from t.d.c. at "c" through "d" to "a" at b.d.c. where the
cycle restarts. The small graph alongside the sketch in Fig. 31 illustrates the variation of
torque with crank, position through two complete revolutions; anyone who has turned a
direct driven hand pump via a crank and hand-wheel will have experienced how the load
builds up in this way for a quarter cycle and falls back to (near) zero for the next quarter
cycle.

56

Fig. 31Crank operated piston pump (valve details not indicated)

57
If the crank has a flywheel attached to it, as it normally will, then the momentum of the
flywheel will smooth out these cyclic fluctuations by slowing down very slightly (too little
to be noticeable) during the "a-b-c" part of the cycle and speeding up during the "c-d-a"
part, as illustrated by the broken line following the first revolution in the graph in Fig. 31.
If the flywheel is large, then it will smooth the fluctuations in cyclic torque to an almost
steady level approximating to the mean value of the notched curve in the figure. The
mean value of half a sine wave, to which this curve approximates, is the peak value
divided by pi (-), (where = 3.142). Therefore:

Therefore, the torque necessary to turn a crank through its first revolution will be about K
(i.e. approximately three) times greater than the mean torque which is needed to
maintain steady running. Many prime movers cannot readily produce three times the
torque needed for running in order to start a pump, and even with those that can, there is
usually a price to pay to achieve this requirement. This is one reason why centrifugal
pumps rather than piston pumps are more commonly used with engines and electric
motors, as they actually need less torque to start them than to run them.
Power can be calculated as the product of speed and torque. Hence if the rotational
speed is n ( rpm) then a measure of power can be obtained as:

3.5.2 Double-acting Piston Pumps and Plunger Pumps

A single-acting pump only discharges water when the piston rises (if mounted vertically)
and the down stroke is utilised simply to displace more water into the working space
ready for the next stroke. It is possible to arrange things so that while one side of the
piston displaces water to discharge it, the other induces more water, so that discharge
takes place on both the up and the down stroke. Such pumps are known as "doubleacting" pumps and are significantly more productive for their size than single-acting
ones.
Fig. 32, diagrams A to H, illustrate various reciprocating displacement pump principles,
and shows various single and double-acting configurations:
Diagram A in the figure shows a conventional single acting pitcher pump, as in Fig. 29.
Diagram B shows how, if the water needs to be delivered to a level higher than the point
at which the pump rod enters the pump, a sealed "lid" is needed to prevent leakage. The
seal is usually achieved by fitting a stuffing box through which the pump rod passes.
Diagram C is a piston pump in which the cylinder carries no valve; in principle it is similar
to the pump in diagram B. It has the advantage that no stuffing box is needed, but it has

58
the disadvantage that the discharge stroke requires the piston to be pushed rather than
pulled, which needs a much stiffer pump rod in order to avoid buckling. There can also
be problems with this kind of pump because the direction of motion of the water is
reversed. This action can only be performed slowly, especially at low heads with large
volumes of water per stroke, or sudden stopping and starting of the mass of the water
will cause "water hammer" (much as when a tap or valve is suddenly closed and causes
a "bang" in the pipes). Water hammer can damage or even burst a water system and is
of course to be avoided. Pumps of this kind are therefore unusual today.
Diagram D is a similar pump to C, except it is a plunger pump rather than a piston pump.
Here a solid plunger, sealed with a large diameter stuffing box or gland packing
displaces the water; this is a more robust pump than C. The main justification for using
plunger pumps is that the piston or plunger seals are less prone to wear through
abrasive solids in the water, and also, where very high pressures and low flow rates are
needed, a smaller plunger or closed piston is possible because a through valve is not
needed through its centre. Therefore the main use today for pumps of this kind is for
pumping small volumes of water up to very high pressures or heads, such as for reverse
osmosis desalination plants, where pressures of the order of 300m of water are required.
The plunger pump also suffers from the flow reversal problem of pump C, but this is less
serious where small flow rates at high heads are involved. However the diagram shows
a pump with air chambers below the inlet valve and above the delivery valve which are
necessary to cushion the shocks caused by sudden reversal of flow direction, as
explained in more detail in Section 3.5.4. Plunger pumps offered a manufacturing
advantage in the past, in that it was sometimes easier to produce a good external finish
on a plunger than inside a cylinder, but modern pump production techniques have
reduced this advantage.
Diagram E indicates one of several methods to obtain large valve openings; this is
important for low head pumps where high flow rates are required and it is necessary to
minimise the hydraulic losses caused by forcing a lot of water through a small opening.
Here the piston is external to a seal rather than internal to a cylinder; another way of
looking at it is that a cylinder is being pulled up and down over a fixed piston.
Diagram F shows a pump that is similar in principle to that in C, but double-acting. Here
when the piston is on the upstroke it induces water into the lower chamber and
discharges from the upper, while on the downstroke water is induced into the upper
chamber and discharged from the lower. The same points that apply to pump C apply to
this design.

59

Fig. 32 Different types of reciprocating displacement pumps

Pump G is known as a differential pump and is also double-acting; the pump rod is of a
large diameter where it enters the upper chamber and if it is sized so that its cross-

60
sectional area is exactly half -he cross sectional area of the chamber, it will therefore
displace half the volume of the chamber on the downstroke (the principle being as for
plunger pump D), but on the up stroke, the other half of the volume will be discharged by
the upward movement of the piston, as for bucket pump B.
Pump H applies a similar differential double-acting principle to pump G, but uses a
closed piston as for pump C, and apart from being a more complicated arrangement, will
be more prone to water hammer due to the flow reversal involved in both chambers.
Of the above configurations, only A, B, E and G are generally appropriate for irrigation
pumping duties.
Although more complicated than single-acting pumps, double-acting pumps have
considerably smoother outputs, and a smoother torque requirement. They were
therefore widely used in conjunction with reciprocating steam engines, but the move to
mechanize with electric motors or high-speed diesel engines has made them rarely used
today, mainly because they have the following disadvantages:
i.
ii.
iii.

they are larger and more complicated and hence more expensive
they usually involve flow reversal which can cause waterhammer so some need
to be run quite slowly or to incorporate air chambers
the drive requires that the pump rod is pushed as well as pulled (at least with the
configurations shown), so there must be no back lash or free-travel in the
transmission (or hammering and wear and tear will result); also the pump rods
must be capable of taking the compressive load on the down- stroke without
buckling.

All this involves heavier more precisely engineered components in the drive train, which
generally adds to the cost. Therefore, the trend in recent years has been to restrict the
use of piston pumps to simple single-acting bucket pumps where their essential
simplicity and low cost provide the justification for choosing them.

3.5.3

Pistons and Valves

Fig. 33 shows details of a typical borehole pump's piston and foot-valve. The simplest
type of seal, commonly used in single-acting bucket pumps, is the leather cup washer as
shown. Suitable grades of leather, commonly impregnated with "neatsfoot oil" boiled
from the hooves of cattle, will function for surprisingly long periods (several years) in
smooth drawn brass cylinders, or in smooth PVC. With the high cost of servicing deep
boreholes, it is worth paying a premium to get a good life out of pump seals. Various
synthetic leather "compound" materials based on plastics have been used for seals;
these are often more consistent in their performance than leather and will often have
bettor wear resistant characteristics.

61

Fig. 33 Vertical section through a borehole pump (with extractable foot valve)

62
Pistons can also be sealed by packings or piston rings. Packings generally need to be
compressed by a certain optimum amount; too much compression, and friction and wear
become excessive, while too little and there is excessive leakage around the piston.
Graphited asbestos is the traditional industrial packing material but graphited PTFE
(poly-tetrafluoroethylene) is now becoming available and offers superior sealing and
wear characteristics. Similar packings are used around pump rod seals for reciprocating
pumps and also sometimes for shaft seals on various rotary pumps. In all these
applications they are generally compressed by screwing down a fitting against them in
order to provide the sealing pressure; this needs regular adjustment.
All reciprocating pumps (and some rotary pumps) depend on check valves (sometimes
Known as non-return valves) which as their name suggests, allow water to flow one way
but not the other. There are basically three categories of check valve:
i.
ii.
iii.

Flexible valves that normally lie in a closed position, but open by being bent or
deformed when pressure is applied
Hinged valves that open like a door
Straight lift valves which rise vertically and evenly from their seats

Fig. 34 shows a typical check valve design of the kind that may be used in a
reciprocating pump. Valves are invariably opened by the difference in water pressure
across them created by piston movement, but they may be closed again either by their
own weight usually in combination with the weight of water trying to flow backwards. In
some cases closing is assisted by a light spring, (as shown in the figure). Valve springs
are usually made of bronze to avoid corrosion problems, but alternatively, valves may be
made from an elastic material like rubber.

63

Fig. 34 Typical pump valve (shown open in cross-section)

The main requirements of valves are a good seal when closed combined with lack of
resistance to flow when they are open, and rapid opening and closing while achieving
good durability. Usually rubber or alternatively precision ground metal mating surfaces
are necessary to ensure there are no leakage gaps when the valve is closed. Effective
sealing is particularly important with foot valves. To offer as little resistance to flow as
possible when open (and to be capable of opening and closing quickly) demands large
port areas with as few changes of flow direction as possible and sharp surfaces that can
cause turbulence minimized. A rule of thumb sometimes used is that the suction valve
should have a port area of at least two-thirds of the piston area, while the discharge
valve (or piston valve) should have an area of at least half the piston area, [14]. Finally,
rapid opening and closing (to minimise back-leakage) depends on light weight for the

64
valve, combined with a short travel; light weight can demand a trade-off with robustness
and durability, while short travel conflicts to some extent with the need for a good
unobstructed passage for water when the valve is open. Therefore all valves are a
compromise in achieving conflicting requirements.
Finally, valves are the main mechanical components of a pump and so are subject to
wear and tear. It is therefore desirable to use pumps in which the valves and their seats
can readily (and inexpensively) be replaced when necessary.

3.5.4

Reciprocating Pumps and Pipelines

A reciprocating pump moves water in a non-continuous manner, so the water is

constantly accelerated and decelerated by the movement of the piston. Very large forces
can be created if long pipelines containing a large mass of water are directly connected
to a reciprocating pump. This is because the pump piston tries to force the water in the
pipeline to move rapidly from rest to speed, and then back to rest; since water is
incompressible it will try and follow the motion of the piston. Therefore reciprocating
pumps need to be isolated from water in long pipelines by methods described shortly, in
order to cushion the water in the pipeline from the motion of the piston.
To gain an appreciation of the damage that can happen and the consequent importance
of isolating reciprocating pumps from pipelines, it is worth running through some simple
calculations to quantify the forces concerned. Fig. 35 illustrates a simple piston pump
(diameter D by stroke S) with a long length (L) of delivery pipeline (diameter d). Newton's
Laws of Motion, state that a force is necessary to accelerate a mass from one velocity to
another (or from rest); this force is numerically equal to the product of the mass and the
acceleration at any moment in time. If we assume the pump piston is driven sinusoidally,
(as it would be if driven by a steadily revolving crank, having a long connecting rod in
relation to the stroke), then the maximum acceleration of the piston (and hence of water
being propelled by it) will be Amax, where:

where w is the angular velocity of the driving crank in radians/second; (2.radians = 1

revolution or 360).
The acceleration of the water in the pump will be magnified for the water in the pipe,
since if the pipe cross sectional area is smaller than that of the pump, a higher velocity
will be needed to pass the same flow of water, and hence a proportionately higher
acceleration to reach the higher velocity. The magnification will be proportional to the
ratio of pump cross sectional area to pipe cross sectional area, which in turn is
proportional to the ratio of their diameters squared; hence the acceleration of water in
the pipeline will be:

65

Fig. 35 Piston pump connected to a pipeline

The force necessary to achieve this acceleration for the water in the pipeline will be
equal to the mass of water flowing multiplied by its acceleration. The mass is the volume
of water in the pipe times its density; hence the accelerating force is:

the maximum force will occur at the moment of maximum acceleration. So, assuming
sinusoidal motion of the piston and the flow, this is numerically such that:

Hence, if m is the mass of water in the pipeline:

66

For example: suppose we consider a pump of 100mm diameter, connected to a pipeline

of 50mm diameter, density of water 1 000kg/m3, a crank speed of60rpm (which is 1 rev/s
or (1 x 2 )rad/s) and a stroke of 300mm (0.3m), then:

In other words, there is a peak reaction force due to the water in the pipe of 46.5N per
metre of pipe, due to sinusoidal acceleration under the conditions described. A 100m
pipeline will therefore, with conditions as specified, experience peak forces of 4 650N
while a 1 km pipeline will experience 46 500N - and so will the pump in both cases; (46
500N is the equivalent of about 4.5 tonne, or 10 000 lb).
Since this force is proportional to the square of the pump speed, doubling the pump
speed to 120 strokes per minute will impose four times the acceleration and hence four
times the force. Even such modest pump speeds (by rotary pump standards) as say a
few hundred rpm will therefore impose impossible accelerative forces on the water in the
pipe line unless it is isolated or cushioned from the motion of the pump piston.
In reality, the situation is not quite as bad, as even steel pipes are flexible and will
expand slightly to take the shocks. But in some respects it can also be worse, because
when valves slam shut, very brief but large shock accelerations can be applied to the
water; these are known as "water hammer" because of the hammering noise when this
happens. Water hammer shocks can damage both a pump, and its prime mover, as well
as possibly causing burst pipes or other problems.
The same problem can occur on suction lines as well as on delivery lines, except that a
sudden drop of pressure caused by high flow velocities can cause "cavitation" where
bubbles of water vapour and dissolved air suddenly form. When the pressure increases
again slightly, the bubbles can then suddenly catch up and the bubble will implode
violently causing water hammer.
When the pump outlet is set close to or above the pipeline discharge level, there is no
great problem because the pipeline can be de-coupled from the pump by feeding into a
small tank which can then gravity feed the pipeline steadily; see Fig. 36 B. Alternatively,
a riser open to the atmosphere in the pipeline near to or over the pump can achieve the
same effect (as in Fig. 36 A); because the pump rod can go down the riser it neatly
avoids the need for a seal or stuffing box.
Where the pump delivers into a pipeline which discharges at a significantly higher level,
it is generally not practical to have a riser open to the atmosphere at or near the pump,
since it obviously would have to extend to a height above the level of the discharge. The
solution generally applied in all such cases where more then a few metres of suction or
delivery line are connected to a reciprocating pump is to place an air chamber or other

67
form of hydraulic shock absorber between the pump and the pipeline (Fig 36 C) and
always as close to the pump as possible to minimize the mass of water that is forced to
follow the accelerations of the piston. Then when water from the pump seeks to travel
faster than the water in the pipeline it will by preference flow into the air chamber and
compress the air inside it. When the piston slows so that the water in the pipeline is
travelling faster than that from the pump, the extra water can flow out of the airchamber
due to a slight drop in pressure in the pipeline and "fill the gap". In otherwords, an air
chamber serves to smooth the flow by absorbing "peaks" in a reciprocating output and
then filling the "troughs" that follow the peaks.

68

Fig. 36 Three methods for isolating reciprocating pumps from pipelines

A. Open riser (also avoids need for pump rod seal) but only suitable where riser can be

69
above pipeline discharge level
B. Gravity pipeline from small tank near pump
C. Air chamber; necessary when discharge is significantly higher than pump
Air chambers are generally vital on long or on large capacity pipelines when using a
reciprocating pump (eg. Fig. 106), but they are well worth their extra cost not only in
reducing wear and tear, but also the peak velocity of water in the pipeline will be reduced
which in turn reduces pipe friction; this reduces the power requirement and saves
pumping energy.
A special problem with air chambers on delivery lines is that the air in the chamber can
gradually dissolve in the water and be carried away, until there is no air left and water
hammer then occurs. Therefore simple air chambers usually require regular draining to
replenish their air by opening a drain plug and an air bleed screw simultaneously,
obviously when pumping is not taking place. Suction line air chambers are usually
replenished by air coming out of solution from the water, although when air-free
groundwater is being drawn, a small air snifting valve may be needed to deliberately leak
in a minute flow of air and prevent the chamber losing its air volume. Industrial air
chambers sometimes contain a sealed rubber bag which will retain its air indefinitely;
these are recommended in situations where regular attention cannot be guaranteed.
Spring loaded hydraulic shock absorbers have also been used in the past instead of air
chambers (Fig. 37). Another useful alternative is to pump into a rubber or other type of
flexible pipe which is less prone to water hammer than a steel pipeline. Care must be
taken with rigid plastic pipelines as they can easily be broken by water hammer,
especially in cold weather when they tend to be more brittle.
A typical size for an air chamber will be around twice the swept volume of the pump,
however it will need to be larger to cater for more severe flow irregularities or long
delivery lines generally.

3.5.5

Reciprocating Borehole Pumps

When groundwater is deep, or the ground is hard to dig, boreholes are generally quicker
and less expensive to construct than dug wells. Most boreholes are lined with either a
100mm or 150mm (4in or 6in) steel rising main, so the pump must be small enough in
diameter to fit down it. Fig. 38 illustrates a typical borehole with a piston pump in it.
Fig. 33 shows a typical borehole piston pump; because bore hole diameters need to be
kept small (large boreholes are very expensive to drill) the best way to obtain a larger
output from a borehole pump is to increase its stroke. The speed of operation of
borehole pumps is usually restricted to about 30 strokes a minute, although a few
operate at up to 50 strokes per minute. Higher pumping speeds tend to buckle the long
train of pump rods by not giving the pump rods and piston sufficient time to fall back on
the down stroke under gravity. Obviously, with any depth of borehole, it is important to
avoid compressing the pump rods, or they may buckle and jam against the sides of the
rising main.
The normal method of installing or removing a pump from a borehole is to raise the
entire rising main using block and tackle or a crane, section by section. The procedure is
to lift a complete section clear of the well; clamp the next section below it and lower the

70
whole assembly so it hangs on the clamp. The top 6m (20ft) section standing clear of the
well can then be unscrewed and the process is then repeated. Since this is a long and
expensive process, it is important to minimize the number of times it is necessary to do
this in the life of a borehole. It is common therefore to install pumps with a piston
diameter slightly smaller than the diameter of the rising main, especially on deep
boreholes. This allows the piston to be pulled up through the rising main by lifting only
the pump rods. In fact, Fig. 33 shows an extractable borehole pump of this kind; this
pumps is designed so that the footvalve can also be removed without removing the
cylinder and rising main. One method, as shown in the figure, is where a loop on top of
the footvalve can be fished for, caught, and used to pull it out. Alternatively, with some
other types of footvalves, the pump rod is disconnected from the drive at the surface,
and the piston can then be lowered onto the footvalve, which has a threaded spigot on
top of it which will screw into the base of the piston if the pump rods are twisted. Then
when the pump rods are raised, both footvalve and piston can be drawn up to the
surface together. The importance of this is that piston and footvalve seals are the
primary wearing parts needing regular replacement. Therefore, the ease with which a
borehole pump can be overhauled is an important consideration for all but the shallowest
of boreholes.

71

Fig. 37 Hydraulic shock absorber can serve as an alternative to an air chamber,

especially with very high pressure pipelines

72

Fig. 38 Schematic cross-section through a borehole

Pump rods can be improvised from galvanized water pipe joined with standard pipe
sockets, but properly purpose-designed pump rods are commercially available, and
although more expensive, are more easily connected or disconnected and are also less
likely to break or to come unscrewed. A broken pump rod at best is difficult to recover

73
and at worst can mean the loss of an expensive borehole. Where wood of suitable (very
high) quality is available, plus appropriate couplings, wooden pump rods have the
advantage that they do not impose any extra weight, which hinders initial start up of a
reciprocating pump in a borehole, because they float in the water of the rising main.
The main attraction of using reciprocating pumps driven by tensile pump rods in
boreholes is that they are essentially simple and the better commercial products have
become highly reliable in operation. Typically, pump rods in tension can pull well over a
tonne, allowing the use of this type of system on boreholes as deep as 300m (1000ft);
this is the standard system used for farm windpumps and quite a few of the more
traditional diesel powered systems use a gearbox and pitman mechanism to drive a
reciprocating borehole pump in exactly the same way.
A common problem with boreholes is that they are often not truly vertical and sometimes
they are curved, making any pump rods scrape the rising main and eventually wear a
hole in it. Sometimes boreholes start off straight and earth movements cause
subsequent distortion of the bore. Also, removing a pump to change its seals, with its
long train of pumprods is a slow and expensive operation. Therefore there has been an
incentive to find other methods of driving borehole pumps than by pump rods.

3.5.6

Hydraulically activated Borehole Pumps

An alternative method for powering reciprocating borehole pumps to pump rods, is with
an hydraulic transmission. Here water under pressure is used to push more water to the
surface.
An example is the Vergnet "Hydro-pompe", illustrated in Fig. 39, which has been quite
successfully used for small water supplies, particularly in the West African Sahel region,
but the output is probably on the low side for most irrigation applications. It works by a
foot pedal which is mounted on a pilot piston to force water down a flexible pipe. The
pump cylinder, which is located below water level in the well or borehole, has a
conventional "suction valve" or foot valve and strainer, a pump chamber, and at the top
of the chamber a discharge check valve with a discharge pipe leading up to the surface
above it. Instead of a piston to displace water in the cylinder there is what Vergnet call a
"Diaphragmatic Hose", this is in fact an elastic bladder which expands to displace water
from the pump chamber when the foot pedal is depressed and pumps water into it. A
similar hydraulic transmission, called the "Hidromite" system, for driving a reciprocating
borehole pump, was developed in Australia mainly for use with windmills. Here the
windmill, (or other prime-mover), drives a master piston located at the surface. This is a
double-acting water pump connected by two hydraulic transmission pipes to a slave
piston at the bottom of the borehole, directly connected by a short pump rod to the actual
pump piston below it, which operates in the conventional way.
Although hydraulic transmission units are quite attractive in some respects compared
with pump rods, they are significantly more complicated and expensive. Their efficiency
is also likely to be lower, due to the extra pipe friction involved in moving the water
needed to power the submerged pump.

74

Fig. 39 The Vergnet hydraulic foot-pump

75
3.5.7

Diaphragm pumps

An alternative to the use of a piston in a cylinder for pumping is to fit one wall of a pump
chamber with a flexible diaphragm which when moved in and out displaces water (see
Fig. 40). Here the left hand valve is analagous to the foot valve of a piston pump and the
right hand one is the delivery valve.
In general, the advantages of a diaphragm pump are:
i.
ii.
iii.
iv.
v.

perfect sealing (except for any shortcomings of the two check valves);
high mechanical efficiency, since flexing a diaphragm involves much less friction
then sliding a piston with seals up and down a cylinder;
no. seal is needed at the pump rod which also reduces friction losses still further
compared with piston pumps;
they are self-priming, hold their prime very well and can often handle a higher
than average.suction head;
they often function well with gritty or muddy water which could damage a piston
pump.

There are however also disadvantages:

i.
ii.
iii.

iv.

diaphragms need to be high quality rubber if they are to last, and are therefore
expensive;
diaphragm pumps are often dependent on specialized spare parts that cannot
easily be improvized in the field;
a diaphragm pump is similar to a large diameter piston pump with a short stroke;
so the pump rod forces are high in in relation to the head and swept volume. This
imposes a high load on transmission components and on the point of attachment
of the pump rod to the diaphragm;
therefore diaphragm pumps (of the kind in Fig. 40) are only suitable for low head
pumping in the 5-10m range.

Fig. 41 shows a foot-operated, double-acting diaphragm pump developed by the

International Rice Research Institute (IRRI) in the Philippines for irrigation purposes.
Unlike traditional devices such as Dhones or Shadoofs, this pump is portable {ay two
men) and can therefore be moved along an irrigation canal in order to flood one paddy
after the other. However it is less efficient than the better traditional water lifters.
Fig. 42 shows a commercially manufactured, double-acting diaphragm pump that is
mostly used for purposes such as dewatering building sites; it has the advantage of
being portable, reasonably efficient and well suited to low heads and can deliver quite
high outputs, so it, or similar designs, could equally be used for irrigating small
landholdings. A pump of this kind was well liked by Ethiopian farmers irrigating small
plots from the Omo River in a training project in which the author was involved [15].

76

Fig. 40 Cross-section of a diaphragm pump

Fig. 41 Schematic drawing of the IRRI foot-operated diaphragm pump

77

Fig. 42 Commercial portable double acting diaphragm pump

One type of diaphragm pump that can be improvized and which reportedly works
reasonably well at low heads is a design based on the use of an old car tyre as the
flexible member (Fig. 43). Worn car tyres are of course widely available. The principle of
this pump is to make a chamber by fitting end-plates into the openings of the tyre so that
one is anchored and the other can be forced up and down. If suitable check valves are
provided, this can make an adequate diaphragm pump. The prospective user should not
underestimate the constructional requirements to make an adequately reliable device of
this kind. For example a typical car tyre of 400mm overall diameter will have an effective
area of 0.126m2; this requires a force of 1 230N per m head; i.e. only 3m lift requires
nearly 3 700N pull to displace any water (this is the equivalent of 376kg force or 830 lb).

78
Robust fixings and connections are therefore needed to prevent such a pump coming
apart, even at quite low heads. One further problem reported with car tyre pumps is that
they do not work well as suction pumps because the internal structure of a tyre can
separate from the outer rubber casing when repeatedly pulled by suction pressures.
Tyre pumps could make a useful high-volume low-head pump however, providing they
are skillfully constructed, to be powered perhaps by two people working a suitably strong
lever, and providing they operate submerged or with limited suction lift.

3.5.8 Semi-rotary Pumps

A form of reciprocating positive displacement pump, using the same principles as a
piston pump, is the semi-rotary pump. Here a pivoted plate, or "bucket vane" can be
reciprocated, like a door on hinges, through about 270 within a circular chamber. It
alternately draws water from one side and then the other through check valves. The
semi-rotary pump is mostly used as a hand-pump, often for pumping kerosene and fuel
oils rather than water, and has only a small capacity. It is also sensitive to any dirt in the
water, which can easily jam it. It is therefore unlikely to be useful for irrigation purposes
and is simply mentioned for completeness.

3.5.9

Gas Displacement Pumps

Water can be displaced by a gas or vapour as readily as by a solid. A number of air and
vapour displacement pumps were manufactured at the beginning of this century. The
former rely on air delivered by an engine-driven compressor, while the latter generally
used steam to displace water directly, rather than through the intermediary of a steam
engine and pump. The Humphrey Pump is an analagous device which uses the gases
generated in an internal combustion engine cycle to displace water directly in much the
same way. Both compressed air and steam displacement pumps suffer from being
inherently inefficient, as well as being massive (and hence expensive) in relation to their
pumping capacity, but in contrast the Humphrey Pump is actually more efficient than
most comparably sized conventional i.e. engine pumping systems, although it is also
quite large.
Fig. 44 shows the principle of the Humphrey Pump, which consists in effect of a
conventional four-stroke i.e. engine cylinder head mounted on top of a pipe which forms
the working space. This is in turn connected to a long horizontal pipe and a riser to the
discharge level. The working space pipe is located over the water source and there are
inlet valves to allow water to enter the system below. The Humphrey Pump cycle is
similar to a standard four-stroke piston engine except that instead of the engine having a
metal piston driving a crank shaft, the water in the working space acts as a piston. All the
Humphrey Pumps so far built could only run on gaseous fuels such as coal gas or
natural gas because of difficulties with vaporising liquid petroleum fuels successfully
under the cool cylinder conditions which occur when cold water acts as a piston.

79

Fig. 43 Detail of the New Alchemy Institute (USA) tyre pump Principle of operation of a
4-stroke Humphrey pump

80

Fig. 44 The Humphrey pump: a liquid piston internal combustion engine and pump
combined

81
The Humphrey Pump's cycle is sequenced by a pressure sensor which controls a simple
linkage to open and close the exhaust and inlet valves on the cylinder head at the
correct times. The water in the long horizontal "U" pipe and riser oscillates, to provide
the induction and compression strokes, before being driven forcibly along the pipe by the
firing stroke. A vacuum created behind the departing column of water causes more water
to be induced through the inlet valves.
Humphrey Pumps were used for irrigation projects in the USA, and Australia in the early
part of this century, with some success, and the University of Reading in England has
developed a modern small scale prototype intended for irrigation pumping with biomass
fuels.
The main advantage of the Humphrey Pump other than good fuel efficiency, is its great
mechanical simplicity. It therefore can readily handle muddy or sandy water and has the
potential for extreme reliability, yet requires very little maintenance. The main negative
features are the need for gaseous fuels and it can only readily operate from water
sources where the water level does not change much.

82

3.6

ROTARY POSITIVE DISPLACEMENT PUMPS

There is a group of devices which utilize the displacement principle for lifting or moving
water, but which achieve this by using a rotating form of displacer. These generally
produce a continuous, or sometimes a slightly pulsed, water output. The main advantage
of rotary devices is that they lend themselves readily to mechanization and to high speed
operation. The faster a device can be operated the larger the output in relation to its size
and the better its productivity and cost-effectiveness. Also, steady drive conditions tend
to avoid some of the problems of water hammer and cavitation that can affect
reciprocating devices.
Centrifugal pumps, which use a different principle and are described later, have in fact
become the most general mechanized form of pump precisely because they can be
directly driven from internal combustion engines or from electric motors. But rotary
positive displacement pumps have unique advantages over centrifugal pumps in certain
specialized situations, particularly in being able to operate with a much wider range of
speeds or heads.
Some types of rotary positive displacement pump have their origins among the earliest
forms of technology (eg.the Archimedean Screw), and even today lend themselves to
local improvization. In the past, industrially manufactured rotary pumps were less
successful than centrifugal ones, possibly because they suffered from a number of
constructional and materials problems. But modern, tougher and more durable plastics
and synthetic rubbers may well be an important factor in encouraging the manufacture of
a number of new types of rotary positive displacement pumps which could be
advantageous in some situations, as will be described.

3.6.1

Flexible Vane Pumps

Here a flexible toothed rotor is used, generally made of rubber, Fig. 45. This is very
simple in concept, being like a revolving door, but it can involve both considerable
friction and significant back leakage. It cannot therefore be considered as an efficient
type of pump. On the positive side, it will readily self-prime and can achieve a high head
at low rotational speeds. Much will depend on the quality of the rotor material and the
type of internal surface of the casing so far as both friction and durability are concerned.
Another similar type, developed recently by Permaprop Pumpen in Germany, has an
endless rubber toothed belt which is driven around two pulleys; (see Fig. 46). As it
curves around a pulley, the teeth on the belt spread apart and increase the volume
between them, thereby drawing in water. The diagram shows how both sides of the
chamber simultaneously pump in opposite" directions, and suitable channels in the
casing direct the water. The advantages claimed by the manufacturers are, inter alia,
that it can run on "snore" indefinitely - (i.e. pumping a mixture of air and water), it will
readily self-prime and suck water up to 8m and lift it a further 45m under the power of a
small portable single cylinder engine. It is therefore a much more versatile pump than
the equivalent centrifugal pump, but it is more complicated and expensive.

83

Fig. 45 A flexible vane pump

Fig. 46 The Permaprop tooth pump

84
3.6.2

Progressive Cavity (Mono) Pumps

None of the rotary pumps so far reviewed lend themselves to being lowered down
boreholes; in fact their main selling point is as suction pumps. However, the "progressive
cavity" alias "progressing cavity" alias "Mono" pump (after its French inventor,
Moineaux), (see Fig. 47) is unique in being a commercially available rotary positive
displacement pump that readily fits down boreholes. This is a great advantage because
positive displacement pumps can cope much more readily with variations in pumping
head than centrifugal pumps. Therefore, any situation where the level may change
significantly with the seasons or due to drawdown, or even where the drawdown is
uncertain or unknown, makes the progressive cavity pump an attractive option. It also
has a reputation for reliability, particularly with corrosive or abrasive impurities in the
water. The reasons for this relate to good construction materials combined with a
mechanically simple mode of operation.
Fig. 47 shows that this pump consists of just a single-helix rotor inserted in a doublehelix stator. A single helix is rather like a simple spiral staircase while a double means
two intertwined helixes. The stator helix is usually made from chromium plated steel or
from stainless steel with a polished surface finish, and is circular in cross section and fits
accurately into one of the two helices of the stator. The stator is usually moulded from
rubber or plastic and the cross-section of its internal helix is oval, similar to two circles
similar to the rotor abutting each other. A feature of the geometry of this type of pump is
that the empty second start of the stator is divided into a number of separated empty
voids, blocked from each other by the solid single start rotor. When the rotor is turned,
these voids are screwed along the axis of rotation, so that when "he assembly is
submerged, discrete volumes of water will be trapped between the single start rotor helix
and the inside of the double start stator in the voids and when the shaft is rotated these
volumes of water are pushed upwards and discharged into a rising main.

85

Fig. 47 Progressive cavity or 'Mono' pump

Schematic cross-section to illustrate principle of Archimedean screw
Pumps of this kind are usually driven at speeds of typically 1000 rpm or more, and when
installed down a borehole they require a long drive shaft which is guided in the rising
main by water lubricated "spider bearings" usually made of rubber. Although friction
forces exist between the rotor and stator, they are reduced by the lubricating effect of the
flow of water, and they act at a small radius so that they do not cause much loss of
efficiency. Progressive cavity pumps therefore have been shown to be comparably
efficient to multi-stage centrifugal pumps and reciprocating positive displacement pumps
under appropriate operating conditions. Their main disadvantage is their need for
specialized components which cannot be improvized and their quite high cost; however,
high cost is unfortunately a feature common to all types of good quality borehole pump
and is usually justified by the need to minimise the frequency of the expensive procedure
of removing and overhauling any pump from a deep borehole.

86
The progressive cavity pump can be "sticky" to start - i.e. it needs more starting torque
than running torque (similarly to a piston pump) to unstick the rotor from the stator and
get the water that lubricates the rotor flowing. This can cause start-up problems if electric
motors or engines are used, but certain improved versions of this kind of pump include
features which reduce or overcome this problem.

3.6.3

Archimedean Screw and Open Screw Pumps

The progressive cavity pump is one of the more recent pump concepts to appear, while
the Archimedean screw is one of the oldest, yet they have a number of similarities.
Fig. 48 illustrates a typical Archimedean screw pump (and an animal-powered version is
shown in Fig. 97). The traditional version of this pump, built since before Roman times
and still used in a similar form in Egypt, is made up of a helix of square cross-section
wooden strips threaded onto a metal shaft and encased in a tube of wooden staves,
bound like a barrel with metal bands.
The Archimedean screw can only operate through low heads, since it is mounted with its
axis inclined so its lower end picks up water from the water source and the upper end
discharges into a channel. Each design has an optimum angle of inclination, usually in
the region of 301 to 40 , depending on the pitch and the diameter of the internal helix.
The principle is that water is picked up by the submerged end of the helix each time it
dips below the surface, and as it rotates a pool of water gets trapped in the enclosed
space between the casing and the lower part of each turn. As the whole assembly
rotates, so the helix itself screws each trapped pool of water smoothly further up the
casing until it discharges from the opening at its top; the water pools move much as a
nut will screw itself up a bolt when prevented from rotating with it. This is also analagous
to the trapped volumes of water screwed between the rotor and stator of progressive
cavity pumps.
Traditional wooden Archimedean screws of the kind just described have been tested and
found to have efficencies in the region of 30%.

87

Fig. 48 An Archimedean screw. Two men are needed if the water head is more than 0.6
metres (See also Fig. 97 for an animal-powered version) (after Schioler [24])
The modern version of the Archimedean screw is the screw pump, Pig. 49. This consists
of a helical steel screw welded around a steel tubular shaft, however unlike an
Archimedean screw, there is no casing fixed to the screw, but it is mounted instead in a

88
close fitting, but not quite touching, semi-circular cross-section inclined channel. The
channel is usually formed accurately in screeded concrete. Because of the clearance
between the screw and its channel, some back-leakage is inevitable, but the total flow
rate produced by a screw pump is so large that the backflow is but a small percentage.
Therefore modern screw pumps can achieve high efficiencies in the region of 60-70%.
Their primary advantage is that the installation and civil workings are relatively simple,
compared with those for large axial flow pumps necessary to produce the same volume
of output (which would need a, concrete sump and elaborate large diameter pipework as
in Fig. 66). Also, the screw can easily handle muddy or sandy water and any floating
debris, which is readily pulled up with the water.
Probably the main disadvantage of screw pumps is that an elaborate transmission
system is needed to gear down an electric motor or diesel engine drive unit from
typically 1500 rpm to the 20-40 rpm which is normally needed. Mechanical transmissions
for such a large reduction in speed are expensive and tend to be no more than 60-70%
efficient, thereby reducing the total efficiency of the screw pump, including its
transmission to about 50-60%. Screwpumps also present a safety hazard by having a
potentially dangerous open rotor and should therefore be fitted with mesh guards.
Finally, they cannot cater for much change in level of the water source, unless provision
is made to raise or lower the entire unit and also the maximum head that can be handled
will not exceed much more than 6m in most cases, and will normally be no more than 4
or 5m for smaller screw pumps.

Fig. 49 Cross-section through an open screw pump

89

Fig. 50 Hydrostatic pressure pumps

90
3.6.4

Coil and Spiral Pumps

These pumps use a similar principle to the Archimedean screw except that they run
horizontally while the Archimedean screw is tilted at about 30. The coil and spiral pump
family, if fitted with a suitable rotating seal, can deliver water to a greater height, typically
5-10m, above their discharge opening. Fig 50 A shows a spiral pump and B shows a coil
pump.
Both these pumps work on the same principle, involving either a spiral or a coiled
passage (in the latter case a coiled hosepipe serves the purpose) rotating on a
horizontal axis. One end of the passage is open at the periphery and dips into the water
once per revolution, scooping up a pool of water each time. Due to the shape of the
spiral or the coil, sufficient water is picked up to fill completely the lower part of one turn,
thereby trapping air in the next turn. The pools of water progressively move along the
base of the coil or of the spirals as the pump turns, exactly like an Archimedean Screw.
However, when acting against a positive head, the back-pressure forces the pools of
water slightly'. back from the lowest position in each coil as they get nearer to the
discharge; so they progressively take up positions further around the coil from the lowest
point. The maximum discharge head of either type of pump is governed, by the need to
avoid water near the discharge from being forced back over the top of a coil by the back
pressure, so this is still a low head device.
The spiral pump has to be designed so that the smaller circumferences of the inner
loops are compensated for by an increased radial cross section, so it would normally be
fabricated from sheet metal; the coil pump is of course much easier to build.
This type of pump was originally described in the literature as long ago as 1806 and has
attracted much fresh interest recently, with research projects on it at the universities of
California (USA), Salford (UK), Los Andes (Colombia) and Dar es Salaam (Tanzania),
[16]. Although historically the coil pump was used as a ship's bilge pump, today it is
finding favour for use in river current powered irrigation pumps by, for example, the
Royal Irrigation Department of Thailand (see Fig. 153) and also similarly by
Sydfynsgruppen and the Danish Boy Scouts for an irrigation project in south Sudan, with
support from Danida and on the Niger near Bamako in Mali, under a project supported
by the German aid agency BORDA. Chapter 4.9 deals with some of the practical
applications of this device.
The advantages of these devices are their inherent mechanical simplicity combined with
the fact that, unlike an Archimedean screw, they can deliver into a pipe to a head of up
to about 8-10 m, making them more versatile. The only difficult mechanical component is
a rotary seal to join a fixed delivery pipe to the rotating output from the coil. They are
ideal for water wheel applications due to the low speed and high torque needed, (which
is where most of the research effort appears to be concentrated).
Their main disadvantage is that their output is small unless rather large diameter hose is
used, being proportional to the capacity of the lower part of one turn of hose per
revolution. A simple calculation indicates that a significant and not inexpensive length of
hose is needed to produce an adequate coil pump, (e.g. just 20 coils of only 1.5m
diameter needs nearly 100mm of hose). Supporters of this concept argue that its
simplicity, suitablility for local improvization and reliability should compensate for these

91
high costs, but this type of pump has so far not been popularized successfully for
general use and it does not exist as a commercial product.

Fig. 51 Paddle-wheel or tread-wheel

92
3.6.5

Paddle-wheels, Treadmills and Flash-wheels

These devices are, in effect, rotary versions of the simple scoop; however instead of one
scoop being moved back and forth, a number are set around the periphery of a wheel,
(Fig. 51). Like the scoop a paddle wheel is only useful for very low lift pumping, such as
flooding paddy fields at no more than about 0.5m height above the water source.
The simplest version is the paddle-wheel in which an operator walks directly on the rim,
turning it so that it continuously and steadily scoops up water and deposits it over a low
bund, (Fig. 51). In its basic form the paddle wheel is not very efficient since a lot of the
water lifted flows back around its edges. Therefore an improved version involves
encasing the wheel in a closely fitting box which not only reduces the back-leakage of
water but also slightly increases the head through which the device can operate.
Paddle wheels have been mechanized in the past, although they are unusual as water
lifting devices today. Many of the windmills used in the Netherlands to dewater large
parts of the country drove large paddle wheels, which when mechanized and refined, are
usually known as flash wheels. Flash-wheels function best with raked back blades, and
the best had measured efficiencies in the range 40-70%. Small straight-bladed paddlewheels are probably only 10 or 20% efficient, but have the virtue of being simple to build
and install in situations were a lot of water needs to be lifted through a small head. They
are occasionally used on traditional windpumps, as shown in Fig. 110.

3.6.6

Water Ladders and Dragon-Spine Pumps

The main disadvantage of the paddle wheel just described is that to lift water through a
greater height a bigger wheel is needed. The water ladder was developed to get around
this problem by taking the paddles and linking them together in an endless belt which
can be pulled along an inclined open wooden trough or flume (Fig. 52). The endless belt
is driven by a powered sprocket, at the discharge end, and passes around a freewheeling sprocket at the lower end. The lower end of the trough or flume is submerged,
so that the moving paddles in the belt, which almost fill the cross section of the flume,
push water up it. In many ways this method of water lifting is analagous to a screw pump
which also pushes water trapped between the blades of a mechanism up a flume. As
with the screw pump there is some back-leakage, but with a well-built unit, this is but a
small fraction of the high flow that is established.
The water ladder is still very widely used on small farms in S E Asia for flood irrigation of
small fields and paddies from open streams and canals or for pumping sea water into
evaporation pans to produce sea salt. In China it is known as a "dragon spine" or dragon
wheel" and in Thailand as "rahad". In most cases it is made mainly of wood, and can
consequently easily be repaired on-farm. It is one of the most successful traditional,
high-flow, low-lift water pumping devices and is particularly applicable to rice production,
where large volumes of water are sometimes needed.
On traditional Chinese water ladders, the upper sprocket is normally driven by a long
horizontal shaft which traditionally is pedalled by from two to eight people working
simultaneously; (Fig. 52). The treadles are spaced on the drive shaft so that one or more
of the operators applies full foot pressure at any moment, which helps to smooth the
torque output and keep the chain of boards tensioned and running smoothly.

93

Fig. 52 Water ladder or Chinese 'Dragon Spine' pump

94
Versions of it have been mechanized by using windmills (see Fig. 111), (in Thailand as
well as China), or a buffalo sweep (China) or with small petrol (gasoline) engines.
Water ladders range in length from 3 to 8m and in width from 150 to 250mm; lifts seldom
exceed 1.0 to 1.2m, but two or more ladders are sometimes used where higher lifts are
required. A rough test made in China with a water ladder powered by two teams of four
men (one team working and one resting) showed an average capacity of 23m3/h through
a lift of 0.9m [1]. Further details of Chinese water ladders are given in Table 6.
Table 6 SPECIFICATIONS OF CHINESE "DRAGON-SPINE" WATER LIFTS
Specifications of
products
Name of
product

Single man
hand -turning
water lift

Two men
treadling water
lift
Four men
treadling water
lift

Wind powered
water lift of
diagonal web
member

Weight Volume
of
of
Factory
water
timber price
lift
used (yuan)
(kg)
(m3)

Length
of trough
(m)

Dimensions
of
intake
(height x
width) (m)

1.5

0.18 x 0.14

18

0.2

46

1.8

0.18 x 0.14

20

0.2

50

2.0

0.18 x 0.14

22

0.2

57

2.3

0.18 x 0.14

24

0.3

64

3.0

0.18 x 0.15

30

0.3

76

3.5

0.18 x 0.15

35

0.3

80

2.3

0.25 x 0.20

50

0.3

93

3.0

0.25 x 0.20

55

0.4

106

3.5

0.25 x 0.19

70

0.5

126

4.1

0.25 x 0.19

75

0.5

138

4.7

0.25 x 0.19

85

0.6

151

5.3

0.25 x 0.19

105

0.7

165

3.5

0.25 x 0.19

335

1.1

609

4.1

0.25 x 0.19

345

1.2

622

4.7

0.25 x 0.19

350

1.2

635

Notes

Wind sail
wheel 4-6m
in diameter,
coupled with a
water
lift

Remarks: The products in the table are made by Chengqiao Water Lift and Agricultural
Tool Plant, Hangjiang Commune, Putian County.
Tests on a traditional wooden water ladder powered by a 2-3hp engine were carried out
in Thailand in 1961 [16]. The trough was 190mm deep by 190mm wide and the paddles
were 180mm high by 150mm wide and spaced 200mm apart; note that the clearance
was quite large, being 20mm each side. The principle findings of this study were:

95
i.
ii.
iii.
iv.

the flow rate is maximized when the submergence of the lower end of the flume
is 100%
a paddle spacing to paddle depth ratio of approximately 1.0-1.1 minimizes losses
and maximizes output
the sprocket speed has to be kept to less than 80 rpm to avoid| excessive wear
and frequent breakage
the average efficiency of this device was 40%

It is possible that if a smaller clearance had beer: used between the paddle edges and
the trough, a higher efficiency 'may have resulted; no doubt the optimum. spacing is
quite critical. If it were too small, friction would become excessive and possibly cause
frequent breakage; of the links, while if too large, brick-leakage becomes excessive and
reduces the overall efficiency.

3.6.7

Chain and Washer or Paternoster Pumps

The origins of this type of pump go back over 2 000 years, and they work on a similar
principle to the water ladder just described except that instead of pulling a series of
linked paddles through an open inclined flume or trough, a series of Linked discs or
plugs are pulled through a pipe (Figs. 53, 84 and 96). As with the ladder pump, they lend
themselves to human, animal or mechanical prime-movers and are most commonly
powered by either a team of two to four people or by a traditional windmill.

96
Fig. 53(a) Chinese Liberation Wheel chain and washer pump - an animal-powered
version is shown in Fig. 96

Fig. 53 (b)A view of a hand-operated liberation pump

As discussed in more detail in the next sections covering the use of human and animal
power, a major advantage of this kind of pump is that it requires a steady rotary power
input which suits the use of a crank drive with a flywheel, which is a mechanically
efficient as well as a comfortable way of applying muscle power. It also readily matches
with engines and other mechanical prime-movers.
The main advantage of the chain and washer pump is that it can be used over a wide
range of pumping heads; in this respect it is almost as versatile as the commonly used
reciprocating bucket pump as it is applicable on heads ranging from 1m to over 100m.
For low lifts, loose fitting washers are good enough to lift water efficiently through the
pipe, since back-flow will remain a small and acceptable fraction of total flow. At higher
lifts, however, tighter fitting plugs rather than washers are necessary to minimize backleakage; many materials have been tried, but rubber or leather washers supported by
smaller diameter metal discs are commonly used. Most chain and washer pumps have a
bell mouth at the base of the riser pipe to guide the washers smoothly into the pipe. With
higher lift units where a tighter fit is needed, this is only necessary near the lower end of
the riser pipe; therefore the riser pipe usually tapers to a larger diameter for the upper
sections to minimize friction (see Fig. 53 ).

97
The capacity of a chain and washer pump is a function of the diameter of the riser pipe
and of the upward speed of the chain. For example, four men are necessary to power a
unit with 6m lift and a 100mm riser tube, [1].
Chain and washer pumps have been, and still are in very widespread use, especially in
China, where industrially manufactured pumps of this kind are commonly used and are
often known as "Liberation Pumps". They represented in development terms in China a
major improvement over more traditional and primitive water lifting techniques and an
interim step to modernisation using powered centrifugal pumps. Two to three million
Liberation Pumps were used in China at the peak of their use in the 1960s [17]. The
following performance characteristics relate to typical chain and washer pumps used in
China:
motive
power
2 men
donkey
3kW(e) motor

pumping
head
6m
12m
15m

discharge
rate
5-8m3/h
7m3/h
40m3/h

efficiency
(pump only)
76%
68%
65%

This indicates that the chain and washer pump is not only versatile, but also rather more
efficient than most pumps. It also has an important characteristic for a positive
displacement pump of generally needing less torque to start it than to run it, which
makes it relatively easy to match to prime-movers having limited starting torque.

98

3.7

RECIPROCATING INERTIA (JOGGLE) PUMPS

This range of pumps depend on accelerating a mass of water and then releasing it; in
other words, on "throwing" water. They are sometimes known as "inertia" pumps.
As with the other families of pumps so far reviewed, there are both reciprocating inertia
pumps, described below, (which are only rarely used) and much more common rotary
types which include the centrifugal pump, described in Section 3.8.

3.7.1

Flap Valve Pump

This is an extremely simple type of pump which can readily be improvized; (see Fig. 54).
Versions have been made from materials such as bamboo and the dimensions are not
critical, so that little precision is needed in building it.
The entire pump and riser pipe are joggled up and down by a hand lever, so that on the
up-stroke the flap valve is sucked closed and a column of water is drawn up the pipe, so
that when the direction of motion is suddenly reversed the column of water travels with
sufficient momentum to push open the flap valve and discharge from the outlet. Clearly a
pump of this kind depends on atmospheric pressure to raise the water, so it is limited to
pumping lifts of no more than 5-6m.

3.7.2

Resonant Joggle Pump

Fig. 55 shows an improved version of the flap-valve pump. Here there is an air space at
the top of the pump which interacts with the column of water by acting as a spring, to
absorb energy and then use it to expel water for a greater part of the stroke than is
possible with a simple flap-valve pump. This uses exactly the same principle as for an air
chamber (see Section 3.5.4).

99

Fig. 54 Flap valve pump

Fig. 55 Joggle pump

100
The joggle pump depends on being worked at the correct speed to make it resonate. An
example of a resonant device is a weight hanging from a spring, which will bounce up
and down with a natural frequency determined by the stiffness of the spring and the
magnitude of the weight. The heavier the weight in relation to the spring stiffness, the
slower the natural frequency and vice-versa. If the spring is tweaked regularly, with a
frequency close to its natural frequency, then a small regular pull applied once per
bounce can produce a large movement quite easily, which is an example of resonance.
In exactly the same way, each stroke of a resonant joggle pump makes a column of
water of a certain mass bounce on the cushion of air at the top of the column.
Depending on the size of the air chamber and the mass of the water, this combination
will tend to bounce at a certain resonant frequency. Once it has been started, a pump of
this kind needs just a regular "tweak" of the handle at the right frequency to keep the
water bouncing. This effect not only improves the overall efficiency but makes it
relatively effortless to use. Dunn [20] reports performance figures of 60 to 100
litres/minute lifted through 1.5 to 6m at a frequency of 80 strokes per minute.
It is worth noting that the performance of some reciprocating piston pumps fitted with
airchambers (as in Fig. 36 C) can be similarly enhanced if the speed of the pump is
adjusted to match the resonant frequency of the water in the pipeline and the "stiffness"
of the trapped air in the air-chamber. This is usually only feasible with short pipelines at
fairly low heads, as otherwise the natural frequency in most practical cases is far too low
to match any reasonable pump speed. If resonance is achieved in such situations the
pump will often achieve volumetric efficiencies in the region of 150 to 200%; i.e.
approaching twice the swept volume of the pump can be delivered. This is because the
water continues to travel by inertial effects even when the pump piston is moving against
the direction of flow, (the valves of course must remain open). As a result, water gets
delivered for part of the down stroke as well as on the up stroke. Well-engineered
reciprocating systems taking advantage of resonance can achieve high speeds and high
efficiencies. Conversely, care may be needed in some situations (such as pumps where
there is a reversal of the direction of flow), to avoid resonance effects, as although they
can improve the output, they can also impose excessive loads on the pump or on its
drive mechanism.

3.8
3.8.1

ROTODYNAMIC PUMPS
Rotodynamic Pumps: Basic Principles

The whole family of so-called rotodynamic pumps depends on propelling water using a
spinning impeller or rotor. Two possible mechanisms are used either alone or in
combination, so that water is continuously expelled from the impeller by being:
i.
ii.

deflected by the impeller blades (in propeller type pumps);

whirled into a circular path so centrifugal force then carries the water away, in the
same way a weight on a string when whirled around and released will fly away.

The earliest practical rotodynamic pumps were developed in the 18th and early 19th
century, (Fig. 56). Type A in the figure simply throws water outwards and upwards. Type
B is actually a suction centrifugal pump and needs priming in order to initiate pumping; a
foot valve is provided to prevent the loss of the priming water when the pump stops. A
circular casing is provided to collect the output from the impeller at the delivery level. A

101
pump of this kind is extremely inefficient as the water leaves the impeller with a high
velocity which is simply dissipated as lost energy. Pump C, the Massachusetts Pump of
1818, had the collector built around a horizontal shaft so that the velocity of the water
could be directed up the discharge pipe and carry it to some height; in some respects
this is the fore-runner of the modern centrifugal pump which today is the most commonly
used mechanically driven type of pump.

3.8.2

Volute, Turbine and Regenerative Centrifugal Pumps

The early pumps just described differed from modern pumps in one important respect;
the water left the pump impeller at high speed and was only effectively slowed down by
friction, which gives them poor efficiency and poor performance. The application of an
important principle, shown in Fig. 57, led to the evolution of efficient rotodynamic pumps;
namely that with flowing fluids, velocity can be converted into pressure and vice-versa.
The mechanism is to change the cross section of the passage through which water (or
any liquid) is flowinq. Because water is virtually incompressible, if a given flow is forced
to travel through a smaller cross section of passage, it can only do so by flowing faster.
However pressure is needed to create the force needed to accelerate the mass of water.
Conversely, if a flow expands into a larger cross section, it slows down to avoid creating
a vacuum and the deceleration of the fluid imposes a force and hence an increase in
pressure on the slower moving fluid. It car. be shown (if frictional effects are ignored)
that if water flows through a duct of varying cross sectional area, then the head of water
(or pressure difference) to cause the change? in velocity from v, to vout, will be H, where:

where g is the acceleration due to gravity.

Fig. 56 Early types of centrifugal pumps

102

Fig. 57 The relationship between pressure and velocity through both a jet and a diffuser
The diagram in Fig. 57 shows how the pressure decreases in a jet as the velocity
increases while the reverse occurs in a diffuser which slows water down and increases
the pressure. Qualitatively this effect, is obvious to most people. From experience, it is
well known that pressure is needed to produce a jet of water; the opposite effect, that
smoothly slowing down a jet increases the pressure is less obvious.
When this was understood, it became evident that the way to improve a centrifugal pun;
is to throw the water out of an impeller at high speed (in order to add the maximum
energy to the water) and then to pass the water smoothly into a much larger cross
section by way of a diffuser in which the cross section changes slowly. In this way, some
of the velocity is converted into pressure. A smooth and gradual change of cross-section
is essential, any sudden change would create a great deal of turbulence which would
dissipate the energy of the water instead of increasing the pressure. There are two main
methods of doing this, illustrated in Fig. 58 by diagrams A and B, and a more unusual
method shown in C.
Diagram A shows the most common, which is the "volute centrifugal" pump, generally
known more simply just as a "centrifugal" pump. Here a spiral casing with an outer snailshell-shaped channel of gradually increasing cross section draws the output from the
impeller tangentially, and smoothly slows it down. This allows the water to leave
tangentially through the discharge pipe at. reduced velocity, and increased pressure.
Diagram B shows the other main alternative, which is the so-called "turbine centrifugal"
or "turbine pump", where a set of smoothly expanding diffuser channels, (six in the
example illustrated) serve to slow the water down and raise its pressure in the same

103
way. In the type Of turbine pump illustrated, the diffuser channels also deflect the water
into a less tangential and more radial path to allow it to flow smoothly into the annular
constant cross-section channel surrounding the diffuser ring, from whore it discharges at
the top.
Diagram. C shows the third, lesser known type of centrifugal pump which is usually
called a "regenerative pump", but is also sometimes called a "side-chamber pump" or
ever, (wrongly) a "turbine pump". Here an impeller with many radial blades turns in a
rectangular sectioned annulus; the blades accelerate the water by creating two strong
rotating vortexes which partially interact with the impeller around the rim of the pump for
about three-quarters of a revolution; energy is steadily added to the two vortexes each
time water passes through the impeller; for those familiar with motor vehicle automatic
transmissions the principle is similar to that of the fluid flywheel. When the water leaves
the annulus it passes through a diffuser which converts its velocity back into pressure.
Regenerative pumps are mentioned mainly for completeness; because they have very
close internal clearances they are vulnerable to any suspended grit or dirt and are
therefore only normally used with clean water (or other fluids) in situations where their
unique characteristics are advantageous. They are generally inappropriate for irrigation
duties. Their main advantage is a better capability of delivering water to a higher head
than other types of single-stage centrifugal pump.

104

Fig. 58 Centrifugal pump types

3.8.3

Rotodynamic Pump Characteristics and Impeller Types

It is not intended to deal with this complex topic in depth, but it is worth running through
some of the main aspects relating to pump design to appreciate why pumps are
generally quite sensitive to their operating conditions.
All rotodynamic pumps have a characteristic of the kind illustrated in Fig. 16, which gives
them a limited range of speeds, flows and heads in which good efficiency can be
achieved. Although most pumps will operate over a wider range, if you move far enough

105
from their peak efficiency with any of these parameters, then both the efficiency and
output will eventually fall to zero. For example, Fig. 16 shows that if you drive the pump
in question with a motor having a maximum speed of say 2000 rpm, there is a maximum
flow which can be achieved even at zero head, and similarly there is a head beyond
which no flow will occur. The design point is usually at the centre of the area of
maximum efficiency.
Since any single rotodynamic pump is quite limited in its operating conditions,
manufacturers produce a range of pumps, usually incorporating many common
components, to cover a wider range of heads and flows. Because of the limited range of
heads and flows any given impeller can handle, a range of sub-sets of different types of
impellers has evolved, and it will be shown later there are then variations within each
sub-set which can fine tune a pump for different duty requirements. The main sub-sets
are shown in Fig. 59, which shows a half-section through the impellers concerned to give
an idea of their appearance.

Fig. 59 Typical rotodynamic pump characteristics

106

Fig. 60 Axial flow (or propeller) pump

It can be seen that pump impellers impose radial, or axial flow on the water, or some
combination of both. Where high flows at low heads are required (which is common with
irrigation pumps), the most efficient impeller is an axial flow one (this is similar to a
propeller in a pipe) - see Fig. 60. Like a propeller, this depends on lift generated by a
moving streamlined blade; since in this case the propeller is fixed in a casing, the
reaction moves the water. Conversely, for high heads and low flows a centrifugal (radial
flow) impeller is needed with a large ratio between its inlet diameter and its outlet
diameter, which produces a large radial flow component, as in the left-most type in Fig.
59. In between these two extremes are mixed flow pumps (see also Figs. 61 and 62)
and centrifugal pumps with smaller ratios of discharge to inlet diameter for their
impellers. The mixed flow pump has internal blades in the impeller which partially propel
the water, as with an axial flow impeller, but the discharge from the impeller is at a
greater diameter than the inlet so that some radial flow is involved which adds velocity to
the water from centrifugal forces that are generated.
Fig. 59 also shows the efficiency versus the "Specific Speed" of the various impeller subsets. Specific Speed is a dimensionless ratio which is useful for characterising pump
impellers (as well as hydro-turbine rotors or runners). Text books on pump/turbine

107
hydrodynamics cover this topic in greater depth. The Specific Speed is defined as the
speed in revolutions per minute at which an impeller would run if reduced in size to
deliver 1 litre/sec to a head of 1m and provides a means for comparing and selecting
pump impellers and it can be calculated as follows:

where n is speed in rpm, Q is the pump discharge in litre/sec and H is the head in
metres.

Fig. 61 Surface mounted mixed flow pump

108

Fig. 62 Submerged mixed flow pump

Fig. 59 indicates the Specific Speeds which best suit the different impeller sub-sets; e.g.
an axial flow impeller is best at flow rates of 500-1 000 litre/sec and has a Specific
Speed of 5 000-10 000, at heads of about 5m. Specific Speed can be converted back to
actual rpm (n) at any given head (H) and flow (Q) as follows:

where n is in rpm, N is the Specific Speed from Fig. 59, H is the head in metres, and Q is
the flow in litres/sec.
The choice of impeller is not only a function of head and flow but of pump size too;
smaller low powered pumps of any of these configurations tend to be somewhat less

109
efficient and they also operate best at lower heads than geometrically similar larger
versions.
Fig. 59 also indicates the effect on power requirements and efficiency (marked "kW" and
"EFF" respectively) of varying the key parameter of head "H", away from the design
point. In the case of a centrifugal pump the small diagram shows that increasing the
head reduces the power demand, while in the case of an axial-flow pump, increasing the
head increases the power demand. Paradoxically, reducing the head from the design
head on a centrifugal pump increases the power demand; the reason for this is that
decreasing the head by, say, 10% can increase the flow by 25% - the efficiency may
also go down by 10%, and since the power requirement is head times flow divided by
efficiency, the new power demand will change from:

the ratio of these is 1:1.25, so the power demand will be increased by 25% in this case.
Therefore, varying the conditions under which a pump operates away from the design
point can have an unexpected and sometimes drastic effect. The use of pumps off their
design point is a common cause of gross inefficiency and wasted fuel.

3.8.4

Axial-Flow (Propeller) Pumps

As already explained, an axial-flow (or "propeller") pump propels water by the reaction to
lift forces produced by rotating its blades. This action both pushes the water past the
rotor or impeller and also imparts a spin to the water which if left uncorrected would
represent wasted energy, since it will increase the friction and turbulence without helping
the flow of water down the pipe. Axial flow pumps therefore usually have fixed guide
vanes, which are angled so as to straighten the flow and convert the spin component of
velocity into extra pressure, in much the same way as with a diffuser in a centrifugal
pump. Fig. 60 shows a typical axial flow pump of this kind, in which the guide vanes, just
above the impeller, also serve a second structural purpose of housing a large plain
bearing, which positions the shaft centrally. This bearing is usually water lubricated and
has features in common with the stern gear of an inboard-engined motor boat.
Axial flow pumps are generally manufactured to handle 'flows in the range 150 to 1
500m3/h for vertically mounted applications, usually with heads in the ranqe 1.5-3.On. By
adding additional stages (i.e. two or more impellers on the same shaft) extra lift up to
10m or so can be engineered.
Because pumps of this kind are designed for very large flows at low heads, it is normal
to form the "pipes" in concrete as illustrated, to avoid the high cost of large diameter
steel pipes. Most axial flow pumps are large scale devices, which involve significant civil
works in their installation, and which would generally only be applicable on the largest
land-holdings addressed by this publication. They are generally mainly used in canal
irrigation schemes where large volumes of water must be lifted 2-3m, typically from a
main canal to a feeder canal.

110
Small scale propeller pumps are quite successfully improvised but not usually
manufactured; ordinary boat propellers mounted on a long shaft have been used for
flooding rice paddies in parts of southeast. Asia. The International Rice Research
Institute (TRRI) has developed this concept into a properly engineered, portable high
volume pumping system, (see Fig. 63); it is designed to be manufactured in small
machine shops and is claimed to deliver up to 180m3/h at heads in the range l-4m. This
pump requires a 5hp (3kW) engine or electric motor capable of driving its shaft at 3
000rpm; its; length is 3.7m, the discharge tube is 150mm in diameter and the overall
mass without the prime mover fitted is 45kg.

Fig. 63 Portable axial flow pump (IRRI)

3.8.5

Mixed-Flow Pumps

The mixed-flow pump, as its name suggests, involves something of both axial and
centrifugal pumps and in the irrigation context can often represent a useful compromise
to avoid the limited lift of an axial flow pump, but still achieve higher efficiency and larger
flow rates than a centrifugal volute pump. Also, axial flow pumps generally cannot
sustain any suction lift, but mixed-flow pumps can, although of course they are not selfpriming.
Fig. 61 shows a surface mounted, suction mixed-flow pump and its installation. Here the
swirl imparted by the rotation of the impeller is recovered by delivering the water into a
snail-shell volute or diffuser, identical in principle to that of a centrifugal volute pump.

111
An alternative arrangement more akin to an axial flow pump is shown in Fig. 62. Here
what is often called a "bowl" casing is used, so that the flow spreads radially through the
impeller, and then converges axially through fixed guide vanes which remove the swirl
and thereby, exactly as with axial flow pumps, add to the efficiency. Pumps of this kind
are installed submerged, which avoids the priming problems that can afflict large surface
suction rotodynamic pumps such as in Fig. 61. The "bowl" mixed-flow pump is sometime
called a "turbine" pump, and it is in fact analagous to the centrifugal turbine pump
described earlier; the passage through the rotor reduces in cross-section and serves to
accelerate the water and impart energy to it, while the fixed guide vanes are designed as
a diffuser to convert speed into pressure and thereby increase both the pumping head
and the efficiency. A number of bowl pumps can be stacked on the same shaft to make
a multi-stage turbine pump, and these are quite commonly used as borehole pumps due
to their long narrow configuration. Mixed-flow bowl pumps typically operate with flows
from 200-12 000m3/h over heads from 2-10m. Multiple stage versions are often used at
heads of up to about 40m.

3.8.6

Centrifugal Pumps

i. Horizontal shaft centrifugal pump construction

These are by far the most common generic type of electric or engine powered pump for
small to medium sized irrigation applications. Fig. 64 shows a typical mass-produced
volute-centrifugal pump in cross section. In this type of pump the casing and frame are
usually cast iron or cast steel, while the impeller may be bronze or steel. Critical parts of
the pump are the edges of the entry and exit to the impeller as a major source of loss is
back-leakage from the exit of the impeller around the front of it to the entry. To prevent
this, good quality pumps, including the one in the diagram, have a closely fitting wear
ring fitted into the casing around the front rim of the impeller; this is subject to some wear
by grit or particulate matter in the water and can be replaced when the clearance
becomes large enough to cause significant loss of performance. However, many farmers
probably do not recognize wear of this component as being serious and simply
compensate by either driving the pump faster or for longer each day, both of which
waste fuel or electricity. Another wearing part is a stuffing box packing where the drive
shaft emerges from the back of the impeller casing. This needs to be periodically
tightened to minimize leakage, although excessive tightening increases wear of the
packing. The packing is usually graphited asbestos, although graphited PTFE is more
effective if available. The back of the pump consists of a bearing pedestal and housing
enclosing two deep-groove ball-bearings. This particular pump is oil lubricated, it has a
filler, dip-stick and drain plug. Routine maintenance involves occasional changes of oil,
plus more frequent checks on the oil level. Failure to do this leads to bearing failure,
which if neglected for any time allows the shaft to whirl and damage the impeller edges.

112

Fig. 64 Typical surface mounted pedestal centrifugal pump

ii . Centrifugal pump installations

Figs. 65 and 66 show two alternative typical low lift centrifugal pump installations; the
simplest is the suction installation of Fig. 65. As mentioned earlier in Section 2.1.5,
centrifugal pumps are limited to a maximum in practice of about 4-5m suction lift at sea
level (reducing to around 2m suction lift at an altitude of 2 000m, and further reduced if a
significant length of suction pipe is involved; otherwise problems are almost certain to be
experienced in priming the pump, retaining its prime, etc. A foot valve is a vital part of
any such installation as otherwise the moment the pump stops or slows down, all the
water in the pipeline will run back through the pump making it impossible to restart the
pump unless the pipeline is first refilled. Also, if water flows back through the pump, it
car: run backwards and possibly damage the electrical system.
If the delivery pipeline is long, it is also important to have another check valve (nonreturn valve) at the pump discharge to the pipeline. The reason for this is that if for any
reason the pump suddenly stops, the flow will continue until the pressure drops enough
to cause cavitation in the line; when the upward momentum of the water is exhausted,
the flow reverses and the cavitation bubbles implode creating severe water hammer.
Further severe water hammer occurs when the flow reverses causing the footvalve to
slam shut. The impact of such events has been known to burst a centrifugal pump's
casing. The discharge check valve therefore protects the pump from any such back
surge down the pipeline.

113

Fig. 65 Surface mounted centrifugal pump installation

Fig. 66 Below-surface (sump) centrifugal pump installation

114

Fig. 67 Various types of centrifugal pump impellers

Fig. 68 Effect of direction of curvature of vanes of centrifugal pump impellers

In many cases there is no surface mounting position low enough to permit suction
pumping. In such cases centrifugal pumps are often placed in a sump or pit where the
suction head will be small, or even as in Fig. 66 where the pump is located below the
water level. In the situation illustrated a long shaft is used to drive the pump from a
surface mounted electric motor; (to keep the motor and electrical equipment above any
possible flood level).

115
iii. Centrifugal pump impeller variations

The component that more than anything else dictates a centrifugal pump's
characteristics is its impeller. Fig. 67 shows some typical forms of impeller construction.
Although the shape of an impeller is important, the ratio of impeller exit area to impeller
eye area is also critical (i.e. the change of cross section for the flow through the
impeller), and so is the ratio of the exit diameter to the inlet diameter. A and B in the
figure are both open impellers, while C and D are shrouded impellers. Open impellers
are less efficient than shrouded ones, (because there is more scope for back leakage
and there is also more friction and turbulence caused by the motion of the open blades
close to the fixed casing), but open impellers are less prone to clogging by mud or
weeds. But shrouded impellers are considerably more robust and less inclined to be
damaged by stones or other foreign bodies passing through. Arguably, open impellers
are less expensive to manufacture, so they tend to be used on cheaper and less efficient
pumps; shrouded impellers are generally superior where efficiency and good
performance are important.
Also in Fig. 67, A and C are impellers for a single-suction pump, while B and D are for a
double-suction pump in which water is drawn in symmetrically from both sides of the
impeller. The main advantage of a double-suction arrangement is that there is little or no
end thrust on the pump shaft, but double suction pumps are more complicated and
expensive and are uncommon in small and medium pump sizes.
The shape of the impeller blades is also of importance. Some factors tend to flatten the
HQ curve for a given speed of rotation, while others Steepen it. Fig. 68 shows the effect
of backward raked, radial and forward raked blade tips; the flattest curve is obtained with
the first type, while the last type actually produces a maximum head at the design point.
Generally the flatter the HQ curve, the higher the efficiency, but the faster the impeller
has to be driven to achieve a given head. Therefore impellers producing the most
humped characteristics tend to be used when a high head is needed for a given speed,
but at some cost in reduced efficiency.
iv. Series and parallel operation of centrifugal pumps

Where a higher head is needed than can be achieved with a single pump, two can be
connected in series as in Fig. 69 A, and similarly, if a greater output is needed, two
centrifugal pumps may be connected in parallel as in Fig. 69 B. The effects of these
arrangements on the pump characteristics are illustrated in Fig. 69 C, which shows the
changes in head, discharge and efficiency that occur as a percentage of those for a
single pump operating at its design point. It is clear that series connection of pumps has
no effect on efficiency or discharge but doubles the effective head. Parallel operation
does not however normally double the discharge compared to a single pump, because
the extra flow usually causes a slight increase in total head (due to pipe friction), which
will move the operating point enough to prevent obtaining double the flow of a single
pump.

116

Fig. 69 Combining centrifugal pumps in series or parallel

3.8.7

Multi-stage and Borehole Rotodynamic Pumps

Where high heads are needed, the primary means to achieve this with a single impeller
centrifugal pump are either to drive the impeller faster or to increase its diameter. In the
end there are practical limits to what can be don in this way, so that either single impeller
pumps can be connected in series, or a more practical solution is to use a multiple

117
impeller pump in which the out put from it from one impeller feeds directly, through
suitable passages in the casing, to the next, mounted on the same shaft. Fig. 70 shows
a 5 stage borehole pump (where limitations on the impeller diameter are caused by the
borehole, making multi-staging an essential means to obtain adequate heads). Fig. 44
includes a three stage centrifugal pump, coupled to a turbine as a prime-mover, as
another example of multi-staging.
Surface-mounted multi-stage pumps are probably only likely to be of relevance to
irrigation in mountainous areas since there are few situations elsewhere where surface
water needs to be pumped through a high head. More important from, the irrigation point
of view is the vertical shaft multi-stage submersible borehole pump which has an integral
submerged electric motor directly coupled to the pump below the pump as in the
example of Fig. 70, However, it is possible to get bare-shaft multi-stage borehole pumps
in which the pump is driven from the surface via a long drive shaft supported by spider
bearings at regular intervals down the rising main; see Fig. 134 (b) or with the motor
arranged as for the centrifugal pump in Fig. 66, but. with a vertically mounted multi-stage
pump in either a sump or well.
In recent years numerous, reliable, submersible electric pumps have evolved; Fig. 70.
Section 4.6 discusses in more detail the electrical implications and design features of
this kind of motor. Extra pump stages can be fitted quite easily to produce a range of
pumps to cover a wide spectrum of operating conditions. The pump in Fig. 70 is a 5
stage mixed-flow type, and the same figure also shows how, simply by adding extra
stages (with increasingly powerful motors) a whole family of pumps can be created
capable in the example illustrated of lifting water from around 40m with the smallest unit
to around 245m with the most powerful; the efficiency and flow will be similar for all of
these options. Only the head and the power rating will vary in proportion to the; number
of stages fitted.
Finally, Fig. 71 and Fig. 134 (a) show borehole installations with submersible electric
pumps. The pump in Fig. 71 has level sensing electrodes clipped to the rising main,
which can automatically switch it off if the level falls too low.

118

Fig. 70 Multi-stage submersible electric borehole pumps

119

Fig. 71 Schematic of complete electric submersible borehole pumping installation

120
3.8.8

Self-priming Rotodynamic Pumps

Rotodynamic pumps, of any kind, will only start to pump if their impellers are flooded
with water prior to start-up. Obviously the one certain way to avoid any problem is to
submerge the pump in the water source, but this is not always practical or convenient.
This applies especially to portable pump sets, which are often important for irrigation, but
which obviously need to be drained and re-primed every time they are moved to a new
site.
If sufficient water is present in the pump casing, then even if the suction pipe is empty,
suction will be created and water can be lifted. A variety of methods are used to fill
rotodynamic pumps when they are mounted above the water level. It is, however, most
important to note that if the suction line is empty but the delivery line is full, it may be
necessary to drain the delivery line in order to remove the back pressure on the pump, to
enable it to be primed. Otherwise it will be difficult if not impossible to flush out the air in
the system. One way to achieve this is to fit a branch with a hand valve on it at the
discharge, which can allow the pump to be "bled" by providing an easy exit for the air in
the system.
The most basic method of priming is to rely on the footvalve to keep water in the system.
The system has to be filled initially by pouring water into the pipes from a bucket; after
that it is hoped that the footvalve will keep water in the system even after the pump is not
used for some time. In many cases this is a vain hope, as footvalves quite often leak,
especially if mud or grit is present in the water and settles between the valve and its seat
when it attempts to close. Apart from the nuisance value when a pump loses its prime,
many pumps suffer serious damage if run for any length of time while dry, as the internal
seals and rubbing faces depend on water lubrication and will wear out quickly when run
dry. Also, a pump running dry will tend to overheat; this will melt the grease in the
bearings and cause it to leak out, and can also destroy seals, plastic components or
other items with low temperature tolerance.
The two most common methods for priming surface-mounted, engine driven suction
centrifugal pumps are either by using a small hand pump on the delivery line as
illustrated in Fig. 72, (this shows a diaphragm priming pump which has particularly good
suction capabilities) or an "exhaust ejector" may be used; here suction is developed by a
high velocity jet of exhaust from the engine, using similar principles to those illustrated in
Fig. 57 and described in more detail in Section 3.8.9 which follows.
Several alternatie methods of priming surface suction pumps may be commonly
improvized. For example, a large container of water may be mounted above the pump
level so water can be transferred between the pump and the tank via a branch from the
delivery line with a valve in it. Then when the pump has to be restarted after the pipe-line
has drained, the valve can be opened to drain the tank into the pump and suction line.
Even the worst footvalves leak slowly enough to enable the system to be started, after
which the tank can be refilled by the pump so as to be ready for the next start.
Alternatively, a large container can be included in the suction line, mounted above the
level of the pump, which will always trap enough water in it to allow the pump to pull
enough of a vacuum to refill the complete suction line. Care is needed in designing an
installation of this kind, to avoid introducing air-locks in the suction line.

121

Fig. 72 Direct-coupled air-cooled diesel engine and pump installation with hand-operated
diaphragm pump for priming
Yet another simple method to use, but only if the delivery line is long enough to carry a
sufficient supply of water, is to fit a hand-valve immediately after the pump discharge
(instead of a non-return valve) so that when the pump is turned off, the valve can be
manually closed. Then the opening of this valve will refill the pump from the delivery line
to ensure it is flooded on restarting.
Sometimes the most reliable arrangement is to use a special "self-priming" centrifugal
pump (Fig. 73). Here, the pump has an enlarged upper casing with a baffle in it. When
the pump and suction line are empty, the pump casing has to be filled with water from a
bucket through the filler plug visible on top. Then when the pump is started, the water in
the casing is thrown up towards the discharge and an eye is formed at the hub of the
impeller which is at low pressure; until water is drawn up the suction pipe the water
discharged from the top of the pump tends to fall back around the baffle and some of the
entrained air carries on up the empty discharge pipe. The air which is discharged is
replaced by water drawn up the suction pipe, until eventually the suction pipe fills
completely and the air bubble in the eye of the impeller is blown out of the discharge
pipe. Once all the air has been expelled, water ceases to circulate within the pump and
both channels act as discharge channels. A check valve is fitted to the inlet of the pump
so that when the pump is stopped it remains full of water. Then even if the foot valve on
the suction line leaks and the suction line empties, the water trapped in the casing of the
pump will allow the same self-priming function as described earlier to suck water up the
suction line. Hence, pumps of this kind only need to be manually filled with water when
first starting up after the entire system has been drained.

122

Fig. 73 Self-priming centrifugal pump

3.8.9

Self-Priming Jet Pumps

An alternative type of self-priming centrifugal pump uses the fact that if water is speeded
up through a jet, it causes a drop in pressure (see Section 3.8.2). Here the pump is fitted
into a secondary casing which contains water at. discharge pressure, (see Fig. 74). A
proportion of the water from this chamber is bled back to a nozzle fitted into the suction
end of the pump casing and directed into the eye of the impeller. Once the pump has
been used once (having been manually primed initially) it remains full of water so that on
start up the pump circulates water from the discharge through the jet and back into the
suction side. As before, air is sucked through and bubbles out of the discharge, while
(until the pump primes) the water falls back and recirculates. The jet causes low
pressure in the suction line and entrains air which goes through the impeller and is
discharged, hence water is gradually drawn up the suction line. As soon as all the air is
expelled from the system, most of the discharge goes up the discharge line, but a
proportion is fed back to the nozzle and increases the suction considerably compared
with the effect of a centrifugal impeller on its own. Therefore, this kind of pump not only
pulls a higher suction lift than normal, but the pump can reliably run on "snore" (i.e.
sucking a mixture of air and water without losing its prime). This makes it useful in
situations where shallow water is being suction pumped and it is difficult to obtain
sufficient submergence of the footvalve, or where a water source may occasionally be
pumped dry.
This jet pump principle can also be applied to boreholes as indicated in Fig. 75. An
arrangement like this allows a surface-mounted pump and motor to "suck" water from
depths of around 10-20m; the diffuser after the jet serves to raise the pressure in the
rising main and prevent cavitation. Although the jet circuit commonly needs 1.5-2 times
the flow being delivered, and is consequently a source of significant power loss, pumps
like this are sometimes useful for lifting sandy or muddy water as they are not so easily

123
clogged as a submerged pump. In such cases a settling tank is provided on the surface
between the pump suction and the jet pump discharge to allow the pump to draw clearer
water.

Fig. 74 Schematic of a surface-suction jet pump

The disadvantages of jet pumps are, first, greater complexity and therefore cost, and
second, reduced efficiency since power is used in pumping water through the jet,
(although some of this power is recovered by the pumping effect of the jet). Obviously it
is better to use a conventional Centrifugal pump in a situation with little or no suction lift,
but where Suction pumping is essential, then a self-priming pump of this kind can offer a
successful solution.

124

Fig. 75 Borehole jet pump installation

125

3.9

AIR LIFT PUMPS

The primary virtue of air lift pumps is that they are extremely simple. A rising main, which
is submerged in a well so that more of it is below the water level than above it, has
compressed air blown into it at its lowest point (see Fig. 76). The compressed air
produces a froth of air and water, which has a lower density than water and
consequently rises to the surface. The compressed air is usually produced by an engine
driven air compressor, but windmill powered air compressors are also used. The
principle of it is that an air/water froth, having as little as half the density of water, will rise
to a height above the water level in the well approximately equal to the immersed depth
of the rising main. The greater the ratio of the submergance of the rising main to the
static head, the more froth will be discharged for a given supply of air and hence the
more efficient an air lift pump will be. Therefore, when used in a borehole, the borehole
needs to be drilled to a depth more than twice the depth of the static water level to allow
adequate submergence.

Fig. 76 Air lift pump (schematic)

126
The main advantage of the air lift pump is that there are no mechanical below-ground
components, so it is essentially simple and reliable and can easily handle sandy or gritty
water. The disadvantages are rather severe; first, it is inefficient as a pump, probably no
better, at best, than 20-30% in terms of compressed air energy to hydraulic output
energy, and this is compounded by the fact that air compressors are also generally
inefficient. Therefore the running costs of an air lift pump will be very high in energy
terms. Second, it usually requires a borehole to be drilled considerably deeper than
otherwise would be necessary in order to obtain enough [submergence, and this is
generally a costly exercise. This problem is obviously less serious for low head
applications where the extra depth [required would be small, or where a borehole needs
to be drilled to a considerable depth below the static water level anyway to obtain
sufficient inflow of water.

3.10

IMPULSE (WATER HAMMER) DEVICES

These devices apply the energy of falling water to lift a fraction of the flow to a higher
level than the source. The principle they work by is to let the water from the source flow
down a pipe and then to create sudden pressure rises by intermittently letting a valve in
the pipe slam shut. This causes a "water hammer" effect which results in a sudden sharp
rise in water pressure sufficient to carry a small proportion of the supply to a
considerably higher level.
They therefore are applicable mainly in hilly regions in situations where there is a stream
or river flowing quite steeply down a valley floor, and areas that could be irrigated which
are above the level that can be commanded by small channels contoured to provide a
gravity supply.
The only practical example of a pump using this principle is the hydraulic ram pump, or
"hydram", which is in effect a combined water-powered prime mover and pump. The
hydraulic ram pump is mechanically extremely simple, robust and ultra reliable. It can
also be reasonably efficient. However in most cases the output is rather small (in the
region of 1-3 litre/sec) and they are therefore best suited for irrigating small-holdings or
single terrace fields, seedlings in nurseries, etc.
Hydraulic rams are described more fully in Section 4.9.? dealing with water powered
pumping devices.

127

3.11
3.11.1

GRAVITY DEVICES
Syphons

Strictly speaking syphons are not water-lifting devices, since, after flowing through a
syphon, water finishes at a lower level than if started. However syphons can lift water
over obstructions at a higher level than the source and they are therefore potentially
useful in irrigation. They also have a reputation for being troublesome, and their
principles are often not well understood, so it. is worth giving them a brief review.
Fig. 77 A to C shows various syphon arrangements. Syphons are limited to lifts of about
5m at sea level for exactly the same reasons relating to suction lift for pumps. The main
problem with syphons is that due to the low pressure at the uppermost point, air can
come out of solution and form a bubble, which initially causes an obstruction and
reduces the flow of water, and which can grow sufficiently to form an airlock which stops
the flow. Therefore, the syphon pipe, which is entirely at a sub-atmospheric pressure,
must be completely air-tight, Also, in general, the faster the flow, the lower the lift and
the more perfect the joints, the less trouble there is likely to be with air locks.
Starting syphons off can also present problems. The simplest syphons can be short
lengths of flexible plastic hose which may typically be used to irrigate a plot by carrying
water from a conveyance channel over a low bund; it is well known that all that needs to
be done is to fill the length of hose completely by submerging it in the channel and then
one end can be covered by hand usually and lifted over the bund, to allow syphoning to
start. Obviously, with bigger syphons, which are often needed when there is an
obstruction which cannot easily be bored through or removed, or where there is a risk of
leakage from a dam or earth bund if a pipe is buried in it, simple techniques like this
cannot be used.

128

Fig. 77 Syphon arrangements

129
In Fig. 77 A, a non-return valve or foot-valve is provided on the intake side of the
syphon, and an ordinary gate valve or other hand-valve at the discharge end. There is a
tapping at the highest point of the syphon which can be isolated, again with a small hand
valve. If the discharge hand valve is closed and the top valve opened, it is possible to fill
the syphon completely with water; the filler valve is then closed, the discharge valve
opened and syphoning will commence.
Diagram B is similar to A except that instead of filling the syphon with water to remove
the air, a vacuum pump is provided which will draw out the air. Obviously this is done
with the discharge valve closed. The vacuum pump can be a hand pump, or it could be a
small industrial vacuum pump. Once the air is removed, the discharge valve can be
opened to initiate syphoning.
Diagram C shows a so-called "reverse" syphon, used for example where a raised
irrigation channel needs to cross a road. Reverse syphons operate at higher than
atmospheric pressure and there is no theoretical limit to how deep they can go, other
than that the pipes must withstand the hydrostatic pressure and that the outflow must be
sufficiently lower than the inflow to produce the necessary hydraulic gradient to ensure
gravity flow.

3.11.2

Qanats and Foggara

Qanats, as they are known in Farsi or Foggara (in Arabic), are "man-made springs"
which bring water out to the surface above the local water table, but by using gravity.
Like syphons they are not strictly water lifting devices, but they do offer an option in lieu
of lifting water from a well or borehole in order to provide irrigation. They have been used
successfully for 2 000 years or more in Iran, and for many centuries in Afghanistan,
much of the Middle East and parts of North Africa.
Fig. 78 shows a cross-section through a qanat; it can be seen that the principle used
exploits the fact that the water table commonly rises under higher ground. Therefore, it is
possible to excavate a slightly upward sloping tunnel until it intercepts the water table
under higher ground possibly at some distance from the area to be irrigated. It is exactly
as if you could take a conventional tube well and gradually tip it over until the mouth was
below the level of the water table, when, clearly water would flow out of it continuously
and without any need for pumping.
Qanats are typically from one to as much as 50 kilometres long, {some of the longest are
in Iran near Isfahan). They are excavated by sinking wells every 50 to 100m and then
digging horizontally to join the bases of the wells, starting from the outflow point.
Traditional techniques are used, involving the use of simple hand tools, combined with
sophisticated surveying and tunnelling skills. Many decades are sometimes needed to
construct a long qanat, but once completed they can supply water at little cost for
centuries. The surface appearance of a qanat is distinctive, consisting of a row of low
crater-like earth bunds (or sometimes a low brick wall) surrounding each well opening;
this is to prevent flash floods from pouring down the well and washing the sides away.
The outflow from a qanat usually runs into a cultivated oasis in the desert, resulting from
the endless supply of water.

130

Fig. 78 Cross-section through a Qanat

Efforts have been made in Iran to mechanize qanat construction, but without great
success, although in some cases qanats are combined with engine powered lift pumps
in that the qanat carries water more or less horizontally from under a nearby hill
possessing a raised water table to a point on level ground above the local water table
but below the surface, where a cistern is formed in the ground. A diesel pump is then
positioned on a ledge above the cistern to lift the water to the surface.

3.12

MATERIALS FOR WATER-LIFTING DEVICES

This is a complex technical subject if discussed fully, but. it is worth briefly setting out
some of the advantages and disadvantages of different materials that are commonly
used, as an aid to appraising the specification of different equipment.
Four main considerations apply for construction materials used for pumping water:
a. strength; stressed components need to be able to function over a long period of
time without either failing through overload or, more likely, through fatigue;
b. corrosion resistance and general ability to coexist under wet conditions;
c. resistance to wear and abrasion is important for components that rub or slide or
which are in contact with flowing water if any particulate matter is likely to be
suspended in the water;
d. cost.
As in most branches of engineering, nature has not been kind enough to offer materials
which simultaneously satisfy all these requirements completely; invariably compromises
are necessary. The important point is to be aware of these and to judge whether they
are the right compromises for the application of interest.
It is worth reviewing briefly the pros and cons of various different materials which feature
frequently in pumps and water lifts; these are also summarised in Table 7.

131
Table 7 RELATIVE MERITS OF MATERIALS FOR PUMPS

Material

Strength

Mild Steel

High

Cast Iron

Moderate

Corrosion
Abrasion
resistance
resistance
in water

Typical application

Moderate
to Good

Low

Shafts Pump rods nuts &

bolts Structural items

Moderate Moderate
to Good

Low

Pump casings

High

V. Good

Good

High

Nuts & bolts Shafts

Impellers Wet. rubbing
surfaces Valve components

Brass

Moderate

Good

Moderate
to Good

High

Impellers Pump cylinders

Wet rubbing surfaces

Bronze/ Gun metal

High to
Moderate

V. Good

Moderate
to Good

High

Impellers Pump pistons Wet

bearings & Rubbing
surfaces Valve parts

Phosphor Bronze

Moderate

Good

Good

High

Plain bearing & Thrust

washers

Aluminium & light

alloys

High to
Moderate

Moderate
to Good

Poor

Poor

Poor

Poor

Low

Lightly loaded structural

items

Moderate

Moderate

Poor

Low

Moderately loaded
structures

Stainless Steel

Soft Woods
Bamboos
Good quality
Hardwoods
Thermoplastics PVC
Polythene, etc

V. Poor

Cost

Moderate Pump casings Irrigation

to High pipes

Moderate to Moderate Moderate Moderate Structures

Good
to Good to Good to High
Moderate

Thermoplastics, Filled High to

plastics & Composites Moderate

V. Good

Moderate Moderate Pipes and components

to Good

Generally Generally Moderate Pump casings Components

Good
Good
to High Bearings

132

Fig. 79 Animal-powered Chinese Liberation Pump mechanism uses steel components

(for strength) to good effect (see also Fig. 96).

133
i. Ferrous Metals

Most ferrous, or iron based, materials are subject to corrosion problems, but to
compensate, they are perhaps the most familiar low-cost "strong material" that is widely
available. Generally speaking iron and steel are best suited for use in structural
components where strength is important but a surface coating of rust will not cause
serious problems.
Ordinary mild steel is one of the most susceptible to corrosion. Iron and steel castings,
except where they have been machined, are generally partially protected by black-iron
oxide which forms when the casting is still hot. There are several methods to protect
steels from corrosion, including conventional paints, various modern corrosion inhibitors
which chemically bond with the surface of the metal and inhibit corrosion, various forms
of plating and metallic coatings such as zinc (galvanizing) and cadmium plating. Various
steel chromium and nickel based alloys, the so-called stainless steels, are also resistant
to oxidation and corrosion, but they are not cheap.
Stainless steels do make a useful alternative to brasses and bronzes, but they are very
difficult to machine and to work and therefore most pump manufacturers prefer nonferrous corrosion resistant alloys. One important application for stainless steel is as nuts
and bolts in situations where mild steel nuts and bolts readily corrode; stainless steel
nuts and bolts are expensive compared with mild steel ones, but cheap in terms of time
saved in the field on items that regularly need to be dismantled in wet conditions for
maintenance or replacement.
A primary mechanism for corrosion of steel in wet conditions is if the steel is in
combination with nobler metals, (e.g. copper), and there is an electrical link between
them while both metals are in contact with water. This can encourage what is known as
electrolytic corrosion, especially if the water has a significant mineral content which will
generally increase its conductivity.
Therefore, ferrous components ought to be well protected from corrosion and generally
are best suited as structural items not having any "high quality" surfaces in contact with
water. An example of a bad use for iron, where it sometimes is applied, is as cast iron
pump cylinders. Here the internal surface will often keep in quite good condition so long
as the pump is worked, but any lengthy period during which it is stopped a certain
amount of oxidation will occur; even a microscopic outgrowth of iron oxide (rust) forming
will quickly wear out piston seals once the pump is started again. Obviously any thin
internal coating or plating of a pump cylinder is not likely to last long due to wear.
However, cast steel centrifugal pump casings are often quite satisfactory, although parts
requiring critical clearances such as wear rings are usually inserts made of a more
appropriate corrosion resistant metal. Similarly, cast steel centrifugal pump impellers are
sometimes used; they are not of the same quality as non-ferrous ones, but are obviously
a lot cheaper. Pumps with steel impellers usually cannot have close clearances and
machined surface finishes, so their efficiencies are likely to be lower.
ii. Non-ferrous Metals

Brass (a copper-zinc alloy) is commonly used for reciprocating pump cylinders. Due to
its high cost, thin, seamless, brass tube is often used as a cylinder liner inside a steel

134
casing, instead of a thick brass cylinder; obviously the steel must be kept from direct
contact electrolytically with the water. Brass has good wear resistance in "rubbing"
situations - i.e. with a leather piston seal, but it is not a particularly strong metal
structurally, especially in tension. So called "Admiralty Brass" includes a few percent of
tin, which greatly improves its corrosion resistance.
The bronzes and gun-metals are a large family of copper based alloys, which are
generally expensive but effective in a wet environment; they usually have all the
advantages of brass, but are structurally stronger, (and even more expensive too).
Bronzes can contain copper alloyed with tin, plus some chromium or nickel in various
grades and traces of other metals including manganese, iron and lead. So-called leaded
bronzes replace some of the tin with lead to reduce costs, which still leaves them as a
useful material for pump components. The inclusion of antimony, zinc and lead in
various proportions produces the form of bronze known as gun-metal, which is a useful
material for corrosion resistant stressed components. A bronze containing a trace of
phosphorus, known as phosphor bronze, is an excellent material for plain bearings and
thrust washers, if run with an oil film against a well-finished ferrous surface such as a
machined shaft. Aluminium-bronze, which is cheaper but less corrosion resistant,
replaces much or all of the scarce and expensive tin with aluminium.
Bronzes are generally among the best materials for making precision components that
run in water and which need good tensile strength, such as pistons, valves, impellers,
etc. Castings with a good finish can readily be obtained, and most bronzes machine very
easily to give a precision surface.
Other materials, such as aluminium and the light alloys are generally not hard or wearresistant enough for hydraulic duties, although by virtue of being very light they are
sometimes used to make portable irrigation pipes; however they are not cheap as pipe
material and can only be justified where the need to be able easily to move pipelines
justifies the cost.
iii. Timber

Timbers exist in a very wide variety of types; their densities can range from around
500kg/m3 (or less) up to 1 300kg/m3. They also offer a very large variation in mechanical
properties, workability, wear resistance and behaviour in wet conditions. Timber is of
course also susceptible to damage by insects, fungus or fire.
The most durable timbers are generally tropical hardwoods such as Greenheart, Iroko,
Jarrah, Opepe, Teak and Wallaba. The durability of many timbers can be improved by
treating them with various types of preservative; the most effective treatments involve
pressure impregnation with either tar or water-based preservatives.
One of the main factors affecting the strength of a wooden member is whether knots are
present at or near places of high stress. Where wood is used for stressed components,
such as pump rods for windpumps or handpumps, it is important that it is finegrained
and knot-free to avoid the risk of failures. Good quality hardwoods like this are not easily
obtained in some countries and, where available, they are usually expensive. Cheap
wood is limited in its usefulness and must be used for non-critical components.

135
Certain woods like 1ignum-vitae have also been used in the past as an excellent plain
bearing material when oil lubricated running against a steel shaft, although various
synthetic bearing materials are now more readily available and less expensive.
Wood is available processed into plywood and chipboards; with these a major
consideration is the nature of the resins or adhesives used to bond the wood. Most are
bonded with urea-based adhesives which are not adequately water resistant and are not
suitable for outside use, but those bonded with phenolic resins may be suitable if applied
correctly and adequately protected from water with paints. Therefore, for any irrigation
device it is essential that nothing but "marine" quality plys and chipboards are used.
iv. Plastics

There is a large and growing family of plastics, which broadly include three main
categories; thermoplastics, which soften with heat (and which can therefore readily be
heated and worked, moulded or extruded); thermosets (which are heated once to form
them during which an irreversible chemical process takes place) and finally various
catalytically-cured resins. As with almost everything else, the better quality ones are
more expensive. Great improvements are continuously being introduced and there are
interesting composite plastics which contain a filler or a matrix of some other material to
enhance their properties at no great cost.
Although plastics are weaker and softer than metals, they generally nave the virtue of
being compatible with water (corrosion is not a problem) and although their raw
materials are not always low in cost, they do offer the possibility of low cost massproduction of pipe or components.
Thermoplastics based on polymerized petro-chemicals are generally the cheapest
plastics; those used in the irrigation context include:
PVC (polyvinylchloride) is commonly used for extruded pipes; it can be rigid or
plasticized (flexible); it is important to note that only certain grades of PVC (and other
plastics) are suitable for pipes to convey drinking water for people or livestock, since
traces of toxic plasticiser can be present in the water passed through some grades. PVC
is relatively cheap and durable, but it is subject to attack by the UV (ultra-violet)
wavelengths in sunlight and should therefore either be buried to protect it from the sun,
or painted with a suitable finish to prevent penetration by UV radiation. PVC is also a
thermoplastic and therefore softens significantly if heated above about 80C; however
this is not normally a problem in "wet" applications.
High density "polythene" (polyethylene) is cheaper and less brittle than PVC (especially
at low temperatures) and is commonly used to make black flexible hose of use for
irrigation, but it is also structurally much weaker than PVC, which is not necessarily a
disadvantage for surface water conveyance at low pressure; however PVC is better for
pressurized pipes.
Polypropylene is in the same family as polythene, but is intermediate in some respects in
its properties between polythene and PVC. Polypropylene is less liable to fracture or to
be sub-standard, due to bad management of the extrusion equipment, than is PVC; i.e.

136
quality control is less stringent, so it can be more consistently reliable than poorly
produced PVC.
None of the above plastics are generally applicable for manufacturing pump components
for which strength and durability are important; these require more expensive and
specialized plastics, such as nylons, polyacetals and polycarbonates. Nylons can be
filled with glass, (for strength), molybdenum disulphide (for low friction), etc. An
expensive specialized plastic of great value for bearings and rubbing surfaces on
account of its low friction and good wear resistance is PTFE (polytetrafluoroethylene);
certain water lubricated plain bearings rely on a thin layer of PTFE for their rubbing
surface, and this proves both low in cost and extremely effective.
There are also various specialized thermoset plastics which find applications as pump
components; these tend to be tougher, more wear resistant and more heat resistant than
thermoplastics, and therefore are sometimes used as bearings, pump impellers or for
pump casings. They are also useful for electrical components which may get hot. Most
"pure" plastics are inclined to creep if permanently loaded; i.e. they gradually deform
over a long period of time; this can be avoided and considerable extra strength can be
gained, through the use of composite materials where glass fibre mat (for example) is
moulded into a plastic. Various polyesters and epoxides are commonly used to make
glass-reinforced plastics (g.r.p. or "fibreglass"); these are used to make small tough
components or, in some cases, to make large tanks. Another example of composite
plastics is the phenolic composites where cloth and phenolic resin are combined to
make a very tough and wear resistant, but readily machinable material which makes an
excellent (but expensive) water lubricated bearing, such as "Tufnol".

137

3.13

SUMMARY REVIEW OF WATER LIFTING DEVICES

Table 5, which introduced this chapter, is sorted into categories of pump types based on
their working principles, but it is difficult to see any pattern when looking through it.
Therefore, Table 8, which concludes this chapter attempts to quantify the characteristics
of all pumps and water lifts in terms of their operating heads, power requirements, output
and efficiency. Finally, Fig. 80 (A, B and C) indicates the different categories of pump
and water lift demarcated on a log-log head-discharge graph (similar to that of Fig. 11).
Obviously there are no hard and fast boundaries which dictate the choice of pump, but
the figure gives a graphic indication of which pumps fit where in terms of head and flow,
and hence of power. Note that Table 8 shows input power requirements, whereas Fig.
80 gives the hydraulic power produced, which will be a lesser figure by the factor of the
pump efficiency. Due to the use of the log-log scales, the smaller devices appear to
occupy a larger area then they would if linear scales had been chosen, however in this
case it would not have been possible to fit sufficient detail in to the corner where the
multiplicity of low-powered, low-head and low-flow devices fit, had a linear scale been
applied.
Table 8 REVIEW OF PUMPS AND WATER LIFTS
Head
Range
(m)

Input
Power
(kW)

Flow
Range
(m3/h)

Efficiency
(%)

5-3

.02

.5

5-15

Scoops and bailers

.04

40-60

Swing basket

.6

.06

10-15

Pivoting gutters & "Dhones"

.3-1

.04

5-10

20-50

Counterpoise or "Shadoof

1-3

.02-.08

2-4

30-60

Rope & bucket and windlass

5-50

.04-.08

10-40

Self-emptying bucket, "Mohte"

5-10

.5-.6

5-15

10-20

Reciprocating bucket hoist

100+

100+

400+

70-80

Continuous bucket pump

5-20

.2-2

10-100

60-80

Persian wheel or "tablia"

1.5-10

.2-.6

5-25

40-70

Improved Persian W. "Zawaffa"

.75-10

.2-1

10-140

60-80

Scoop wheels or "Sakia"

.2-2

.2-1

15-160

60-80

Waterwheels or "Noria"

.5-8

.2-1

5-50

20-30

Piston/Bucket pumps

5-200+

.03-50+

2-100+

40-85

Plunger pumps

40-400

.50-50+

2-50+

60-85

Category and Name

I

DIRECT LIFT DEVICES

Reciprocating/Cyclic
Watering can

Rotary/Continuous

II DISPLACEMENT PUMPS
Reciprocating/Cyclic

138
Diaphragm pumps/IRRI pump

1-2

.03-5

2-20

20-30

"Petropump"

5-50

.03-5

2-20

50-80

Semi-rotary pumps

1-10

.03-.1

1-5

30-60

Gas or Vapour displacement

5-20

1-50+

40-400+

n/a

5-10

.05-.5

2-20

25-50

10-100

.5-10

2-100+

30-70

Archimedean screw

.2-1

.04

15-30

30-60

Open screw pumps

2-6

1-50+

40-400+

60-80

Coil and spiral pumps

2-10

.03-.3

2-10

60-70

Flash-wheels & Treadmills

.2-1

.02-20

5-400+

20-50

Water-ladders

5-1

.02-1

5-20

50-70

Chain (or Rope) and Washer

5-20

.02-1

5-30

50-80

Inertia and "Joggle" pumps

2-6

.03

1-3

20

Flap valve pump

2-6

.03

1-3

20

Resonating joggle pump

2-6

.03

2-4

50

Rotary/Continuous
Flexible vane pumps
Progressive cavity (Mono)

III VELOCITY PUMPS

Reciprocating/Cyclic

Rotary/Continuous
Propeller (axial flow)

5-3

10-500+ 100-500+

50-95%

Mixed flow pumps

2-10

150-500+

50-90%

Centrifugal pumps

4-60

.1-500+

1-500+

30-80%

Multi-stage mixed flow

6-20

50-500+

10-100

50-80%

Multi-stage centrifugal

10-300

5-500+

1-100

30-80%

Jet pump Centrifugals

10-30

5-500+

50-500

20-60%

IV BUOYANCY PUMPS
Air lift

5-20

V IMPULSE PUMPS
Hydraulic ram

10-100

VI GRAVITY DEVICES
Syphons, Qanats or Foggara

1-6

139

Fig. 80a Typical head and discharge capacities for different types of pumps and waterlifting devices (on a log-log scale) (and continued in Figures 80b and 80c)

140

Fig. 80b DISCHARGE

141

4. POWER FOR PUMPING

Fig. 81 indicates the most feasible linkages between different energy resources and
prime movers. It shows how all energy sources of relevance to small or medium scale
irrigation pumping originate from renewable energy resources or from fossil fuels; the
arrows then show all the routes that can apply from an energy resource to produce
pumped water. In some cases similar components can be used within systems
energized in completely different ways; for example electric motors are necessary either
with a solar photovoltaic pumping system or with a mains electrical system, so the
motor-pump sub-systems of both types of system can have a lot in common.
The details of the components in Fig. 81 are discussed through the following section, but
it is first worth reviewing a few generalities relating to the combination of prime movers
and pumps.

4.1

PRIME MOVERS AS PART OF A PUMPING SYSTEM

4.1.1 Importance of "Cost-Effectiveness"

Almost every aspect of an irrigation pumping system consists of compromises, or tradeoffs, between the capital (or first) cost of the system and the running (or recurrent) costs.
Farmers tend to purchase cheaper systems with higher running costs, resulting in the
widespread use of other than the most efficient and cost-effective systems. No doubt this
is because farmers are generally short of capital and often in any case regard large
capital investments as being inherently more risky than incurring regular running costs,
(which in time may mount up to a large sum).
There is therefore a good case for institutional users with an interest in improving
agricultural techniques, such as agricultural credit agencies, aid agencies and
governments, to try to assist in overcoming these problems by providing suitable
financial inducements to encourage the use of more cost-effective irrigation systems.
In the final analysis, the combination of components to make a pumping system will
depend on the cost effectiveness of the total chosen system under whatever specific
technical, agricultural and financial conditions happen to be prevalent. That is "costeffectiveness" in the very broadest sense, including not just the first costs and running
costs for the system, but factors like the convenience and ease of use as perceived by
the farmer; (i.e. including, in economists' jargon, the "opportunity costs"). In reality, the
selection decision is usually limited to what is known to be available and affordable and
yet is capable of the required pumping duty.
The selection process is discussed in more detail in the next section, but certain
considerations inherent in the economics and hence in the relative cost-effectiveness of
different system choices are important in relation to any discussion of linking
components to make up pumping systems. To this end, some of the cost attributes of
different prime-mover options are compared in Table 9 . Here the main categories of
irrigation pumping system are reviewed in terms of first costs, recurrent costs and then in
terms of two factors relating more to their "effectiveness" (i.e. their productivity and their
general availability). It is clear from this table that no single method offers both low first

142
costs and low recurrent costs and yet is also among the most productive (otherwise it
would be universally applicable and the other options would be of little interest). High
productivity depends either on relatively high first costs (when using renewable energy
equipment generally) or on relatively high running costs combined with moderately high
first costs when using fossil-fuelled equipment. In the end, successful selection depends
on the choice of the best trade-off between the availability of finance, the capability of
maintaining and financing the recurrent costs of the system and the performance or
productivity that is expected.

143

Fig. 81 Linkages between energy resources and appropriate prime movers

144
Table 9 COST ATTRIBUTES OF PRIME MOVERS
Prime
mover
Cost

Petrolfuelled Coal
Human Animal Solar Wind Hydro Biomass Mains1
fuelled
fuelled electricity engines
engines
engines
* or

First
Costs

Capital

**

Shipping

***** **** ****

***

****

***

****

*****

**

****

****

**

**

****

Installation
Recurrent
Costs

* or
*

**

***

***

***

*****

**

***

Fuel

*****

***

NIL

NIL

NIL

**

***

***

**

Spares

****

****

***

Maintenance

****

**

****

*****

****

***

***

***

****

****

****

****

**

***

****

**

Attendance2 *****

** to ** to
Productivity

Opportunity *****
Costs

**

***

***

**** ****
*

Key * = Low
** = Low to Moderate
*** = Moderate
**** = Moderate to High
***** = High
Electricity can vary from low to high depending on whether a mains connection is
already available or not.

"Attendance" implies level of human intervention needed.

It is possible that potentially useful systems are usually not considered simply because,
not being conventionally used at present, they are unfamiliar and therefore are not
known or understood well enough by potential buyers; this is where it is hoped that
publications such as this may encourage some attempts to try new methods, preferably
by institutions or individuals with the resources to underwrite the risks inherent in
experimenting with new or unfamiliar technologies.
i. Cost-effectiveness and efficiency

Generally, a cost-effective system needs to be technically efficient; i.e. a relatively high

output is needed in relation to the energy input. This is just as true for renewable energy
powered systems as for fossil fuelled systems. In the former case, the energy
resource, if it is solar energy, wind or water power is notionally cost-free, but the capital
cost of the system is closely linked to the efficiency. This is because for a given
pumping requirement, if you halve the efficiency of the system you must double the
"cross section" of the energy resource to be intercepted; i.e. you need twice the area of

145
solar collector or twice the rotor area of a windmill, or a turbine capable of passing twice
the flow rate of water. This tends to require a system that is twice as large and therefore
usually twice as expensive.
In all cases there is an ultimate technical efficiency that can be approached but never
quite achieved, (Fig. 82 A). Pursuing the cause of better efficiency is usually worthwhile
up to a point, but thereafter it brings diminishing returns as increasing complication,
sophistication and cost is required to achieve small further gains in efficiency. However it
usually requires a mature technology to be at the level where further improvements in
efficiency are counter-productive, and in any case, new manufacturing processes and
materials or increases in recurrent costs (due to inflation) sometimes allow
improvements to become cost-effective in the future which were not justifiable in the
past.
The influence of efficiency on costs is illustrated in Fig. 82 B, which shows how low
efficiency generally causes high costs and that there is an optimum range of efficiency
for most technologies where reasonably low costs are achieved, but above which
diminishing returns set in. In the case of renewable energy systems these costs will be
largely attributable to the capital cost and hence to financing the investment, while in the
case of fossil fuelled devices a large proportion of the costs will relate to running and
maintaining the system.

Fig. 82 A. How diminishing returns eventually defeat the benefits of seeking increased
efficiency beyond certain levels
B. The influence of efficiency on costs
ii. Combining system components with differing efficiencies

Virtually all pumping system components achieve an optimum efficiency at a certain

speed of operation. Some components like pipes and transmission systems are most
efficient (in terms of minimizing friction and hence losses) at very low rates of
throughput, but they are then least productive and they will therefore have a point of
"optimum cost-effectiveness" where there is a good compromise between their

146
productivity and their efficiency. Prime movers invariably have an optimum speed of
operation; this is as true of humans and animals as it is of diesel engines or windmills.
Fig. 83 shows three sets of curves; first, efficiency against speed for two prime-movers,
(in this example electric motors would fit the speeds and efficiencies shown); second,a
curve for a typical pump and lastly, for the combination of the prime-movers with the
pump. It must be remembered that the efficiency of a combination of two components is
numerically the product (i.e. the multiplied result) of their individual efficiencies eg. a
30% efficient engine (0.3) with a 50% efficient pump (0.5) has a combined efficiency of:

Fig. 83 Illustration of how correct speedmatching of a prime mover to a pump can be

more important than the efficiency of the prime mover. Here prime mover 'A' is less
efficient than 'B' but has a better speed match to the pump hence the less efficient
prime mover 'A' provides a more efficient system
The important point contrived in Fig. 83 is that the prime-mover with the highest optimum
efficiency is not in this case the best one to use with a particular pump. In the example
motor "A" has a best efficiency of 58% while motor "B" achieves 66%, yet, because the
optimum efficiency of motor "A" occurs at a speed which coincides well with the optimum
efficiency of the pump, the combined efficiency of that combination is better if the motor
is direct-coupled to the pump; (motor "B" will drive the pump at a speed greater than its
optimum, as at 1 500 rpm the pump has an efficiency of only 35%) so the best
efficiencies of the two alternative combinations are:

147

This illustrates how it is generally more important to ensure that the design speeds of
components match properly than to ensure that each component has the highest
possible peak efficiency.

4.1.2

Transmission Systems

Components often do not match effectively; i.e. their optimum speeds of operation are
different. In such situations it generally pays, and it is sometimes essential, to introduce
a speed changing transmission. Also, in many situations the prime-mover cannot readily
be close to the pump, and some method is therefore necessary for transmitting its output
either horizontally or vertically to the water lifting device.
i. Transmission principles

Power can be transmitted from a prime mover to a pump in a number of ways; the most
common is a mechanical connection, which can either rotate (shafts, belts or gears) or
reciprocate (pump rods or levers). Where power has to be transmitted some distance,
then electricity, hydraulic pressure or compressed air can be used, since it is difficult to
transmit mechanical power any distance, especially if changes of direction or bends are
needed.
In all transmissions there is a trade-off between the force or torque being transmitted by
the system (which demands robustness to resist it) and the speed of operation (which
tends to cause wear and reduced life). Power, which is what is being transmitted, can be
defined as the product of force and velocity. Mechanical systems that run at slow
rotational or reciprocating speeds need larger forces to transmit a given amount of
power, which in turn require large gear teeth, large belts or large pump rods (for
example) and these inevitably cost more than smaller equivalents. Where mechanical
power is transmitted some distance any reciprocating linkages need to be securely
anchored; (even a 5m farm windpump can pull with a reciprocating force peaking at
about 1 tonne). For this reason, most modern commercial systems involving lengthy
mechanical links tend to use high speed drive shafts (for example surface mounted
electric motors driving a rotodynamic pump located below the water (or below flood
level) as in Figs. 66 or 134 B). A high speed drive shaft can be quite small in section
because its high speed results in low torque. However a high speed drive needs to be
built with some precision and to have good (and expensive) bearings to carry it and to
align it accurately so as to prevent vibrations, whirling of the shaft, premature wear and
other such problems.
Electrical, hydraulic or pneumatic transmissions all have a common requirement
demanding that their voltage, or pressure of operation ideally needs to be high to
minimize the cross section of cable or of pipe needed to transmit a given power flow
efficiently. High voltage cable (or high pressure pipes), need to be of a good quality and
inevitably cost more per metre for a given cross section. Therefore, with all
transmissions there is a trade-off between efficiency and cost; cheap transmissions often
reduce the capital costs but result in high recurrent costs due to their lower efficiency

148
and greater maintenance and replacement needs, and vice-versa. It is therefore
advantageous to match prime-movers and pumps of similar speeds to avoid the cost
and complication of speed-changing transmissions.
ii. Mechanical transmissions

The most common need for a mechanical transmission is to link an engine or an

electric motor with a pump. Generally such prime-movers are used with centrifugal or
other rotodynamic pumps which run at the same speed as the engine or motor; in such
situations they can be direct-coupled with a simple flexible drive coupling as in Figs. 72
and 105. Speed changing of up to about 4:1 can readily be achieved with vee-belts as
shown in Figs. 84, 99 and Fig. 106. Fig. 84 shows a two stage vee-belt drive where the
total speed change can be as much as 4:1 on each stage. In this situation the total
speed change is the product of the ratios for each stage. Where multiple vee belts are
needed on one drive stage, as in Fig. 99 (showing four in use), it is best to use matched
sets from a supplier, and always to renew all belts simultaneously so that they all share
the load effectively; a more modern and convenient type of belt is the so-called poly-vee,
which is similar to a whole lot of small vee-belts fixed together edge-to-edge. Flat belts
(made of leather) used to be common and they are coming back, sometimes today
made of synthetic materials, as they are more efficient with less friction than a set of vee
belts.

Fig. 84 Two stage speed reduction transmisson used in China to connect an electric
motor to a chain and washer pump

149
If a speed change greater than about 4 or 5 to 1 is needed, then an alternative to
multiple stages of belts (which introduce problems with belt adjustment) is to use
gearboxes. A right angle drive may be created to drive a vertical shaft borehole pump
(for example) either by using a 90 geared well head, or by using a twisted flat belt. To
be successful twisted belt drives need to have a generous distance between the pulleys
in relation to their diameters or excessive wear will occur.
Other mechanical transmissions commonly used are reduction gearboxes with a
pitman drive, similar in most respects to the windpump transmission of Fig. 109 They
consist of a rotary drive shaft which drives a single or pair of larger gear wheels via a
small pinion; the large gear wheels drive a reciprocating cross-head or pitman slider via
two connecting rods. The pump rod is connected to the cross-head or pitman.
Mechanisms of this kind can be used to connect a diesel engine or an electric motor to a
reciprocating piston pump. Other mechanical right-angle drives are illustrated by
reference to Figs. 94, 95 and 96, (the large size necessary for making a strong enough
drive from traditional materials is well-illustrated in Fig. 95).
When budgeting for a pumping system, it is important to know that the mechanical
transmission can cost as much or often more than the prime-mover, especially if a
geared or reciprocating well-head is used. The high cost is due to the mechanical
requirements for reliable operation being demanding and the volume of production
usually being much lower than for engines or electric motors.
An effective method of transmitting mechanical shaft power any distance is via a high
speed rotating shaft. This needs to be steadied by bearings at quite close intervals to
prevent the shaft "whirling" like a skipping rope, a phenomenon which causes intense
vibration and destruction of the shaft. Vertical drive shafts down boreholes, as much as
100m deep and running at 1 500rpm or more are commonly and successfully used,
although submersible electric multi-stage pumps are becoming a more popular solution.
iii. Electrical, hydraulic or pneumatic transmission

The use of a diesel-generating set (or wind-electric, solar-electric or hydro-electric unit)

as a prime mover allows considerable flexibility in transmission (literally) since electric
cable is all that is needed to link the prime mover to a motor-pump unit, (which can even
be submerged down a borehole as in Figs. 71 or 134 A).
Other options, which are technically feasible, but more rarely used are hydraulic or
pneumatic transmissions in which either a liquid (water or oil) or air are pumped through
pipes to drive a pump. Examples of hydraulic transmissions are given with the jet pump
in Fig. 75, or the positive displacement hydraulically activated pumps of Fig. 39. The air
lift pump of Fig. 76 is an example of a pumping system which requires pneumatic
transmission. Pneumatic diaphragm pumps are commercially available and tend to be
most commonly used for construction projects, with an air supply from mobile engine
driven air-compressors. They are not normally used for on-farm irrigation but there is no
technical reason why they would be unsuitable. However, hydraulic and pneumatic
transmissions tend to be inefficient and therefore such a system may have high running
costs.

150
4.1.3 Fuels and Energy Storage
Power sources need energy, whether it is fuel for an engine, wind for a windpump or
sunshine for a solar pump. The main difference is that the provision of fuel can usually
be arranged by the user, but nobody can make the wind blow or make the sun shine on
demand. There is therefore an obvious qualitative difference between wind and solar
powered devices which will only function under certain weather conditions and the rest
which generally can be made to operate at any pre-planned time.
Although the apparent randomness of wind or solar availability would appear to be a
serious disadvantage, in reality the energy available over a period of a few days in a
given location at a given time of the year does not vary much from year to year. The
problem is more one of covering a mismatch that can occur between the rate at which
energy is available and the duty cycle the farmer would like to impose. This can often be
overcome either by choice of technique or by including a storage facility.
In most cases, where the output required is water, the most cost-effective solution is to
introduce a storage tank between the pump and the field; (in some cases the field itself
can act as a storage tank). The other principal method for small scale energy storage is
to use lead-acid electrical batteries, but this becomes prohibitively expensive except
when small amounts of energy of less than about l-2kWh need to be stored. The costs of
tanks for storing water relate to their volume, while the costs of batteries relate to their
energy capacity; therefore, at low heads when large volumes of water may need to be
stored, but which involve little energy, electrical battery storage can be cheaper (and
less demanding in terms of land utilisation) than storage tanks. However, before
considering the substitution of batteries for storage tanks, it must be remembered that
batteries would need replacing a lot more often than the storage tank and would also
need much more maintenance.

151

4.2

HUMAN POWER

4.2.1 Human Beings as Power Sources

In the whole small-scale pumping field it is generally difficult to make precise statements
on pumping performance which are generally correct; nowhere is this more true than in
the field of human powered waterlifting devices and pumps. This is partly because
human capabilities are very variable, but also because there is a multiplicity of pumps
and water lifts of widely varying efficiency.
i. Efficiency as prime movers

People (and animals) derive their power from the calorific content of their food. Even
when physically inactive the human body requires energy to run its basic metabolic
functions, i.e. to power the heart and circulate blood, to work the lungs and digestive
system, etc. Energy for muscle power is then an extra requirement on top of this. A
typical food energy requirement is around 2400kcal, 10MJ or 2.8kWh per 24 hrs. Table
10 indicates the calorific values of various staple foods (after Leech [6]).
A person's muscular work capability per day is in the region of 200-300Wh/day. Human
beings therefore have an average overall efficiency in the region of 7-11% for converting
food energy to mechanical "shaft energy". This figure includes the basic metabolic
energy requirement; the efficiency of the muscles for short but strenuous efforts can be
as high as 20 or 30% [12] & [19], which compares well with internal combustion engines.
Table 10 THE CALORIFIC VALUES OF VARIOUS STAPLE FOODS
Staple Crop

Energy Content
MJ/kg

Kg/day to
provide 10MJ

Dayak rice

10.4-11.4

0.92

Iban rice

13.3

0.75

Tanzania rice

8.2

1.22

Maize (Africa)

4.2

2.38

Millet (Africa)

3.8

2.63

Sweet potato (Africa)

10.1

0.99

Cassava (Africa)

15.0

0.67

Yams (Africa)

9.5

1.05

Groundnuts (Africa)

7.2

1.39

After Leech, reference [6].

Kraatz, [12] (quoting Wood), gives a calculation of the food required by a man to
generate the energy needed to irrigate a crop. On the assumptions of a rice crop
needing 850mm of water in 120 days, with a yield of 600kg of rice from a 0.2ha plot, with
a 50% efficient water lifting device lifting the necessary water through 3m head, the
marginal cost of "fuelling" the human prime-mover for the irrigation pump was calculated

152
to be 35kg of rice or 6% of the expected total yield. An additional 35kg would be needed
to cover the basic metabolism of the person concerned, giving a total of 70kg or 12% of
the rice produced.
The Intermediate Technology Development Group's Water Panel [20] gave a rule of
thumb of a food requirement of 0.5kg of rice per MJ of hydraulic work, plus 0.012kg of
rice per day per kg body weight. In the earlier example, the hydraulic requirement was
50MJ, which under the rule of thumb just quoted demands 25kg of rice for pumping
effort, and a 60kg man would additionally need (0.012 x 60 x 120 = 86kg) of rice, giving
a total requirement of 111kg of rice, or 18% of the total crop produced.
Allowing for losses of rice, possible worse yields than that assumed, and the food
requirements of the farmer's dependants, it is easy to see how hard it is to generate a
surplus when cultivating staple crops on small land holdings. For example, if he has
three dependants and loses just 20% of his crop through various forms of wastage, the
farmer and his family will need to retain 60 to 90% of the harvest, depending on the
method of estimating rice requirements used. Slightly worse wastage or a larger family
would result in barely sufficient food for pure subsistance.
ii.

Productivity

Contrary to popular belief, human muscular energy is not cheap. The poor are forced to
use human power, usually because they cannot afford anything better, since the cash
investment required is minimized and therefore it is more "affordable" than other options.
As will be shown, almost any other source of power will pump water more cheaply
unless only very small quantities are required.
The human work capability is around 250Wh/day, so it takes four days' of hard labour
to deliver only one kWh - which a small engine could deliver in less than one hour while
burning less than one litre of petroleum fuel. So the farmer with a small mechanized
pumping system has the equivalent of a gang of 20 to 40 men who will work for a "wage"
or running cost equivalent to say 1 litre of fuel per hour; not surprisingly, any farmer who
can afford it will sooner choose to employ an engine rather than 20 to 40 men. This
argument can be turned on its head to show the high price of human muscle power, if
the "opportunity cost" or a real wage cost is assigned to human muscular labour; eg.
assuming a daily wage rate of US $1.00/day gives an energy cost of about $4.00/kWh.
Although this is a low wage for hard labour, even in some of the poorer countries, it
represents an energy cost that is significantly more expensive than even new and exotic
power sources such as solar photovoltaic panels.
There is an opportunity cost caused by diverting people from more important work to
pumping water; the best asset people have is brains rather than muscle; therefore, if
agricultural productivity is to improve and economic standards are to be advanced, it is
essential to introduce more productive power sources for all except the very smallest of
land-holdings.
iii.

Power capability

Muscle power can handle quite large "overloads" for short periods, but the power
capability diminishes if more than a few minutes of activity are required. The power

153
availability is also a function of the build, age, state of health and weight of an individual;
finally the ability to produce power depends on the nature of the device being worked
and the muscles that can readily be utilized. Table 11 A indicates power outputs that
may be expected for individuals of 20, 35 and 60 years age respectively over periods of
operation ranging from 5 minutes to 3 hours, after Hofkes [21], presumably with devices
allowing much of the body to operate. Table 11 B shows actual results measured by the
Blair Research Institute in Zimbabwe, [22].
Therefore, although the actual output from any human powered pump is not precisely
predictable, an approximate prediction can reasonably be made. Fig. 85 gives a set of
curves indicating the capability of from one to four people each providing 240Wh per day
of useful work through a pumping device with an efficiency of 60%. These curves
probably represent what is generally achievable under favourable circumstances and
indicate, for example, that the daily output per person if lifting water 5m, is about 12m3.
Table 11 POWER CAPABILITY OF HUMAN BEINGS
Age

Human power by duration of effort (in watts)

Years

5 min

10 min

15 min

30 min

60 min

180 min

20

220

210

200

180

160

90

35

210

200

180

160

135

75

60

180

160

150

130

110

60

PUMP 1. Pumping head 10.54m

time to fill 20 litre can
Age

Wt (kg)

Time (secs)

Mean Power
(watts)

34

173

12

14

54

74

28

14

54

77

27

16

50

69

30

18

55

59

35

20

68

70

29

29

82

100

21

33

65

55

32

47

75

48

43

PUMP 2. Pumping head 6.35m

Time to fill 20 litre can
Age

Wt (kg)

Time (secs)

Mean Power (watts)

33

51

29

11

31

48

28

154
11

31

68

18

10

55

60

20

10

33

66

19

14

37

55

22

PUMP 3. Pumping head 2.14m time to fill 20 litre can

Age

Wt (kg)

Time (secs)

Mean Power (watts)

17

61

6.9

10

27

44

9.5

14

64

36

6.6

19

54

39

11.0

38

57

34

12.0

155

Fig. 85 The number of people required to provide a specified quantity of water at

different lifts. The curves have been derived by assuming that a single person can
provide 60 watts of power for 4 hours per day and that the efficiency of the pump is 60%
iv. Ergonomics

The actual useful output from a person depends a lot on the way the water lift or pump
works; the most powerful muscles are the leg and back muscles while the arm muscles
are relatively weak, so conventional hand pumps are less effective at "extracting work"
from a person then a device like a bicycle. Moreover, the "ergonomics" of the design are
important; the operator needs to be comfortable and not contorted into some difficult
position, so the device should require a relaxed posture, with the user well-balanced,
and it should function best at a comfortable speed of operation. Utilization of the leg
muscles will also often allow the operator to throw his or her weight behind the effort in
order to gain further pedal pressure. Wilson [23] reported that a rotary hand pump was
improved in output by a factor of three, (300%), by converting it from hand operation to

156
foot operation. The same article also promoted the bicycle as a supreme example of
effective ergonomics; it uses the right muscles in the right motion at the right speed and
applies human power through a light but strong and efficient mechanism. Wilson makes
the very valid point that what is needed is a pump which is as well designed, strong,
efficient and easy to use as the bicycle. He quotes dynamometer tests as indicating that
the average cyclist works at 75W when cycling at 18km/h; if this output could be
produced while pumping, the following flowrates should be realisable at various lifts,
assuming a water lifting device of only 50% efficiency:
Head:
Flow:

0.5
27.5

1.0
13.8

2.5
5.5

5.0
2.2

10.0 m
1.1 m3/h

For this reason the most effective irrigation pumps are in fact foot operated. Also an
irrigation pump requires to be operated perhaps for several hours and therefore
efficiency and ease of use are crucial. Hand operated devices are easier to install and
can be lighter and smaller (since no one has to stand or sit on them and the forces that
can be applied will not be so great anyway). Where pumps are used for water supply
duties rather than irrigation, efficiency is less of a stringent requirement since any
individual user will generally only operate the pump for a few minutes per day to fill a few
small containers.
Therefore the criteria for defining a good human powered irrigation pump are
significantly different for those for a water supply pump and it may be a mistake to use
pumps for irrigation duties that have only been proved successful in water supply.

4.2.2 Traditional Water Lifting Devices

Many of traditional water lifting devices are particularly designed for low lift irrigation, and
they are often foot-operated since it no doubt became apparent that this was the best
method of harnessing human power.
The least-cost solution has always been a bucket or bag of water lifted when necessary
on a rope; Fig. 86. The best that can be said for this technique is that with small plots the
water can at least be applied with precision to individual plants, so at least efficient
conveyance and distribution can partially compensate for the inefficiency of the actual
water lifting. At low heads, the use of buckets and scoops, (see also section 3.3.1) led to
the development of the swing-basket, (Fig. 18), which can use two people and functions
more rapidly, although only through very low pumping heads such as from canals into
paddies. However, it is not an ergonomic device in that a lot of muscular effort goes into
twisting the body, there is much spillage, and also water is lifted much higher than
necessary. Nevertheless, two young boys using this technique, for example in
Bangladesh, can complete 2 000 swings without a rest, according to Schioler [91].

157

Fig. 86 Rope and bag water lift from a dug well (example from The Gambia)

158

Fig. 87 Relative performance of the swing basket and the Dhone (after Khan [25])
An improvement, obtained at the price of some slight complexity, is the use of
suspended or pivoted devices such as the supported scoop (Fig. 19) and some which
are also balanced such as the dhone (or dhoon) see-sawing gutters, (Fig. 20) or the
counterpoise lift (or shadoof), (Fig. 21). These are no longer portable since they need to
be installed on a site, and they require a supporting structure which has to be attached
securely to the ground, but they are far more efficient than such primitive devices such
as buckets or swing baskets, as indicated by the performance curves taken from Khan,
[25], in Fig. 87. This shows that a single dhone will lift 7.5 litre/sec at a lift of 0.75m (115
gall/min at 30 inches), which reduces to about 2 litre/sec (30 gall/min) at 1.5m (or 60
inches) head. Therefore the dhone will move more than twice as much water as a swing
basket at low lifts, moreover using the power of only one person rather than two. Khan
makes the point that many Bangladesh farmers try and use a single stage dhone at too
high a lift, and lose a lot of performance as a result; the optimum lift per stage is
approximately lm (40in).
Table 12, adapted from Khan, [25], indicates how widely the dhone is used in
Bangladesh, a country with very large areas offering the possibility of shallow lift
irrigation, and it also shows clearly how much of an improvement the dhone is over the
swingbasket. The same table also indicates the characteristics relating to the "dugwell",
a counterpoise lift and shallow hand-dug well, and the "Mosti", which is a cast iron
handpump mounted on a tubewell.

159
Despite being a considerable improvement on the swingbasket, the dhone is probably
not as efficient as the various rotary devices, described in Section 3.6, although it
represents a good compromise between retaining simplicity and low-cost while achieving
a useful output.
Table 12 COMPARISON OF VARIOUS WATER LIFTS IN BANGLADESH1
Dhone

Swing

Dugwell2

Mosti3

ha x 1000

392

65

percentage of total

35

0.8

Surface

Surface

Pumped

Pumped

Max discharge m3

7.5

2.3

0.6

0.8

Pumping head m

0-1.5

0-1.8

0-4.5

1.5-6.0

Capital cost US$

20.00

1.33

10.00

80.00

1.6-2.0

.4-.6

0.3

0.2

Area irrigated

Water source

Working life years Command

area
ha of dry season paddy

20

Adapted from Khan, [25]

Hand dug well with counterpoise bucket lift

Manually operated shallow tubewell for irrigation

Uses industrially made headpump (No. 6. Carbiron pump)

Rotary devices tend to be easier to work as they generate a smooth output and they
therefore often can be driven with a comfortable pedaling motion of either the arm or,
better, the legs. The various flash-wheels and ladder or dragon spine pumps are
generally leg-powered, (see Sections 3.6.5 to 3.6.7 plus Figs. 51 and 52), and they can
often readily be powered in this way by several operators. While today Archimedean
screws are usually hand operated (Fig. 48), in Roman times they were walked on rather
like the treadwheel in Fig. 51, which no doubt was easier for the operator and more
productive.
It is quite possible that further useful improvements could be made with some of the
simple traditional water lifts. This is an area where study and experimentation, perhaps
as part of technical educational programmes or by NGOs working in the field, might yield
useful results.
In many cases human powered devices offer the best means to initiate small scale lift
irrigation because of their low first cost. However, in the longer term it is to be hoped that
small farmers will be assisted to advance towards more productive pumping techniques,
which inevitably require some mechanization.

160
4.2.3

Handpumps

Handpumps, although less productive than footpumps, are the most common form of
industrially manufactured manually operated water lift, and for that reason are very
widely used. The classic design of piston pump or bucket pump is shown in Figs. 29 and
30. Most of these pumps were developed for use by a family to provide water for
themselves and their livestock, rather than for irrigation. The problem when pumps of
this kind are used for irrigation is the intensity with which they are used compared with
their use for water supply; instead of pumping for a few minutes per day they have to be
used for several hours per day, which naturally tends to shorten their useful lives
considerably and also to increase the incidence of breakages.
The forces involved in driving a piston pump with a lever have already been discussed in
Section 3.5.1. So far as handpumps are concerned, Hofkes [21] indicates the following
maximum heads as being generally suitable for comfortable operation of various
common sizes of handpump:
cylinder diamete
(mm)
(in)
51
2
63
2i
76
3
102
4

maximum head
(m)
(ft)
25
75
20
60
15
45
10
30

The load at any given head can be reduced by shortening the stroke of the pump, but
the above recommendation presumably applies with typical strokes in the region of 150
to 300mm.
Hofkes [21] and McJunkin [26] give the nomograph of Fig. 88 as a method for
determining handpump discharge. The method for using this is to rule a pencil line
between the stroke length that applies (250mm or 10in in the example) and the expected
pumping frequency (40 strokes per minute in the example). Then if another pencil line is
ruled from where the first line crosses the "pivot line" through the appropriate cylinder
diameter (76mm or 3in in the example), the discharge is given on the left (46 litre/min
or 12 US gall/min in the example). No allowance is made for "slippage" or leakage of
water which will result in the discharge being less than the swept volume, so the result of
using this nomograph is the maximum flow that might be expected; it may therefore be
more realistic to reduce the result obtained by 10-20%.

161

Fig. 88 Nomograph for calculating hand pump discharge (after Hofkes [21]and
McJunkin[26]).
There have been many failures of handpumps in the field, especially in water supply
projects where no particular individual readily takes responsiblity for the pump. In some
countries communal handpumps developed a bad reputation as a result. Mention should
be made of the major international UNDP/World Bank Global Handpump Project which
seeks among other things to identify good handpumps for development use (mainly for
village water supplies); this project has achieved a lot in this field and yielded a number
of publications of relevance, notably [27],
Fig. 29 shows some of the key features of a typical handpump. Some of the weak points
are the lever and fulcrum mechanism and the pump column itself; these can crack
through metal fatigue or due to the use of poor quality castings. The pump in Fig. 29 is
good in having a bracing strut to support the pump body, but it is bad in having the pivot
bolt for the handle passing through the middle of the most highly stressed part of the
pump lever. A better method of pivoting the hand lever is to have the pivot bearing
passing through a lug below the lever arm to avoid weakening the arm at that point.

162

Fig. 89 Rotary drive hand pump (The Gambia)

163

Fig. 90 Rower pump (part sectioned) (after Klassen [28])

A further problem with hand lever pumps is wear and tear resulting from "hammering" in
the drive train; this can be caused by worn pivots and bearings causing backlash, which
will cause impacts if the operator lifts the handle too rapidly so it tries to overtake the
piston and pump rod on the down stroke and then suddenly takes up the load again and
also by users causing the hand lever to hit the end stops at the end of its travel. Also the
need for the operator to constantly raise and lower a heavy lever (plus their arm(s))
wastes energy. Therefore rotary-drive pumps, in which the piston is driven by a crank
from a rotating drive wheel are often easier to work (see Figs. 31 and 89). Here a
flywheel smooths the fluctuations and thereby makes the pump easier to operate,
especially for long periods, because the cyclic loading involved in accelerating discrete
cylinder-volumes of water up the rising main will be absorbed by the flywheel's
momentum and therefore not be felt by the operator. Rotary drive pumps of this kind
have a further advantage as they cannot suffer from damage common with lever pumps
from the operator hammering the end stops by moving the lever to far, or by the pump
rod momentarily going "slack" due to the operator lifting the hand lever to rapidly.
The main disadvantage with rotary drive pumps is that they are generally relatively
heavy and expensive due to their massive flywheel and crank mechanism, plus the
supporting column that is needed. However, for pumping large quantities of water, as
are required for irrigation, the improved ergonomics of the rotary drive is probably more

164
advantageous than it is with water supply duties where no individual is likely to need to
pump for more than a few minutes at any one time.
Ideally, any piston pump used for irrigation needs to use more than just the arm
muscles. An attempt to produce such piston pump has been made in Bangladesh by the
Mennonite Central Committee, with help from Caritas, according to Klassen; [28]. This
pump is known as the "Rower Pump" because it is inclined at around 30 to the
horizontal and operated with a rowing action (see Fig. 90). It is basically a simple and
cheaply constructed pump, using 2" PVC pipe as the pump cylinder. The Rower Pump is
claimed to pump 50% more water than a standard UNICEF No.6 lever hand pump used
in MOSTI installations, (Manually Operated Shallow Tubewell for Irrigation), and the
improvement is proportionately greater with higher lifts. However the main performance
advantage of the rower pump is attributed by Klassen not so much to the action which is
used (although this is claimed to make it easier to operate and might therefore allow
longer periods of use per day) but because of a suction air chamber which smooths the
flow into the pump (see Sections 3.5.4 and 3.7.2 for explanations of the use of airchambers). Without the air chamber the performance falls to a level almost identical with
that of the UNICEF pump, while if an air-chamber is fitted on the suction side of the
UNICEF pump its performance is also enhanced to a similar level to that of the Rower
Pump. Average performance quoted by Klassen for these two pumps, when pumping
through heads in the range 5-6.5m (17-21ft), is as follows:
MCC Rower Pump

UNICEF No. 6 Pump

without air chamber

with air chamber

without air chamber

with air chamber

1.75

2.8

1.8

2.7 m3/h

7.7

12.3

8.0

11.7 US gall/min

This work is a good example of the kind of useful innovative developments that can be
successfully pursued by NGOs and educational institutions in this important field, as
suggested earlier. It would be interesting to see the effect of adding air-chambers to the
performance and ease of use of other types of handpump in use in the field.

4.2.4

Handpump Maintenance

The problem with industrially manufactured handpumps, as opposed to "self-build"

devices like dhones or counterpoise lifts, is that they are dependent on spare parts
which cannot easily be locally improvised. They also tend to need a certain amount of
preventive maintenance if premature failure of components or impaired performance is
to be avoided. However, they probably represent one of the only routes readily available
to channel development funds into the widespread deployment of water-lifting
equipment, since the alternative of using traditional hand-made devices is usually only
practicable where there is an existing tradition make them and where the conditions for
their use are right (eg. very low lifts from surface water).
Many water supply programmes using large numbers of handpumps have suffered
serious difficulties with pump failures which have generally been attributed to poor
maintenance. In some cases the failures have also been due to the use of poor pump

165
designs which lack the capacity to survive intensive use. The kinds of problems
experienced with hand pumps include:
a. poor quality of pump design and manufacture. This has partly resulted from
manufacturers trimming the weight (and hence the cost) of components and
generally degrading well tried designs in seeking to offer acceptably low bids in
the absence of proper specifications to the procurement agencies;
b. iron and steel plain bearings and journals with poor fits and large clearances are
provided; these properly should require very frequent lubrication which is
impossible to provide, so rapid wear occurs;
c. the great variety of pumps in use leads to difficulties in finding the right spare
parts;
d. very limited record keeping and feedback from the field makes it difficult to
analyse the reasons for failures and to introduce remedial measures;
e. limited maintenance skills and equipment make it difficult for local people to
undertake even basic overhaul operations, while lack of transport and poor
communications make it difficult to summon help from a central source.
Attempts have been made to overcome some of these problems by introducing either a
centralized system in which a maintenance team tours around repairing and maintaining
a few dozen pumps in a district (this has often proved ineffective and expensive in
practice). The other option being advocated is a "two tier" maintenance system, in which
a central agency carries out the original installation, and provides a source of spare
parts, training, transport, etc., but local people are trained to carry out routine repairs and
maintenance.

166

4.3

ANIMAL POWER

The advantages of animal power over human power are twofold. First; draft animals are
five to ten times more powerful than humans, so they can pump more water in a shorter
time which tends to make the irrigation operation more efficient and productive. Second;
by freeing the operator from having to work the water-lifting device, he can often manage
the water distribution system more effectively. In effect, the use of an animal provides
the equivalent power of several people, generally at a fraction of the cost.
Some 200 million draft animals are deployed in developing countries [29], and these
have an aggregate power capacity of about 75 000 000kW and a capital value in the
region of US $20 billion. The majority of these animals are used in southern and south
east Asia; 80 million draft animals are in use in India alone. Any mechanization
programme to replace these animals would obviously have to be extremely large; but
there is more immediate scope for improving the efficiency with which animals are used.
Animal powered irrigation is almost exclusively practised using traditional water lifting
techniques pre-dating the industrial era. Although some attempts to produce improved
mechanisms for the utilisation of animal power have been made in certain areas during
this century, there is little (if any) tendency to introduce animal powered water lifting
anywhere where it has not been traditionally practised, although there seems good
reason to believe it could usefully replace human labour for irrigation in many parts of
the world where it is not already used. Instead, the trend has been either to make the
quantum leap to full mechanisation using engines or mains electricity, or to make no
attempt to improve on human power. An interesting exception is the "Water Buffalo
Project" in Thailand where the use of buffalo powered persian wheels is being
encouraged, [24].
The main disadvantage of animal power is that animals need to be fed for 365 days of
the year, yet the irrigation season usually only extends for 100 or at most 200 days of
the year. In areas where water is close to the surface, and hence irrigation by animal
power is feasible, there are usually high human population densities and a shortage of
land. Since draft animals consume considerable volumes of fodder, a significant
proportion of the available land can be absorbed simply to support the draft animals.
Therefore it probably would be difficult to justify the use of animals for irrigation pumping
alone, but generally there are other economic applications for them, such as transport,
tillage, and post-harvest duties like threshing or milling which allows them to be
employed more fully than if they were used exclusively for irrigation. In India and other
countries where animal powered water lifting is widely practised, it is normal for the
same animals to be used for transport and for tilling the land. They are often fed with
agricultural residues, or are allowed to graze on fields left fallow for a season as part of a
crop rotation.
Animals also serve other purposes; they are a non-monetary form of collateral which is
important in the village economy in many regions, they produce byproducts such as
leather, meat and milk and of course in the Indian sub-continent in particular, their dung
is widely used as a cooking fuel. So mechanization of irrigation pumps in lieu of draft
animals is not necessarily a straightforward alternative.

167
4.3.1

Power capabilities of various species

Fig. 91 gives the approximate water lifting capability of different numbers of oxen,
assuming 60% efficiency from animal to water lifted, as might be achieved by one of the
better types of water lifting device.

Fig. 91 The number of oxen required to provide a specified quantity of water at different
lifts. The curves have been calculated by assuming than an ox can provide 350 watts of
power for 5 hours per day, and the efficiency of the water-lifting device is 60%
Table 13 POWER AND DRAWBAR PULL OF VARIOUS ANIMALS
animal
weight (kg)
draft force (kg) typical speed (m/s)

power (W)

heavy horse

700-1200

50-100

0.8 -1.2

500-1000

light horse

400-700

45-80

0.8 -1.4

400-800

168
mule

350-500

40-60

0.8 -1.0

300-600

donkey

150-300

20-40

0.6 -0.8

75-200

cow

400-600

50-60

0.6 -0.8

200-400

bullock/ox

500-900

60-80

0.5 -0.7

300-500

camel

500-1000

80-100

0.8 -1.2

400-700

buffalo

400-900

60-100

0.5 -1.0

600-1000

The typical power capabilities of various commonly used draft animals are given in
Table 13, based on [1] [30] [24] and other sources.
Draft animals obviously require rest just as humans do, so it is common practice to work
them on about three hour shifts, with a rest in between. 10 or 12 hours per day in total
may be worked when necessary.
Hood, [31], discussing typical draft horses as used in the USA at the end of the last
century advocated the use of efficient mechanisms for coupling horses to water-lifts. He
suggested that a team of two horses with an efficient water lifting device can typically lift
0.22 acre.feet per day through a head of 15ft, (270m3/day through 4.6m). He also cited
various experiments completed at that time in Madras, India, (using bullocks), as an
example for American farmers to emulate. In one of these, a device known as the
Stoney Water Lift, was tested and shown to have an animal-to-water mechanical
efficiency of about 80%. This device was, in effect, an engineered version of the
traditional circular mohte with two balanced buckets, (see Fig. 93). Using a single
Nellore bullock in a series of tests, it lifted between 2000 and 2500 Imp.Gall/hr through
22-23ft (9-11.25m3/h through 6.7-7.0m). It was pointed out that although a high
instantaneous efficiency was achieved by this two bucket lift, in the tests it was found
that the animals were only actually performing useful work for 60% of the duration of the
tests, so the utilisation was not as high as may theoretically be feasible; clearly the
management of an animal powered water lift has a major influence on the daily
productivity.
An important point emphasised by Hood is that all draft animals work best when
subjected to a steady load which matches their pulling capability, (although a horse, for
example, can throw about one third of its weight into pulling for short periods - i.e.
approaching three times the force it can sustain for a long time). Therefore, devices are
needed which shield the animal from any cyclic loadings such as is experienced when a
pump is driven by a crank, (see explanation of cyclic torque requirements of piston pump
in Section 3.5).

4.3.2

Food Requirements

Birch and Rydzewski, [30], show that a cow in Bangladesh can be fed solely with forage
and agricultural residues, (although the latter have a value equivalent to US $0.30 per
day (1980) which is not a trivial sum) and that the same animal requires the residues
from 0.77ha of double-cropped agricultural land. Because of the general shortage of
land and residual fodder in Bangladesh, it is unusual to use animals for water lifting in
that country.

169
The same authors carried out a similar calculation in relation to Egypt, where one
hectare of land is needed to produce sufficient residue to support one animal, and the
daily diet is valued at US $0.50.

4.3.3 Coupling animals to water lifting systems

The original method of using animals to lift water was some device such as the mohte,
(Figs. 22 and 92). Here animals walk in a straight line down a slope away from the well
or water source while hauling water up in a bag or container. Traditional mohtes used a
leather bag to collect the water, but in recent years more durable materials such as
rubber truck inner-tubes (or more rarely steel oil barrels) have been used.

Fig. 92 Cross-section view of a Mohte (after Schioler [24])

170

Fig. 93 Circular Mohte utilizing two buckets with flap-valves in bottom

The mohte is simple to implement; the only mechanical component is a pulley and the
only structure is the frame to hold the pulley. However this key mechanical component is
more complicated than might be expected due to the load on it, and demands
considerable craftsmanship [24]. The instantaneous efficiency of the mohte is high while
the animal(s) are pulling; a disadvantage is the need to reverse the animals back up the
ramp to lower the bucket. Sometimes two teams of animals are used, so that one is led
back to the top while one team descends; in order to do this, a man is needed at both
ends of the ramp to harness and un-harness the animals at the end of each cycle. Using
two sets of animals and, generally, two men and a boy, practically doubles the output
(Molenaar, [1]), but even so inevitably no action occurs during the harnessing and unharnessing process.
The downhill slope helps use the weight of the animal partially to balance the weight of
the full water container; it also applies a reasonably constant load to the animal for most
of the journey down the ramp, from when the container leaves the surface of the water to
when it starts to be tipped.
An improved version of the mohte, sometimes used in parts of India and (a few) in Sri
Lanka, is the "circular mohte"; Fig. 93. This involves attaching the animals to a sweep so
that they can walk in a circle thereby allowing the animals to work continuously with less
supervision. Because their weight can no longer partially balance the load, as they have
to walk on level ground, two buckets are used so that the empty one descends while the
full one comes up; this at least balances the weights of the buckets and means that the
water being lifted is the only large out-of-balance force to be handled. The main.problem
with this device is that the load on the sweep is cyclic, as the pull on the sweep by the

171
chain will not be felt by the animals when the chain acts parallel to the sweep and it will
reach a maximum when the chain is at right angles to the sweep; therefore the animals
will have a tiring sinusoidal load to cope with. Also, the various pulleys and supporting
posts need to be robust and well anchored, as the forces are quite large.
The Persian wheel, (Figs. 23, 94 and 95) is a great improvement on the mohte, as its
chain of buckets imposes an almost constant load on the drive shaft to the wheel.
Persian wheels are usually driven by some form of right angle drive, such as in Figs. 94
and 95. The first is the most common, where the drive shaft from the secondary gear is
buried and the animals walk over it; this has the advantage of keeping the Persian wheel
as low as possible to minimise the head through which water is lifted. The second
example is a traditional wooden Persian wheel mechanism where the animal passes
under the horizontal shaft. The sweep of a Persian wheel carries an almost constant
load and therefore the animal can establish a steady comfortable pace and needs little
supervision.
In Egypt, the sakia (see Section 3.4.3 and Fig. 26) is commonly used for low head
applications instead of the Persian wheel, and it is driven in a similar manner via a
sweep. It has the advantage of applying a constant load to the animal, but is more
efficient at very low heads.
The next development was for the sweep drive gear to be "industrialised" and
manufactured in large numbers from iron or steel to include an engineered set of gears.
Fig. 96 shows a mule harnessed to such a device in order to drive a chain and washer
(or paternoster) pump which are widely used in China, where millions of liberation
pumps, many of which were animal powered, were produced as an intermediate stage
between human pumping and full mechanisation. The liberation pump includes a
mechanism driven by a sweep that is elegantly simple and made from steel castings;
(see Figs. 53 and 90). As detailed in Section 3.6.7, the liberation pump is capable of
achieving an efficiency of approaching 70% in the animal powered form illustrated, and it
is compact enough to fit into quite a narrow well.

Fig. 94 A bullock-driven Persian wheel of the conventional chain and bucket type which
is widely used in many parts of the world, particularly the Near East and Southern Asia

172

Fig. 95 Camel-driven Persian wheel showing over-head drive mechanism (ref. Schioler
[24])

173

Fig. 96 Animal-powered Chinese Liberation Pump A hand-operated version is shown in

Fig. 53 and a motorized one in Fig. 84 ,
Another not unusual concession to modernisation is the use of an old motor vehicle
back-axle embedded in a concrete pillar as a means of obtaining a right angle drive from
an animal sweep; an example of this is indicated in Fig. 97, where a donkey is shown
linked to an Archimedian screw, (see also Section 3.6.3 and Fig. 48). Although in the
example illustrated the matching problem is not always solved satisfactorily, according to
Schioler, [24], Again this is a compact and potentially efficient mechanism, as the
Archimedean screw applies a completely steady load to the animal. It is clear that the
same mechanism improvised from a car back axle could equally easily be coupled to a
sakia, to a Persian wheel or to a chain and washer pump.
A new development in this field has been the appearance of a prototype animal
powered, but industrially manufactured double-acting diaphragm pump from Denmark.
This has the advantage of having the sweep direct coupled to the pump, which acts as a
suction pump. Therefore it is only necessary to bury the pipes carrying the water rather
than a drive shaft and this device can be located up to 80m away from the water source,
which may in some cases be an advantage. This "Bunger" sweep pump is claimed to lift
100m3/day using two animals.

174

Fig. 97 The back axle from a car used as an animal-driven power transmission for an
Archimedean screw pump (see also Fig. 48)
Although a sweep driven device avoids the problem of reversing animals as with a
mohte, it suffers from the disadvantage that by forcing an animal to walk in a circle, even
though the load may be steady, the tractive effort or pull is reduced to 80% of that which
is feasible when the same animal walks in a straight line; Hood; [31]. A mechanism
which can apply a steady load to the animal, but in a straight line, would be better. The
simplest device of this kind is the tread-wheel; in some respects the principle is
analagous to a mohte, with the animal "walking on the spot" inside a large wheel, rather
than up and down a ramp. This principle was taken further at the end of the nineteenth
century in Europe and the USA through the use of "Paddle wheel" animal engines, in
which a horse would be harnessed on an inclined endless belt which it would drive with
its feet. The disadvantage of these animal carrying devices is that a mechanism is
needed which not only has to transmit the maximum draw-bar pull of the animal, but
additionally has to carry the full weight of the beast. Therefore a massive and robust
construction is necessary which inevitably is expensive, and this probably is an example
of where the search for maximum efficiency produces diminishing returns and therefore
is counter-productive.

175

4.4

INTERNAL COMBUSTION ENGINES.

The almost universal power source in the 5-500hp range, in areas having no mains
electricity, is the internal combustion (i.c.) piston engine in either of its two main forms;
the petrol (gasoline) fuelled spark ignition engine (s.i .) or the diesel fuelled compression
ignition engine, (c.i.).
The main reasons for the widespread success of the i.c. piston engine are its high
power/weight ratio, compact size and instant start-up capability, which led to its general
adoption for powering motor vehicles in particular, and small isolated machinery and
boats generally. Mass production, and decreasing fuel prices (until 1973) made small
engines based on designs used for motor vehicles both cheap to buy and inexpensive to
run. Since the 1973 and 1979 oil crises the trend of declining petroleum prices has
generally reversed, although during the mid-1980s we have again a period of declining
oil prices which may even last a few years. But even though crude oil prices have again
declined (in US dollar terms), petroleum fuels frequently remain in short supply in many
developing countries due to inability to finance sufficient oil imports. Almost everywhere
the most serious supply shortages are most prevalent in the rural areas, in other words,
precisely where farmers are in need of power for irrigation pumping. It is this actual or
potential fuel supply problem which makes the alternatives to petroleum fuelled engines,
reviewed later in this chapter, worth considering at all; the great versatility of the i.c.
engine makes it exceedingly difficult to improve on it from any other point of view.

4.4.1 Different Types of I.C. Engines

This is not the book to describe the technicalities of small engines in detail; there are
many standard textbooks on this subject and manufacturers and their agents will
normally provide detailed information in the form of sales literature. However, when
comparing the many available engines, it soon becomes apparent that there is a variety
of different types of engine available even at one particular power output.
i. Diesel versus petrol/kerosene

The two main categories of engine to choose from are c.i. or diesel engines and s.i. or
petrol (gasoline), kerosene or 1.p.g. fuelled engines; (1.p.g. - liquified petroleum gas
supplied in cylinders and usually primarily propane or butane). C.i. engines ignite their
fuel by the heating effect when a charge of air is compressed suddenly enough; finely
atomized fuel droplets are sprayed at very high pressure into the cylinder through the
injector nozzle at the appropriate moment when the temperature is high enough to cause
ignition. Spark ignition engines, on the other hand, work by mixing vapourized fuel with
air, compressing the mixture and then igniting it at the correct moment by the electrical
discharge of a spark plug set into the cylinder. Diesel engines therefore need to be
heavier and more robust in construction to allow the high pressures needed to cause
compression ignition to be sustained, and they also require a high pressure metering
and injection pump to force the fuel in the right quantities at the right instant in time
through the injector nozzle. The diesel injection pump and nozzles are built to a high
level of precision and are therefore expensive components despite being massproduced; they also depend on clean fuel and careful maintenance for reliable operation.

176
Therefore, petrol/gasoline and kerosene engines are inevitably lighter, more compact
and usually cheaper than diesels. Although diesel engines are inherently more
expensive to manufacture, they compensate for this by being more efficient, more
reliable and more long-lasting (but more complicated to maintain in good running order).
The main reasons for their better efficiency are firstly, the higher compression ratio
allows a diesel to "breathe more deeply", to draw in more air per stroke in relation to the
size of combustion space; secondly, fuel injection allows the diesel readily to run on a
leaner fuel/air mixture than the equivalent s.i. engine. A spark ignition engine cannot be
designed to run at such a high compression ratio, or the fuel/air mixture would ignite
prematurely causing "knocking" or "pinking". Another less well known advantage of the
diesel is that diesel fuel is 18% "richer" in energy than gasoline per litre (mainly due to its
higher density); Table 14 indicates the calorific value of the three main petroleum fuels.
Since fuel is generally bought by the litre (or some equivalent volume measure such as
gallons), rather than by its weight or by its energy value, you can buy 18% more energy
per litre of diesel than with petrol (gasoline).
Table 14 COMPARISON OF DIFFERENT PETROLEUM-BASED ENGINE FUELS
Units

Petrol/ gasoline

Paraffin/ kerosene

Diesel oil/ gasoil

MJ/1

32

36

38

MJ/kg

44

45

46

kWh/1

10

11

kWh/kg

12

12

13

hp.h/US gall

45

51

54

hp.h/lb

4.1

4.2

4.5

Therefore the diesel is generally to be preferred as a power source, in terms of efficiency

and reliable operation for long periods per day. For pumping applications, however, the
choice of petrol or diesel relates largely to the scale of pumping required. Where a small,
lightweight, portable system is needed which will only be used for one or two hours per
day, and where simplicity of maintenance is important, and where "affordability" matters
-i.e. the farmer has only the minimum capital to invest, then an i.e. gasoline or kerosene
engine may be best and for that reason is frequently used.
It should be explained that the kerosene engine is similar to a petrol (gasoline) engine;
indeed most kerosene engines need to be started and warmed up on petrol, because
kerosene will not vapourize adequately in a cold engine. Many kerosene engines have a
separate compartment in their fuel tank for a small supply of petrol and a tap to switch
the fuel supply from petrol to kerosene once the engine is warm; it is also important to
switch back to petrol a few moments before stopping the engine so that the carburettor
float chamber is refilled with petrol ready for the next time the engine has to be started.
Some farmers start kerosene engines simply by pouring petrol into the air intake, but this
practice is not to be recommended as it can cause a fire. The advantage of a kerosene
engine is that kerosene is normally available for agricultural purposes in an untaxed,
subsidised or lightly taxed form and it also contains approximately 10% more energy per
litre than petrol. The latter also usually carries a motor fuel tax in most countries as it is
mainly used for private cars; therefore fuel costs for kerosene are generally much lower
than for gasoline. Kerosene is also much less dangerous to store in quantity as it is

177
much less easily ignited. The kerosene supply is also used for lighting and cooking fuel
in many rural households and is therefore a more generally useful fuel.
Table 15 compares the general attributes of the three main i.e. engine options. It should
be noted that diesels are sub-divided into two main categories; "low speed" and "high
speed". The former run at speeds in the 450-1200 rpm range and tend to be much
heavier and more expensive in relation to their power rating than the latter which
typically run at speeds in the 1200-2500 rpm range. The slow speed diesel tends to have
a much longer operational life and to be better suited to continuous operation, or long
duty cycles, but its initial purchase cost is much higher.
Table 15 COMPARISON OF SMALL I.C. ENGINES
Petrol/
gasoline

Paraffin/
kerosene

High speed

Low speed

Average fuel to shaft

efficiency (%)

10-25

10-25

20-35

20-35

Weight per kW of rated power

(kg)

3-10

4-12

10-40

20-80

Operational life (typical)

2000-4000h

2000-4000h

4000-8000h 8000-20000h

Running speed (rpm)

2500-3800

2500-3800

1200-2500

450-1200

0.5-4

3-6

2-10

6-24

1-3

1-3

2-15

2-15

Typical daily duty cycle (h)

Typical power ratings useful
for small to medium irrigation
(kW)

Diesel oil/gasoil

ii. Relationship between size, speed and durability

A general characteristic of all i.c. engines is that the smaller and lighter they are for a
given power output, the lower will be their initial purchase price (which correlates to
some extent with the weight) and the shorter will be their useful life. This is because a
high power/weight ratio is normally achieved by running an engine at high speed; the
faster an engine runs, the more air/fuel mixture it can consume and the greater will be
the energy delivered. However, a faster machine will wear out quicker simply because its
moving and rubbing components travel further in a given number of hours of use. There
is therefore a tradeoff between heavy, expensive and slow engines on the one hand,
and cheap and fast ones of the same power rating. Therefore, small, lightweight engines
are recommended for such duties where portability and low first cost are important. In
most cases, especially if the engine is part of a fixed installation and to be used for
lengthy duty cycles, it will generally be worth investing in a suitably heavy and slower
machine in the interests of achieving better reliability and a longer operational life. In
general light s.i. engines are restricted to duties requiring less than 500 hours running
time per season.
iii. Derating

If an engine is run continuously at its Rated Power, premature wear will occur. All
engines therefore require to be derated from the manufacturer's rated power (which is

178
the maximum power output the engine can achieve for short periods). Small engines are
usually derated to about 70-80% of their rated power; eg. a 5kW rated engine will be
necessary to produce a continuous 3.5-4.0kW.
The main reason for derating an engine is to prevent premature wear, but also the
optimum efficiency for most engines is achieved at a speed corresponding to about 7080% of its speed for maximum power. Therefore, derating an engine usually improves its
specific fuel consumption (the fuel required per unit of output).
Further derating is necessary at high altitudes or at high ambient temperatures;
recommendations to this effect are usually made by the manufacturer. Typically a further
10% derating is recommended for each 1 000m above sea level, plus 1% for each 5C
temperature rise above 16C at the engine air intake. Therefore at 2 000m altitude and
an ambient temperature of 26C it would be necessary to derate an engine by say, 0.8
(generally) times 0.8 (for altitude) times 0.98 (for temperature) which totals .63 or 63% of
rated power which would be the correct load to apply. Therefore a 2 O00W load would
require an engine nominally rated at 2 000/0.63 or 3.2kW (4.3bhp) under those
conditions.
Excessive derating is to be avoided, as (particularly with diesels) running at a fraction of
the design power tends to cause coking of the cylinder. Also, the engine efficiency will of
course be much poorer than normal under such conditions.
iv. Four-stroke or two-stroke

Both s.i. and c.i. engines can be designed to run so that ignition takes place either every
other revolution, (four-stroke or four-cycle) or every revolution (two-stroke or two-cycle).
The four-stroke s.i. engine tends to be more efficient as the "non-firing" revolution gives
more time for inducing a fresh charge of fuel and also for effectively driving out the
exhaust gases from the previous firing stroke (s.i. two-strokes tend to be less well
scavenged of exhaust). Two-stroke diesels do not suffer an efficiency penalty in the
same way, but are not generally available in the small size range of relevance for small
scale irrigation. The two-stroke s.i. engine tends to be high revving and lightweight; it
usually has fewer components than a four-stroke and therefore is cheaper to
manufacture; typical applications are as moped engines. They are less suitable for
irrigation pumping than four-strokes as they use more fuel and wear out more quickly.
Most s.i. two-strokes use the downward movement of the piston into the crankcase to
displace the air/fuel mixture into the cylinder; in which case it is not possible to lubricate
the engine with a separate oil supply and the lubricant is mixed with the petrol, (twostroke mix). This removes the need for oil changes, but is wasteful of lubricant, tends to
cause a smoky exhaust, causes the need for more frequent "decokes" (de-carbonization
of the cylinder head) and introduces a risk of damage caused by an inexperienced
operator failing to mix sufficient lubricating oil with the fuel, or using the wrong type of oil.
For these reasons, two-stroke s.i. engines are tending to be phased out and replaced by
four strokes.
v. Air or water cooling

About one third of the heat produced when the fuel is burnt has to be dissipated through
the walls of the cylinder and through the cylinder head; the two methods generally used

179
for removing this heat and preventing the cylinder overheating are either by surrounding
the cylinder with a water jacket which has water circulated through it and a separate
radiator, or by having many cooling fins on the cylinder (to increase its effective surface
area), and blowing air over the fins with a fan driven off the engine. A few small, lowpowered, and old fashioned low-speed stationery engines have a water jacket with an
open top and keep cool simply by boiling the water, which needs to be topped up from
time to time, but most modern liquid cooled engines have their coolant circulated by a
pump just like car engines.
Each method of cooling has its pros and cons. Water cooled engines tend run slightly
quieter and their engine temperature is more easily regulated through the use of a
thermostat, than with air cooling. However, with water cooled engines, internal corrosion
can occur and also water can leak out, evaporate or freeze. This last problem can be
prevented by the use of anti-freeze (ethylene glycol) mixed with the cooling water in
winter; most anti-freezes also contain corrosion inhibitor and are therefore useful to add
to the cooling water even in climates where freezing is not likely to occur. Loss of coolant
generally causes severe engine damage if the engine is allowed to continue running in
that condition; various safety devices are available either to warn of overheating (caused
by loss of coolant or for any other reason) and in some cases to automatically cut off the
fuel supply and stop the engine. Air cooled engines obviously cannot lose their coolant,
but it is important to ensure that their cooling fins do not get clogged with dust or dirt and
that any cooling fan (when fitted) is clean and functioning correctly.
vi. Cylinder arrangment

The smallest engines usually have a single cylinder, mounted vertically above the
crankshaft, as this is convenient for access to the main engine components and also
allows an oil sump to be conveniently located where oil can drain down to it from the
cylinder. A large flywheel is needed to smooth the output from a single cylinder engine
as excessive vibration can cause problems with parts resonating and fatiguing and nuts
and bolts working loose. Single cylinder low speed diesels need particularly heavy
flywheels because they have large heavy pistons and connecting rods which run at low
speeds; traditional designs are "open flywheel" (Fig. 98) while the more modern style of
high-speed diesel engine usually has an internal enclosed flywheel (Fig. 99).
With larger sizes of engine it becomes feasible to have two or more cylinders. Twin
cylinder engines have a smoother power output because the cylinders fire alternately
and partially balance each other. Multi-cylinder engines therefore run more smoothly and
quietly than ones of the same power with fewer cylinders, but the more cylinders there
are the more components are involved, so obviously a multi-cylinder engine will be more
expensive and more complicated to overhaul and maintain.

180

Fig. 98 Open flywheel low-speed single-cylinder diesel engine

Fig. 99. Belt driven 3 cylinder Lister diesel engine coupled to a centrifugal pump via
multiple 'V' belts

181
vii. Special features and accessories

Most small pumping systems have a hand crank starter or a pull chord (recoil) starter
(the former is more common with small diesels and the latter with small petrol engines).
Diesels often include a decompression valve to aid starting, in which a cylinder valve can
be partially opened to release the pressure when the piston comes up on the
compression stroke, allowing the engine to be hand wound up to a certain speed, when
the decompression valve is suddenly closed and the momentum of the flywheel(s)
carries the machine on sufficiently to fire the engine and start it off. Larger engines often
have an auxilliary electrical system and a battery with an electric starter; (spark ignition
engines generally need a battery to run the ignition, although very small s.i. engines use
a magneto which generates and times the spark from the rotation of the engine).
Engines fitted with electrical systems or engines coupled to generators can have various
electrical controls to warn of or to prevent damage from overheating, loss of lubricant,
etc. Some of these options, although sophisticated, are not expensive and are therefore
a sound investment.
A vital, and often neglected accessory is the air filter, especially in dusty climates.
Usually there is a choice of paper element filter (as used on most cars) where the paper
element needs to be regularly replaced when clogged with dust or torn or, alternatively,
an oil bath filter. The latter is slightly more expensive, but much more effective and
practical in an agricultural context, since at a pinch even old engine oil can be used to
refill it. A worn or malfunctioning air filter can greatly reduce the useful life of an engine,
a fact which is often not fully understood by farmers judging from the number of engines
that can be found running without any air filter at all.
viii. Installation

The smallest engines are supplied mounted on skids or in a small frame and therefore
need no installation other than coupling their pump to the water conveyance system. But
larger machines, and many diesels, need to be properly installed, either on a concrete
pad or on a suitable trolley or chassis. Most manufacturers will provide a detailed
specification, when necessary, for the foundations of any engine driven pumping system;
this should be accurately adhered to.
The engine often needs to be installed in a small lockable building for security. It is
essential, however, that any engine house is well ventilated and that the exhaust is
properly discharged outside. This is not only to avoid the serious danger of poisoning the
user with exhaust gases, but also to ensure the engine does not overheat. Similarly,
engines with direct coupled centrifugal pumps sometimes need to be installed in a pit in
order to lower them near enough to the water level to avoid an excessive suction lift.
Considerable care is needed with such installations to ensure that neither exhaust nor oil
fumes will fill the pit and poison anyone who enters it; carbon monoxide in i.c. engine
exhaust emissions can, and frequently does, cause fatal accidents. Care is also needed
to ensure that the water level will never rise to a level where it could submerge the
engine.

182
4.4.2 Efficiency of Engine Powered Pumping Systems
This is a controversial subject where little reliable data on actual field performance exists
in the literature. A few field tests have been carried out on "typical" irrigation pumping
systems, and in some cases surprisingly poor efficiencies were achieved. For example,
Jansen [32] reported on tests on three kerosene fuelled small pumping sets in the 2-3
bhp range in Sri Lanka; total system efficiencies in the range 0.75-3.5% were recorded
although engine/pump efficiencies (without a pipeline) were in the range from 2.6-8.8%.
Excessive pipeline friction losses obviously caused the very poor system performance.
Unfortunately there is good reason to assume such losses are quite common.
Many of the reasons for poor performance can be corrected at little cost once it is
recognized that a problem exists, but unfortunately, it is easy to run an inefficient
pumping system without even realizing it, because the shortfall in output is simply made
up by running the engine longer than would otherwise be necessary.
Fig. 100 indicates the principle components of a small engine pumping system and the
range of efficiencies that can typically occur for each. Some explanation of these may
give an insight into how such poor total system efficiency can sometimes occur, and by
implication, what can be done to. improve it.
Firstly, fuel spillage and pilferage could perhaps result in, anything from 0-10% of the
fuel purchased being lost; i.e. 90-100% of fuel purchased may be usefully consumed as
indicated in the figure. Spillage can occur not only when transferring fuel, but due to
leaky storage or very frequently due to either a leaky fuel line on the engine or leaky
joints especially on the high pressure lines of a diesel engine; well established fuel
stains on the ground ought to give warning that something is wrong with fuel
management. As the value of fuel increases, pilferage becomes increasingly a problem.
One standard 200 litre drum of fuel is worth typically US $50-100; a small fortune in
many developing countries.

183

Fig. 100 Principal components of a small engine pumping system and their efficiencies
Most basic thermodynamics textbooks claim that s.i. engines tend to be 25-30%
efficient, while diesels are 30-40% efficient. Similarly, manufacturers' dynamometer
tests, with optimally tuned engines, running on a test-bed, (often minus most of their

184
accessories), tend to confirm this. However, such figures are optimistic in relation to field
operations. The difference between theory and reality is greatest with the smallest sizes
of engines, which are inherently less efficient than larger once, and the text-book figures
quoted above really only apply to well tuned engines of above 5-10kW power rating. The
smallest tow-stroke and four-stroke s.i. engines also tend to vary in quality, engine to
engine, and can easily be as poor and only 10% efficient at around 1kW power rating.
Small diesels (the smallest are generally about 1.5 to 2kW) will probably be better than
20% efficient as engines, but components like the injection pump, cooling fan and water
circulating pump (all parasitic energy consumers) can reduce the fuel-to-useful-shaftpower efficiency to around 15% (or less) for the smallest engines. This large drop is
because the parasitic accessories take proportionately more power from very small
engines. Obviously engines are also only in new condition for a small part of their, lives,
and on average are worn and not well tuned, which also undermines their efficiency.
Hence depending on size, type and quality plus their age and how well they are
maintened, engines may in reality be at best 35% efficient and at worst under 10%
efficient.
The next source of loss is any mechanical transmission from the engine to the pump. In
some cases the engine is direct coupled to the pump, in which case the transmission
losses are negligible, but if there is any substantial change of speed, such as a speedreducer and gearbox to drive a reciprocating borehole pump, then the transmission can
be as low as 60% efficient, particularly with small systems of less than 5kW, where geaz
box losses will be relatively large in relation to the power flow.
As discussed in the previous section, the most common type of pump used for irrigation
with a small engine will be a centrifugal pump, either direct coupled or belt driven,
usually working on suction; (eg. Figs. 72 and 99). If properly installed, so that the pump
is operating close to its optimum head and speed, the pump efficiency can easily exceed
60% and possibility be as high as 80% with bigger pumps. However there is a lot of
scope for failing to achieve these figures; bad impeller designs, worn impellers with
much back-leakage, operating away from the design flow and head for a given speed
will all have a detrimental effect and can easily singly, or in combination, pull the
efficiency down to 25% or less. Given a reasonably well matched and well run system,
pump efficiencies will therefore be in the 40-80% range, but they could easily be worse
(but not better) than this range.
It is often not appreciated that the choice of delivery pipe can have a profound effect on
system performance. Engines can deliver very high volumes at low head, so pipe friction
can grossly increase the total head across the pump, particularly with long delivery lines
at low heads. When this happens, it is possible for the total head to be several times the
static head, which multiplies the fuel requirement proportionately. Fig. 4 allows this to be
quantified; eg. even a small portable petrol engine pump will typically deliver over 360
l/min or 6 l/s through 5m head (600W hydraulic output). The friction loss for each 100m
of delivery pipe with this flow and head will approximately as follows:
pipe nominal diameter:
friction head (m/100m):

2" (50mm)
20m

21/2"(65mm)
5m

3" (80mm)
2m

from this, the pipe line efficiency for various lengths of pipe with the above diameters is
as follows:

185
total pipe length

efficiencies in % for 5m static head and pipe

diameters shown
2in

21/2in

3in

10m

71

91

96

50m

33

67

83

100m

20

50

71

Quite clearly, when pumping at such low heads, it is easy to achieve total heads that are
several times the static head, giving rise to pipeline efficiencies as poor as 20% when
100m of 2" pipe is used; (i.e. in that case the pumped head is five times the static head
so that five times as much fuel is needed compared with a 100% efficient pipeline). The
situation gets proportionately less serious at higher static heads, because it is the ratio of
pipe friction head to static head that matters; for example, at 20m static head the above
example would give the same friction head of 20m which although unacceptably high
would at least imply 50% rather than 20% pipeline efficiency. It is a common mistake to
use pipework which is too small in diameter as a supposed "economy" when larger pipe
can often pay for itself in saved fuel within months rather than years. Also, some pumps
which have a 2in discharge orifice may actually need a 3in pipeline if some distance is
involved, yet uninformed users will usually use a pipe diameter to, match the pump
orifice size and thereby create a major source of inefficiency. Incidentally, inch pipe sizes
are quoted here simply because they are in fact more commonly used, even in countries
where every thing else is dimensioned in metric units.
The performance curve in Fig. 101 indicates how centrifugal suction pumps can also
suffer reduced performance as the suction head increases, mainly due to cavitation,
particularly when the suction head is a large fraction of the total head. The figure shows
how at 10m total head, 6m suction head causes a 20% drop in output compared with 3m
suction head. This is without any reduction in power demand, so the former system is
20% less efficient than the latter, simply due to suction losses. There is therefore a
potentially large fuel cost penalty in applying excessive suction lifts (apart from the usual
priming problems that can occur).

Fig. 101 Effect of increasing the suction head on the output for a typical engine pump set

186
Pipe losses, whether due to friction or suction can therefore cause increased head and
reduced flow and will typically have an efficiency in the 30-95% range, as indicated in
Fig. 100.
The efficiency factors discussed so far apply to the hardware, while it is being run and
water is being usefully applied to the field. Inevitably the system needs to run when
water is not being usefully applied. For example, when starting up, any engine will often
be run for a few minutes before the farmer can arrange the discharge to reach the
correct part of the field, and water will be wasted. Similarly when rearranging the
distribution system to deliver water to another part of the field, some wastage may occur.
There is therefore a factor relating to the type of water distribution system and to the
management skill of the farmer at applying the pumped water for as large a fraction of
the time it is being pumped (or the engine is running) as possible. Even moderately bad
management could cause 20% loss compared with ideal usage, and really bad
management can be much worse, so this efficiency is taken as ranging from 80-100%
for the purpose of Fig. 100.
When all the worst efficiencies suggested in Fig. 100 are compounded, they yield a
theoretical worst total efficiency of 0.5% (which Jansen [32] and others have confirmed)
while the best factors in Fig. 100 compounded together give 27%. The "best" figure is
only even theoretically feasible however for a larger diesel (over 5kW), driving a pump at
10-20m (or higher) head, which is rather higher than usual for most irrigation
applications. Most smaller engine pumped irrigation systems therefore in practice
probably achieve 5 to 15% total efficiency, with larger diesels operating at higher heads
being towards the top end of this range and small kerosene or petrol engines at low
heads being at (or below) the bottom. The operator should probably be satisfied if a
small diesel system achieves 10-15% efficiency and a small petrol or kerosene fuelled
system achieves 5-10%. It may be very worthwhile for any reader to investigate the
actual efficiency of any engine pumping system they may be responsible for, (by
comparing fuel consumption against hydraulic energy output) so that if it is below par
steps may be taken to find the causes and to correct them.

187

4.5

EXTERNAL COMBUSTION ENGINES

The difference between internal and external combustion engines, as their names
suggest, is that the former burn their fuel within the power cylinder, but the latter use
their fuel to heat a gas or a vapour through the walls of an external chamber, and the
heated gas or vapour is then transferred to the power cylinder. External combustion
engines therefore require a heat exchanger, or boiler to take in heat, and as their fuels
are burnt externally under steady conditions, they can in principle use any fuel that can
burn, including agricultural residues or waste materials
There are two main families of external combustion engines; steam engines which rely
on expanding steam (or occasionally some other vapour) to drive a mechanism; or
Stirling engines which use hot air (or some other hot gas). The use of both technologies
reached their zeniths around 1900 and have declined almost to extinction since.
However a brief description is worthwhile, since:
i.
ii.
iii.

they were successfully and widely used in the past for pumping water;
they both have the merit of being well suited to the use of low cost fuels such as
coal, peat and biomass;
attempts to update and revive them are taking place.

and therefore they may re-appear as viable options in the longer term future.
The primary disadvantage of e.c. engines is that a large area of heat exchanger is
necessary to transmit heat into the working cylinder(s) and also to reject heat at the end
of the cycle. As a result, e.c. engines are generally bulky and expensive to construct
compared with i.c. engines. Also, since they are no longer generally manufactured they
do not enjoy the economies of mass-production available to i.e. engines. They also will
not start so quickly or conveniently as an i.c. engine; because it takes time to light the
fire and heat the machine to its working temperature.
Due to their relatively poor power/weight ratio and also the worse energy/weight ratio of
solid fuels, the kinds of applications where steam or Stirling engines are most likely to be
acceptable are for static applications such as as irrigation water pumping in areas where
petroleum fuels are not readily available but low cost solid fuels are. On the positive
side, e.c. engines have the advantage of having the potential to be much longer-lasting
than i.c. engines (100 year old steam railway locomotives are relatively easy to keep in
working order, but it is rare for i.c. engines to be used more than 20 years or so. E.c.
engines are also significantly quieter and free of vibrations than i.c. engines. The level of
skill needed for maintenance may also be lower, although the amount of time spent will
be higher, particularly due to the need for cleaning out the furnace.
Modern engineering techniques promise that any future steam or Stirling engines could
benefit from features not available over 60 years ago when they were last in general
use. Products incorporating these new developments are not yet on the market, but R&D
is in hand in various countries on a limited scale; however it will probably be some years
before a new generation of multi-fuel Stirling or steam powered pumps become
generally available.

188
4.5.1 Steam Engines
Only a limited number of small steam engines are available commercially at present;
most are for general use or for powering small pleasure boats. A serious attempt to
develop a 2kW steam engine for use in remote areas was made by the engine
designers, Ricardos, in the UK during the 1950s (see Fig. 157). That development was
possibly premature and failed, but there is currently a revival of interest in developing
power sources that can run on biomass-based fuels (as discussed more fully in Section
4.10). However, small steam engines have always suffered from their need to meet quite
stringent safety requirements to avoid accidents due to boiler explosions, and most
countries have regulations requiring the certification of steam engine boilers, which is a
serious, but necessary, inhibiting factor.
The principle of the steam engine is illustrated in Fig. 102. Fuel is burnt in a furnace and
the hot gases usually pass through tubes surrounded by water (fire tube boilers). Steam
is generated under pressure; typically 5 to 10 atmospheres (or 5-10bar). A safety valve
is provided to release steam when the pressure becomes too high so as to avoid the risk
of an explosion. High pressure steam is admitted to a power cylinder through a valve,
where it expands against a moving piston to do work while its pressure drops. The inlet
valve closes at a certain point, but the steam usually continues expanding until it is close
to atmospheric pressure, when the exhaust valve opens to allow the piston to push the
cooled and expanded steam out to make way for a new intake of high pressure steam.
The valves are linked to the drive mechanism so as to open or close automatically at the
correct moment. The period of opening of the inlet valve can be adjusted by the operator
to vary the speed and power of the engine.

189

Fig. 102 Schematic arrangement of a condensing steam engine

In the simplest types of engine the steam is exhausted to the atmosphere. This however
is wasteful of energy, because by cooling and condensing the exhausted steam the
pressure can be reduced to a semi-vacuum and this allows more energy to be extracted
from a given throughput of steam and thereby significantly improves the efficiency.
When a condenser is not used, such as with steam railway locomotives, the jet of
exhaust steam is utilised to create a good draught for the furnace by drawing the hot
gases up the necessarily short smoke stack. Condensing steam engines, on the other

190
hand, either need a high stack to create a draught by natural convection, or they need
fans or blowers.
Steam pumps can easily include a condenser, since the pumped water can serve to cool
the condenser. According to Mead [13], (and others) the typical gain in overall efficiency
from using a condenser can exceed 30% extra output per unit of fuel used. Condensed
steam collects as water at the bottom of the condenser and is then pumped at sufficient
pressure to inject it back into the boiler by a small water feed pump, which is normally
driven off the engine. A further important advantage of a condensing steam engine is
that recirculating the same water reduces the problems of scaling and corrosion that
commonly occur when a continuous throughput of fresh water is used. A clean and
mineral-free water supply is normally necessary for non-condensing steam engines to
prolong the life of the boiler.
The most basic steam engine is about 5% efficient (steam energy to mechanical shaft
energy - the furnace and boiler efficiency of probably between 30 and 60% needs to be
compounded with this to give an overall efficiency as a prime-mover in the 1.5 to 3%
range). More sophisticated engines are around 10% efficient, while the very best reach
15%. When the boiler and furnace efficiencies (30-60%) plus the pump (40-80%) and
pipework (40-90%) are compounded, we obtain system efficiencies for steam piston
engine powered pumps in the 0.5 to 4.5% range, which is worse, but not a lot worse
than for small s.i. internal combustion engines pumping systems, but allows the use of
non-petroleum fuels and offers greater durability.

4.5.2 Stirling Engines

This type of engine was originally developed by the Rev. Robert Stirling in 1816. Tens of
thousands of small Stirling engines were used in the late nineteenth and early twentieth
century, mainly in the USA but also in Europe. They were applied to all manner of small
scale power purposes, including water pumping. In North America they particularly saw
service on the "new frontier"; which at that time suffered all the problems of a developing
country in terms of lack of energy resources, etc.
Rural electrification and the rise of the small petrol engine during and after the 1920s
overtook the Stirling engine, but their inherent multi-fuel capability, robustness and
durability make them an attractive concept for re-development for use in remote areas in
the future and certain projects are being initiated to this end. Various types of directaction Stirling-piston water pumps have been developed since the 1970s by Beale and
Sunpower Inc. in the USA, and some limited development of new engines, for example
by IT Power in the UK with finance from GTZ of West Germany is continuing.

191

Fig. 103 Rider-Ericsson hot air pumping engine (Stirling cycle) circa 1900
Stirling engines use pressure changes caused by alternately heating and cooling an
enclosed mass of air (or other gas). The Stirling engine has the potential to be more
efficient than the steam engine, and also it avoids the boiler explosion and scaling
hazards of steam engines. An important attribute is that the Stirling engine is almost
unique as a heat engine in that it can be made to work quite well at fractional
horsepower sizes where both i.c. engines and steam engines are relatively inefficient.
This of course makes it of potential interest for small scale irrigation, although at present
it is not a commercially available option.
To explain the Stirling cycle rigorously is a complex task. But in simple terms, a displacer
is used to move the enclosed supply of air from a hot chamber to a cold chamber via a
regenerator. When most of the air is in the hot end of the enclosed system, the internal

192
pressure will be high and the gas is allowed to expand against a power piston, and
conversely, when the displacer moves the air to the cool end, the pressure drops and
the power piston returns. The gas moves from the hot end to the cold end through a
regenerator which has a high thermal capacity combined with a lot of surface area, so
that the hot air being drawn from the power cylinder cools progressively on its way
through the regenerator, giving up its heat in the process; then when cool air travels
back to the power cylinder ready for the next power stroke the heat is returned from the
regenerator matrix to preheat the air prior to reaching the power cylinder. The
regenerator is vital to achieving good efficiency from a Stirling engine. It often consists of
a mass of metal gauze through which air can readily pass, [33], [34 ].
Some insight into the mechanics of a ' small Stirling engine can be gained from Fig. 103,
which shows a 1900 vintage Rider-Ericsson engine. The displacer cylinder projects at its
lower end into a small furnace. When the displacer descends it pushes all the air through
the regenerator into the water cooled volume near the power cylinder and the pressure
in the system drops, then as the displacer rises and pulls air back into the hot space, the
pressure rises and is used to push the power piston upwards on the working stroke. The
displacer is driven off the drive shaft and runs 90 out.of phase with the power piston. An
idea of the potential value of engines such as this can be gained from records of their
performance; for example, the half horsepower Rider-Ericsson engine could raise
2.7m3/hr of water through 20m; it ran at about 140 rpm (only) and consumed about 2kg
of coke fuel per hour. All that was needed to keep it going was for the fire to be
occasionally stoked, rather like a domestic stove, and for a drop of oil to be dispensed
onto the plain bearings every hour or so.

4.6

ELECTRICAL POWER

If a connection is available to a reliable mains electricity supply, nothing else is either as

convenient or more cost-effective for powering an irrigation pump. Unfortunately, the
majority of farmers in developing countries do not have mains electricity close at hand,
and even those that do often find that the supply is unreliable. Electricity supply
problems tend to be particularly prevalent during the irrigation season, because irrigation
pumping tends to be practised simultaneously by all farmers in a particular district and
can therefore easily overload an inadequate rural network and cause "brownouts"
(voltage reductions) or even "blackouts" (complete power cuts). Therefore there is a
major inhibition for many electricity utilities in encouraging any further use of electricity
for irrigation pumping in developing countries where the electrical supply network is
already under strain.
The real cost of extending the grid is very high, typically in the order of $5 000-10 000
per kilometer,of spur. Although connections in many countries have in the past been
subsidised, whatever the pricing policy of the utility, someone has to pay for it and the
tendency today is to withdraw subsidies. Therefore, although an electric motor
considered in isolation is an extremely inexpensive and convenient prime-mover, it is
only useful when conbnnected to a lot of capital-intensive infrastructure which needs to
carry a substantial electrical load in order to be self-financing from revenue.
A further problem for developing countries in considering the mains electricity option is
the high foreign exchange component in the investment; this is typically from 50-80%,
according to Fluitman, [35] (quoting a World Bank source). Electricity generation in rural

193
areas of developing countries tends to be by petroleum-fuelled plant (usually diesel
generators) so this also is a burden on the economy. In fact a large fraction of many
developing countries' oil imports goes to electrical power generation. The attractiveness
of rural electrification as an investment for development is therefore being questioned
much more now then it used to be; (eg. [35]). However it is not proposed here to deal
with policy implications or macro-economic effects of the widespread use of electricity for
irrigation pumping, other then to point out that it cannot be seen as a universally
applicable solution to the world's irrigation pumping needs, because most countries will
not be able to afford to extend a grid to all their rural areas in the forseeable future. Even
where such an option can be afforded, it is still necessary to question whether it is the
most cost-effective solution for irrigation pumping bearing in mind the high infrastructural
costs.

4.6.1 Sources and Types of Electricity

Batteries produce a steady flow of electricity known as "direct current" or DC.
Photovoltaic (solar) cells also produce DC. Electrical generators to produce DC are
sometimes known as "dynamos"; they require commutators consisting of rotating brass
segments with fixed carbon brushes. Alternators are almost universally used today for
the generation of electricity from shaft power. Alternators are simpler and less expensive
than DC generators, but they produce a voltage which reverses completely several times
per revolution. This type of electrical output, which is almost universally used for mains
supplies, is known as "alternating current" or AC.
AC mains voltage normally fluctuates from full positive to full negative and back 50 times
per second (50Hz or 50 cycles/sec) or in some cases at 60Hz. The current fluctuates
similarly. Sometimes the current and voltage can be "out of step", i.e. their peaks do not
coincide. This discrepency (or phase difference) is quantified by the "power factor"; the
output of an AC system is the product of the amps, volts and the power factor. When the
amps and volts are in perfect phase with each other, the power factor is numerically 1.
When the power factor is less than one (it frequently is 0.9 and sometimes less) then the
power available is reduced proportionately for a given system rating. The rating of AC
equipment is therefore generally given not in watts or kilowatts (kW), but in volt-amps or
kilovolt-amps (kVA). The actual power in kW will therefore be the kVA rating multiplied
by the power factor.
Another important principle to be aware of is that it is considerably more economic to
transmit electricity any distance at high voltages rather than low. A smaller cross-section
of conductor is needed for a given transmission efficiency. This is analagous to water
transmission, where higher pressures and smaller flow rates allow smaller pipes to be
used for equal hydraulic power. However, electricity is potentially lethal at AC voltages
much above 240V and at DC voltages much above 100V (it can of course kill at
considerably lower voltages depending on the circumstances and state of health of the
victim) and insulation becomes more difficult. Therefore, for safety reasons, 240V AC or
about 110V DC are usually the maximum voltages used at the end-users' supplies and
for electrical appliances.
The reason AC is generally used for mains applications rather than DC are that it has a
number of important advantages:

194
a. AC generators and motors are much simpler, less expensive and less
troublesome, since they do not require commutators;
b. AC voltages can be changed efficiently and with a high degree of reliability, using
transformers, but it is a technically much more difficult problem to change DC
voltages; therefore AC can easily be transmitted efficiently at high voltages and
then transformed to low, safer voltages close to the point of use;
c. as a result of the advantages of AC, it has become the internationally used
standard for mains supplies and virtually all mass-produced electrical
applicances are designed for AC use.
It is sometimes necessary to convert AC to DC or vice-versa, for example to charge
batteries (which are DC) from the AC mains or to run an AC appliance designed for the
mains from a DC source such as a battery or a solar photovoltaic array. AC can quite
readily be converted to DC by using a rectifier; these (like transformers) are solid-state
devices which require no maintenance and are relatively efficient. A battery charger
usually consists of a combination of a transformer (to step mains voltage down to battery
voltage) and a rectifier to convert the low voltage AC to DC. Converting DC to AC is
more difficult; traditionally an inefficient electro-mechanical device called a rotary
converter was used; this is a DC motor direct coupled to an AC alternator. The modern
alternative is an electronic, solid state device called an inverter. Inverters are relatively
inexpensive for low power applications (such as powering small fluorescent lights from
low voltage batteries), but they become expensive for such higher powered applications
such as electric motors for pumping. The quality and price of inverters also varies a lot; if
a good quality AC output is essential (and high efficiency of conversion) a more
complicated and expensive device is needed. Cheap inverters often produce a crude AC
output and are relatively inefficient; they can also seriously interfere with radio and TV
reception in the vicinity.

4.6.2

AC Mains Power

Mains electricity is generally supplied as alternating current (AC) either at 220 to 240V
and 50Hz frequency or at 110V and 60Hz frequency for low power connections,
(including domestic ones) of up to about 10kW. The 220-240V 50Hz standard is normal
in Europe while the 110V 60Hz standard is in use in the USA; either might be used in
other parts of the world, although 220-240V is more common, especially in Asia and
Africa.
When AC is supplied through two wires, it is known as single-phase. The two wires are
not "positive" and "negative" but are "live" and "neutral"; there should always also be a
third wire included for safety - the "earth" or "ground". The latter is normally connected
to the casing of any appliance or motor so that if any internal fault causes the casing to
come into contact with the live supply, the leakage current will flow to earth (ground) and
trip out the system or blow a fuse. Therefore if an electric pump keeps tripping or
blowing fuses it is as well to have it checked to see if there is a short-circuit.
Mains power is normally generated as "three-phase", in which the alternator transmits
three "single phase" AC outputs down three wires. Each phase is shifted by one third of
a revolution of the alternator, so the voltage peaks in the three conductors do not
coincide, but are evenly spaced out. The three phases, if equally balanced, will cancel
each other out if fed through three equal loads, but in practice they are not usually

195
perfectly balanced so there is normally a fourth return conductor called the neutral. A
single phase AC supply is simply a connection to one of the three "lives" of a threephase source with a return to its neutral. For this reason it is important in many cases
not to confuse the live and neutral; also it is the live which should be protected by fuses
or contact breakers.
At higher power levels, usually above 5kW, and always above about 25kW, it is normal
to use three-phase AC. This is supplied mostly at 415V line to line (Europe) or 190 or
440V (USA).

4.6.3

Electric Motors

An electric motor seems almost the ideal prime mover for a water pump. Power is
supplied "at the flick of a switch", and water is produced at a constant rate until the
motor is turned off. Electric motors have relatively long service lives and generally need
little or no servicing.
The cheapest and simplest type of electric motor is the squirrel cage induction motor
which is almost universally used for mains electric power applications; see Fig. 104 (a)
and Fig. 105. Here there are no electrical connections to the rotating "squirrel cage", so
there are no brushes or slip rings to wear or need adjustment. Motors of this kind are
available in either three-phase or single-phase versions. They run at a fixed speed
depending on the frequency of the power supply and the number of poles in their stator
windings. The most general type (which is usually the cheapest) runs at a nominal 1
500rpm at 50Hz (1 800 at 60Hz), but other speeds are available. It is normal to directcouple a motor to a centrifugal pump where possible (eg. Fig. 105). Non-standard speed
motors may be used where this does not suit the pump, or alternatively a belt speed
reduction arrangement may be used, such as in Fig. 106.
A problem with induction motors is that they normally need over three times as much
current to start as they do once running at rated speed and power. This means that the
peak current that can be supplied must be significantly higher than that needed for
operation, which often causes not just technical but also financial problems, as some
electricity tariffs are determined by the maximum current rating on a circuit. Recently
electronic starting devices have become available which limit the starting current while
the motor runs up to speed and which in some cases also improve the overall efficiency
of an electric motor.

196

Fig. 104 The four main types of electric motors

Fig. 105 Direct coupled electric motor and centrifugal pump

197

Fig. 106 Electric motor powered, belt driven piston pump (Climax) (note air chambers
provided to prevent water hammer)
Induction motors are typically 75% efficient for a 300W (0.5 hp) size and may be around
85% efficient at 10kW size (subject to having a unity power factor). They are not
generally made in sizes significantly smaller than 100-200W.
For very small scale applications, the so-called "universal motor" is most commonly
used. The universal motor (Fig. 104 (b)) is the "classic" electric motor with a brushed
commutator and wound armature. Fixed field coils produce the magnetic flux to run the
motor. Motors of this kind can use either an AC or a DC supply and they are typically
used for very small-scale power applications (such as in power tools, and domestic
appliances, for example). They are more efficient than would be possible with a very
small induction motor and their starting current is smaller in relation to their running
current. They suffer however from needing periodic replacement of brushes when used
intensively, as for pumping duties.
There are small-scale electrical power applications independent of a mains supply,
which use a DC source such as a photovoltaic array, or batteries charged from a windgenerator. In these applications a permanent magnet DC motor is the most efficient
option (Fig. 104 (c)). In these, permanent magnets replace the field coils; this offers
higher efficiency, particularly at part-load, when field windings would absorb a significant

198
proportion of the power being drawn. Permanent magnet DC motors can be 75-85%
efficient even at such low power ratings as 100-200W, needed for the smallest solar
pumping systems. Most permanent magnet motors have brushed/commutated
armatures exactly like a universal motor, which in the pumping context is a major
drawback particularly for submersible sealed in motors. However brushless permanent
magnet motors have recently become available (Fig. 104 (d)). Here the magnets are
fixed to the rotor and the stator windings are fed a commutated AC current at variable
frequency to suit the speed of rotation; this is done by sending a signal from a rotor
position-sensor which measures the speed and position of the shaft and controls
electronic circuitry which performs the commutation function on a DC supply. Motors of
this kind are mechanically on a par with an induction motor, and can be sealed for life in
a submersible pump if required, but they are still produced in limited numbers and
involve a sophisticated electronic commutator which makes them relatively expensive at
the time of writing. With the increasing use of solar pumps they are likely to become
more widely used and their price may fall.
Submersible pump motors, whether AC induction motors or DC brushless permanent
magnet motors, are commonly filled with (clean and corrosion inhibited) water as this
equalises the pressure on the seals and makes it easier to prevent ingress of well water
than if the motor contained only air at atmospheric pressure. Filling motors with water is
obviously only possible with brushless motors, or short-circuits would occur. Another
advantage of water filled motors is that they are better-protected from overheating.

4.6.4 Electrical Safety

AC electrical voltages over about 110V and DC over 80V are potentially lethal,
especially if the contact is enhanced through the presence of water. Therefore, electricity
and water need to be combined with caution, and anyone using electricity for irrigation
pumping should ensure that all necessary protection equipment is provided; i.e. effective
trips or fuses, plus suitable armoured cables, earthed and splashproof enclosures, etc.
Also all major components, the motor, pump and supporting structure should be properly
earthed (or grounded) with all earth connections electrically bonded together. It is vital
that electrical installations should either be completed by trained electricians or if the
farmer carries it out, he should have it inspected and checked by a properly qualified
person before ever attempting to use it; (in some countries this is in any case a legal
requirement). It is also prudent to have some prior knowledge what action to take for
treating electric shock; most electrical utilities can provide posters or notices giving
details of precautions with recommendations on treatment should such an unfortunate
event occur.

199

4.7

WIND POWER

4.7.1 Background and State-of-the-Art

i.

Background

The wind has been used for pumping water for many centuries; it was in fact the primary
method used for dewatering large areas of the Netherlands from the 13th century
onwards; [36]. Smaller windpumps, generally made from wood, for use to dewater
polders, (in Holland) and for pumping sea water in salt workings, (France, Spain and
Portugal), were also widely used in Europe and are still used in places like Cape Verde;
Fig. 107.
However the main type of windpump that has been used is the so-called American farm
windpump; (Fig. 108). This normally has a steel, multibladed, fan-like rotor, which drives
a reciprocating pump linkage usually via reduction gearing (Fig. 109) that connects
directly with a piston pump located in a borehole directly below. The American farm
windpump evolved during the period between 1860 and 1900 when many millions of
cattle were being introduced on the North American Great Plains. Limited surface water
created a vast demand for water lifting machinery, so windpumps rapidly became the
main general purpose power source for this purpose. The US agricultural industry
spawned a multitude of windpump manufacturers and there were serious R&D
programmes, some sponsored by the US government, [37], to evolve better windpumps
for irrigation as well as for water supply duties.

200

Fig. 107 Wooden indigenous windmill pump for pumping sea water into salt pans on the
Island of Sal, Cape Verde

201
Other "new frontiers" such as Australia and Argentina took up the farm windpump, and
to this day an estimated one million steel farm windpumps are in regular use [38], the
largest numbers being in Australia and Argentina; [39], [40]. It should be noted that the
so-called American Farm Windpump is rarely used today for irrigation; most are used for
the purpose they were originally developed for, namely watering livestock and, to a
lesser extent, for farm or community water supplies. They tend therefore to be applied at
quite high heads by irrigation standards; typically in the 10 to 100m range on boreholes.
Large windpumps are even in regular use on boreholes of over 200m depth.
Wind pumps have also been used in SE Asia and China for longer than in Europe,
mainly for irrigation or for pumping sea water into drying pans for sea salt production.
The Chinese sail windpump (Fig. 110) was first used over a thousand years ago and
tens if not hundreds of thousands, are still in use in Hubei, Henan and North Jiangsu
provinces [41]. The traditional Chinese designs are constructed from wire-braced
bamboo poles carrying fabric sails; usually either a paddle pump or a dragon-spine
(ladder pump) is used, typically at pumping heads of less than lm. Many Chinese
windmills rely on the wind generally blowing in the same direction, because their rotors
are of fixed orientation. Many hundreds of a similar design of windpump to the Chinese
ones are also used on saltpans in Thailand, (Fig. 111).

Fig. 108 All-steel 'American' farm wind pump

202

Fig. 109 Gearbox from a typical back-geared 'American' farm windmill

203

Fig. 110 Chinese chain windmill

204

Fig. 111 Thai windpump (after Schioler [24])

205

Fig. 112 'Cretan' type of windmill used on an irrigation project in Southern Ethiopia (after
Fraenkel [15])
Some 50 000 windpumps were used around the Mediterranean Sea 40 years ago for
irrigation purposes, [42], These were improvised direct-drive variations of the metal
American farm windpump, but often using triangular cloth sails rather than metal blades.

206
These sail windmills have a type of rotor which has been used for many centuries in the
Mediterranean region, but today is often known as "Cretan Windmills" (see Fig. 112).
During the last 30 years or so, increased prosperity combined with cheaper engines and
fuels has generally led farmers in this region to abandon windmills and use small
engines (or mains electricity where available). However Crete is well known as a country
where until recently about 6 000 windpumps were still in use [91], mostly with the cloth
sailed rig. The numbers of windpumps in use in Crete are rapidly declining and by 1986
were believed to be barely one thousand.
Another branch of wind energy technology began to develop in the late 1920s and early
1930s, namely, the wind-generator or aero-generator. Many thousands of small wind
generators, such as the Australian Dunlite (Fig. 113), were brought into use for charging
batteries which could be used for lighting, and especially for radio communication, in
remote rural areas. Such machines can also provide an alternative to a photovoltaic
array for irrigation pumping in suitably windy areas, although they have not so far been
applied for this purpose in any numbers.
Large wind turbines for electricity generation have been (and are being) constructed, the
largest being a 5MW (5 000kW) machine under development in West Germany.
However, more modest, but still quite large medium sized machines are being installed
in large numbers for feeding the local grid notably in the state of California (where over
10 000 medium sized wind generators have been installed in little more than 3 years for
feeding the grid) and in Denmark. Fig. 114 shows a typical modern 55kW, 15m diameter
Windmatic wind turbine, from Denmark. Machines of this size may in future be of
considerable relevance for larger scale irrigation pumping than is feasible with more
traditional mechanical windpumps, (see Gilmore et al [43], and Nelson et al [44]).
ii.

State-of-the-Art

There are two distinct end-uses for windpumps, namely either irrigation or water supply,
and these give rise to two distinct categories of windpump because the technical,
operational and economic requirements are generally different for these end uses. That
is not to say that a water supply windpump cannot be used for irrigation (they quite often
are) but irrigation designs are generally unsuitable for water supply duties.
Most water supply windpumps must be ultra-reliable, to run unattended for most of the
time (so they need automatic devices to prevent overspeeding in storms), and they also
need the minimum of maintenance and attention and to be capable of pumping water
generally from depths of 10m or more. A typical farm windpump should run for over 20
years with maintenance only once every year, and without any major replacements; this
is a very demanding technical requirement since typically such a wind pump must
average over 80 000 operating hours before anything significant wears out; this is four
to ten times the operating life of most small diesel engines or about 20 times the life of a
small engine pump. Windpumps to this standard therefore are usually industrially
manufactured from steel components and drive piston pumps via reciprocating pump
rods. Inevitably they are quite expensive in relation to their power output, because of the
robust nature of their construction. But American, Australian and Argentinian ranchers
have found the price worth paying for windpumps that achieve high reliability and
minimum need for human intervention, as this is their main advantage over practically
any other form of pumping systems.

207

Fig. 113 2kW Dunlite wind electricity generator

208

Fig. 114 55kW Windamatic wind electricity generator

209
Irrigation duties on the other hand are seasonal (so the windmill may only be useful for a
limited fraction of the year), they involve pumping much larger volumes of water through
a low head, and the intrinsic value of the water is low. Therefore any windpump
developed for irrigation has to be low in cost and this requirement tends to overide most
other considerations. Since irrigation generally involves the farmer and/or other workers
being present, it is not so critical to have a machine capable of running unattended.
Therefore windmills used for irrigation in the past tend to be indigenous designs that are
often improvized or built by the farmer as a method of low-cost mechanization; (eg. Figs.
110, 111 and 112). If standard farm windpumps (Fig. 108) are used for irrigation, usually
at much lower heads than are normal for water supply duties, there are quite often
difficulties in providing a piston pump of sufficient diameter to give an adequate swept
volume to absorb the power from the windmill. Also most farm windpumps have to, be
located directly over the pump, on reinforced concrete foundations, which usually limits
these machines to pumping from wells or boreholes rather than from open water. A
suction pump can be used on farm windmills with suction heads of up to about 5-6m
from surface water; (see Fig. 115 for typical farm windpump installation configurations).
Most indigenous irrigation windpumps, on the other hand, such as those in China, use
rotary pumps of one kind or another which are more suitable for low heads; they also do
not experience such high mechanical forces as an industrial windpump, (many of which
lift their pump rods with a pull of over 1tonne, quite enough to "uproot" any carelessly
installed pump).
Attempts have been made recently to develop lower cost steel windpumps that
incorporate the virtues of the heavier older designs. Most farm-windpumps, even though
still in commercial production, date back to the 1920s or earlier and are therefore
unecessarily heavy and expensive to manufacture, and difficult to install properly in
remote areas. Recently various efforts have been made to revise the traditional farm
windpump concept into a lighter and simpler modern form. Figs. 116 shows the "IT
Windpump", which is half the weight of most traditional farm windpump designs of a
similar size, and is manufactured in Kenya as the "Kijito" and in Pakistan as the
"Tawana". The latter costs only about half as much as American or Australian machines
of similar capability. It is possible therefore that through developments of this kind, costs
might be kept low enough to allow the marketing of all steel windpumps that are both
durable like the traditional designs, yet cheap enough to be economic for irrigation.

210

Fig. 115 Typical farm windpump installation configurations

A. borehole to raised storage tank
B. well to surface storage tank
C. surface suction pump

211

Fig. 116 IT windpump, made in Kenya as the 'Kijito' and in Pakistan as the 'Tawana'

212

4.7.2
i.

Principles of Wind Energy Conversion

Power available in the wind

The power in the wind is proportional to the wind speed cubed; the general formula for
power in the wind is:

where P is the power available in watts, p is the density of air (which is approximately
1.2kg/m3 at sea level), A is the cross-section (or swept area of a windmill rotor) of air flow
of interest and V is the instantaneous free-stream wind velocity. If the velocity, V, is in
m/s (note that lm/s is almost exactly 2 knots or nautical miles per hour), the power in the
wind at sea level is:

Because of this cubic relationship, the power availability is extremely sensitive to wind
speed; doubling the wind speed increases the power availability by a factor of eight;
Table 16 indicates this variability.
Table 16 POWER IN THE WIND AS A FUNCTION OF WIND SPEED IN UNITS OF
POWER PER UNIT AREA OF WIND STREAM
wind
speed
power
density

m/s

2.5

7.5

10

15

20

30

40

km/h

18

27

36

54

72

108

144

mph
kW/m2

6
.01

11
.08

17
.27

22
.64

34
2.2

45
5.1

67
17

90
41

hp/ft2

.001

.009

.035

.076

.23

.65

2.1

5.2

This indicates the very high variability of wind power, from around 10W/m2 in a light
breeze up to 41 000Wm2 in a hurricane blowing at 144km/h. This extreme variability
greatly influences virtually all aspects of system design. It makes it impossible to
consider trying to use winds of less than about 2.5m/s since the power available is too
diffuse, while it becomes essential to shed power and even shut a windmill down if the
wind speed exceeds about 10-15m/s (25-30mph) as excessive power then becomes
available which would damage the average windmill if it operated under such conditions.
The power in the wind is a function of the air-density, so it declines with altitude as the
air thins, as indicated in Table 17.
Table 17 VARIATION OF AIR DENSITY WITH ALTITUDE
altitude (ft)
a.s.l. (m)

0
0

2 500
760

5 000
1 520

7 500
2 290

10 000
3 050

213
density correction factor

1.00

0.91

0.83

0.76

0.69

Because the power in the wind is so much more sensitive to velocity rather than to air
density, the effect of altitude is relatively small. For example the power density of a 5m/s
wind at sea level is about 75 watts/m2; however, due to the cube law, it only needs a
wind speed of 5.64m/s at 3 000m a.s.l. to obtain exactly the same power of 75 watts/m2.
Therefore the drop in density can be compensated for by quite a marginal increase in
wind velocity at high altitudes.
ii.

Energy available in the wind

Because the speed of the wind constantly fluctuates, its power also varies to a
proportionately greater extent because of the cube law. The energy available is the
summed total of the power over a given time period. This is a complex subject (Lysen
[45] gives a good introduction to it). The usual starting point to estimate the energy
available in the wind at a specific location is some knowledge of the mean or average
wind speed over some predefined time period; typically monthly means may be used.
The most important point of general interest is that the actual energy available from the
wind during a certain period is considerably more than if you take the energy that would
be produced if the wind blew at its mean speed without variation for the same period.
Typically the energy available will be about double the value obtained simply by
multiplying the instantaneous power in the wind that would correspond to the mean wind
speed blowing continuously, by the time interval. This is because the fluctuations in wind
speed result in the average power being about double that which occurs instantaneously
at the mean wind speed. The actual factor by which the average power exceeds the
instantaneous power corresponding to the mean windspeed can vary from around 1.5 to
3 and depends on the local wind regime's actual variability. The greater the variability the
greater this factor.
However, for any specific wind regime, the energy available will still generally be
proportional to the mean wind speed cubed. We shall discuss later in this section how to
determine the useful energy that can be obtained from a wind regime with respect to a
particular windmill.
iii.

Converting wind power to shaft power

There are two main mechanisms for converting the kinetic energy of the wind into
mechanical work; both depend on slowing the wind and thereby extracting kinetic
energy. The crudest, and least efficient technique is to use drag; drag is developed
simply by obstructing the wind and creating turbulence and the drag force acts in the
same direction as the wind. Some of the earliest and crudest types of wind machine,
known generically as "panamones", depend on exposing a flat area on one side of a
rotor to the wind while shielding (or reefing the sails) on the other side; the resulting
differential drag force turns the rotor.
The other method, used for all the more efficient types of windmill, is to produce lift. Lift
is produced when a sail or a flat surface is mounted at a small angle to the wind; this
slightly deflects the wind and produces a large force perpendicular to the direction of the
wind with a much smaller drag force. It is this principle by which a sailing ship can tack at
speeds greater than the wind. Lift mainly deflects the wind and extracts kinetic energy

214
with little turbulence, so it is therefore a more efficient method of extracting energy from
the wind than drag.
It should be noted that the theoretical maximum fraction of the kinetic energy in the
wind that could be utilized by a "perfect" wind turbine is approximately 60%. This is
because it is impossible to stop the wind completely, which limits the percentage of
kinetic energy that can be extracted.
iv.

Horizontal and vertical axis rotors

Windmills rotate about either a vertical or a horizontal axis. All the windmills illustrated so
far, and most in practical use today, are horizontal axis, but research is in progress to
develop vertical axis machines. These have the advantage that they do not need to be
orientated to face the wind, since they present the same cross section to the wind from
any direction; however this is also a disadvantage as under storm conditions you cannot
turn a vertical axis rotor away from the wind to reduce the wind loadings on it.
There are three main types of vertical axis windmill. Panamone differential drag devices
(mentioned earlier), the Savonius rotor or "S" rotor (Fig. 117) and the Darrieus wind
turbine (Fig. 118). The Savonius rotor consists of two or sometimes three curved
interlocking plates grouped around a central shaft between two end caps; it works by a
mixture of differential drag and lift. The Savonius rotor has been promoted as a device
that can be readily improvized on a self-build basis, but its apparent simplicity is more
perceived than real as there are serious problems in mounting the inevitably heavy rotor
securely in bearings and in coupling its vertical drive shaft to a positive displacement
pump (it turns too slowly to be useful for a centrifugal pump). However the main
disadvantages of the Savonius rotor are two-fold:
a. it is inefficient, and involves a lot of construction material relative to its size, so it
is less cost-effective as a rotor than most other types;
b. it is difficult to protect it from over-speeding in a storm and flying to pieces.
The Darrieus wind turbine has airfoil cross-section blades (streamlined lifting surfaces
like the wings of an aircraft). These could be straight, giving the machine an "H"-shaped
profile, but in practice most machines have the curved "egg-beater" or troposkien profile
as illustrated. The main reason for this shape is because the centrifugal force caused by
rotation would tend to bend straight blades, but the skipping rope or troposkien shape
taken up by the curved blades can resist the bending forces effectively. Darrieus-type
vertical axis turbines are quite efficient, since they depend purely on lift forces produced
as the blades cross the wind (they travel at 3 to 5 times the speed of the wind, so that
the wind meets the blade at a shallow enough angle to produce lift rather than drag).
The Darrieus was predated by a much cruder vertical axis windmill with Bermuda
(triangular) rig sails from the Turks and Caicos Islands of the West Indies (Fig. 119). This
helps to show the principle by which the Darrieus works, because it is easy to imagine
the sails of a Bermuda rig producing a propelling force as they cut across the wind in the
same way as a sailing yacht; the Darrieus works on exactly the same principle.

215

Fig. 117 Savonius Rotor vertical-axis windpump in Ethiopia. It was found to be less costeffective than the 'Cretan' windmill of Fig. 112 (See ref. [15])

216

Fig. 118 Typical Troposkien shaped Darrieus vertical axis wind turbine

Fig. 119 Turks and Caicos islands vertical-axis sail rotor (after UNESCAP [51])
There are also two main types of Darrieus wind turbine which have straight blades; both
control overspeed and consequent damage to the blades by incorporating a mechanism

217
which reefs the blades at high speeds. These are the Variable Geometry Vertical Axis
Wind Turbine (VGVAWT) developed by Musgrove in the UK and the Gyromill Variable
Pitch Vertical Axis Wind Turbine (VPVAWT), developed by Pinson in the USA. Although
the Musgrove VGVAWT has been tried as a windpump by P I Engineering, all the
current development effort is being channelled into developing medium to large
electricity grid-feeding, vertical-axis wind-generators, of little relevance to irrigation
pumping.
Vertical axis windmills are rarely applied for practical purposes, although they are a
popular subject for research. The main justification given for developing them is that they
have some prospect of being simpler than horizontal axis windmills and therefore they
may become more cost-effective. This still remains to be proved.
Most horizontal axis rotors work by lift forces generated when "propeller" or airscrew like
blades are set at such an angle that at their optimum speed of rotation they make a
small angle with the wind and generate lift forces in a tangential direction. Because the
rotor tips travel faster than the roots, they "feel" the wind at a shallower angle and
therefore an efficient horizontal axis rotor requires the blades to be twisted so that the
angle with which they meet the wind is constant from root to tip. The blades or sails of
slow speed machines can be quite crude (as in Fig. 107 ) but for higher speed machines
they must be accurately shaped airfoils (Figs. 113 & Fig. 114 ); but in all three examples
illustrated, the principle of operation is identical.
v.

Efficiency, power and torque characteristics

Any wind turbine or windmill rotor can be characterized by plotting experimentally

derived curves of power against rotational speed at various windspeeds; Fig. 120 A.
Similarly the torque produced by a wind rotor produces a set of curves such as in Fig.
120 B.
The maximum efficiency coincides with the maximum power output in a given
windspeed. Efficiency is usually presented as a non-dimensional ratio of shaft-power
divided by wind-power passing through a disc or shape having the same area as the
vertical profile of the windmill rotor; this ratio is known as the "Power Coefficient" or Cp
and is numerically expressed as:

the speed is also conventionally expressed non-dimensionally as the "tip-speed ratio"

( /?/. ). This is the ratio of the speed of the windmill rotor tip, at radius R when rotating at
radians/second, to the speed of the wind, V, and is numerically:

218
When the windmill rotor is stationery, its tip-speed ratio is also zero, and the rotor is
stalled. This occurs when the torque produced by the wind is below the level needed to
overcome the resistance of the load. A tip-speed ratio of 1 means the blade tips are
moving at the same speed as the wind (so the wind angle "seen" by the blades will be
45) and when it is 2, the tips are moving at twice the speed of the wind, and so on.
The cp versus curves for three different types of rotor, with configurations A, B, C, D, El,
E2 and F as indicated, are shown in Fig. 121. The second set of curves show the torque
coefficients, which are a non-dimensional measure of the torque produced by a given
size of rotor in a given wind speed (torque is the twisting force on the drive shaft). The
torque coefficient, Ct, is defined as:

where T is the actual torque at windspeed V for a rotor of that configuration and radius
R.

219

Fig. 120 The power (A) and torque (B) of a wind rotor as a function of rotational speed
for difference wind speeds

220

Fig. 121 The power coefficients (Cp) (above) and the torque coefficients (Ct) of various
types of wind turbine rotor plotted against tip-speed ratio ( ) (after Lysen/CWD [45])

221
vi.

Rotor solidity

"Solidity" () is a fairly graphic term for the proportion of a windmill rotor's swept area
that is filled with solid blades. It is generally defined as the ratio of the sum of the width,
or "chords" of all the blades to the circumference of the rotor; i.e. 24 blades with a chord
length (leading edge to trailing edge) of 0.3m on a 6m diameter rotor would have a tip
solidity of:

Multi-bladed rotors, as used on windpumps, (eg. rotor "B" in Fig. 121) are said to have
high "solidity", because a large proportion of the rotor swept area is "solid" with blades.
Such machines have to run at relatively low speeds and will therefore have their blades
set at quite a coarse angle to the plain of rotation, like a screw with a coarse thread. This
gives it a low tip-speed ratio at its maximum efficiency, of around 1.25, and a slightly
lower maximum coefficient of performance than the faster types of rotor such as "D", "E"
and "F" in the figure. However, the multi-bladed rotor has a very much higher torque
coefficient at zero tip-speed ratio (between 0.5 and 0.6) than any of the other types. Its
high starting torque (which is higher than its running torque) combined with its slow
speed of rotation in a given wind make it well-suited to driving reciprocating borehole
pumps.
In contrast, the two or three-bladed, low-solidity, rotors "El" and "F" in Fig. 121, are the
most efficient, (with the highest values for Cp), but their tips must travel at six to ten times
the speed of the wind to achieve their best efficiency. To do so they will be set at a slight
angle to the plain of rotation, like a screw with a fine thread and will therefore spin much
faster for a given windspeed and rotor diameter than a high solidity rotor. They also have
very little starting torque, almost none at all, which means they can only start against
loads which require little torque to start them, like electricity generators (or centrifugal
pumps) rather than positive displacement pumps.
All this may sound academic, but it is fundamental to the design of wind rotors; it means
that multi-bladed "high-solidity" rotors run at slow speeds and are somewhat less
efficient than few-bladed "low solidity" rotors, but they have typically five to twenty times
the starting torque.
vii.

Matching rotors to pumps

High solidity rotors are typically used in conjunction with positive displacement (piston)
pumps, because, as explained in Section 3.5, single-acting piston pumps need about
three times as much torque to start them as to keep them going. Low solidity rotors, on
the other hand, are best for use with electricity generators or centrifugal pumps, or even
ladder pumps and chain and washer pumps, where the torque needed by the pump for
starting is less than that needed for running at design speed. Table 18 indicates the
relative characteristics and Cp values for various typical wind rotor types so far
described.

222
Table 18 COMPARISON BETWEEN DIFFERENT ROTOR TYPES
Type

Performance
characteristic

Manufacturing
requirements

Horizontal axis
Cretan sail or flat Medium starting torque Simple
paddles
and low speed
Cambered plate High starting torque and Moderate
fan (American) low speed
Moderate speed Low starting torque and Moderate, but with
aero-generator
moderate speed
some precision
High speed aero- Almost zero starting
Precise
gen.
torque and high speeds
Vertical axis
Panemone
Medium starting torque Simple
and low speed
Savonius rotor
Medium starting torque Moderate
and moderate speed
Darrieus rotor
Zero starting torque and Precise
moderate speed
VGVAWT or
Zero or small starting Precise
Gyromill
torque and moderate
speed
*

Cp

Solidarity
6

t.s.r.*
(Optimum)

0.05 to
0.15
0.15 to
0.30
0.20 to
0.35
0.30 to
0.45

50% 1.52.0
50 to 80%

1.5-2.0

5 to 10%

3-5

under 5%

5-10

under
0.10
0.15

50%

.4-.8

100%

.8-1

0.25 to
0.35
0.20 to
0.35

10% to
20%
15% to
40%

1-1.5

3-5

2-3

t.s.r. = tip-speed ratio (X)

Fig. 120 A and B shows the load lines for a positive displacement direct-driven pump
superimposed on the wind rotor output curves. The dotted line on Fig. 120 A indicates
the locus of the points of maximum power; the system will only function continuously
when the operating point is to the right of the line of maximum power, as under that
condition any slight drop in wind speed causes the machine to slow and the power
absorbed by the shaft to increase, which results in stable operation. The operating point
can only remain to the left of the maximum power locus under conditions of increasing
windspeed. It can be seen that the positive displacement pump requires more or less
constant torque of 10Nm in the example, once rotation has been established, but it
needs over three times as much torque to start it for reasons explained in Section 3.5.
The torque curves in Fig. 120 B indicate that 5m/s windspeed is needed to produce the
torque required to start the windpump rotating, but once rotation has commenced, the
windspeed can fall to 3m/s before the operating point moves to the left of th maximum
powere locus and the windpump will stop. Note that the broken line a'-a represents a
transient condition that only occurs momentarily when the windpump starts to rotate.
To extract the maximum power from a windpump at all times would require a load which
causes the operating point to follow close to the locus of maximum power; (Fig. 120).
The figure also indicates that that the operating point will always be where the windpump
rotor curve for the windspeed prevailing at a given moment coincides with the pump load
line. In the example, the operating point is shown for a windspeed of 5m/s; in this
example, it can be seen that only about two-thirds of the maximum power that could be

223
produced in this wind speed is used by the pump, because its load line diverges from the
cubic maximum power curve. This discrepancy is a mis-match between the prime-mover
(the windmill rotor) and the load (the pump). The proportion of the power available from
the rotor in a given windspeed which is usefully applied is known as the "matching
efficiency", and is analysed in detail in Pinilla, et al [46], The figure illustrates how this
mis-match becomes progressively worse as the wind speed increases. This mis-match
is actually less serious than it may seem, since the time when the best efficiency is
needed is at low windspeeds when, fortunately, the best efficiency is achieved. When a
windmill is running fast enough to be badly matched with its pump, it means that the
wind is blowing more strongly than usual and the chances are that the output, although
theoretically reduced by bad matching, will be more than adequate, as the extra speed
will compensate for the reduction in efficiency.
It may be thought that centrifugal pumps would match better with a windmill than positive
displacement pumps, but in practice their efficiency falls rapidly to zero below a certain
threshold running speed at a fixed static head. In otherwords, centrifugal pumps do not
readily run with adequate efficiency over as wide a speed range as is necessary to
match most windmills rotors and they are therefore not generally used with windmills
(except with intermediate electrical transmission which can modify the relationship
between the pump and windmill speeds).
When generators are used as a load, instead of pumps, a much better match can be
obtained. Wind generators therefore tend to have a better matching efficiency over their
whole range of operating speeds than windpumps; the interested reader is referred to a
text on this subject, such as Lysen [45].
There is considerable scope for improving the overall performance of wind pumps by
developing methods of improving the rotor-to-pump match over a wider range of
windspeeds; a certain amount of, work is being carried out in this field and if successful
could result in considerably more effective windpumps in the future. But in the meantime
the main problem is to choose the most appropriate pump size for a given windmill in a
given wind regime and location. Fig.122 shows how the pump load line can be altered
simply by changing the mean pump rod pull, either by changing the stroke (by
lengthening or shortening the crank) or by changing the diameter of the pump being
used. A longer stroke and/or a larger pump will increase the pump rod force, and
increase the mean torque requirement and hence the slope of the load line, and viceversa. In Fig.122 it is clear that increasing the load increases the hydraulic output at
higher speeds, but it also increases the value of Vs, the starting windspeed. Therefore,
pump "C" in the diagram will start in a much lighter wind than the other pumps, but
because of the shallower load line the output will be much smaller in high winds. There
is therefore an important trade-off between achieving starting in adequately light winds
and achieving a good output.

224

Fig. 122 The trade-off between starting windspeed and output for differently loaded
windpumps

225

Fig. 123 The operating characteristic of a windpump showing how the power output and
matching efficiency vary with windspeed
The operating characteristic of a typical windpump, given in Fig. 123, shows how if the
start-up windspeed is Vs a windpump can run down to a slightly lower windspeed V (as
explained earlier, assuming the use of a piston pump). It reaches its best match with the
rotor at windspeeds close to Vmin (in theory at 0.8VS) [46] which is the "Design
Windspeed", and then increases its output almost linearly with windspeed to V (its rated
windspeed). At still higher windspeeds means must be introduced to prevent it speeding
up further, or the machine may be over-loaded and damaged or destroyed; various
methods for doing this are discussed in the next section below. At very high
windspeeds, the only safe course of action is to make the windmill "reef", "furl" or "shutdown"; the figure shows how this process commences at a windspeed Vf (furling speed)
and is completed at windspeed vsd (shut-down).
viii. Methods of storm protection and furling

Windmills must have a means to limit the power they can deliver, or else they would
have to be built excessively strongly (and expensively) merely to withstand only
occasional high power outputs in storms. Sailing ships "take in canvas" by wholly or
partially furling the sails (manually) when the wind is too strong, and Cretan sail

226
windmills and other such simple traditional designs generally use exactly the same
technique; fewer sails are used in high winds or else the sails are partially rolled around
their spars. Metal farm windmills, however, have fixed steel blades, so the solution most
generally adopted is to mount the rotor offset from the tower centre (Fig. 124) so that the
wind constantly seeks to turn the rotor behind the tower. Under normal conditions the
rotor is held into the wind by a long tail with a vane on it. This vane is hinged, and fixed
in place with a pre-loaded spring (as illustrated), then when the wind load on the rotor
reaches a level where the force is sufficient to overcome the pre-tension in the spring,
the tail will start to fold until the wind pushes the rotor around so that it presents its edge
to the wind, as in Fig. 124 This furling process starts when the rated output is reached
and if the windspeed continues to rise, it increases progressively until the machine is
fully furled. Then when the wind drops, the spring causes the tail vane to unfold again
and turn the rotor once again to face the wind. On commercial farm windmills, this action
is normally completely automatic.
Wind-generators and other windturbines with high speed, low-solidity rotors often use a
mechanism which changes the blade pitch; e.g. the Dunlite machine of Fig. 113 which
has small counter-weights, visible near the rotor hub, which force the blades into a
coarser pitch under the influence of centrifugal force when the rotor reaches its furling
speed, against the force of a spring enclosed in the hub. Alternatively air-brake flaps are
deployed to prevent overspeed. Larger windturbines do not use tail vanes to keep them
facing the wind, as they cannot stand being yawed as fast as might occur if there is a
sudden change in wind direction. Instead they usually have a worm-reduction gear
mechanism similar to that in a crane, which inches them round to face the wind; this can
be electrically powered on signals from a small wind direction vane, or it can use the
mechanism visible on the Windmatic in Fig. 114, used on large windmills for several
centuries, where a sideways mounted windrotor drives the orientating mechanism every
time the main rotor is at an angle other than at right angles to the wind direction.

227

Fig. 124 Typical windpump storm protection method in which rotor is yawed edge-on to
the wind (plan view)

228
4.7.3

The Wind Resource

It is not intended to attempt a detailed discussion of the complex subject of the causes
and behaviour of the wind or how it is measured and analysed; useful references are
Lysen [45], Park [46], Golding [47] and the W.M.O. [48].
The main fact to be aware of is that although the wind is extremely variable and
unpredictable on a minute-by-minute or an hour-by-hour basis, the actual average
windspeed at a given location over any given month of the year will not differ much from
year to year. So if mean monthly windspeed figures are available, taken over a number
of years, a reasonable prediction of the performance of a windpump should be possible.
A variety of methods may be used to do this, some of which are described in the next
section.
i.

The minimum wind requirements for windpumps

It has shown that, typically, windpumps require an average "least windy month"
windspeed of about 2.5m/s to begin to be economically competitive, (eg. Fraenkel, [40]).
Because of the cube relationship between windspeed and energy availability, which is
true for any optimally matched wind pump, and wind regime, the economics of
windpumps are very sensitive to windspeed. Therefore, windpumps are one of the most
cost-effective options (compared with engines or any other prime-movers) for pumping in
locations with mean wind speeds exceeding about 4m/s, but, conversely, they are not at
all cost-competitive where mean windspeeds are significantly below 2.5m/s. Fig. 125
rather crudely indicates the world's windspeed distribution pattern. This is considerably
simplified and a much more detailed treatment of this topic is available from the World
Meteorological Organization; [48]. It can be seen that much of the world, with the
exception of the centres of the major land-masses, and the equatorial forested regions,
is suitable for deploying windpumps. As a rule of thumb, areas that are free of trees (i.e.
savannah grasslands, semi-deserts and deserts) tend to be windy and suitable for
windpumps, while forested and wooded areas not only have less wind but trees make
siting of windmills difficult unless very high towers can be used.

229

Fig. 125 Annual mean wind speeds (approximate indication)

Various studies on the potential market for windmills in different parts of the world, eg.
[49], plus numerous country-specific studies of meteorological data, suggest that quite a
number of developing countries probably have areas with adequate wind speeds for the
use of windpumps. Some of these are listed below; those where windpumps are already
known to be in at least moderately widespread use are marked (+):
Algeria, Argentina(+), Cape Verde Islands(+), Chile, China (+), Cyprus (+), Ecuador,
Egypt, Ethiopia, India, Jordan, Kenya(+), Libya, Madagascar, Malta, Mauritania,
Mauritius, Morocco (+), Mozambique, Namibia(+), Oman, Pakistan, Peru(+), Senegal,
Somalia, Sudan, Syria, Tanzania, Thailand(+), Tunisia(+), Uruguay, Zambia,
Zimbabwe(+).
There are also many small islands which are not listed but which invariably have
adequate wind regimes due to the proximity of the open ocean.
ii.

Variation of wind speed with height

The speed of the wind increases with height. The rate of increase is dependent partly on
the height and partly on the nature of the ground surface. This is because rough ground,
with many uneven trees, bushes or buildings, causes turbulence, while a flat and
unobstructed surface like the sea or a flat grassy plain allows the air to flow smoothly
which results in higher windspeeds nearer to the surface. The relationship between
windspeed and height can be estimated as follows:

230

where V is the wind velocity at height H and Vr is a reference wind velocity measured at
height Hr. The exponent "a" is a function of the surface roughness, as follows, [50]:
type of terrain
smooth sea or sand
low grass steppe
high grass and small bushes
woodlands and urban areas

a
0.10
0.13
0.19
0.32

For example, if there is high grass and small bushes, and a mean reference windspeed,
for example, of 5m/s recorded at the standard meteorological office recommended
height of 10m, this can be adjusted to obtain the mean windspeed at 20m windmill hub
height as follows:

from which V= 5.7m/s A gain of 0.7m/s from mounting a windmill at 20m rather than at
10m may sound small in relation to the cost of the extra high tower required, but the
energy available at those two heights will be related to the cube of the velocities
(assuming optimally matched pumps in each case) and will therefore be:

this shows that a 48% increase in energy availability can be gained in terrain of that type
from using a 20m tower instead of a 10m tower (or a windmill with a smaller rotor could
be used to gain the same energy - in this case the rotor area could be reduced so that a
windmill with 20% smaller rotor diameter on a 20m tower would be used compared with
one on a 10m tower).
iii.

Effects of obstructions

Any obstruction to the wind has a wake extending up to 20 or 30 diameters (of the
obstruction) downwind. The wake is depleted of wind energy compared with the
surrounding wind, and is turbulent. For example a large mango tree (or similar rounded,
well-leafed tall tree) can have a wake which even 200-300m downwind has 10% less

231
wind energy than either side of it. Sharp edged and irregular obstructions such as rock
outcrops, cliffs and escarpments, or large buildings can cause violent turbulence which,
apart from depleting the energy available, can cause damage to a windmill located
nearby.
Therefore it is normal to recommend that windmills are mounted so that the rotor is at
least 200-300m from any significant obstruction to the wind. Ideally if obstructions like
trees or buildings are nearby, the rotor should be mounted on a high enough tower so its
bottom edge is a clear 5m or more above the highest point of the obstruction. In reality it
is often impossible to avoid obstructions, so the least that can be done is to try and
locate the windmill so that it is unobstructed from the direction of the prevailing wind.
Because the wind lifts to go over an obstruction, siting windmills with large obstructions
nearer than 100-200m downwind should also be avoided.
iv.

Wind measurements and wind records

Sizing of a windmill for a particular pumping duty is commonly done, where windmills are
already in general use, on the basis of the experience of other nearby users of
windpower. If a nearby site looks better or worse in terms of exposure to the wind, then
some suitable allowances must be estimated to compensate, without too much risk of
serious misjudgement. Usually in such situations where windpumps are reasonably
common, the suppliers will be able to recommend a suitable size to suit the duty
requirement on the basis of past experience.
It is less easy to pioneer the use of windmills in any particular area; in such situations it
is necessary to obtain some estimate of the local wind regime. One way to do this is by
obtaining data from the nearest meteorological station or airport and making allowances
on the basis of the relative exposure of the proposed site compared with the measuring
station. Usually the protocol for obtaining this data is to request it from the head office of
the Meteorological Department or the Civil Aviation Authority. Unfortunately, however,
most small rural meteorological stations were not set up primarily to log wind data, and
more often than not, they have incorrectly sited anemometers. Often the anemometers
are on 2m tall masts and are surrounded by trees or buildings; any readings from such a
site are next to useless for wind energy prediction purposes yet, unfortunately, they are
often logged, sent to head office and incorporated in the national data-base, where they
distort the apparent wind regime so far as its value for wind power is concerned.
Therefore, when using data from a local rural meteorological station, it is strongly
recommended that a visit should be made to the station to check whether the data were
measured in an acceptable manner to make them of value. Data from international
airports (or from major meteorological stations) are usually reliable as the anemometers
are normally located at the W.M.O. recommended height of 10m and they will be
unobstructed. This is especially true at airports where wind behaviour is of considerable
interest from the point of view of aircraft safety. Most such stations log wind speed and
direction data continuously on either paper charts or on magnetic tape.
The most basic format for national wind statistics is as in Fig. 126 which shows average
wind speeds collected over a number of years, for each month, for a selection of
meteorological stations in India. The Indian irrigation season typically occurs in the dry
period from January to May, so by inspection it is possible to identify places with

232
seemingly adequate wind regimes during this period; i.e. with monthly means preferably
exceeding 2.5-3m/s.

Fig. 126 Typical presentation of long term wind data as monthly averages
Another format commonly used for the presentation of wind data is given in Fig. 127,
which shows a so-called "wind rose"; this includes information on the wind direction as
well as its strength. The figure in the middle is the percentage of calm, (defined as less
than 3 mph) and the "petals" of the rose are aligned with the points of the compass, (N,
NNW, NW, WNW, W, etc.) and shows both the percentage of time the wind blows from
each direction; (the concentric circles are 5% intervals) and the mean windspeeds (given
in this case in "bins" of 3-8, 9-15, 16-38 and over 39mph. In the example shown the wind
is clearly predominantly northeasterly and southwesterly. Wind roses are of most interest
for comparing different locations. They are difficult to analyse from the point of view of
predicting wind energy availability; raw wind data or even monthly mean wind data are
more useful.

233

Fig. 127 Wind rose

If wind records are not available from a sufficiently close or representative existing
meteorological station, then it is necessary to set up an anemometer and log wind
records for at least one year and preferably two to three years. Obviously this is not a
recommendation so much for the small farmer wanting one small windpump, but rather
for institutional users contemplating a large investment in wind power which needs to be
soundly based on objective wind records. Ideally, about three years of records are
required to obtain reasonably representative averages, as mean monthly wind speeds

234
can vary by 10-20% or so from one year to the next. The need for this is of course
greater in areas which are thought to be "marginal" for the use of windpumps; it will be
shown in the next section that commercial windpumps begin to become economically
competitive with engines or other sources of power for water lifting in windspeeds above
about 2.5m/s. Therefore, in places which are decidedly windy; i.e. with mean wind
speeds almost certainly in excess of say 4m/s, there is no great risk in guessing the
mean wind speed when ordering a windpump, and it probably is not worth the trouble
and cost of carrying out long term wind measurements in advance. It is possible to "fine
tune" the original guess, if it turns out not to be sufficiently accurate, simply by changing
the stroke or the pump size. Subsequent windmills can be ordered on the basis of
experience with the first one. However, if it turns out that there is just not enough wind,
no amount of tampering with the stroke or pump sizes will adequately correct that
misjudgement.
The most simple method of measuring mean wind speeds is to install a cup-counter
anemometer which simply totalizes the kilometres (or miles) of wind run just as a car
odometer totalizes kilometres of road run. By noting the time when each reading is
taken, and dividing the difference between two readings by the time interval, it is
possible to determine the mean wind speeds over the time period. A mechanical cupcounter anemometer of meteorological office quality costs around US $300 (without a
mast - the mast can be improvised with 2" water pipe and guy wires). Ideally such
instruments should be read three times per 24 hours, in the early morning, at mid-day
and in the evening to allow the diurnal pattern of wind to be recorded. This allows the
mean wind speeds for the mornings, afternoons and night periods to be separated.
Failing this, an early morning and an evening reading should be taken each day to allow
day and night averages to be calculated. Once-a-day or once-a-week readings,
providing they are consistently and accurately logged, are however a lot better than
nothing, although they will not show diurnal patterns at all.
The effort involved in analyzing raw, continuously recorded data is formidable, so with
the recent upsurge of interest in windpower, numerous electronic data loggers have
come onto the market which can record wind data in a form that is convenient for "wind
energy prospectors". A commonly used approach is to log the frequency with which the
wind speed is measured to be blowing within a series of pre-defined speed "bins", such
as 0-5 km/h, 5-10 km/h, 10-15 km/h, and so on. If more accurate results are wanted,
then narrower bins may be defined to improve the resolution, but this of course requires
a more sophisticated logger or more analytical work afterwards.
The most useful starting point for any sophisticated attempt to predict the performance of
a windmill in a given wind regime is to create or obtain a velocity-frequency histogram
which shows the percentage of the time that the wind blows at different speeds; (as in
Fig. 128). This has been constructed from hourly wind data by adding up how many
hours in the year, on average, the wind was recorded as having been blowing at a
velocity within each pre-defined "bin"; for example, in Fig. 128, the bins are at 1 mph
intervals, and there were about 5 hourly records of 0 mph, 100 hourly records of 1 mph,
and so on. Clearly an electronic data-logger that automatically measures and records
the frequency of wind speeds in predetermined bins makes this task much easier.
It is also quite common to present wind data as a velocity-frequency curve. These are in
effect fine resolution velocity-frequency histograms. The wind regime of a given site is

235
characterized by the velocity-frequency curve which will have a similar shape every year
and will not vary very much from one year to the next. Velocity frequency curves can be
synthesized by a sophisticated mathematical process using what is known as a Weibull
Probability Distribution Function, which, providing certain parameters are correctly
selected, will produce a passable correlation with natural empirically measured wind
regime curves; the analysis required is beyond the scope of this book and is dealt with in
ref. [45] among others.

Fig. 128 Wind velocity-frequency histogram

Therefore, the best information that is ever usually available will be an hour-by-hour wind
frequency distribution curve for the site. Ideally, this should be combined with data giving
the monthly mean wind speeds and the percentage of calm per month. For irrigation
pumping it is of critical importance to consider the wind regime during the month of
maximum water demand; annual averages are not good enough for this.
v.

Windpump manufacturers' performance claims

The easiest method of estimating the performance to be expected from a windpump is to

use manufacturers' data as printed in their brochures. For example, Fig. 129 shows a
table and performance curves for the "Kijito" range of windpumps, based on the "IT
Windpump" and made in Kenya, (see Fig. 116). The table indicates the average daily
output to be expected at different pumping heads for the four sizes of Kijitos in three
different average speeds, defined as "light" 2-3m/s, "medium" 3-4m/s and "strong" 45m/s while the curves reproduce these results just for the "medium" wind speeds. It is
interesting to note how sensitive windpumps are to windspeed; the smallest machine
with a 12ft (3.7m) rotor will perform in a 5m/s wind almost as well as the largest machine

236
(24ft or 7.3m) does in a 3m/s wind; this is because there is 4.6 times as much energy
per unit cross-section of a 5m/s wind as in a 3m/s wind as a result of the cube law.

Fig. 129 Manufacturers' performance data for the Kenyan-made 'Kijito' windpump range
based on the I.T. windpump (see also Fig. 116 )
A problem with manufacturers' performance claims is that some brochures include
inaccurate, unreliable or even incomplete data. For example, it sometimes is not clear
what wind speed applies for the manufacturer's claimed outputs. There is a tendency for
manufacturers to quote performance figures for unusually high average wind speeds, no
doubt because this makes the performance look more impressive, and they then give
rules of thumb which in some cases do not seem accurate for reducing the outputs to
more realistic levels for more common wind speeds.
The difficulty inherent in monitoring the performance of a windpump often prevents users
from actually checking whether they are getting what they were promised; if a cup
counter anemometer and a water meter is available, it is possible to measure the wind
run and the water output over fixed periods of time (either 10 minute intervals or daily
intervals - short-term and long-term - is recommended). The mean wind speed and
mean water output over these periods can be logged and then plotted as a scattergram.
When enough points are obtained on the scattergram, a "best-fit" curve can be drawn to
obtain the performance characteristic. It is recommended that institutional users and
others who need to procure numerous windpumps should seek performance guarantees
from their supplier and should make some attempt to verify whether the performance is
being achieved, for example, in the manner just suggested. In some cases sub-optimal

237
performance can occur simply because the wrong pump size or wrong stroke has been
used, and considerable improvements in performance may result from exchanging the
pump for the correct size or from altering the stroke.

238
4.7.4
i.

Windpump Performance Estimation

General principles

To size a windpump for irrigation purposes will usually require an estimate to be made of
the week by week or month by month average output. One method for making such an
estimate is to combine data on the known performance of the windpump at various
hourly average wind speeds with data from a wind velocity distribution histogram (or
numerical information on the number of hours in the month that the wind blows within
pre-defined speed "bins"). This is illustrated by Table 19, which gives the expected
output of a windpump in various windspeeds, and the statistical average number of
hours that the wind blows within each speed range, (or speed "bin" is the favoured
jargon). Hence, the total output for each speed bin is obtained by multiplying the output
per hour at that speed and the number of hours at which that speed is likely to recur. By
adding together the output for each speed bin we arrive at the total annual output. The
importance of doing this monthly is that quite often the least windy month will have a
mean windspeed of only around 60 to 70% of the annual mean wind speed, so the
available wind energy in the least windy month can be as little as 20% of what can be
expected for a mean windspeed equal to the annual average wind speed. Therefore if
annual averages are used, a considerable margin of safety is necessary to allow for
"least windy month" conditions, (assuming irrigation water is needed in the least windy
month or in a month with a mean windspeed below the annual average.
Table 19
DATA

CALCULATION OF WINDPUMP OUTPUT USING "BINNED" WINDSPEED

Annual output of water for a given wind regime

Annual duration
Output rate
hours
m3/hr
m/sec

Wind speed
m.p.h.
7
8
9
10
11
12
13
14
15
15 plus

3.15
3.6
4.05
4.5
4.95
5.40
5.85
6.30
6.75

600
500
500
400
500
450
450
300
300
1,700
Total 5,700

0.3
1.4
2.3
3
3.7
4.2
4.7
5.2
5.7
6

total output
m3
180
700
1,150
1,200
1,850
1,890
2,115
1,560
1,710
10,200
Annual
Total 22,555 m3

A graphic way of achieving the same result as used in the Table 19 is illustrated in Fig.
130. Here the wind velocity frequency histogram, (preferably for a month at a time)
shown in (a) is multiplied by the windpump performance characteristic shown in
idealised form in Fig. 130 (b); in this example a windpump is assumed which starts at a
nominal 2m/s, produces 3kW of output at its rated speed of 5m/s and is fully furled at

239
9m/s. A windpump performance histogram using similar speed "bins" to diagram (a) is
constructed over the windpump performance characteristic. Finally the windpump's
generalised performance histogram in (b) is multiplied by the wind velocity distribution
histogram in (a) to arrive at a performance prediction histogram (c). To illustrate the
mechanics of multiplying two histograms together, the first speed bin in (a) is 200h, the
first in (b) is zero kW, so their product in (c) is zero kWh; the second is 600 x 0 = 0, the
third is 750hours x 0.5kW = 375kWh, the fourth 800h x 1.5kW = 1200kWh, and so on.

Fig. 130 Example of how to calculate the energy output of a windmill by using the
velocity frequency distribution of the wind regime (a) and multiplying it by the windmill
performance characteristic (b) to obtain the output (c)
ii.

Simple "rule-of-thumb" approach

The problem with the above methods is that the result is only as good as the wind and
performance data used, which all too often are unreliable. Therefore, if only approximate
data are available, a simpler rule of thumb may be more appropriately used than
attempting any detailed analysis, (and is often useful anyway as a cross check on other
methods). This method was originally proposed in [45] and [51], and provides a
reasonable method of estimation.
The rule of thumb assumes that a windpump system will, on average, be 17% efficient in
converting wind energy into hydraulic output, which in many cases is probably not a bad
estimate (the losses in a windpump system and the total efficiency that can be expected
are discussed in more detail in the next section of this chapter). The average hydraulic
output power for a windpump of about 17% average efficiency will be:

because the density of air at sea level is approximately 1.2kg/m3, it follows that:

240

If .1V3 is multiplied by the time interval in hours applying to the average windspeed used,
then the output can be calculated. For example, if V is the daily mean windspeed (based
on 24 hours), then the daily hydraulic energy output will be:

if this is multiplied by the rotor area in square metres, it gives the daily hydraulic energy
output. Dividing the number of watt-hours per day by 2.725 converts this to a daily "cubic
metres-metre" product, or m3.m (i.e. output in cubic metres times head in metres).
Although an unusual method of expressing energy, this can readily be converted to a
daily output of water at any particular pumping head; eg.:

which is 36.7m3/day if lifted through 10m, or 3.67m3/day through 100m, etc.

iii.

Overall system efficiency

Table 20 indicates the efficiency factors relating to windpumps,and shows that between
7 and 27% of the value obtained for the energy in the wind (using the mean wind speed)
can be converted to hydraulic energy. Because the speed of the wind continuously
varies, the actual energy in the wind is considerably greater than if the wind blew
continuously at the mean wind, speed. The exact proportion by which the energy
available exceeds the figure that can be calculated based on the mean depends on the
shape of the wind velocity distribution curve, but typically it can be twice the figure
arrived at using the mean. This explains why there is an "efficiency" factor of 180-250%
used to allow for the underestimate which would otherwise occur if the mean wind speed
were simply cubed with no other allowance.
Table 20 FACTORS AFFECTING WINDPUMP SYSTEM EFFICIENCY
factor
rotor to shaft
shaft to pump
pump
wind regime % energy capture
actual wind energy to wind energy calculated using mean wind speed
TOTAL EFFICIENCY OF ENERGY CONVERSION BASED ON
MEAN WIND SPEEED

typical efficiency
25 - 30
92 - 97
60 - 75
30-50
180 - 250
7 - 27 %

241
This result indicates that the figure of 17% overall efficiency used to produce the 0.1V3
rule of thumb described above may vary with different windpumps and wind regimes by
plus or minus 10% of the total energy available. Therefore, the rule of thumb coefficient
used of 0.1, and hence the predicted output using this method, may generally be
expected to be within the range of 0.6 to 1.6 times the result obtained by using 0.1 and
the mean wind speed.
iv.

Wind requirements for economic operation

A convenient method for costing and comparing windmills is to estimate the total
installed cost of a system as a function of rotor area (e.g. in dollars per square metre of
rotor). Since the energy output is a function of the wind regime combined with the
system efficiency and its rotor area, the unit cost of the rotor area will give an indication
of the cost of energy in a given wind regime if a uniform system efficiency is assumed,
(or upper and lower limits such as indicated in the previous section may be used).
Wind pump prices correlate quite closely with the area of rotor being purchased. Prices
for industrially manufactured windpumps, (not including shipping and installation) are in
the US $200-400/m2 range, while wind generators tend to be about three times as
expensive per unit of rotor area (for small machines) or twice as expensive for larger
machines. Windpumps made in developing countries can be significantly cheaper; for
example the "Tawana", Pakistan-made version of the I T Windpump costs about $130/m2
(in 1985). "Do-it-yourself" windpumps, built in the village, such as are used in Thailand
are very much cheaper still, but the cost depends very much on what assumptions are
made on the value of the construction labour.
Combining these costs with the previously given performance assumptions will show
that at current prices, most windpumps need mean winds speeds in the region of 2.5 to
3m/s to begin to be economically attractive, and wind generators need 3.5 to 5m/s.
Moreover the cube law ensures that the economics of windmills improve dramatically at
higher windspeeds, making them a most economically attractive technology in most wind
regimes having mean windspeeds exceeding say 5m/s.

242

4.8

SOLAR POWER

Small-scale irrigation pumping is one of the more attractive applications for solar power.
Solar radiation tends to be at its most intense when the need for pumped water is
greatest and the energy supply is available at the point of use, making the farmer
independent of fuel supplies or electrical transmission lines. The main barriers at present
to the wider use of solar pumps are their high costs combined with general unfamiliarity
of the technology. A cheap and reliable solar pump, which could quite possibly be
developed within the next few years, will have the potential to revolutionize Third World
agriculture. This is why it is worth being aware of the potential of this promising new
technology, even though it is probably not yet economically viable for irrigation.
Many solar pumps have been built and operated, so their technical feasibility is proven,
but the technology is still immature with production running in dozens or hundreds of
units per year rather than the thousands that must be manufactured before costs drop.
Also, solar pumps tend to become economically viable in water-supply applications
sooner than for irrigation, due to the much higher value that can be placed on drinking
water; in fact an economic case can already be made for using solar pumps for village
water supplies given favourable operating conditions. Solar pumps for irrigation are only
currently economically viable at very low heads where the power demand is extremely
small. Nevertheless, significant technical developments coupled with cost reductions are
being achieved and it can be expected that reliable and economically viable solar
irrigation pumps will be available within the medium to long term.

4.8.1

Background and State-of-the-Art

There are two main methods for converting solar energy into power for driving a pump.
Solar thermodynamic systems depend on using the heat of the sun to power an engine
(steam or Stirling cycle), while solar photovoltaic systems convert solar radiation directly
to electricity by means of photocells, and hence they can power an electric pump. The
most feasible routes for applying solar energy to pumping water are shown in Fig. 131.
Paradoxically, solar thermodynamic experiments date back over a century, yet today
such systems remain immature compared with the much more recent solar photovoltaic
alternative based on technology discovered as recently as the 1950s.

243

Fig. 131 Feasible options for solar-powered pumping systems

The first successful solar thermodynamic systems were developed in France during the
mid-nineteenth century; see Butti and Perlin, [52] and Daniels, [53]. By 1900 most of the
development effort had shifted to the USA, where several people were seeking to
develop commercially viable solar pumping systems. However, although several solar
steam engines were successfully demonstrated, they tended to cost several times as
much as a conventional steam engine of the same power (although their fuel costs were
of course free), which limited their commercial success. This work culminated in the
construction in 1914, by an American called Frank Shuman, of what even today remains
one of the largest and, for that matter most technically successful solar thermodynamic
pumping systems. Developed in the USA, but installed at Meadi in Egypt to pump
irrigation water, this unit incorporated hot water storage and could therefore drive an
irrigation pump for 24 hours per day. After a few teething troubles, the Meadi solar pump
was shown to develop 55 hp (40kW) and it could pump up to 1 300 m3/h (360 litre/sec).
Under Egyptian conditions at that time the plant could pay back in competition with a
steam engine in 2 years and could completely pay for itself in 4 years. Enormous interest
ensued, and ambitious plans to install similar solar pumps in other parts of the world
were initiated, but the First World War broke out and Shuman, the driving force behind
solar pumps at that time, died before it ended. The cheap oil era that followed led to a
lack of interest in solar power for pumping until the oil price increases of the 1970s.
Numerous researchers initiated laboratory research work on solar thermodynamic
systems during the 1970s, and SOFRETES, a French company, began manufacturing a
low temperature solar pumping system, nominally capable of about 500W-1000W of
pumped output, which was distributed with French government support to a number of
developing countries in the late 1970s on a pilot demonstration basis. Also, a number of
quite large solar thermal installations, some for pumping water and some for electricity
generation, were completed in the USA and in Africa. Unfortunately, the majority of
these latterday solar thermodynamic systems performed poorly and many also proved
troublesome to operate reliably under field conditions. Today the French company that
produced the solar thermodynamic systems is no longer in business and no other
manufacturer has a proven solar thermodynamic system commercially available.

244
In 1979, the UNDP (United Nations Development Programme) initiated a Global Project
specifically to evaluate small scale solar pumps for irrigation, with the World Bank as the
executing agency; (project UNDP GLO/78/004 followed by GLO/80/003). The UNDP
Project field trials were held in Mali, Sudan and the Philippines, and it was necessary to
select solar pumps which were at least nominally commercially available and which
showed reasonable prospects of being able to function adequately in the field, when bids
were invited from suppliers all over the world, it transpired that only one solar
thermodynamic solar pump was available for procurement by the project in 1980, and
this only barely fulfilled the minimum requirements to justify field testing it. In contrast, no
less than eleven solar photovoltaic pumps were short-listed from a still broader choice.
As these projects proceeded the emphasis moved from the use of solar pumps for
irrigation to their use for water supply duties, where they are more immediately
economically viable. Numerous reports were produced for this project; eg. [54],[55] and
[56], plus a "Handbook on Solar Water Pumping", [57], which is intended to help those
seeking more detailed information on selecting and sizing solar pumps.
Therefore, solar thermodynamic small scale systems have not really succeeded, despite
a century of development with early signs of promise, but the much newer photovoltaic
technology, although still expensive, works reasonably successfully. Photovoltaic
systems use photo-electric cells, which can convert sunlight directly to electricity. This
technology had its origins at the Bell Laboratories in the USA in the early 1950s. Solar
photovoltaic cells were originally developed to power space satellites, which in common
with many rural parts of the earth's surface, need a small independent power supply.
This technology therefore started at the frontiers of science on an "expense no object"
basis. Therefore, the earliest terrestrial solar power cells, which became available in the
1960s, were prohibitively expensive for such applications as irrigation pumping.
The considerable cost of solar photovoltaic cells, which are generally made from a highly
purified form of silicon (an abundant element) relates mainly to the high overheads
associated with the investment in sophisticated manufacturing facilities needed, rather
then to the cost of raw materials. There are therefore major economies of scale relating
to volume of production, so cost;: have consequently been declining steadily over the
last few years. By 1984, solar modules (the basic "building brick" for a solar array) could
be purchased on the world market for US $8.00 per rated peak watt. Solar photovoltaic
systems have their power capability defined in "peak watts" which is the electrical output,
that would be produced in direct sunshine of 1 000W/m2 strength when the solar module
has a cell temperature of 28C. Many predictions of radical reductions in solar cell costs
"by the mid 1980s" were made in the buoyant period for renewable energy development
of the late 1970s; in the event these have not yet happened, but nevertheless, significant
and useful cost reductions have been achieved (50% in real terms between 1980 and
1985). It can reasonably be expected that the present technology solar cells will come
down in price in real terms by another 50% as production volume increases, and this will
be enough to make them worthy of serious consideration for general small scale
irrigation pumping at heads up to about 6m. Possible new solar cell technologies could
conceivably result in even greater cost reductions in five to ten years' time, which would
make this technology decidedly attractive compared with the other options, (most of
which are getting more expensive in real terms).

245
4.8.2

Principles of Solar Energy Conversion

Because solar thermodynamic pumping systems, although experimented with for over a
century, remain generally immature, it is proposed to deal only briefly with them and to
concentrate on the currently more promising photovoltaic technology.
i.

Thermodynamic systems

Thermodynamic systems can be further sub-divided into three main categories,

operating at low, medium and higher temperatures with a "Rankine" or steam engine at
the lower or medium temperatures and a Stirling engine at the higher temperatures.
As indicated in Fig. 131, thermodynamic pumping systems always include a solar
collector, which collects solar heat and transfers it to a working fluid; an engine, which
takes heat from the working fluid and converts this to shaft power and a transmission
system to connect the engine to a pump.
Low temperature thermodynamic systems use what are called "flat plate collectors",
which usually consist of a flat absorber panel containing passages through which the
working fluid circulates. These are generally painted matt black and mounted in a
shallow insulated housing behind a glass window of similar area to the panel. The
absorber panel is mounted at an angle to the ground of approximately the latitude of the
location, and facing towards the equator. This has the effect of maximizing the amount of
solar energy received by the absorber.
The glass window readily admits light, but inhibits the outflow of the heat which is
produced; this is commonly known as "the greenhouse effect". This effect can be
enhanced with double glazing, but this also increases the costs, since the glass is one of
the more expensive components. Similarly, the absorber panel can be coated with a
special "selective surface" which also enhances the efficiency of absorption of solar
radiation, but at increased cost compared with regular black paint. Flat plate collectors of
this kind can only achieve working temperatures of up to about 80C, but more usually
will operate at about 60C. Flat plate collectors are therefore used with low-boiling point
working fluids such as ammonia, the freons (fluoro-hydrocarbon fluids which are used in
air conditioning and refrigeration circuits) or even butane (the latter poses some fire
hazard if it leaks). However, even with low boiling point fluids, the efficiency of a rankine
engine will be very low, because the thermodynamic efficiency is, for fundamental
physical reasons, a function of the temperature difference between the hot vapour
admitted to the engine and the cool vapour that is exhausted. The exhausted vapour can
only be cooled to the condenser temperature which will generally be the temperature of
the pumped water which is usually used to cool the condenser. Therefore it is only
possible to achieve a temperature difference (sometimes known as "delta T") of about
30C with flat plate collectors.
If higher temperatures are required it becomes necessary to focus, or concentrate, the
sunlight on a smaller area than it would naturally fall on. This is normally achieved by
using a parabolic mirror. The simplest arrangement is a line focussing parabolic trough
collector, which focuses the light falling on it into a line, as in Fig. 132; greater
temperatures and concentration factors can be achieved with a "point focus" parabolic
dish concentrator. These offer improved thermodynamic efficiency in terms of a greater

246
"delta T" (line focussing achieves 100-200C while point focussing achieves 200-500C).
Unfortunately focussing, concentrating solar collectors have two major disadvantages
(which apply equally to the use of focussing with photovoltaics as well as with
thermodynamic systems):
a. they need to be aimed at the sun so that the sun's rays are focussed on the heat
absorber; the greater the concentration, the greater the accuracy with which the
device must be aligned with the sun. To do this requires a mechanical means to
make the collector track the sun;
b. they can only focus direct rays (beam radiation) and therefore cannot make use
of scattered light (diffuse radiation) which is available to flat plate collectors. In
areas with haze, atmospheric dust, high humidity or scattered clouds there is
much diffuse radiation, often amounting to as much as 30% or more of the total
solar irradiation, which is not accessible to a concentrating collector but which
can be used by a flat plate collector.

Fig. 132 Losses in a typical solar thermodynamic pumping system

247
Therefore, although a concentrator improves the thermodynamic efficiency of a solar
pumping system, and thereby reduces the area of solar collector needed to achieve a
given power output, the added complication increases the cost per square meter of solar
collector and also reduces the available solar energy by limiting usage to direct beam
radiation. Analysis by Halcrows and I T Power [54] [55], indicates that a compromise of
using low concentrating factor line focussing collectors appears to be the most costeffective approach with current solar collector costs, but that future technical
developments could make point-focus systems competitive if a clever, low cost and
reliable parabolic dish and sun tracking mechanism is developed. But this would only
apply in locations (such as deserts) with a high proportion of direct beam radiation.
Fig. 132 is a schematic diagram of a solar thermal pumping system, showing the energy
flow and the principal losses. A good solar collector will typically collect 60% of the solar
energy falling on it. Of this 60%, about 7% will be converted to shaft power and the rest
is lost as rejected condenser heat, feed pump and expander losses, and finally perhaps
50% of what is left, i.e. 3.5% of the original input, will be converted into useful hydraulic
pumped power. This is an example of the better types of small solar thermal systems,
many that have been developed are not nearly as good as this in practice.
ii.

Photovoltaic power systems

Systems of this kind depend on a property of certain semi-conductors to generate

electricity when exposed to light. The physical principles are explained in broad detail in
[57] or in depth in [58]. A number of different photovoltaic cell materials exist, but the
only commercially proven power cells currently available are based on thin wafers cut
from purified crystals of silicon. The majority of these are sliced from single crystals of
silicon and are known as "mono-crystalline" cells. However, a recent process involves
the growth of several crystals simultaneously, which when sliced yield so-called "polycrystalline" cells. Both technologies are competitive at present.
Silicon is an abundant element, but the purification, crystalization and slicing process
requires high technology and is expensive. Work is in hand on a variety of cell
technologies that promise in some cases improved efficiency and in others lower costs,
but none are as yet at the production stage. Some of the so-called "thin film"
technologies under development promise substantial reductions in costs which could
greatly enhance the scope for using when they become commercially available.
Normal practice with mono- or poly-crystalline silicon cells is to slice them into wafers,
which may be circular or square and are typically 100mm across, but quite thin (about
1mm or less). The front and rear surfaces are doped with impurities to create the
necessary semi-conductor properties. The rear surface is metalised, while the front has
a fine pattern of conducting metal collectors plated on to it which are usually connected
by two thicker conductors; see Fig. 133. Light falling on the cell produces a potential
difference (or voltage) between the conductors on the top surface and the metalised rear
surface. Each cell can develop about 0.4 volts when under load, (0.6V on open circuit),
so by connecting a string of cells together in series, a higher voltage can be created.

248

Fig. 133 Construction of a silicon photovoltaic cell

Solar cells are delicate, and expensive, and need to be safely enclosed behind an
appropriate window to let the light in. Therefore strings of cells are generally mounted in
glass fronted frames known as "modules". Toughened glass is usually used for the cover
(this is sometimes given a matt finish to minimise reflection) and the cells are
encapsulated behind the glass in plastic with either a metal or glass rear surface. The
frame is commonly made of light-alloy and a good seal is needed between the frame
and the glass.
Typically, cells are mounted in strings of 36 per module, giving a nominal voltage of 1416V. 100mm diameter cells typically produce about 1W of electrical output when
exposed normal to full strength sunlight amounting to 1000W/m2, so a standard module
of 36 x 100mm cells will usually be rated at 35W (recent types with larger and/or more
efficient cells are rated at 40W and more).
Solar modules are the most expensive part of a solar photovoltaic system, costing in
1985 typically US $7.00/W(p) (per peak watt); i.e. a standard 40W module currently
costs around $280. Fortunately, good quality modules are highly reliable and should last
20 years or more, providing they are not subject to physical damage to the glass covers.
A serious worry, once their high intrinsic value becomes more widely known and a
"second-hand market" develops, is the possibility of theft.
Modules are mounted on a supporting structure in a frame and this assembly is known
as a photovoltaic array which can simply be connected by electric cable to the
motor/pump sub-system. Examples of four typical types of solar photovoltaic pumping
systems are shown in Fig. 134.

249
In a few cases arrays have been developed which track to follow the sun; this of course
increases the amount of sunlight intercepted as the array will always face the direct
beam radiation. However, the complication and expense of providing mechanical
tracking has not generally been found worthwhile so it is not common practise. A few
portable, or semi-portable small scale solar pumping systems exist in which the arrays
can be manually orientated by moving them bodily; here some significant advantage at
little effort or cost is gained by facing the array(s) south east in the morning and south
west in the afternoon when in the northern hemisphere (or NE and NW respectively in
the southern hemisphere).
Most arrays are generally designed to carry the modules at a fixed tilt angle which
maximises the amount of sunlight received over the year. It so happens that the
theoretical angle for this to be achieved coincides with the angle of latitude of the
location. This is because that is the angle which will place the array perpendicular to the
sun at solar noon on the equinoxes, and will therefore make the divergence of the sun's
rays from perpendicular to the plain of the array a minimum at all other times. In practice,
the actual angle for the array is sometimes varied so as to bias the optimization to suit a
season having more cloud cover; for example, in areas with a marked rainy season it
may be advantageous to incline the array more normal to the sun in that season, and
sacrifice a little solar energy in the dry season when perhaps more energy than
necessary can be intercepted. Also, although mounting the array at zero angle (i.e.
horizontal) is the optimum angle for the equator, it is normally recommended that arrays
be mounted at angles of at least 10 so that there will be good rainwater run-off, which
helps to keep them clean. It is often advantageous to have an array with a capability of
having its angle adjusted manually every few months, as this can improve the output by
10% or more, over the year.

250

Fig. 134 Examples of solar pump configuration

A. submerged motor/pump set
B. submerged pump with surface motor
C. floating motor/pump set
D. surface motor with surface mounted pump

251

Fig. 135 Performance characterists of silicon photovoltaic cells

252
The performance characteristics of silicon solar cells are shown in Fig. 135. It can be
seen that the efficiency of energy conversion is in the region of 10%; e.g. in peak
sunlight (which yields approximately 1000 W/m2) lm2 of silicon cell surface will yield
about 100W of electrical output.
Unfortunately, the efficiency of solar cells declines as they get hotter as shown in Fig.
135 A. Most solar cells attain temperatures in the range 50-60C in full sunlight
conditions, but their power rating is generally quoted by manufacturers for an operating
temperature of 25C, which coincides with laboratory conditions for a standard method
of testing. They therefore will never achieve their rated output in practise under normal
sunshine, but sizing methodologies, such as are described later, take account of this
discrepency.
Fig. 135 B indicates the voltage-current (V-I) characteristic of a solar cell, with 1000
W/m2 irradiance. At the one extreme you can short-circuit the cell and get a maximum
current Isc (the short-circuit current) of about 30mA/cm2 of cell, but the voltage will be
zero. At the other extreme, on open circuit, there is zero current, but an open-circuit
voltage, Voc of about 0.55-0.60V per cell, (regardless of its size). The maximum power
occurs when the load applied is such that the maximum value of the product of V and I is
obtained (electrical power in watts is numerically volts times amps). This occurs near the
"knee" of the V-I curve at around 0.4V per cell, as indicated by the dotted curve showing
power. As the efficiency is exactly proportional to the power output, the efficiency curve
is identical in shape to the power curve.
The sun mostly shines with a strength of significantly less than 1000W/m2, (which is a
mid-day intensity), so Fig. 135 C indicates how, if the irradiance is reduced by say 80%
to 200W/m2, the current density declines proportionately, but the voltage stays the same.
The power from a cell is therefore almost linearly proportional to the intensity of
illumination from the sun. Because a voltage is produced even in low levels of irradiance,
solar photovoltaic systems can function in low levels of sunlight, depending on the
nature of the electrical load and what the threshold power for starting happens to be.
A few manufacturers have produced solar photovoltaic systems with concentrating
collectors, using mirrors or lenses to intensify the strength of sunlight on the cells. This
allows a smaller area of cells to be used, but the added expense of mirrors or lenses
seems to negate any saving on cells, especially bearing in mind the other disadvantages
of concentrating collectors discussed earlier.
iii. Photovoltaic motor-pump sub-systems

Photovoltaic pumping systems require, in addition to the array, a so-called "sub-system"

which consists of at least an electric motor powering a pump. Fig. 134 illustrates various
commonly used configurations that can comprise a photovoltaic pumping sytem. Fig.
136 shoes schematically the necessary (and optional) components of a solar
photovoltaic pumping system and how they interact.
Because the output from an array is DC (direct current), either a DC electric motor needs
to be used, or alternatively an inverter can be used to synthesize AC so as to allow a
conventional mains power AC electric motor to be used. The disadvantage of using an

253
inverter is its cost and the power loss inherent in its use, but the advantage is that many
relatively low cost, mass-produced, standard electric pumps can be used.

Fig. 136 Schematic arrangement of a photovoltaic solar pumping system. (Note optional
linkages shown with broken lines and essential components shown with heavy outline)
Where DC motors are used, the most efficient types are permanent magnet motors,
where the flux is provided by magnets rather than an electro-magnetic field coil. A
problem with DC motors is that carbon brushes are generally needed and these are a
potential source of trouble as they wear and need regular replacement. Brushless,
maintenance-free, DC motors have recently been developed however, which have
electronic circuitry to perform the same function as a commutator and brushes.
Much of the discussion in Chapter 4.6 in the context of mains (or generating set)
electricity applies to the motors and pumps of photovoltaic electric systems. For
example, combined submersible motor pump units may be used, (either the borehole
type or surface water type) which need either AC or brushless DC motors since they are
sealed and therefore not suited to brushed motors. Surface suction pumps direct
coupled to motors may be used, or surface mounted motors driving submerged pumps
via a long shaft. Also, a transmission may be used to convert the rotary output of a motor
to a slower reciprocating output and to drive a standard piston pump.
For irrigation duties, generally at low heads, the most common and useful system is
either a submersible motor pump unit, usually suspended from a float for pumping
surface water, or a fixed installation usually using a surface mounted motor driving a
centrifugal or multi-stage turbine pump via a long shaft. Surface-suction centrifugals also
have a role, but self-priming capabilities are essential with a solar power source or the
user may need to reprime the pump everytime a cloud passes by.

254
Because the motor-pump combination will have an optimum performance at a certain
voltage and current, the photovoltaic system designer tries to match a motor-pump unit
to the array so that it will electrically load the array under all typical irradiance levels so
as to keep the voltage and current levels as near to the "knee" of the photovoltaic
module characteristic as possible. There is an optimum characteristic curve for any array
where the output is maximised under all solar conditions, so the aim is to provide a
motor and pump with a characteristic such that with varying currents caused by changes
in irradiance level, the voltage requirement follows the array maximum characteristic
curve as closely as possible. With good design, centrifugal pumps can be sized to load a
system very close to this optimum right across the operating range. Positive
displacement pumps, on the other hand, need varying voltage to change their output and
therefore they inherently fail to match well with a solar power system. Another difficulty
with positive displacement pumps is that they generally need more torque (and hence
current) to start than to keep going, which makes starting in low levels of irradiance
problematic unless some artificial method to overcome the problem is used.
A general solution to matching arrays and motor-pump units (and to allow for variations
in their load lines when operating off the design head, for example), is to use an
electronic power conditioning device called a Maximum Power Point Tracker (MPPT).
This can convert a DC input at one voltage to a DC output at another voltage. It is
possible to sense the output current and voltage and provide a control function which will
automatically adjust the output voltage to obtain maximum power for the irradiance level
and pump load applying at any time, generally using a microprocessor to make the
logical decisions. This approach results in extra cost for the MPPT, plus a small parasitic
power loss inherent in any extra component, but in many cases, particularly with positive
displacement pump systems, it greatly improves performance. The gains from an MPPT
are less worthwhile in the case of well-matched low head direct-driven centrifugal pumps
of the kind which tend to be most often used for irrigation duties.
A controversial technical question is whether batteries should be used with solar pumps.
While they are not essential, and therefore not often used, they can provide a useful
element of energy storage, allowing the system to work when the user wants to use it
rather than when it happens to be sunny. Batteries also provide power conditioning
capability as an alternative to a MPPT in that they can deliver a different current to that
received. However their life under tropical conditions tends to be limited to four or five
years, and they need regular topping up with distilled water unless more expensive
sealed batteries are used.
As always, each component causes a loss and an inefficiency; the power flow through a
typical photovoltaic pumping system is illustrated in Fig. 137. This indicates that, at best,
the following efficiencies can be expected:
photovoltaic modules
motor-pump unit, cables and controls

11%
45%

Therefore the actual hydraulic output, ignoring pipe losses will be about 5% of the power
of the sun arriving through the window of the array. Pipe losses, even with a well
designed low head system, may be another 10% of the hydraulic output, or 0.5% of the
total input, giving at best an overall efficiency of 4.5%.

255

Fig. 137 Losses in a typical solar photovoltaic pumping system

Tests, such as are reported in ref. [59], have shown that the best systems do achieve
this level of efficiency, but many systems are not nearly as good being typically only 23% efficient or worse. Efficiency is an important factor in selecting a solar pump,
because with the high price of photovoltaic arrays low efficiency implies a need for a
proportionately larger and more expensive array. A measure of this can be gained by an
example of comparing the use of a 70% efficient motor with an 80% efficient one for a
500W(p) system. At this power level, 10% extra array size to make up for the less
efficient motor will cost approximately US $300-400. Since the difference in cost
between an efficient motor and an inefficient one is usually much less than this, it is
obviously worth buying the best motor available. This decision is of course usually made
by the system manufacturer and not the user, but the informed user can at least
question the choices made by manufacturers by studying their specifications before
buying.

4.8.3

The Solar Energy Resource

The average value of solar irradiance just outside the earth's atmosphere is 1 353 W/m2.
Solar radiation is attenuated as it passes through the atmosphere so that the maximum
level normally recorded at sea level is around 1 000W/m2. This is made up of two
components, the direct beam radiation from the sun and the diffuse radiation consisting
of light scattered by the atmosphere from haze, the sky itself and clouds.

256
The combination of direct beam and diffuse radiation is called the global radiation; the
level of global radiation varies through the day because the path length the light has to
take through the atmosphere is greatest when the sun appears low in the sky in the early
morning and late afternoon and least with the sun overhead at solar noon. For the same
reason there are variations with both season and latitude. The total solar radiation
received in a day can vary from about 0.5 kWh/m2 on a sunny northern winter day to
over 6.0 kWh/m2 on a sunny day in the tropics. Obviously on a cloudy day the solar
energy reaching ground can be almost zero. On a really clear, cloudless sunny day
diffuse solar energy reaching the ground may be 15 to 20% of the global irradiance, but
on an overcast day it can only be 100%.
Because of the variability of solar irradiation, the performance of solar pumping systems
can be strongly influenced by location, time of year and weather. However, just like with
wind energy, solar energy tends to be consistently available at any particular location
from one year to the next; in other words, analysis of past solar energy records allows
accurate prediction of solar energy and hence of system sizing. To some extent it is
easier to achieve accuracy with solar energy prediction than with wind energy prediction
since it is less affected by local site conditions, (providing the site is not actually shaded
by trees), and any errors in solar energy measurement have less influence on system
performance than comparable percentage errors in wind speed recording.

4.8.4

Performance Estimation

Accurate sizing of a solar pump array is necessary to minimize the array size for a given
duty and thereby to achieve the most cost-effective system possible. If there is doubt
about the sizing, a technically acceptable approach is to use a larger than necessary
system in order to guarantee an adequate output, but the cost will tend to increase in
proportion with power rating; 10% over-size therefore means 10% more expense than
necessary.
Solar pumps used for irrigation should be sized for the "critical month"; that is when the
system is most heavily loaded in relation to energy available. This is usually the month of
maximum irrigation water demand, which fortunately is rarely one of the least sunny
months, since crop irrigation water demand and solar energy tend to be well correlated.
'By contrast, a solar pump for drinking water supply duties is likely to need to deliver a
constant daily water output, which would in this case make the least sunny month the
critical month for sizing purposes.
Fig.138 gives an example of mean monthly solar energy availability plotted on the same
page as mean monthly gross irrigation water requirement for the Lake Tchad region of
Africa. It can be seen from this that the month of June is not the sunniest month, but it is
the critical month in terms of solar irrigation pump sizing, for that is when the water
demand is largest in relation to the energy availability. The sunniest month, April, is not a
critical month as it is too early in the growth cycle for the crop in question (i.e. cotton),
while the fields are fallow in the second sunniest month of March.
Determination of the critical month, the mean daily water demand and the mean daily
solar irradiation for that month are the starting points to size a solar pump (and hence
to estimate its likely cost). Irradiation figures can be obtained from most national
meteorological departments, and they are also published covering the whole world, for

257
each month in references such as [57] [59] (as well as in some of the brochures of
several of the larger manufacturers of solar photovoltaic power systems). Such
published data are more region-specific than location-specific so there is a good chance
of making an adequately accurate judgement simply on the basis of regional irradiation
figures.

Fig. 138 Solar energy availability compared with crop irrigation water demand (Lake
Chad region)
References such as [57] and [60] give a detailed explanation of rigorous methods for
solar photovoltaic system sizing, but for most purposes a reasonable estimate can be
made using a rule-of-thumb approach, such as the one explained below. In any case,
most photovoltaic system manufacturers maintain computer models which they use for
sizing systems; these generally include a world-wide data-base, so when requested to
provide a quotation they can minimize the system size so as to price competitively.
Therefore the potential purchaser should request quotations from several sources and
compare the sizings and prices quoted. Care should be taken, however, to avoid simply
taking the cheapest system and thereby possibly obtaining one which is undersized for
the duty required.
Solar pump economics are discussed in more detail in Chapter 5, but in general terms,
solar pumps begin to be economically competitive for applications where:

258
a. the peak daily head-flow product is under 150m m (eg. 60m /day through 2.5m
3

head)
and where
b. the mean daily irradiation is greater than 4.2 kWh/m2 (or 15MJ/m2) in the critical
month.
A rule-of-thumb method may be used to obtain an approximate idea of the size and cost
of a photovoltaic array, as follows:

i.

estimate the peak daily hydraulic output (in kWh) required, using either Fig. 13 or
the following formula:

where Q is the output in m3/day and H is the head in metres. eg. 8mm of water on 0.3ha
pumped through a head of 10m; the volume is 24m3/day which requires (24 x 10)/367 =
0.654kWh(hyd)/day;
ii.

iii.

assume a figure for the sub-system efficiency; i.e. the efficiency of conversion of
electricity to hydraulic output (wire to water). This is sometimes given in
manufacturers' brochures or in technical studies such as [55] or [56]. However
for small low head (2-5m) systems a figure of around 30% may be appropriate,
while 40% can be achieved by more powerful and higher head units (say 5-20m
head);
divide the daily hydraulic output by the efficiency (say 35%) to arrive at the daily
electrical energy demand for the system; eg. in the same example, 0.65kWh/0.40
= 1.625kWh(e) (assuming a sub-system efficiency of 40%);

259
Fig. 139 World map giving average annual Clearness Index for solar distribution

iv.

v.

vi.

refer to the map in Fig. 139 and note the approximate Clearness Index for the
location in question; eg. if we take Dakar in Senegal, West Africa, the map shows
Dakar in the zone between 60 and 70%, at say 63%. The latitude of Dakar is
approximately 15N, so reference to the mean global extraterrestrial radiation
chart in Fig. 139 gives a mean of 9.73kWh/m2. This needs to be multiplied by the
relevant Clearness Index of 63%, which yields 0.63 x 9.73 = 6.1 kWh/m2 per day
on average. To allow for below average months and the inaccuracy of this
estimating technique it is prudent to reduce this value by 20% for solar pump
sizing purposes to give, say 0.8 x 6.1 = 4.9kWh/m2. To summarise this; take the
extraterrestrial irradiation for the latitude, multiply by the Clearness Index given
on the map, and finally reduce by 20%.
Take the daily electrical energy requirement (calculated in (iii) above) and divide
it by the daily irradiation (as calculated in (iv) above) and multiply the result by 1
200 to get the approximate peak watt rating required for the array. In the above
example it is (1.625/4.9) x 1200 = 398W(p).
Finally, arrays come in nominal 35 or 40W(p) modules, so it is then necessary to
divide the answer by 35 or 40 (as appropriate) and round up the answer to an
integer (whole) number of modules, e.g., with 40W modules we get 398/40 =
9.9, therefore 10 x 40W(p) modules would be needed, giving a true array rating
of 400W(p).

Typical total total solar pumping system costs (1985) can be estimated on the basis of
US $15-25 per W(p) of array, depending on the supplier and the specification of the
system. Lower powered systems in the region of 200 to 500W(p) will tend towards the
upper end of this price range, while higher powered systems, of say 2 000W(p) or more
will tend towards the lower end. These figures are of course very approximate, as is the
sizing methodology outlined, but with the example worked through we can see that the
above example requiring a 400W(p) system, is likely to cost in the region of US $6 000-8
000. These costs are likely to come down during the next few years by perhaps 25 to
50%, as the volume of photovoltaic manufacturing increases.

260

4.9
4.9.1
i.

HYDRO POWER
Background and State-of-the-Art

The use of water power for irrigation

The main shortcoming of water power as an energy resource for irrigation is that it is
only available for convenient use in a limited number of locations having suitable flows
and heads to engineer an effective site. Also, most regions with hydro-potential tend also
to have plentiful rainfall, which often makes irrigation a low priority as a hydro-power
application. However, despite these drawbacks, there are many areas where rainfall is
seasonal (eg. where water can usefully be applied during the hot dry summer months to
gain an extra crop or to increase yields and, similarly, in mountainous areas, where
snow-melt water often provides some limited hydro-potential in the dry summer season).
However, steep hydraulic gradients in hilly country generally allow any valley floor areas
to be gravity irrigated by digging contoured canals, which take water from a stream and
direct it along the side of a valley at a lesser hydraulic gradient then the main stream.
This practice is of course widespread in much of the central Asian massif. When gravity
does the work, pumping techniques are generally not needed.
Therefore, the main applications for the use of water power to pump water will be in
lowland areas, inevitably using low head water drops as a power source, to irrigate land
which would not be accessible to gravity water flow. An obvious important application,
common in China, is to extend the command area available to a gravity irrigation
scheme to take in higher land above a dam. In some arid regions, or regions with dry
seasons, there are large perennial rivers where the river current can be utilized to lift
water which would otherwise flow past parched fields. Even in some wetter and more
mountainous areas there are situations where water power could allow irrigation of
terraces or plateaux that are inaccessible to gravity flow; this can be of importance
where flat land with soil suitable for cultivation is scarce.
Given a suitable site in proximity to a suitable need, hydro-power has a number of
important and fundamental attractions:
a. it is generally available 24h/day
b. it is a relatively concentrated energy source
c. the available energy is easily predictable in areas where river flow data are
available
It follows from these points that hydro-powered systems tend to have a high power to
size ratio and hence a favourable power to cost ratio, they also tend to be mechanically
simple and robust, and therefore have long working lives and require limited and simple
maintenance. As a result hydro-power can be one of the most economic sources of
power for those fortunate enough to have a suitable resource available.
ii.

General principles

All water power applications involve removing energy from falling or flowing water. The
power P available from a flow of Q litre/sec falling through a head of H m is:

261

Where 9.8 is the gravitational constant (in metric units of m/s2), and if Q is in m3/s, then
the above formula gives P in kW. In otherwords there is 9.8kW available for a flow of
lm3/s per meter drop. The actual output will be reduced by multiplying by the system
efficiency; eg. a device with an efficiency of 50% will convert half the available power.
The efficiency of hydro-powered systems is high. A good small turbine will typically be
70% efficient and even water wheels and other more primitive devices tend to be 30 to
60% efficient, as indicated in Table 21.
Table 21 EFFICIENCY OF HYDRO POWERED SYSTEMS
type of device

Undershot water wheel

Vertical shaft water mill
Poncelet undershot or Breast
wheel
Overshot water wheel
Impulse turbine (eg Pelton)
Reaction turbine (eg Francis,
Prop.)
Water pressure engine
Turbine-pump
Hydraulic ram (hydraulic output)
River current converter

efficiency
factor
(typical)
0.30-0.40
0.20-0.35
0.50-0.65
0.50-0.70
0.70-0.85
0.60-0.80
0.60-0.80
0.35-0.50
0.30-0.60
0.25-0.30

Note that in all cases above, except turbine-pumps and hydraulic ram pumps, we are
considering the conversion simply of water power to shaft power; to pump water then
requires the addition of a pump or water lifting device which in turn will have further
inherent inefficiencies. The exception has an efficiency figure relating to the hydraulic
output divided by the hydraulic input, since there is no measurable intermediate
production of shaft power.
It can be seen that the two key parameters required to estimate the power potential of a
water resource are head and flow. Techniques for doing this are discussed in most
references on micro-hydro power, such as [61] and [62], The static head is the most
straightforward to measure; methods which can be used include:
- surveyor's levelling equipment;
- spirit level, wooden pegs and straight edge;
a length of hosepipe with a pressure gauge (lm of water = O.lkg/cm2 or .098bar or or 1.42p.s.i.
- or 9.81kPa); the hosepipe is completely filled with water with its open end submerged in the
upper water source and the other end fitted with the pressure gauge is held at the lower level to

262
measure the actual static head;
- barometer/altimeter (for higher head applications); (i.e. surveying barometer/altimeter)
- accurate existing large-scale maps to take advantage of a previous survey
The flow rate can vary a lot from season to season; obviously the key period of interest
to farmers will be the irrigation season. With larger rivers, especially those having
existing hydro-electric schemes on them, discharge measurements will already have
been carried out and been recorded and will usually be available from the relevant river
authority. However, in most cases involving small-scale equipment for hydro-powered
water lifting it will only be of interest to consider quite small streams, rivers or canals, for
which no data will be available. There is a variety of methods of stream gauging varying
in sophistication and accuracy, such as:
a. dam the stream and measure the overflow by measuring the time to fill a
container;
b. construct a small rectangular weir from wood or concrete and use the method
given in Fig.140
c. in streams where it is not practicable to build a weir, measure the depth of the
stream at equal intervals across its width (at a point where the stream is straight
and uniform) as in Fig. 141. Then time a float drifting down the centre of the
measured section. This gives the speed of the current at the centre of the
stream; the mean velocity will be 0.6-0.85 of this figure. A rough, rocky stream
bed requires a factor of 0.6, while smooth muddy surfaces require a factor of
0.85. The flow is calculated from the product of the mean velocity and area of
stream cross-section using appropriate units:
e.g. velocity in m/s. x cross-section in m2 gives flow rate m3.
Obviously a current meter or a ship's log, if available, can more conveniently be
used to measure the current velocity instead of timing a float.
d. dilution salt gauging is a new technique in which a known quantity of salt is
poured into the centre of the stream at one point. An electrical conductivity meter
is then used downstream from that point to measure the dilution. The data can
then be used to determine the flow rate;
e. use of automatic stream gauging instruments; a calibrated weir is used to derive
the day by day river flow on an electric or clockwork water level recorder.

263

Fig. 140 Flow measurement with rectangular weir (after BYS Nepal)

264

Fig. 141 Method of stream gauging without the need to build a weir
In some cases it may be of interest to use river or canal current powered devices, in
which case it is necessary to measure the current velocity either by timing the passage
of a float thrown into the current between two points on land which can be aligned with it
as it floats past, or a calibrated current metering device (or ship's log) can be used.
It is extremely difficult to estimate flow or current velocity by eye; human judgement is
peculiarly bad at assessing these kinds of factors and very inaccurate results are likely.

4.9.2
i.

Use of Turbines for Water Lifting

General types of turbine

As with pumps, there are families of different types of turbine to deal with different types
of situation. Briefly, these can be characterized as follows:
a.
b.
c.

low head
medium
head
high head

propeller/Kaplan

Banki/Francis

Pelton/Turgo

A typical small fixed blade propeller turbine is shown in Fig. 142 A. These are only
adequately efficient over a narrow range of flows. Where variable flow and power is
needed, adjustable gates are provided and the turbine runner may have fixed or
adjustable pitch blades. The latter is known as a Kaplan turbine and is more efficient
over a wider range of flows than a fixed pitch propeller turbine. But the complication of

265
adjustable pitch runner blades is expensive, and therefore is only normally applied for
larger-scale installations.
There is a middle head range of turbines, some of which involve mixed flow, rather like a
mixed flow pump in reverse. Turbines of this kind run full of water, and are known
generically as Reaction Turbines. The most widely used medium-head reaction turbine
resembles a centrifugal pump (except the flow travels radially inwards from a spiral
casing) and is a Francis turbine; Fig. 142 B. Reversed centrifugal pumps have actually
been used as a cheap form of turbine, and can be quite acceptably efficient over a
narrow range of flows and speeds at a given head. Another type of medium head turbine
is known as the Banki (alias Ossberger, Mitchell or Cross-Flow) turbine. Here a jet of
water impinges on a set of curved blades mounted between two discs, travels through
the centre of the rotor and emerges from the far side again. Turbines of this kind do not
run full of water and are known as impulse turbines in that they derive their rotation from
deflecting a jet of water. High head applications are generally dealt with by another type
of impulse turbine known as a Pelton Wheel (Fig. 142 C). With this a high speed jet of
water is directed at a series of buckets set around the rim of the wheel. Each bucket has
a central splitter which splits the jet in two and directs it almost back in the direction it
came from and outwards from the rim of the wheel.

266

267
Fig. 142 Some of the main types of hydro turbine for low, medium and high heads
The reasons for using different turbine types at different heads and flows relate to the
efficiency and speed of rotation that is required. Low head types of turbine, like propeller
turbines have a high specific speed; this means they tend to rotate faster in relation to
the velocity of the water travelling through them, which is important at low heads if
electrical machinery, or rotodynamic pumps, are to be driven. However high heads
would cause equipment of this kind to run at unacceptably high speeds, (and other
problems would also occur) so lower specific speed types of turbine become necessary
as the head increases. The interested reader should consult a standard text book on
hydro-turbines as further discussion is beyond the scope of this work.
The shaft of a turbine can be connected directly to a generator, which then can be used
to power electric pumps for irrigation, or it can be directly coupled to an appropriate
centrifugal or other rotodynamic pump. The route via electricity is of interest in that
irrigation water pumping is. highly seasonal, and a system which produces electricity can
in many cases perform useful duties other than irrigation, such as providing light at night,
post harvest processes, etc. However electricity generation inevitably involves a higher
level of engineering sophistication and investment than is inherent in powering a simple
pump by a simple turbine. Also, any hydro-electric plant needs to be well protected from
flooding, while a simple turbine powering a pump is much less liable to damage and can
therefore be more simply installed. There are also losses of efficiency inherent in
converting shaft power to electricity, transmitting the electricity and then converting it
back to shaft power; this could absorbe 25 to 40% of the energy converted.
The chosen approach must depend on the size of the hydraulic resource, and the
potential for satisfying other economic applications, plus of course the financial and
organizational resources of the user community.
ii.

Chinese turbine-pumps

The Chinese have taken the combination of turbines and pumps to the logical conclusion
of producing a large range of integrated turbine-pump units. Although this is a relatively
recent development (originating in the early 1960s) some 60 000 turbine pumps were
reportedly in use [63] irrigating 400 000ha by 1979.
Chinese turbine-pumps are generally for low head applications, where the hydro-power
source will often be a canal jump in an irrigation scheme or a weir on a canalized river
giving a head in the region of 1 to 15m. Therefore the turbine most commonly used is a
fixed pitch propeller turbine which is appropriate for low heads. This is generally
mounted with a vertical shaft surrounded by fixed gates; it therefore tends to be at its
most efficient only over a narrow range of flow rates. A centrifugal pump impeller is
mounted on the same shaft as the turbine, back-to-back as in Fig. 143. Where a high
head water supply is needed, multi-stage centrifugal pumps may be connected to the
turbine as in Fig. 144. In some cases an extension drive shaft can be fitted, as in Fig.
145; this allows the turbine to be used as a general power source at times when there is
no demand for irrigation water. For example it can readily be applied to powering a small
rice mill, oil expeller, generator, etc. This clearly can greatly enhance the economic value
to be gained from the installation.

268
A large variety of different sizes and models of turbine pumps are made by numerous
small manufacturers in China, and attractively low prices have been quoted for the
export market. Table 22 indicates the range of sizes of single stage turbine pumps
manufactured by a typical production unit, the Youxi Turbine-Pump Plant, Fujian, and
the (1981) export prices quoted for these by the Fujian Provincial Agricultural Machinery
Import & Export Corporation.
Because it can be misleading to make judgements on the economics of Chinese
equipment on the basis of Chinese export prices converted at the prevailing exchange
rate, some limited Chinese data on the relative costs of alternative pumping systems
within the Chinese economy are of interest. It can be seen from Table 23 that turbine
pumps are significantly cheaper than either electrically energized pumps or engine
pumps, as is to be expected. Whether the same differentials would apply elsewhere is
open to question, but the turbine-pump is almost certain to be economically attractive
wherever there is the need for irrigation water combined with suitable installation sites.
Table 24 indicates the performance of a selection of small low, medium and high head
turbine-pumps, including estimates of shaft power and efficiency. The overall efficiency
is in the range 32 to 50% for the models considered; this implies that the turbines and
pumps considered individually rather than in combination have efficiencies in the 56 to
71% range, assuming roughly equal efficiency for each. Other models exist with claimed
overall efficiencies as high as 58%.
Because fixed pitch propeller turbines only have a narrow operating range where high
efficiency can be achieved, it is important that they are accurately sized to suit the flow
and head. Where varying flow conditions occur, it is usual to install several small units
rather than one large one. This means each unit can always be run near to its optimum
flow condition by shutting them down one by one to cater for reduced flows. A good
design strategy is to install two units, one with twice the flow capacity of the other; then
both are used under the maximum flow condition, the larger can take 2/3 of the
maximum flow on its own, and the smaller can take 1/3 of maximum flow on its own.
This allows efficient operation at 1/3, 2/3 and full-flow.
Turbine pumps are typically installed on a concrete platform built into a weir, as in Fig.
146. Therefore, although the turbine-pump unit is inexpensive, depending on the site,
civil workings are likely to represent the largest cost-element. Pipes also will be
expensive, but then for a given flow they will be equally expensive regardless of the
choice of pumping system.
Higher heads or higher flows are commonly catered for by connecting turbine pumps in
series or parallel.

269

Fig. 143 Single stage high lift turbine pump (cross-section)

270

Fig. 144 Multi-stage high lift turbine pump (cross-section)

271

Fig. 145 Turbine pump fitted with extension drive shaft (show part-sectioned)

272
Table 22 TYPICAL SIZES AND PRICES OF CHINESE TURBINE-PUMPS
Type
20-6
30-6
40-6
60-6
60-16

Head
(m)
1-4
1-4
1-4
1-4
1-6

Lift
(m)
6-24
6-24
6-24
6-24
16-96

Flow
(1/s)
88-180
210-420
370-740
684-1354
650-1620

Discharge
(1/s)
6-12
21-42
37-72
70-142
25-57

Weight (net)
(kg)
59
155
290
985
1374

Price (net)
(US$)
89.00
127.00
198.00
571.00
1237.00

Table 23 COMPARISON OF IRRIGATION COSTS IN CHINA

Type of system

Lift
(m)
30
30
70

Electric
Engine
Turbine pumps

Irrigation cost
(US $/ha)
35-90
70-130
5-15

Table 24 SOME PERFORMANCE DATA OF SMALL TURBINE PUMPS

Turbine pump
model
High yield
High yield
High yield
Low head ZD
680
Low head ZD
680
Low head ZD
680
Medium head
ZD 540
High head ZD
440
High lift
High lift

10-0
20-4
20-6
30-6
10-6
10-6
20-6
20-6
30-10
30-16
Z20-6
Z20-6
G20-6
G20-6
40-6
40-6
40-17
40-17

Working Working Delivery Delivery Shaft

Overall
head
flow
head
flow
power efficiency
(m)
(1/a)
(m)
(1/a)
(watts)
(%)
0.5
15
3
0.8
47
32.8
0.5
60
2
7.8
203
1.0
84
6
6.0
570
42.8
1.0
190
6
16.0
1,340
50.6
1.0
20
6
1.2
130
4.0
40
24
2.4
1,000
36.0
1.0
81
6
6.4
560
47.0
6.0
198
36
15.5
8,180
6.0
446
60
19.2 19,250
6.0
446
96
11.2 19,250
43.0
5
139
30
10
4,620
14
232
84
16 21,600
42.0
11 14,400
12
172
72
14 30,900
38.0
20
222
120
0.5
238
3
20.9
5.0
750
30
66.1
1.0
333
17
3.3
6.0
815
102
8.5

273

Fig. 146 Typical turbine pump installation

274

Fig. 147 Schematic diagram of hydram installation

4.9.3

The Hydraulic Ram Pump (or Hydrant)

The hydraulic ram pump, or hydram, concept was first developed by the Montgolfier
brothers, better remembered for their pioneering work with hot-air balloons, in France in
1796. Essentially, a hydram (shown schematically in Fig. 147), is an automatic pumping
device which utilizes a small fall of water to lift a fraction of the supply flow to a much
greater height. In other words, as with the turbine pump, it uses a larger flow of water
falling through a small head to lift a small flow of water through a higher head. The main
virtue of the hydram is that it has no substantial moving parts, and is therefore
mechanically extremely simple, which results in very high reliability, minimal
maintenance requirements and a long operational life.
Its mode of operation depends on the use of the phenomenon called water hammer and
the overall efficiency can be quite good under favourable circumstances. Fig. 147
illustrates the principle; initially the waste valve (1) will be open under gravity, and water
will therefore flow down the drivepipe (2) from the water source (3) (having been drawn
through a strainer (4) to prevent debris entering the hydram). As the flow accelerates,
the hydraulic pressure under the waste valve and the static pressure in the body of the
hydram (5) will increase until the resulting forces overcome the weight of the waste valve
and it starts to close. As soon as it starts to close, and the aperture decreases, the water
pressure in the valve body builds up rapidly and slams the waste valve shut. The moving
column of water in the drivepipe is no longer able to exit via the waste valve so its
velocity must suddenly decrease; this continues to cause a considerable rise of pressure

275
which forces open the delivery valve (6) to the air-chamber (7). Once the pressure in the
air chamber exceeds the static delivery head, water discharges through the delivery pipe
(8). Air trapped in the air chamber is simultaneously compressed to a pressure
exceeding the delivery pressure. Eventually the column of water in the .drivepipe comes
to a halt and the static pressure in the casing then falls to near the static pressure due to
the supply head. The delivery valve will then close, due to the pressure in the air
chamber exceeding the pressure in the casing. Water will continue to be discharged
through the check valve (9), after the delivery valve has closed, until the compressed air
in the air chamber has expanded to a pressure equal to the delivery head. At the same
time, as soon as the delivery valve closes, the reduced pressure in the casing of the
hydram allows the waste valve to drop open, thereby allowing the cycle to start again.
The air chamber is a vital component, as apart from improving the efficiency of the
process by allowing delivery to continue after the delivery valve has closed, it is also
essential to cushion the shocks that would otherwise occur due to the incompressible
nature of water. If the air chamber fills with water completely, not only does performance
suffer, but the hydram body, the drivepipe or the air chamber itself can be fractured by
the resulting water hammer. Since water can dissolve air, especially under pressure,
there is a tendency for the air in the chamber to be depeleted by being carried away with
the delivery flow. Different hydram designs overcome this problem in different ways. The
simplest solution requires the user to occasionally stop the hydram and drain the air
chamber by opening two taps, one to admit air and the other to release water. Another
method on more sophisticated hydrams is to include a so-called snifting valve which
automatically allows air to be drawn into the base of the air chamber when the water
pressure momentarily drops below atmospheric pressure at the moment after flow
commences when the waste valve reopens. It is important with such units to make an
occasional check to see that the snifting valve has not become clogged with dirt and is
working properly.
This cycling of the hydram is timed by the characteristic of the waste valve. Normally it
can be either weighted or pre-tensioned by an adjustable spring, and an adjustable
screwed stop is generally provided which will allow the maximum opening to be varied.
The efficiency, which dictates how much water will be delivered from a given drive flow,
is critically influenced by the valve setting. This is because if the waste valve stays open
too long, a smaller proportion of the throughput water is pumped, so the efficiency is
reduced, but if it closes too readily, then the pressure will not build up for long enough in
the hydram body, so again less water will be delivered. There is often an adjustable bolt
which limits the opening of the valve to a predetermined amount which allows the device
to be tuned to optimize its performance. A skilled installer should be able to adjust the
waste valve on site to obtain optimum performance for that particular hydram and site.
Therefore, it can be seen that the output of a hydram is constant, 24 hrs/day, and cannot
readily be varied. A storage tank is usually included at the top of the delivery pipe to
allow water to be drawn in variable amounts as needed.
i.

Installation requirements

In a typical hydram installation, a supply head is created either by digging a small

contoured diversion canal bypassing a river, or in some cases, particularly with small
streams, it is normal simply to create a weir and to install the hydram directly below it.

276
Where greater capacity is needed, it is common practice to install several hydrams in
parallel. This allows a choice of how many to operate at any one time so it can cater for
variable supply flows or variable demand.
The size and length of the drivepipe must be in proportion to the working head from
which the ram operates. Also, the drivepipe carries severe internal shock loads due to
water hammer, and therefore normally should be constructed from good quality steel
water pipe. Normally the length of the drivepipe should be around three to seven times
the supply head. Also, ideally the drivepipe should have a length of at least 100 times its
own diameter (eg. the design length for a 100mm (4in) drivepipe is about 10m, and for a
150mm (6in) drivepipe it is about 15m). The drivepipe must generally be straight; any
bends will not only cause losses of efficiency, but will result in strong fluctuating
sideways forces on the pipe which can cause it to break loose.
The hydram body needs to be firmly bolted to a concrete foundation, as the beats of its
action apply a significant shock load. It should also be located so that the waste valve is
always above flood water level, as the device will cease to function if the waste valve
becomes submerged. The delivery pipe can be made from any material capable of
carrying the pressure of water leading to the delivery tank. In all except very high head
applications, plastic pipe can be considered; with high heads, the lower end of the
delivery line might be better as steel pipe. The diameter of the delivery line needs to
allow for avoiding excessive pipe friction in relation to the flow rates envisaged and the
distance the water is to be conveyed. It is recommended that a hand-valve or checkvalve (non-return valve) should be fitted in the delivery line near the outlet from the
hydram, so that the delivery line does not have to be drained if the hydram is stopped
for adjustment or any other reason. This will also minimize any back-flow past the
delivery valve in the air-chamber and improve the efficiency.
ii.

Choice of hydram design

Traditional hydram designs, such as Fig. 148, developed a century ago in Europe, are
extremely robust. They tend to be made from heavy castings and have been known to
function reliably for 50 years or more. A number of such designs are still manufactured in
Europe and the USA in small numbers. The hydram in Fig. 148 differs from the
schematic diagram of Fig. 147 in having its waste valve on the same side as the drive
pipe, but its principle of operation is identical.
Lighter designs, fabricated using a welded sheet steel construction . (Fig. 149) were
developed first in Japan and are now in production in other parts of SE Asia including
Taiwan and Thailand. These are cheaper, but only likely to last a decade or so as they
are made from thinner material which will probably eventually corrode; nevertheless they
offer good value for money and are likely to perform reliably for a respectably long time.
However, hydrams are mostly intended for water supply duties, operating at higher
heads and lower flow rates than are normal for irrigation. Therefore it is likely that the
most useful hydrams for irrigation purposes will be the larger sizes having 100-150mm
(4"-6") drivepipes. Some simple designs that can be improvized from pipe fittings have
also been developed by aid agencies, (such as in Fig. 150) and some interesting
versions have also been quite crudely improvized using scrap materials, such as in
southern Laos using materials salvaged from bombed bridges and with old propane
cylinders as air chambers. Needless to say, such devices are very low in cost; the pipes

277
in the end cost considerably more than the hydram. They are not always as reliable as
traditional designs, but nevertheless are usually quite adequately reliable and are easy
to repair when they fail.

Fig. 148 Traditional (Blakes) European hydram design Fig. 149 South-East Asian type of hydram

Fig. 150 Low-cost hydraulic ram using standard pipe fittings

Table 25 HYDRAM INPUT CAPACITY
Nominal diameter of drive ins. bore
pipe
mm. bore

11/4 11/2 2 21/2 3

32

40

50 65

80 100 125 150 175

8
200

278
Volume of driving water litres per
required to operate the ram minute

from

12

27 45

68 136 180 364 545

770

to

16

25

55 96 137 270 410 750 1136 1545

Table 26 HYDRAM PERFORMANCE

Supply
Head
(m)
1.0
1.5
2.0
2.5
3.0
3.5

Delivery head (m)

5
144

7.5
77
135
220
280

10
65
96
156
200
260

4.0
5.0
6.0
7.0
8.0
9.0
10.0

15
33
70
105
125
180
215

20
59
54
79
100
130
150

30

255
310

173
236
282

115
155
185
216

20
36
53
66
87
100

Litres pumped in 24 hours per 1/min of drive

water.

12.0
14.0
16.0
18.0
20.0

40
12
19
33
40
65
75

50

60

80

100 125

15
25
32
51
60

20
24
40
46

12
16
27
32

12
18
20

12
14

86
118
140
163
187

69
94
112
130
149

53
72
94
109
125

36
50
64
82
94

23
36
48
60
69

16
23
34
48
55

212
245

168
187

140
156

105
117

84
93

62
69

295

225
265

187
218
250
280

140 113
167 132
187 150
210 169
237 188

83
97
110
124
140

iii. Performance characteristics

Table 25 (from [64]) indicates the input capacity of different sizes of hydram; it is this
input which determines the delivery flow at a given delivery head. The lower limit
indicates the minimum input flow required for practical operation, while the upper limit
represents the maximum possible flow a hydram can handle. Table 26 [64] then
indicates the probable litres pumped in 24 hrs for each litre/min of drive water.
Although the costs of hydrams are apparently low, as soon as high flow rates are
needed at lower heads, the size of hydram and more particularly of drive pipe, begins to
result in significantly higher costs. Therefore hydrams are best suited to relatively low
flow rates and high head applications, (perhaps terraced tree nurseries in mountainous
regions) while a turbine-pump, as discussed in the previous section, appears more
attractive for the lower heads and high flow rates that are more common for irrigation of
commercial crops on low-land farms.

279
4.9.4

Water Wheels and Norias

The undershot waterwheel is probably both the most obvious and the oldest method of
extracting energy from rivers. In many cases the device simply dips into the river and is
turned by the movement of the current; see Fig. 151. In the example illustrated, from
Vietnam, the entire structure is made of bamboo and bamboo tubes with one end closed
are mounted around the rim of the wheel. The bamboo tubes dip into the river and reemerge filled with water, which they to near the top, where the water pours out into a
trough. Devices of this kind are quite widely used in SE Asia, including China, Japan and
Thailand as well as Vietnam, and are known as "norias". Fig. 152 shows a Chinese
version and illustrates the principle and method of construction quite clearly. The noria is
similar in many ways to the Persian wheel and was discussed earlier in Section 3.4.1.
The biggest shortcoming of the noria is that they need to be of a diameter somewhat
greater than the head; this makes them fine for low head applications, but they get large
and cumbersome for higher lifts. The example illustrated in Fig. 151 is 10m in diameter
and is claimed to be able to irrigate about 8ha [65]. There is a small weir just visible in
the illustration which creates a head of about 100mm at the base of the wheel which
significantly improves the performance. The 10m diameter Vietnamese noria turns at the
rate of about 1 revolution in 40 seconds and delivers water, typically, at the rate of 7
litre/sec. The cost of these Vietnamese norias is quoted as being the equivalent of US
$225-450, [65]. Although the noria is attractive in being relatively inexpensive and also
being capable of manufacture in the village, the sites where they may be used are
limited and they are particularly prone to damage by floods. Therefore annual repair
costs quoted for Vietnamese norias [65] can be as much as 30-50% of the capital cost of
the installation. However, under Vietnamese conditions the same source indicated that
some sample installations which were surveyed produced a return for the users in terms
of value of grain production in the range 24-60% over the estimated costs.

280

Fig. 151 Bamboo water wheels, Vietnam

281
Fig. 152 Chinese type of small-scale Noria

Fig. 153 Water wheel driven coil pump (see also Fig. 50)
A modern version of the Asian noria is the floating coil pump, versions of which have
been tested by the Sydfynsgruppen (supported by DANIDA) at Wema on the Tana River
in Kenya, by the Danish Boy Scouts on the White Nile at Juba in southern Sudan and by
the German agency BORDA on the River Niger near Bamako in Mali [66] (see Fig. 153).
The principle of the coil pump is explained in Section 3.6.4 and also illustrated in Fig. 50.
These recent experimental river current powered irrigation pumping systems consist of
floating undershot water wheels (mounted on floating pontoons made from empty oil
barrels), and by using a coil pump it is possible to use quite a small diameter water
wheel and to use it to lift water to a height considerably greater than its diameter (around
10-15m is possible from a 2-3m diameter rotor).
The claimed performance of the prototype floating coil pumps was 6.6 litre/sec against a
delivery head of 5m with a river current velocity of 1.2 m/s on the Tana River at Wema.
For reasons which are not known, the similar unit tested on the Nile only pumped 0.7
litre/sec against 5m head with the same river current velocity of 1.2m/s. Quite high
current velocities in the range from l-2m/s (2-4 knots) are necessary for devices of this
kind. The Kenyan machine was made mainly from glass-reinforced plastic (fibreglass)
and cost US $4 000 in 1979, but the much less efficient Sudan prototype used steel and
wood, with flexible plastic pipe for the coil pump and only cost $350 In both cases the

282
considerable length of flexible pipe required accounted for about one third of the total
cost.
The Royal Irrigation Department of Thailand has developed a similar floating undershot
waterwheel device to those just described, but in this case the wheel is mechanically
linked to a conventional piston pump by means of sprockets and chains. It requires a
minimum current velocity of lm/s with a river flow of at least 0.6m3/s and it is claimed to
be capable of pumping from 0.3-1.5 litre/sec to heads of 60m-15m respectively. The cost
was quoted at $1 450, not including the delivery piping and header tank.
A general problem with water wheels is that if the drive is taken from the main shaft, the
costs become high in relation to the power available, because large slow-moving (and
therefore expensive) mechanical drive components are needed to transmit the high
torque involved. Also, with shafts only turning at between 1-5 rpm (which is typical of
waterwheels) either very large pump swept volumes are required or a lot of gearing up is
necessary to drive a smaller pump at an adequate speed; either way the engineering is
expensive in relation to the power. Therefore waterwheels, although apparently offering
simple solutions, are not always as easy to adapt for powering mechanical devices such
as pumps as might at first be expected.

4.9.5
i.

Novel Water-Powered Devices

The Plata Pump

An unconventional alternative to a small water wheel for powering a small pump is the
Plata Pump (Fig. 154) . This device was invented in New Zealand and some efforts have
been made to commercialize it internationally, although it is not believed to be very
widely used. In it, a series of small turbine rotors are mounted on a single shaft along the
axis of a cylindrical duct, approximately 2.5m long by 0.5m in diameter. It is a bit like a
multi-rotored propeller turbine, although there are no diffusers or other static blades to
control rotation of the fluid. The shaft drives two opposed single-acting piston pumps via
a crank. The Plata Pump is intended to be mounted in a low dam or weir, so that it
slopes at a slight angle and water runs downhill through it. It is not supposed to run full
of water, but works best when running around 1 to J full; perhaps because when running
full the flow rotates and thereby causes loss of efficiency.

283

Fig. 154 Cut-away view showing general arrangement of a Plata pump installation
The Plata Pump is designed to operate on heads of 0.25-lm; usually this is engineered
by placing the Plata Pump at an appropriate angle on a stream bed and building up a
weir with rocks or other material to create the necessary head. The overall efficiency of
the Plata Pump has been measured as being in the 6-30% range at delivery heads from
6-90m; [66], The best efficiency was recorded at 24m head. Typical performance with a
working flow of 85 litre/sec is 1.3 litre/sec at 6m, 0.25 litre/sec at 24m, 0.11 litre/sec at
38m. A working flow of 153 litre/sec was necessary to allow 90m delivery head to be
reached.
The price of the Plata Pump in kit form was about US $2 000 in 1980, [66], so it seems
to be relatively expensive in relation to its performance in comparison, for example, with
Chinese turbine pumps described earlier. It also appears to be less efficient and robust
than the turbine pump.
ii.

The river current turbine

The energy of motion (or kinetic energy) available in river or canal currents is, exactly as
for wind, proportional to the cube of the velocity. The relationship between power and
velocity is:

where p is the density of water (1000kg/m3 for fresh water), A is the area of cross section
of current in m2 and V is the mean velocity through the cross-section in m/s. From this
the following power densities can be calculated, as shown in Table 27:

284
Table 27 POWER DENSITY IN WATER CURRENTS AS A FUNCTION OF WATER
VELOCITY
velocity (m/s)
(knots)
power density (kW/m2)

0.5

1.0

1.5

2.0

1.0
0.06

2.0
0.5

3.0
1.7

4.0
4.0

It is interesting to compare this with Table 16 giving the equivalent result for wind; similar
power is experienced in water at about 1/9 the velocity in wind needed to achieve the
same power, due to the much higher density of water. Since a mean wind speed of 9m/s
would be considered most attractive for the economic use of wind power, it is clear that
even currents of lm/s may be more than adequate as a power source if they could be
exploited efficiently. Moreover, a major problem with exploiting wind energy is its great
variability, but river currents generally flow steadily 24 hours per day, [67].
The traditional and most obvious technique, undershot water wheels, are an inefficient
means to exploit currents, since the bulk of the machine is external to the water at any
one time and therefore provides no power. Therefore the Intermediate Technology
Development Group (UK) undertook a programme, financed by the Netherlands
Government, to develop turbines which would operate efficiently entirely submerged to
extract shaft power from river currents; (the author must declare an interest, since the
device in question was developed under his technical supervision).
The main device tested is a vertical axis cross-flow turbine similar to the Darrieus
windmill in principle (see Section 4.7.2). It has the advantage that the vertical drive shaft
conveniently comes through the surface, so that any mechanical components can be
located on the deck of a pontoon above water. Unlike a water wheel this device can
intercept a comparatively large cross-section of current using little material other than a
rotor to do so; moreover it turns relatively fast in relation to the current (13.5rpm in this
case with a lm/s current) which reduces the gearing needed to drive a pump at a
reasonable speed. The concept is illustrated in Fig. 155 , while an actual working
irrigation pumping unit, installed on the white Nile near Juba in southern Sudan, worked
reasonably reliably, was fitted with a 3m diameter vertical axis rotor (3.75m2 crosssectional area) powering a centrifugal pump via a two stage toothed belt speed
increaser. It was tested and found to pump approximately 3.5 litre/sec through a head of
5m with a current of 1.2m/s. The rotor efficiency is 25 to 30% (as with a small windmill)
and an overall system efficiency of 6% has been achieved, including pipe losses as well
as pump and transmission losses. The prototype, which cost the equivalent of US $5
000 to build, has been used successfully to irrigate a 6ha vegetable garden. Smaller,
simpler, low-cost versions were also tested near Juba. Costs are difficult to determine
with prototypes, but the device appears to be potentially economically attractive.

285

Fig. 155 IT river current turbine pump

The potential for using the river to pump its own water has been demonstrated, but
further work will be needed to optimize this device and adapt it for commercial
production. However, the considerable power potential in many river, canal and for that
matter tidal currents should ensure that this at present little known and little thought of
power source gains increasing recognition, and will eventually be made use of. This is
particularly because many large rivers (such as the Nile, Euphrates, Zambesi, Indus),
flow through regions which are arid or which have several months of dry season.

286

4.10

BIOMASS AND COAL (THE NON-PETROLEUM FUELS)

4.10.1 General Description

i.

The availability and distribution of fuels

The developed world runs mainly on petroleum oils because, even at today's prices, oil
is a cheap fuel and is far more convenient to use, transport and store than any
alternative. But, the world's supply of easily recovered petroleum is fast diminishing and
we are running towards an era when supply will no longer exceed the demand. Already,
many developing countries can no longer afford to import sufficient oil for their present
needs, let alone to satisfy expanded future energy demand.
The main fuel alternatives to petroleum are solid fossil fuels such as coal, lignite and
peat plus the "biomass fuels", which are derived from recently living rather than
fossilized organic material. Estimates on world coal reserves vary, but there is general
agreement that the amount of coal available is in the order of ten times the total for oil in
energy terms, so although oil shortages due to physical depletion can be expected within
decades, coal should remain available for centuries. In fact it is likely that a need to
control atmospheric pollution will constrain the use of coal long before the reserves are
exhausted.
The main problem with biomass as a fuel is its uneven distribution, being most
available in forested regions where it is little needed. Considered as a global resource
there is no shortage of biomass [68], since:
a. the world's annual rate of use of energy is only about 10% of the annual natural
rate of photosynthetic energy storage (i.e. solar energy stored in plant matter);
b. stored biomass on the earth's surface at present is approximately equivalent to
the entire proven fossil fuel reserves (oil, coal and natural gas).
a subsistence diet. This issue of food versus fuel is an important one, but where the fuel
crop is used for irrigation pumping to produce food crops, especially in a small-scale "onfarm" process, the same objections need not apply.
ii.

The nature and calorific value of fuels

All fuels involve the combustion of carbon and usually hydrogen with atmospheric
oxygen to produce mainly carbon dioxide and water, plus heat. As far as heat production
is concerned, the relative merits of various fuels are best summarised in terms of the
calorific value of total heat released when
The main problem with all fuels, but particularly with biomass, is distribution. Even
energy-intense and efficiently marketed commercial fuels like petroleum are often in
short supply in many remote areas of developing countries, so it is not surprising that
more bulky and less valuable biomass fuels present even greater problems of
distribution. Fuelwood supplies in the Third World do not coincide at all well with
population distribution, so that there are areas of scarcity and areas of surplus. In most
cases the fuels in question are too bulky and low in commercial value to allow such a
simple solution as transporting them from areas of surplus to areas of scarcity.

287
The size of the areas suffering from a deficit of biomass fuels is increasing due to
population pressure, combined in some cases, such as the African Sahel region, with
apparent climatic changes. There is a narrow gap between a sustainable rate of
firewood foraging and the catastrophic rate of removal of vegetation which leads to
desertification. Lack of firewood leads to increasing use of dung as a fuel (in India it is
perhaps the primary fuel for cooking), and the use of dung for fuel rather than fertilizer
leads to a further decline in soil quality and contributes to the same problem. It has been
estimated [70] that almost 3 billion people will face cooking fuel shortages by the turn of
the century. Despite this grave situation, most of the R&D on biomass utilization has
focussed on its use as a petroleum substitute (eg. alcohol for powering cars in Brazil), to
address the "oil crisis" of the rich, rather than the "wood crisis" of the poor. Some of
these latter developments, involving the large scale production of biomass fuels, give
rise for concern as to the impact they might have on the capability of countries to feed
their populations. For example. Brown [71] indicates the land requirements in Brazil to
fuel a car compared with those to feed a person, are as indicated in Table 28.
Table 28 LAND REQUIREMENTS IN BRAZIL TO PRODUCE GRAIN FOR FOOD OR
FOR FUEL ALCOHOL
grain crop use
subsistence diet
affluent diet
medium sized car
large (US) car

quanitity
required
180kg/yr
700kg/yr
2 800kg/yr
6 600kg/yr

cultivated
land area
0.1 ha
0.4 ha
1.4 ha
3.2 ha

This shows that even a medium sized car doing only 12 000km per year needs a land
area to produce its alcohol fuel sufficient to feed 14 people on
Biomass, in the form of firewood, charcoal, agricultural residues, or dried animal dung is
already the main energy resource for the majority of the poorer half of humanity, over 2
000 million people. They currently use 15% or more of total world energy, entirely in the
form of biomass. Certain poorer countries depend on biomass fuels at present for over
90% of their energy needs (i.e. mainly for cooking fuel). It has been estimated [69] that
53% of energy use in Africa, 17% in Asia and 8% in Latin America is currently met from
biomass resources. Therefore biomass is already a huge and vital economic resource,
although it is usually informally exploited mainly for cooking purposes and still remains
rarely used for the production of shaft power via heat engines, and even less so for
irrigation pumping.
they are burnt; see Table 29 and 30, [6], This can be expressed in energy per unit
weight or per unit volume, but in the end, what matters most is the energy per unit cost.
Also of importance with biomass fuels if grown as a fuel crop is their productivity, which
obviously effects their cost-effectiveness. This must clearly depend on many factors
such as location, soil quality, etc. The most productive areas for photosynthetically
produced material are the forested regions; agricultural land is typically only half as
productive per hectare in terms of biomass generation. Table 31, after Earl [72], shows
some typical production rates, while Table 32 gives calorific values for a variety of fuels,
both fossil and biomass. Fossil fuels tend to be more consistent in their properties then
biomass. The calorific value of biomass fuels is particularly influenced by the moisture

288
content of the fuel. For example, air dried wood, which normally has a moisture content
of 25% will produce 50%-100% more heat than "green" freshly cut wood with a moisture
content of 50%. Oven drying of wood increases the heat yield further (but requires
expenditure of heat); fuel, once oven-dried, has to be kept in warm dry conditions before
being burnt or it will reabsorb atmospheric moisture. Therefore there is a good case to
be made for utilization of waste heat from any biomass engine system to help dry its own
fuel supply.
Table 29 PRINCIPAL CROP RESIDUES IN DEVELOPING COUNTRIES (WORLD
TOTALS) (after Leech, [6])
Type of residue

Energy value
(GJ/yr x 10)
1,060
790
185
110
110
35
35
20

Sugar cane bagasse

Rice hulls
Coconut husk
Cotton husk
Groundnut shells
Coffee husks
Oil palm husk
Oil palm fibre

Level of
utilization
high
low
low
high
high
low
high
high

Table 30 TYPICAL CEREAL CROP RESIDUES (after Leech, [6])

Crop
Rice
Wheat
Maize
Sorghum
Barley
Millet

Typical yield
(tonne/ha)

Residue
(tonne/ha)

2.5
1.5
1.7
1.0
2.0
0.6

5.0
2.7
4.25
2.5
3.6
2.0

Calorific value of residue

(GJ/ha)
90
49
76
45
65
36

(kWh/ha)
25,000
14,000
21,000
12,000
18,000
10,000

Table 31 PHOTOSYNTHETIC CARBON PRODUCTION RATES (after Earl, [72])

Type of land

A. Forest
Temperate deciduous
Conifer and mixed
Temperate rain forest
Tropical rain forest
Dry woodlands
Sub-total:

Net primary
production
(tonne/ha)

Total world annual

natural production
(giga-tonne)

10
6
12
15
2

8
9
1
15
3
36

289
B. Non-Forest
Agricultural
Grasslands
Tundra
Deserts
Sub-total:
Total:

4
3
1
1

6
8
1
3
18
54 Gt

Table 32 RELATIVE HEAT VALUE OF VARIOUS FUELS (APPROXIMATE VALUES)

Calorific value/unit wt.
(MJ/kg)
(BTU/lb)
A: Fossil fuels
Petrol/gasoline
Fuel oil
Paraffin/kerosine
Diesel/gas oil
Coal tar/asphalt
Anthracite coal
Bituminous coal
Lignite (brown) coal
Peat
Coke
Natural gas (methane)
Coal gas
Propane (cylinder gas)
Butane (cylinder gas)
B: Bio-mass fuels
Wood (oak)
Wood (pine)
Wood (acacia)
Charcoal
Sunflower stalks
Wheat straw
Beef cattle manure
Methanol (methyl alcohol)
Ethanol (ethyl alcohol)
Bio-gas (65% methane)
Wood gas (typical)
(producer gas)
Vegetable oil

Calorific value/unit vol.

(MJ/m3) (BTU/ft3)

44
44
45
46
40
35
33
30
20
28
56
9
48
47

19,000
19,000
19,500
20,000
17,000
15,000
14,000
13,000
9,000
12,000
24,000
4,000
21,000
20,000

32,000
39,000
36,000
38,000
40,800
56,000
42,900
37,500
18,200
22,400
40*
20*
90*
120*

860,000
1,050,000
970,000
1,020,000
1,100,000
1,500,000
1,150,000
1,010,000
490,000
600,000
1,020
490
2,400
3,100

18
20
16
28
20
18
14
20
28
20

8,000
9,000
7,000
13,000
9,000
8,000
6,000
8,600
12,000
8,600

14,400
10,000
11,000
11,000
10,000.

19,000
28,000
23*

390,000
270,000
300,000
300,000
270,000

500,000
700,000
600

39

16,500

5*
32,000

140
860,000

290
Since these fuels are normally gaseous, the calorific value per unit is relatively low
compared with liquid and solid fuels.

Table 33 POTENTIAL BIO-MASS VALUES OF SELECTED CROPS

Species

Location

Annual
dry matter yield
(ton/acre)

Sunflower
Forage sorghum
Hybrid corn
Water hyacinth
Sugar cane (average)
Sugar cane (experiment)
Sudangrass
Bamboo
Eucalyptus
Eucalyptus
Eucalyptus
American sycamore
Algae(pond)
Tropical rainforest
(typical)
Subtropical deciduous
forest (typical)

USSR
Puerto Rico
USA (Miss)
USA (Fla)
USA (Fla)
USA (Cal)
USA (Cal)
SE Asia
USA (Cal)
India
Ethiopia
USA (Ga)
USA (Ca)

Heat value

Tonne oil
equiv. per
ha./yr
(tonne/ha) (106 BTU/ (GJ/ha/yr)
acre/yr)

13.5
30.6
6
16
17
32
16
5
20
17
21
3.7
39
18

30
69
13
36
39
72
36
11
45
39
48
8
88
41

200
460
90
240
260
480
240
70
300
260
320
60
580
270

530
1,210
250
630
680
1,250
630
210
790
678
834
160
1,520
710

12
28
6
14
16
29
15
5
19
16
19
4
36
17

11

24

160

420

10

The production rate of cultivated biomass materials is even more variable than its
quality. Table 33 indicates examples of measured yields of potential fuel crops.
iii. The potential for using biomass for fuelling irrigation pumps

Biomass ought to be an ideal energy resource for irrigation, since the whole point of
irrigation is to produce more biomass, usually for food rather than for fuel. A simple
calculation confirms that it is possible, at least in theory, to grow more than enough
biomass to fuel an engine and produce an additional food crop, even without considering
using food-crop residues. For example, considering the irrigation system sized in Fig. 13,
where 3ha is irrigated with 8mm of water on average per day and a pumping head of
10m; assuming that irrigation is necessary on 200 days per year, Fig. 13 indicates an
average shaft power requirement of 13kWh/day or in this case 2 600kWh per year. If the
power system can produce shaft power from fuel at 10% efficiency (which is possible at
this power size) then the gross fuel requirement is for 26 000kWh or 94 GJ. It can be
seen from Table 30 that this requirement could often be met simply from cereal crop
residues for 3ha, but if a fuel crop was grown because, for example the residues were
needed for other purposes such as cooking fuel, then just 0.1 to 0.2ha of eucalyptus (for
example) would produce this fuel requirement. Therefore, the entire irrigation energy

291
demand could be met from a fuel crop occupying in this example less than one tenth of
the area to be cultivated. And 10m lift with 200 days per year irrigation is a more
demanding irrigation energy need than would apply in many cases. Therefore, the
reason biomass is not more often used for irrigation pumping may relate more to
ignorance of this option or perhaps to lack of opportunity than to any technical constraint,
or perhaps it is because the more knowledgable farmer who may know of this option is
so often the more prosperous one who can in any case at present afford the greater
convenience of using diesel power. The rest of this chapter sets out to review the
biomass fuelled options and their advantages and disadvantages.
iv.

Technical options for using biomass fuels for irrigation pumping

All biomass fuels are ultimately burnt so as to power an appropriate internal or external
combustion engine. There is, however, a plethora of options available for preparing,
processing or modifying raw biomass for more effective use as a fuel, as shown in Fig.
156. Generally these involve a trade-off between enhancing the properties of the
biomass as a fuel on the one hand, and extra cost combined with losses of some of the
original material. No particular route can be said to offer advantages over any other,
rather there are "horses for courses"; some are better than others in specific applications
and situations.

292

Fig. 156 Routes for processing biomass fuels

It can be seen from Fig. 156 that there are three primary categories of biomass raw
materials:
a. solid woody ligno-cellulose material;

293
b. wet vegetation, residues and wastes;
c. oil-bearing seeds and resins.
In most cases these need some kind of processing before use; at the very least drying.

4.10.2

The Use of Solid Fuels

As indicated in Fig. 156, solid fuels may be treated in a number of ways and either
burned as a solid to power an external combustion engine such as a steam engine, or
they may be pyrolised to yield either combustible gas (which may be used for an internal
combustion engine), or the volatiles may also be condensed to yield a limited quantity of
liquid fuels.
Because, as explained earlier, moisture content of the fuel has a profound effect on its
calorific value, it pays to use any waste heat from the system to pre-dry moist fuels.
i.

Direct combustion

Historically a huge range of solid fuel furnaces and boilers existed, but today only a few
manufacturers make them for small power systems, although medium sized furnaces
and steam plant for use in tropical agro-industrial process plants such as sugar
refineries, which are much larger than is appropriate for powering small scale irrigation
pumps, are readily available.
Perhaps the most appropriate and simple arrangement for small systems is for fuel to be
simply fed by hand into a furnace, containing a boiler to generate steam. Fig. 157 shows
as an example, the 2kW experimental Ricardo steam engine developed in the early
1950s. A similar type of unit of 1900 vintage is shown in Fig. 103, for firing a small
Stirling cycle hot air engine. Furnaces of this kind will typically burn 2-3kg/h of wood per
kilowatt of shaft output, with small steam engines.
There are some difficulties in designing a furnace which will handle any fuel; quite
different grate arrangements are needed to cope with particulate fuels such as sawdust
or rice hulls, as compared with large lumps such as logs or coal. Therefore it is important
to use equipment able to accept the proposed fuel. For example, there are furnaces
designed especially , to handle fuels like sawdust or rice hulls, which would clog up a
conventional grate arrangement; in one such type known as a "Kraft Furnace", the
furnace and storage hopper are combined so that the outer surface of the mass of
sawdust burns and the partially burned gases are drawn through a multitude of small
passages into a secondary combustion chamber, where combustion is completed.

294

Fig. 157 2kW Ricardo steam engine

295

Fig. 158 The three main types of gasifier

ii.

Gasification

The purpose of gasification is to convert some of the energy of an inconvenient solid fuel
into a more convenient gaseous fuel. The main advantage of gaseous fuels is that they
can generally be used with internal combustion engines and not just with rare steam or
Stirling engines; [73].
The first commercially successful i.e. engine powered by a gasifier was built by Lenoir in
France in 1860 and ran on coal, [74] and the technique was widely used at the beginning
of this century. It was again widely used during the 1939-45 World War, when some 700
000 gasifiers were in use for powering motor vehicles due to shortages of petroleum;
[75]. The subsequent cheap oil era led to a great decline in their use, but nevertheless a
number of manufacturers in various countries still make gasifier or producer gas units.
The process involves the heating of a solid, carbonaceaous fuel to drive off inflammable
volatiles and to produce carbon monoxide (CO) from a reaction between the carbon and
the carbon dioxide generated by primary combustion. Moisture in the fuel, plus
carbohydrates in the biomass also react with carbon to yield further carbon monoxide
plus free hydrogen, and some of the free hydrogen reacts with carbon to produce
methane. Any source of heat may be used to gasify biomass fuels, but usually the
heating process is by partial combustion; i.e the fuel is burnt to heat itself. The chemical
make up of producer gas is typically:

296
and its calorific value will be approximately 5MJ/m3. Because of the high proportion of
inert nitrogen and carbon dioxide, this is only one eighth of the energy per unit volume of
natural gas, such as methane. The calorific value can be enhanced by injecting steam
into the gasifier; this yields more hydrogen, but with small systems it is difficult to do this
in an optimal manner without actually extinguishing the primary combustion; sometimes
there is an advantage therefore if the raw fuel is slightly moist which achieves a similar
effect.
Although the calorific value of producer gas is low, the quantity of air required for
combustion is also low, so that the thermal value of a stoichiometric mixture, as is
required to be induced into an engine for optimum combustion, of producer gas and air
is better than might be expected, as indicated by the following relationships:
Fuel gas

thermal value

natural gas (methane)

Producer gas

(MJ/m3)
40
5

thermal value of
air-fuel mix
(MJ/m3)
3.5
2.5

A producer gas generator is usually a vertically mounted cylinder (see Fig. 158 ) which is
generally loaded with fuel from the top. The fuel falls under gravity to replace burnt and
gasified material in the lower fire zone. There are three main types of gasifier, as shown
in the figure:
a. up-draught, in which air is introduced at the bottom and the feedstock has the
products of gasification passing through it;
b. down-draught, in which the volatiles are drawn downwards through the hottest
part of the combustion zone; this has the effect of cracking any tars and complex
chemicals and thereby gasifying them more effectively than with an up-draught
unit;
c. cross-draught; these have a small intensely hot zone fed from an air nozzle or
tuyere.
The first of the above options is simplest, but it produces gas with a lot of carry-over of
tar and volatiles which can rapidly damage an internal combustion engine. Therefore
down-draught gasifiers are more commonly used for powering i.e. engines as their
output is easier to clean, particularly if burning raw biomass fuels containing a lot of
volatiles (tar-problems for up-draught units are reduced by using pre-pyrolized fuels like
charcoal or coke, but of course much of the original energy content of the feedstock is
lost in producing the charcoal or coke). Cross-draught units produce very intense heat in
a small area, which results in effective gasification of volatiles and tars, but there are
often problems with the nozzle burning out unless it is water cooled (which is a
complication); therefore they are less common for use with small engines.
Although down-draught units are preferable for use with i.e. engines, they are less
capable of drying the fuel (unlike an up-draught unit the hot gases do not pass through
the unburnt fuel), neither can they handle small or particulate fuels so well as these fall
through the grate and clog it.

297
A typical small producer gas irrigation pumping system is illustrated schematically in Fig.
159 (after [76]). Here a down-draught gasifier is used, with wet coke as a primary filter
and cotton waste as a secondary filter for the gas.
iii. Producer gas cleaning

Before producer gas can be used in an internal combustion engine, it needs to be

effectively cooled and cleaned of impurities including ash, unburnt fuel dust, tar and
acidic condensates as otherwise any significant carry-over of these materials will quite
rapidly destroy the engine. Obviously the more ash and tar in the original fuel, the more
of a problem there is in cleaning the output gas. Therefore fuels having an ash content
greater than 5-6% are not recommended for use in producer gas units for i.e. engines.
Also, the high performance gasifiers necessary to run small i.e. engines tend to be
sensitive to inconsistencies in the fuel, so that regularly sized, low ash fuels are best.
Charcoal provides one of the best fuels for gasifiers being almost pure carbon in itself,
but coconut shells and maize cobs are both relatively effective gasifier fuels.
The methods used for gas cleaning vary; water or air coolers are generally used to
reduce the gas temperature to near ambient conditions, and cyclones, spray scrubbers,
filters packed with a wet matrix of wood-wool, steel swarf, coir fibre and other materials
have been tried. Ineffective gas cleaning remains the "Achilles Heel" of small gasifier
systems, being a major cause of premature engine failure.

Fig. 159 Small producer gas irrigation pumping system (ref. Damour [76])

298
iv.

Engines for use with gasifiers

After cleaning, the producer gas is mixed with air metered in the appropriate quantity
and the resulting mixture can then be induced into the inlet manifold of most standard
i.e. engines. Spark ignition engines are capable of running exclusively on producer gas,
but diesel engines on the other hand will not fire when run purely on producer gas, and
need to be run with at least a small amount of diesel fuel so that the timed injection fires
the mixture at the appropriate moment. Therefore they can be run as pilot fuel engines in
which diesel is used to start up and continues to be used in quantities normally
necessary just for idling, with producer gas making up the main part of the fuel supply. In
practice it is possible to run diesels on about one third diesel fuel and two thirds
producer gas. The experimental unit illustrated in Fig. 159 [76] is claimed to have
actually achieved an average of 88% replacement of diesel, but this may be partly due to
unusually careful operation.
The low calorific value of producer gas compared with petroleum fuels generally leads to
a marked reduction in power output, often by as much as 30 to 50% below the rated
power using petroleum. An approximate idea of the fuel requirements using producer
gas compared with conventional diesel operation, is as follows:
Gasifier fuel required to produce 1kWh of shaft power
charcoal
1-1.3kg

hardwood chips
2-3kg

diesel fuel
0.3-0.5kg

Typical producer gas system costs, for units made in Europe or North America (and
excluding the engine) are in the region of US $2 500 for a unit capable of fuelling a small
engine of about 2kW rating up to around US $4 500 for a unit capable of sustaining a
10kW engine. Less sophisticated units, costing only a few hundred dollars, are
manufactured in the Philippines and Brazil. The Indian unit of Fig. 159 [76] has a gasifier
costing 8 000 rupees, which is approximately US $800, and is sized for a 3kW engine.
v.

Operation and maintenance of gasifiers

A draught is needed to get a gasifier going; this is usually created by starting the engine
on a petroleum fuel and then introducing a burning rag or other source to the gasifier
fire-zone. Care is needed to ensure that there is no residual gas from the last time the
system was run which could explode. The producer gas unit needs to be refuelled before
the fuel in the hopper reaches a level of less than about 300mm above the fire zone, or
gas production may not be reliable.
Care is needed with gasifiers, since; firstly producer gas is extremely toxic due to the
carbon monoxide present and therefore a unit must never be used in enclosed
conditions where producer gas could build up; secondly there is a significant risk of
explosion and/or fire when opening the unit to refuel it. Opening the gasifier to refuel
often causes a small blow-back explosion, but the experienced operator can open the
hopper and refuel safely.

299
The gasifier, and gas cleaning system, must be regularly cleaned out and any leaks
must be repaired immediately. Experience with automotive gasifiers during the war
suggested that as much as one hour per day on average is needed to clean and prepare
a gasifier for operation.

4.10.3 The Use of Liquid Biomass Fuels

There are two main categories of liquid fuels which are relevant to powring small
engines for irrigation pumping; alcohols and oils, plus a third (latex or sap) which may
come into use in the future.
i.

Alcohols

There are two varieties of alcohol that can be used to run engines; ethanol (ethyl
alcohol) and methanol (methyl alcohol). The former is the only type which at present has
any prospect to be produced economically from biomass feedstocks (the latter, although
once known as "wood alcohol" as this was originally the main route to its production, is
usually produced by an industrial process at high pressures and temperatures from
natural gas as the process for distilling it from wood is inefficient and unproductive).
Ethanol, which is the type of alcohol found in wines and other drinks, can be produced
by the bacteriological fermentation of natural sugars or other carbohydrates such as
starches, either from purposely grown fuel crops or from wastes and residues. Starches
first require hydrolyzation, usually with acids, to change them into fermentable sugars.
Current activities to produce fuel alcohol focus on the use of sugar cane, maize and
cassava [77], processed on a large scale. There is as yet no technically and
economically viable small-scale process for fuel-alcohol production, so the use of alcohol
by farmers must depend on any national programmes in their countries. They may also
be in a position to grow the fuel crops for the programme, so such programmes could be
more important for farmers than simply providing an alternative fuel.
There are a number of problems inherent in the large-scale production of fuel alcohol.
Obviously the food versus fuel argument, as outlined earlier, is important; secondly,
large-scale ethyl alcohol production produces large volumes of "distillery slops", which
pose a serious disposal and pollution problem. Finally, the product is only marginally
economic at present compared with gasoline and the main justification must generally be
import substitution rather than cheaper fuel. Nevertheless a number of alcohol fuel
programmes have been initiated. By far the largest is in Brazil, where the goal is to
produce 12 billion litres of ethanol per annum by 1985, mainly from sugar cane. The
USA also has a major ethanol production programme, based largely on using up maize
surpluses, which has the goal of an overall 10% substitution for gasoline by the 1990s. A
number of other countries have initiated power alcohol activities, including Thailand, the
Philippines, New Zealand, Australia, Kenya, Zambia, Zimbabwe, Nicaragua, Paraguay,
and Fiji [77].
R&D is in hand to develop alcohol production processes that could work effectively and
economically using woody waste materials and residues, which if successful might
create a much more promising future for biomass-based alcohols.

300
ii.

Oils from biomass

There are two types of vegetable oil that show promise as fuel for internal combustion
engines, these are expressed oils from seeds and the saps or latex from succulent
plants and various trees such as the rubber tree.
Some successes have been reported with running diesel engines on vegetable oils.
Tests have been run on seed oils from peanut, rape, soybean, sunflower, coconut,
safflower and linseed [78]. Sunflower oil, in particular, shows promise as a fuel for diesel
engines. The main problems relate to the much higher viscosity of vegetable oils
compared with diesel gasoil; this makes it difficult to start a diesel on vegetable oil, but
once warm it will run well on it. Tests have shown that the performance is little effected,
but fuel consumption on vegetable oil is slightly higher due to its lower calorific value. A
major problem with unmodified vegetable oils has been a tendency for engines to coke
up rapidly, leading to reduced power and eventually engine seizure if no corrective
action is taken.
Chemical treatment of vegetable oils to turn them into an ethyl or methyl ester has been
found to overcome most of these problems and to actually give a better engine
performance than with diesel oil, combined with less coking than with diesel [78]. Also,
blends of sunflower oil with diesel fuel appear to reduce or eliminate some of the
problems experienced with pure sunflower oil.
Large scale processing of vegetable oil can crack the oil in much the same way as crude
oil, to produce veg-gasoline as well as veg-diesel. During the Second World War, China
developed an industrial batch cracking process for producing motor fuels from vegetable
oils, mostly tung oil [79]. The China Vegetable Oil Corporation of Shanghai was able to
produce 0.6 tonne of veg-diesel, 250 litres of veg-gasoline and 180 litres of vegkerosene per tonne of crude vegetable oil.
It is possible to extract vegetable oil "on-farm" on a small scale and to consider using
this to reduce diesel fuel requirements, although obviously any vegetable oil needs to be
extremely well filtered before it can be used in an engine. A more likely approach would
be production on a small-industry basis, in which the extraction unit procured seed from
a district for oil production on a more economic scale. Typical seed yields (sunflower)
are 700-1800kg/ha. It is possible to express between 0.30-0.43 litres of oil per kg of
seed, depending on the technique. Small-scale presses will produce the lower level of
yield while large screw presses and solvent extraction are needed to achieve the upper
level. This implies that from 210-770 litre/ha can be produced. The development of more
efficient on-farm oil seed expellers could make this a potentially viable process in many
areas. In fact a combination of efficient cultivation and efficient oil extraction could yield
in excess of 1 tonne/ha of vegetable oil. It has been argued that if a mechanized farmer
used 10% of his land for sunflower cultivation, he could become energy self-sufficient
[79] (although it is not explained by the reference how he overcomes the need to blend
his oil with diesel to avoid gumming and coking of the engine).
The energy ratio for the production of vegetable oils as fuel is much more favourable
than for alcohol production, and the process is simpler and less capital intensive.

301
Therefore, the use of oil-seed as a feed-stock to produce diesel fuel certainly looks
technically feasible. However the economics remain more doubtful with present diesel
fuel prices, since the value of refined vegetable oils on the international market is
generally 50-100% above that of diesel fuel, although this price differential has not
always applied and may not actually reflect the true cost (as opposed to the price) of
vegetable oils.
iii.

Latex bearing plants

Some recent investigations have shown a potential diesel fuel can be obtained from the
combustible rubbery sap or latex of various trees and succulent plants. Some of these
plants actually produce hydrocarbons, similar to but molecularly more complex than
petroleum oils. Some promising plants for this purpose are the Euphorbia species [80].
These grow well in semi-arid areas on marginal and barren lands which generally will not
support food crops. Professor Calvin of the University of California who has studied this
possibility projects a yield of 10 barrels of oil per acre with existing wild species and that
this yield could perhaps be doubled through seed selection and genetic improvements to
produce a plant developed for oil production [81]. The oil or emulsion which is tapped off
has too high a viscosity for immediate use as a diesel fuel, and contains gums and other
complex chemicals which would coke the engine prematurely, so it needs to be refined.
Although at present this source of fuel remains unproven, it shows considerable promise
for the future, particularly if it can be cultivated on lands of little use for food or other
crops.

4.10.4

Gas from Biomass: Biogas

At present biogas is the most immediately practicable means for powering a

conventional internal combustion engine from biomass. It lends itself to small-scale onfarm use and there is considerable experience with this technique in a number of
countries.
Biogas is produced naturally by a process known as anaerobic digestion, the action of
bacteria on water-logged organic materials in the absence of air. Biogas occurs
naturally as "marsh gas", an inflammable gas which bubbles out of stagnant marshes or
bogs. The same process occurs in the digestive system of cattle.
Biogas consists of about 60% methane, a non-toxic and effective fuel gas similar to
many forms of natural gas; the remaining 40% is mainly inert carbon dioxide with traces
of hydrogen, hydrogen sulphide, etc. Raw biogas has a calorific value of about 23 MJ/m3,
which is considerably better than producer gas (see Table 32). The carbon dioxide can
be removed by bubbling raw biogas through slaked lime (calcium hydroxide) but this
process requires regular replacement of the lime. After this treatment biogas
approximates to pure methane with a calorific value of about 40MJ/m3.
Biogas is an attractive fuel for use in i.e. engines since it has no difficult pollutants that
can damage them (like producer gas does). Moreover, biogas has good anti-knock
properties and can safely be used with high compression ratio spark ignition engines as
the sole fuel. When used with diesel engines it needs to be used as a supplementary
fuel because a small quantity of diesel fuel needs to be injected to fire the mixture (the
injection pump determines the ignition timing). Biogas can be used to reduce diesel fuel

302
requirements by from 50 to 80% with minimal modification to the engine being
necessary. To make the best use of biogas requires a spark ignition engine with a
compression ratio approaching that of a diesel. Some special biogas engines have been
built, which run on 100% biogas more efficiently than with an unconverted gasoline
engine [82].
An important further advantage of this process, especially in the context of irrigation
pumping, is that the digested sludge makes a good fertilizer, so that unlike the situation
where when biomass is totally burnt, it is possible to return much of the original material
to the land and thereby improve the soil quality and displace the use of chemical
fertilizers. The anaerobic digestion process makes the nitrogen and various other
chemicals more accessible to plant growth than the normal aerobic (in air) composting
process. Also, unlike artificial fertilizers, the sludge left over from the biogas process
contains humus which can improve the soil structure. This process also is useful as a
method for treating sewage or disposing of other unpleasant or potentially dangerous
organic wastes as well as for producing fertilizer and fuel gas. Anaerobic digestion is a
standard sewage treatment process which kills most water-borne pathogens harmful to
people and converts the effluent to a relatively innocuous and odourless liquid which can
easily be sprayed or poured onto the fields.
Anaerobic digestion is quite widely used for large scale city sewage plants, but it is also
increasingly being applied on farms. The first reasonably widespread farm use was in
France during the Second World War, when farmers built concrete digesters to produce
methane to replace petroleum fuels which were unobtainable for them at that time. More
recently efforts have been made to popularise the use of biogas in Asia, mainly in China,
but also in India, Nepal and some of the SE Asian countries. Commercial farm biogas
units have also gone into production in various countries, including the USA, UK,
Australia and Kenya as well as the main users of the technology, China and India.
Although the widespread use of biogas only started in China in the early 1970s, within
ten years some seven million biogas units had been installed [83], with the majority
being in Sichuan Province. The technology has been less successful at spreading in
India, although some 80 000 digesters are believed to be in regular use there.
Experience in India has been that the larger biogas units used by richer farmers and by
institutions have generally been more successful than smaller "family-sized" units.
Figs. 160 and 161 illustrate the two main types of small-scale biogas digester, developed
originally in China and in India respectively. The Chinese type of digester consists of a
concrete-lined pit with a concrete dome, entirely below ground. It is completely filled with
slurry, and once gas begins to form, it collects under the dome and forces the level of
the slurry down by up to about lm. The gas pressure is consequently variable depending
on the volume of gas stored, but by using a simple manometer on the gas line it is
possible to measure the gas pressure and thereby gain an accurate indication of the
amount of gas available. The Indian type of digester (Fig. 161) is more expensive to
construct because it has a steel gas holder, on the other hand it is less likely to leak than
the Chinese design which requires high quality internal plastering to avoid porosity and
hence gas leaks. With the Indian design, gas collects under the steel gas holder, which
rises as it fills with gas. The height of the gas holder out of the pit indicates how much
gas is available, and the pressure is constant.

303

Fig. 160 Fixed dome biogas digester (China)

Fig. 161 Biogas digester with floating gas holder and no water seal (India)
i.

The biogas production process

The biogas process requires an input material provided as a liquid slurry with around 510% solids. It is important to use materials which breakdown readily; highly fibrous
materials like wood and straw are not easily diegested by the bacteria, but softer

304
feedstocks like dung and leaves react well to the process. Also some feedstocks are
more productive than others as indicated by Table 34, and some producers of feedstock
are more productive than others as indicated by Table 35.
Table 36 (after Meynell [84]), gives the principal operating parameters of typical
continuous biogas digesters (it is also possible to run batch digesters in which each
digester is loaded, completes its cycle and is then unloaded, but this needs several
units, probably three, to ensure gas is always available). For optimum performance the
internal temperature of the digester needs to be in the mid-30s centigrade and certainly
over 25C, moreover temperature conditions need to be as steady as possible. The
digestion process generates a small amount of heat, but in cooler climates or seasons
the unit needs to be well insulated and may need heating when cold spells occur. The
average retention time for solids for the complete process is normally 20 to 40 days.
With continuously operating (as opposed to batch) digesters, the actual digester size has
to be equal to the design retention time in days multiplied by the daily input rate; i.e. with
a 30 day retention time and lm3/day of input, the digester volume needs to be 30m3. The
longer the retention time and the warmer the digester, the more complete the process
and the more energy per kg of volatile solids is obtained, however the larger and more
expensive the digester needs to be. Hence the sizing and retention time are usually a
compromise between keeping costs reasonable and obtaining complete digestion. The
loading rate and the moisture content are related; . the slurry needs to be kept to around
85 to 95% liquid. Too thin a slurry takes up more volume and needs a bigger digester
than necessary, while too thick a slurry limits mixing and tends to solidify and clog up the
unit. Another important criterion is the carbon/nitrogen ratio; for efficient digestion the
process requires between 20 and 30 parts of carbon to be present per part of nitrogen.
Certain carbon-rich materials like leaves or grass benefit, therefore, from being mixed
with nitrogen-rich substances such as urine or poultry droppings. Alternatively ammonia
or other nitrogen rich artificial chemicals may be added to a digester running on mainly
vegetation to obtain a batter ratio and help the process. Finally, the output to be
expected will be in the order of 0.1-0.7m3 per kg of volatile solids input per day.
Table 34 BIOGAS YIELD FROM VARIOUS FEEDSTOCKS
Feedstock

Sewage sludge
Pig dung
Cattle dung
Poultry droppings
Poultry droppings and paper pulp
Grass

Gas yield per unit mass of

feedstock
(m3/kg)
0.3-0.7
0.4-0.5
0.1-0.3
0.3-0.5
0.4-0.5
0.4-0.6

energy yield
(MJ/kg)
6-17
8-11
2- 6
6-11
8-11
8-14

Table 35 QUANTITIES OF EXCRETA FROM VARIOUS SPECIES

Source of waste

Volatile solids yield per Biogas yield Energy

animal
(m3/day)
yield
(kg/day)
(MJ/day)

305
Humans (inc. cooking wastes)
Pigs
Cattle
Poultry (x100 birds)

0.1
0.6
4.0
2.2

.03-.07
.24-.30
.40-1.2
.07-1.1

.6-1.7
4.8-6.6
8-24
13-24

Table 36 PRINCIPAL OPERATING PARAMETERS FOR FARM BIOGAS DIGESTERS

Operating temperature
Retention time
Loading rate (volatile solids)
Operating moisture content
Specific gas production
Feedstock carbon/nitrogen ratio

30-35C
20-40 days
2-3 kg/m3 per day
85-95%
0.1-0.7 m3/kg per day
20-30

Because biogas digesters have the capability of storing at least a 12 hour supply of gas,
an engine can be used that draws gas at quite a high rate. In fact the size of engine is
not critical since it is only the number of hours it will run that are governed by the
digester gas capacity. Transporting biogas is technically difficult. In China it is quite often
piped several hundred metres through plastic tubes. Unlike propane or butane, it is not
possible to compress biogas into a liquid at normal temperatures and the only way to
transport it as a gas are either in high pressure cylinders, which of course require a high
pressure compressor to charge them, or in a plastic bag. Fig. 162 shows how small twowheel tractors in China are powered from a bag of biogas carried on an overhead rack.
The unit in the figure is towing a trailer tank full of biogas digester sludge and it also
carries a pump driven off the engine for pumping the sludge onto the field via a spraying
nozzle. An interesting option for irrigation by biogas power is to combine the digester
sludge and the irrigation water in order to perform three functions simultaneously; i.e.
irrigation, the application of fertilizer and waste disposal.

306

Fig. 162 Chinese two-wheel tractor running on biogas and being used to pump digester
slurry on to the field
ii.

Sizing example

Biogas typically has a calorific value as a fuel for running small engines of about
6.4kWh/m3, so it is quite straightforward to estimate the daily volume of biogas needed to
perform a given pumping requirement. A worked example of how to do this is given in
Table 37, which indicates how a 3ha small-holding could be irrigated using biogas
generated from the wastes from 20-30 pigs, 5-10 cattle, 500-700 poultry or a community
of 80-200 people. The production rate of biogas can be enhanced by mixing vegetation
with the animal wastes, although extra nitrogen, which could be in the form of urine, may
need to be introduced to balance the excess carbon present in the vegetable wastes.
Table 37 SIZING EXAMPLE TO RUN AN IRRIGATION PUMP ON BIOGAS
Requirement: 8mm of water per day pumped through a head of 6m (i.e. 240m3/day)
Engine;
s.i. engine assumed 10% efficient (fuel to hydraulic power
Biogas:
calorific value assumed at 6.4kWh/m3
The energy requirement to lift 240m3/day of water through 6m is:

Assuming a system efficiency of 10%, the fuel energy requirement is:

307

Hence the daily biogas requirement will be:

This requires a biogas digester of about 5-10m3 capacity, loaded with from 10-60kg/day of input
material (volatile solids), which can in turn be obtained from, for example:
20-30 pigs
5-10 cattle
500-700 poultry
80-200 people
The addition of vegetable wastes, providing it did not unduly upset the carbon/nitrogen ratio, could
allow the same volume to be produced from possibly two thirds to three quarters the number of
livestock or people.
The example therefore shows that this process needs significant inputs of waste material
to yield even quite modest amounts of pumped water. Therefore, looked at just in energy
terms, the economics tend to be at best marginal in comparison with petroleum fuels,
however when the fertilizer value plus the waste disposal benefits are factored in, the
process frequently comes out as being economically worthwhile.
It is difficult to generalize on the economics of biogas since - many factors that are
locality-specific are involved. However there is no doubt that the process offers
significant economies of scale. For example a survey [85] of biogas units in India found a
payback period, using a 10% discount rate, of 23 years for a 1.7m3 (60 cu.ft.) plant which
improved to 7, 4 and 3 years respectively with 2.8, 5.7 and 8.6 m3 units (100, 200 and
300cu.ft). The sizes of plants needed to run small engines are much bigger than this and
are therefore likely to be more cost-effective.
The country which has made by far the most use of biogas production in agriculture is
China, where, particularly in Sichuan province, there are several million working biogas
units. This development took place almost entirely within the last 10-15 years. Various
studies (eg. [86]) have indicated that the value of the fertilizer output usually surpasses
the value of the energy produced by the process in China. The waste disposal and
sanitation aspects of the process are also important justifications for its use.

308

5. THE CHOICE OF PUMPING SYSTEMS

5.1
5.1.1

FINANCIAL AND ECONOMIC CONSIDERATIONS

Criteria for Cost Comparison

The ultimate criterion for choosing an irrigation pumping system is to obtain the most
"cost-effective" system; this does not necessarily mean the "cheapest" system, since low
first cost often results in high running costs. To arrive at a realistic assessment of the
true cost-effectiveness is not easy, particularly as many of the parameters required for
such an analysis will often be uncertain or variable, and many "costs" and "benefits" do
not readily lend themselves to financial quantification at all (eg. reliability, availability of
spare parts or maintenance skills, ease of use and vulnerability to theft are some factors
which have cost implications, but which are difficult to factor in to any cost analysis).
Nevertheless, the objective of most methods of financial or economic analysis is to arrive
at a figure for the true "life-cycle costs" of a system (i.e. the total costs of everything
relating to a system over its entire useful life) which are often compared with the "lifecycle benefits", which are the total benefits generated by the system in its lifetime. Since
the length of life to be expected for different options will vary, it is necessary to find a
technique for reducing the life cycle costs and benefits to those over, say, one year or to
find some other method of presentation which allows ready comparison between
options. Some methods for achieving this are discussed in more detail later.
While a financial or economic appraisal of options generally represents the primary
criterion for selection, this should not necessarily be used as the sole method for
ranking. Obviously any clearly uneconomic type of system is to be rejected, but,
generally, there will be a number of possible options which in purely economic or
financial terms show similar costs. The final selection from such a "short-list" generally
needs to be based on technical or operational considerations. It is a mistake to use
economic or financial analysis as the sole arbiter for choosing between options.
It is important to distinguish between economic and financial assessment of
technologies. An economic assessment seeks to look at "absolute" costs and benefits
and therefore considers costs and benefits as they would be if unaffected by taxes,
subsidies or other local influences; the object is to arrive at a valuation of the technology
in pure terms, excluding any local financial conditions. The value of an economic
assessment is more for policy makers and those who need to compare technologies.
The farmer, on the other hand, should do a financial assessment which takes account of
conditions within his local economy, such as subsidies and taxes and the local market
price of the final product harvested as a result of irrigation. The economist and the
farmer may therefore come to different conclusions as to what is "cost-effective".
There are two main cost factors relating to any system; its first cost (or capital cost) and
its recurrent costs or operation and maintenance costs (O&M costs). So far as the first
cost is concerned, the over-riding consideration from the individual farmer's point of view
is whether it is affordable as a cash payment and if not, whether he can obtain finance
on acceptable terms. The institutional user will no doubt have access to finance and will
be more concerned to ensure that an adequate return on the investment will be
obtained. The O&M costs can vary considerably both within and between technologies

309
and will no doubt increase with time due to inflation. Generally speaking, all options for
pumping water represent a trade-off between capital costs and recurrent costs; low first
cost systems usually have high recurrent costs and vice-versa.
Clearly, an investment in extra costs at the procurement stage may save recurrent costs
throughout the life of the system, but it is difficult to compare a low cost engine having
high running costs with a high cost solar pump having virtually zero running costs.
Therefore, a major problem which may often inhibit the choice of high capital cost
systems is the difficulty involved in comparing the cost of such an investment with the
savings it could produce for many years into the future.

5.1.2

Calculation of Costs and Benefits

In all cases it becomes necessary to use some method to compare the notional value of
money in the future with its value today, so that such things as the trade-offs between
spending more on the first cost in order to reduce the recurrent costs can be properly
assessed. All the methods used rely on what is called Discounted Cash Flow (DCF). The
general principle behind this is that money available in the future is worth less than if it
were available now; this is not necessarily because of inflation, but because it is
assumed that money available now can be invested and will therefore be worth more in
the future than the original sum. For example, investing $100 now at 10% will yield $110
in one year's time; therefore $100 today is notionally worth 10% more than the same
sum if made available in one year's time.
Discounted Cash Flow therefore takes all future payments and receipts relating to a
predictable cash flow, and discounts them to their present value using an appropriate
interest rate, or discount rate. Going back to the previous example, $110 received in one
year's time is said, therefore, to have a "present value" (or PV) of $100 if discounted to
the present at a 10% discount rate.
When capital costs are annualized, then first costs can be properly equated with
recurrent costs as far into the future as is desired. The way this is done is to note that it
is unimportant to an investor whether he has, say, $100 today which he plans to leave in
a deposit account yielding 10% interest, or whether he simply has $110 promised to him
in one year's time. It follows from this that, for example, a payment of "C " in different
numbers of years' time will be discounted back to the present, assuming a discount rate
of "d" in the following way:

Each of these factors gives the PV of a payment Cr in l,2,3,...,n years time.

Table 38 gives calculated values for these factors for discount rates of 2, 5, 10, 15 and
20% for all years from 1 to 25. To find the Present Value of, say $1 000 to be paid in 10
years time at a discount rate of 10%, the relevant PV factor is looked up from the table
(0.39 in the example) and multiplied by the sum of money in question to give, in this

310
example 0.39 x $1,000 = $390. If it can be anticipated that $1 000 needs to be paid, say,
every 5 years (for example to replace an engine), then the PV factors for years 0, 4, 9,
14, 19 and 24 at the relevant discount rate are looked up, added together and multiplied
by the sum in question; in this case the calculation for a 10% discount rate and 25 year
period would be:

A complete cash flow over some predetermined future period can be planned and when
multiplied by the relevant PV factors for each year can be reduced to a complete Present
Value.
Another effect to be considered is that of inflation. The real purchasing power of money
tends to decline, so that for example a litre of diesel fuel in one year's time may be
anticipated to cost say 5% more than it does today. So if we wish to compare the cost
over an extended period, we need to allow for the likely increase in cost of fuel.
Therefore if we expect to spend $100 this year, we must plan to spend $105 next year
and so on. This results in a need to increase the actual amount of currency required to
achieve a given real value. This can be anticipated by assuming an inflation rate "i", and
as a result the progression of PV values discounted back to the present becomes:

Whether or not inflation is included, the actual life cycle cost of a system will be the sum
of the present values of all its initial costs, plus all its future costs; i.e. the life cycle cost
is obtained by taking the capital cost and adding on the anticipated discounted present
values of the future recurrent costs.
For most purposes of comparison it is easiest and quite sufficient to ignore inflation and
work entirely in "present day money", especially as future inflation is not readily
predictable over long periods of time. Therefore Table 38 uses zero inflation rate and no
attempt has been made to provide a similar set of tables to cater for various rates of
inflation.
A regular payment or benefit "Ca" which recurs every year can be simultaneously inflated
at a rate "i" and discounted at "d". To avoid calculating each year separately as above
and having to add them together, the series can be summed by a general formula for a
whole string of regular payments, as follows:
Table 38 PRESENT VALUE FACTORS UP TO 25 YEARS
Year
0
1

0.02
1.00
0.98

0.05
1.00
0.95

Discount Rate
0.10
1.00
0.91

0.15
1.00
0.87

0.20
1.00
0.83

311
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

0.96
0.94
0.92
0.91
0.89
0.87
0.85
0.84
0.82
0.80
0.79
0.77
0.76
0.74
0.73
0.71
0.70
0.69
0.67
0.66
0.65
0.63
0.62

0.91
0.86
0.82
0.78
0.75
0.71
0.68
0.64
0.61
0.58
0.56
0.53
0.51
0.48
0.46
0.44
0.42
0.40
0.38
0.36
0.34
0.33
0.31

0.83
0.75
0.68
0.62
0.56
0.51
0.47
0.42
0.39
0.35
0.32
0.29
0.26
0.24
0.22
0.20
0.18
0.16
0.15
0.14
0.12
0.11
0.10

0.76
0.66
0.57
0.50
0.43
0.38
0.33
0.28
0.25
0.21
0.19
0.16
0.14
0.12
0.11
0.09
0.08
0.07
0.06
0.05
0.05
0.04
0.03

0.69
0.58
0.48
0.40
0.33
0.28
0.23
0.19
0.16
0.13
0.11
0.09
0.08
0.06
0.05
0.05
0.04
0.03
0.03
0.02
0.02
0.02
0.01

0.15
1.00
1.15
0.62
0.44
0.35
0.30
0.26
0.24
0.22
0.21
0.20
0.19
0.18
0.18

0.20
1.00
1.20
0.65
0.47
0.39
0.33
0.30
0.28
0.26
0.25
0.24
0.23
0.23
0.22

Table 39 ANNUALIZATION FACTORS UP TO 25 YEARS

Year
0
1
2
3
4
5
6
7
8
9
10
11
12
13

0.02
1.00
1.02
0.52
0.35
0.26
0.21
0.18
0.15
0.14
0.12
0.11
0.10
0.09
0.09

0.05
1.00
1.05
0.54
0.37
0.28
0.23
0.20
0.17
0.15
0.14
0.13
0.12
0.11
0.11

Discount Rate
0.10
1.00
1.10
0.58
0.40
0.32
0.26
0.23
0.21
0.19
0.17
0.16
0.15
0.15
0.14

312
14
15
16
17
18
19
20
21
22
24
24

0.08
0.08
0.07
0.07
0.07
0.06
0.06
0.06
0.06
0.05
0.05

0.10
0.10
0.09
0.09
0.09
0.08
0.08
0.08
0.08
0.07
0.07

0.14
0.13
0.13
0.12
0.12
0.12
0.12
0.12
0.11
0.11
0.11

0.17
0.17
0.17
0.17
0.16
0.16
0.16
0.16
0.16
0.16
0.16

0.22
0.21
0.21
0.21
0.21
0.21
0.21
0.20
0.20
0.20
0.20

Having produced a Present Value for some cash flow stretching into the future, it is then
important to be able to reduce this to an annual equivalent sum in present day money; in
other words, how much money do you have to save each year to finance the cash flow?
The annual equivalent "Cr" of a capital sum "C" is similar to paying an annuity and can
be calculated as follows:

Table 39 gives calculated annualization factors for various discount rates and periods;
for example, the annualized payment due to pay off a sum of $1 000 (in present day
value money) over ten years at a discount rate of 10% can be calculated by taking the
relevant factor (0.16 in this case) and multiplying it by the capital sum to arrive at $160/yr
for this example. Therefore is you borrow $1 000 today under those financial conditions,
the repayments, or the cost of financing the loan will total $160 per year. Even if you pay
$1 000 cash for some item with a 10 year estimated life, the "opportunity cost" resulting
from not having the cash available to earn interest in a bank will still be $160 in the case
considered, so it is correct to say that it costs $160 per $1 000 tied up over 10 years at a
10% discount rate. Table 39 can readily be used to calculate annualized values of other
capital sums over a range of periods up to 25 years for the same discount rates as for
the PV factors of Table 38.

313
Table 39 can also be used in reverse to calculate the Present Value of a regular cash
flow; using the same example, the Present Value of $160/yr over a 10 year period at a
10% disount rate is exactly $1 000. The method to use Table 39 to obtain the Present
Value of a cash flow is to look up the factor relating to the period and discount rate
required, and to divide the annual payment by the factor. For example, the Present
Value of $80/yr over a period of 25 years at a discount rate of 15% is the factor 0.03
divided into $80, i.e.:

This is a lot easier than taking the PV factor for every year from year 0 to 24 from Table
38, adding them and multiplying by 80, which should give the same result.
Having explained the methods of equating future costs and payments to the present, it
remains to use these techniques to arrive at a means for comparing the relative
economic merits of different systems. There are in fact four commonly used techniques
for making economic appraisals:

a. Life-cycle-costs and Unit-output costs

The total present value of all costs for various systems sized to do a certain job can be
compared (i.e. it is assumed that similar benefits will be achieved regardless of the
choice of technology, so the problem is simplified to finding the least-cost solution). This
can be confusing if different options have significantly different life expectations (in
years), so the lifecycle costs are usually annualized as described above to give a more
general means for comparison which includes an allowance for the expectation of life of
each system. The annualized life cycle costs can then readily be converted to unit output
costs; i.e. the cost of unit quantity of water through a given head, as described above.
Where different heads may be involved, the "unit hydraulic output costs" can also be
determined by taking the cost of the head-flow product, but care is needed to ensure that
you are still comparing like with like as some systems' costs will change significantly as a
function of head. Life-cycle costs are purely a criterion for comparison as they do not
indicate whether a specific water pumping system is actually economically viable (for
example whether the value of additional crops gained from pumping irrigation water may
actually exceed the cost of pumping the water).

b. Net Present Value (NPV) or Net Present Worth

This is be obtained by determining cash flows for the benefits expected as well as for the
costs, using exactly the same methods as described above. The benefits will generally
be the marginal income gained from irrigation and will be given a positive value and the
costs, calculated as before will be made negative. To obtain the Net Present Value the
present values of the total costs and benefits cash flow are added together. To be
worthwhile a positive Net Present Value is required, and the more positive the better; i.e.
the summed present values of the benefits must exceed those of the costs. The answer
obtained will differ depending on the choice of discount rate, and the period considered
for analysis, so this is not therefore purely dependent on choice of technology.

c. Benefit/Cost Ratio

314
A variation on the concept of Net Present Value is to calculate the total life-cycle benefits
and the total life-cycle costs and then to divide the former by the latter to obtain the
Benefit/Cost Ratio. If this ratio is greater than one, then the benefits exceed the costs
and the option is worthwhile. The same criticisms apply to this as to the Net Present
Value approach.

d. Internal Rate of Return

Internal Rate of Return is difficult to calculate, but provides a criterion for comparison
independent of any assumptions on discount rates or inflation. It is therefore a purer
method for comparing technologies. The Internal Rate of Return can be defined as the
discount rate which will give a Net Present Value of Zero (or a Benefit/Cost ratio of 1);
i.e. it is the discount rate which exactly makes the benefits equal the costs. To calculate
Internal Rate of Return requires finding the discount rate to achieve a Net Present Value
of zero; this is usually determined by trial and error, by recalculating the Net Present
Value for different discount rates until the correct result is achieved. It is tedious to do
this manually, but various standard micro-computer spread sheets are widely available
which make it a relatively easy task. The advantage of the Internal Rate of Return as a
selection criterion is that it is, in effect, the discount rate at which an option just breakseven. If the Internal Rate of Return is higher than the actual discount rate, then the
option can be said to be economically worthwhile. Obviously, the higher the Internal
Rate of Return of an option, the more attractive it is as an investment, since it basically
says whether you do better to leave your money in the bank or invest it in an irrigation
pumping system, or whatever.
The choice of discount rate used for analysis effectively reflects the analyst's view of the
future value of money; a high discount rate implies that money available now is much
more useful than money available in the future, while a low discount rate is more
appropriate when longer term considerations lie behind an investment decision. The
implications of this are that low discount rates favour the use of high-first-cost lowrecurrent-cost systems, which cost a lot now in order to save money in the future, while
high discount rates make low-first-cost high-recurrent-cost systems look good be making
the high future costs less important.
The length of period selected for analysis (n) can affect the answer, so it is normal to use
a long period in order to minimize this effect; this will normally need to be 15 to 20 years
or more because at normal discount rates, costs more than about 20 years into the
future become discounted to such small levels that they cease to affect the results very
much. Obviously, short-life equipment will need replacement during the period under
analysis, and this is taken account of by adding the discounted value of the capital costs
of future replacements to the life cycle costs.
i.

Shadow pricing

Economists recognize that there are so-called "opportunity costs" associated with cash
transactions; for example, although for most analytical purposes the exchange rate of a
local currency will be taken at the official rate, in practice this does not always reflect its
real value in terms of purchasing power. The opportunity cost of using foreign currency
is therefore often higher than the exchange rate would suggest. It is therefore legitimate
in a comparative analysis to penalize options involving a lot of foreign exchange to a

315
greater extent than would apply simply by using the prevailing exchange rate. The
normal method of doing this is to multiply the actual financial cost by a so-called
"shadow price factor"; where there is a shortage of a commodity (eg. frequently, diesel
fuel) it will have a shadow price factor greater than unity, conversely where there is a
surplus (such as being able to use unskilled labour), then a shadow price factor of less
than one may be applied to unskilled labour wages. For example the shadow price of
diesel fuel may in some rural areas of developing countries be as much as four times the
real price, while the ready availability of unskilled labour may allow it in reality to be
costed at as little as 70% of its real wage level for economic comparisons. Tables of
shadow prices specific to different countries, and even to regions of countries have been
developed, but the concept of shadow pricing is complex and is probably best left to
economists. However the principles of shadow pricing should at least be borne in mind,
at least since items with a high opportunity cost may go into regular short-supply and
cause operational problems.
II.

A procedure for a cost appraisal of an irrigation pumping system

Fig. 163 outlines a method that can be used to compare the costs of alternative water
lifting techniques. This step-by-step approach is based on life-cycle-costing of the whole
system. It takes into account all the identifiable costs, but ignores benefits gained by the
users of water.
An integrated approach which considers the system as a whole, from the water source,
to the point of application on the field is recommended; i.e. including the water source
costs (such as well digging) and water distribution costs (such as digging ditches or
purchasing pipes or sprinklers).
iii.

Example; comparison of Engine, Wind and Solar Pumps

A simple worked example is included as Table 40. The first step is to determine the
hydraulic energy requirements. Suppose we wish to irrigate 0.5ha to a depth of 10mm
while pumping through a static head of 4m in the month of maximum water demand.
Reference to Fig. 13 in Chapter 2 indicates that this requires a nett daily hydraulic
energy output of 0.545kWh (factoring down by ten from the scale used in the figure).
Alternatively, the following relationship, also explained in Chapter 2, may be used:

Allowance must now be made for distribution losses; for convenience it is assumed that
all three systems being compared involve the same distribution efficiency of 60%. Then
the gross hydraulic energy requirement is (0.545/0.6) = 0.91kWh/day for all three. It
should be noted that in reality, different distribution efficiencies might occur with different
types of system, which would result in different energy demands.
The next step is to determine the design month; this is generally the. month of maximum
average water demand if the power supply is uneffected by climatic conditions, (eg. for
engines), but where the energy resource is the wind or the sun, it becomes necessary to
compare the energy demand with the energy availability and the design month will be
the month when the ratio of energy demand to energy availability is highest. Supposing

316
in this case that the design month does coincide with the month of maximum water
demand in all three examples, then using the assumptions in Table 40, the bases for
which are discussed in more detail in the relevant earlier chapters, we arrive at the sizing
for the systems. The engine in this case will be of the smallest practical size, but the
wind and solar pumps need to be suitably sized.
The next step is to estimate the installed capital cost of the system, generally by
obtaining quotations for appropriately sized equipment. Some "typical" values, valid at
the time of writing, have been used; namely $200/m2 of rotor for a windpump and $15/Wp
for a solar pump. The product of size and cost factor gives the installed capital cost,
which for simplicity is assumed to include the water source and distribution system in all
three cases. Some storage facility is likely to be needed in all cases; a secure lock-up
holding two 200 litre oil drums is assumed for the diesel while a low cost compacted soil
bund having a cement lining is assumed for water storage for the wind and the solar
systems, holding 40 and 30m3 respectively. In the case of petroleum fuelled systems, the
cost of storage includes the notional investment in stored fuel (assuming on average that
the storage is at 50% capacity and amortizing a continuous investment in 100 litres of
fuel).

317

Fig. 163 Step-by-step procedure for a cost appraisal of a water pumping system

318
Table 40 ANALYSIS OF UNIT WATER COSTS FOR FOUR TYPES OF IRRIGATION
PUMPING SYSTEM
Operational requirement:
10mm of water lifted 4m to cover 0.5ha (peak irrigation demand). Annual requirement
averages 67% of peak for five months, which is 5,094m3/year.
Assumed water source costs are identical in all four cases, for simplicity, although in
practice this may not always be correct.
Peak daily hydraulic energy requirement is 0.92kWh/day
Total irrigation demand: 5,094m3/yr in all cases.
Financial parameters (all cases) D = 10% N = 25 yrs i = 0

Parameter
Price of fuel
delivered to field
Critical month
mean irradiation
Critical month
mean windspeed
Sizing assumption

Assumption and result

Wind
Solar
-

Gasoline
50c/litre

Diesel
40c/litre

5.8kWh/m2

3.5m/s

lkW
3%
effic.

minimum
0.1V3
W/m2
size
available
i.e. 2kW &
8% effic.
1.2 1/day 9.7m2 rotor
diesel fuel
area

Requirement to
3.11/day
produce peak daily
gasoline
water output
Requirement to
2.11/day
0.8 1/day
ditto
produce mean daily
gasoline
diesel fuel
water output
$330 (engine
$1,500
$1,940
Capital cost
& pump)
(engine & ($200/m2)
assumption (total
pump)
power system and
pump)
Storage tank
2 x 200 1 2 x 200 1 fuel 40 m3
capacity
fuel in
in secured water tank
secured
shed
shed

Notes

35% mean
motorpump
effic.

540Wp
array
ditto

$8,100
($15/Wp)

30m3
water
tank

Cost of storage

$280

$250

$600

$450

Life of system

3yrs

7 yrs

20 yrs

15 yrs

Life of storage

15 yrs

15 yrs

15 yrs

15yts

Lifecycle system
costs
Lifecycle storage
costs

$1,224

$2,865

$2,405

$10,444

$347

$310

$744

$558

Cost includes
estimated average
fuel inventory cost as
well as shed cost.
Hence gasoline
storage costs more on
average.

319
Total lifecycle
capital costs (at
present value)
Annualised system
costs
Annualised storage
costs
Annual 0 & M
costs
Annual fuel costs
Total annual cost:
Average unit cost
of water

$1,571

$3,175

$3,149

$11,002

$135

$315

$265

$1,166

$38

$34

$82

$61

$220

$200

$50

$50

$156
$549
l0.8c/m3

$49
$598
11.7c/m3

$397
7.5c/m3

$1,277
25.1c/m3

The actual useful life of the systems and storages is assumed, as indicated in the table,
as are financial parameters for the discount rate and the period for analysis. Hence the
life-cycle costs may be determined by working out the present values of the first system
and all subsequent replacement ones (using factors from Table 38) and adding them all
together. Table 40 shows the system and storage life-cycle costs separately, but they
could also be combined. In order to arrive at a comparable annualized cost for the
capital investment in each system, the factors of Table 39 are used to convert the lifecycle cost; in this case a 25 year period and 10% discount rate gives a factor of 0.11,
which the life-cycle cost needs to be multiplied by. The storage system was dealt with
similarly.
The combined system and storage annualized costs represent the annual investment or
"finance" costs. Different systems also have recurrent costs consisting of O&M
(operational and maintenance) costs, and sometimes fuel costs. When the finance, O&M
and fuel costs are added, we obtain the gross annual cost of owning and operating each
system.
Where an identical useful output is to be produced, then the gross annual cost is
sufficient for ranking purposes. In reality, however, different cropping strategies may
apply for different irrigation systems, resulting in different crop irrigation water demands
and different benefits (in terms of the market values of the crops). Therefore it is useful
to divide the gross annual cost by the gross annual irrigation water demand to arrive at
an average unit cost for water from each option.
In this example, the windpump comes out marginally better than the petrol engine but
the decision probably ought to be made between them on other than economic grounds
as there is so little to choose between them in simple unit cost terms. In this example,
the solar pump does not seem economically competitive. It must be stressed that this is
but one simple example which should not be blindly used to draw any conclusions on the
relative merits of engine, wind and solar pumps generally. Even varying totally nontechnology dependent parameters such as the discount rate, the period of analysis, the
water demand or the head could significantly change the results and rankings obtained,
and so could changing the technical performance and/or cost parameters, which would
have an even more profound effect.

320
5.1.3

relative economics of different options

A procedure similar to that just described has been followed to analyse a representative
selection of the types of water lifting systems described earlier in this paper.
Most studies attempting this kind of analysis use a single assumption for each and every
parameter and compound these to arrive at a single answer, as in the example just
given, often presented as a single curve on a graph for each option. The trouble with this
approach is that errors are compounded and may not cancel out, so the result could be
very misleading. In an attempt to minimize this problem, the approach in this case has
been to choose a low and a high parameter at each and every decision point; i.e. a
plausible pessimistic and a plausible optimistic one. Two sets of calculations are then
completed for each technology, to produce a pessimistic and an optimistic result, which
when graphed give two curves. It is then reasonable to assume that the real result is
likely to lie between the two curves and the results are therefore presented as a broad
band rather than a thin line. Therefore, where the broad band for one technology lies
wholly above or below another it is reasonable to assume the lower one is the cheaper
option, but obviously many options overlap considerably and in such situations other
considerations than water cost should dictate the decision.
Table 41 lists all the systems considered and gives the principle assumptions used for
calculating the cash flows. The capital cost assumptions are intended to include the
entire system as defined in the previous section; i.e. not just a prime-mover and pump,
but all the necessary accessories that are necessary and appropriate for each type of
technology and scale of operation.
Table 41 COST AND PERFORMANCE ASSUMPTIONS USED FOR COMPARISON OF
ALTERNATIVE PUMPING METHODS
Capital Cost
Solar PV (hi) present
Module $10/Wp

Life Maintenance
years per pump
15

Motor and pump

$(500 + 1.5Wp)

7.5

b.o.s. $(1500 + 2.0Wp) 15

Wind (hi)

Operating
cost

$50+$0.05/Wp NIL
p.a.

Performance
assumptions

Notes

Motor/pump
Subsystem
efficiency = 35%

Sized for the design

month. Irradiation
levels of 20 MJ/m2
in the design month
examined, Peak
water requirement
in design month
assumed to be 2 x
average water
requirement. Wp is
the array rating in
peak Watts.

321
$400/m2 of swept
rotor area

20

$50 + 2.5 x
area p.a.

NIL

Meaning Hydraulic Sized for a design

power=0.1V3W/m2 mean wind speed
(in the design
(= mean wind
month) of 3 m/s.
speed)
Peak water
requirement = 2 x
average water
requirements.

$400 p.a.

80 c/litres

7.5

$200 p.a.

40 c/litre

Overall (hydraulic/ P is the shaft power

fuel) efficiency = in kW. Minimum
operation 0.25
0.03 + 0.007P
hours per day.
Efficiency in first
half hour assumed
efficiency = 0.13 +
to be half of
0.007P
'steady-state'
efficiency.

$200 p.a.

80 c/litre

efficiency = 2%

Size of engine = 1.0

kW. Number of
engines chosen to
meet the demand.

$100 p.a.

40 c/litre

efficiency = 6%

Efficiency
assumptions at
start-up as for
diesel.

$20 p.a.

0.03c/MJof
delivered
gas

$20

$0.75 per
Hydraulic output = Assumed to work
animal day 200Wper animal continuously for 8
hours per day at a
hydraulic output of
200W i.e. 587m per
day

Diesel (hi)
$(1900 + 8.6P)

Diesel (lo)
$(950 + 4.3P)

Kerosene (hi)
$600

Kerosene (lo)
$200

Biogas
Gas holder $(137+ G) 5
b.o.s. $(91 + 1.89G)
pump hi $600 pump lo
$200

Oxen (lo)
animal $250

10

pump $100

G is the energy
content of the gas
produced in MJ per
day. The biogas unit
is sized to provide
the daily input
energy
requirements of the
spark ignition
engine.

322
Oxen (hi)
animal $125

10

pump $100

$20
$1.25 per
animal day

Human (hi)
$200 per pump

Human (lo)
$20 per pump

Turbine (lo)
$200 per pump

15

$200 for civil works

30

Turbine (hi)
$200 per pump
15
$2000 for civil works 30
Hydram (hi)
$3000 per hydram
Hydram (lo)
$1000 per hydram

$1 per man Output of each

day
pump = 37W
hydraulic

$20p.a.

$0.30 per
man day

$20p.a.

Output 350W per

pump
These pumps are
assumed to operate
over a 24 hour
period.

$20p.a.

30 $5 p.a.
30

$5 p.a.

Mains (hi)
$10000 for connection 30
+ $(265 + 0.75P) per
pump

$5 p.a.

Mains (lo)
No connection charge
$(265 + 0.75) per
pump

Pump gives rated

output for 4 hours
per day (i.e. 54m
per day)

30 $20 p.a.

$0,042 per
MJ of
electricity =
15c/kWh

Output 100W

P is the rated output

power of the pump.
The pump size is
determined by
assuming the daily
water requirements
is to be provided in
6 hours.

$0.02 per
MJ of
electricity
7c/kWh

Because of the large numbers of options to be analysed, calculations were carried out
on a computer and the results printed out graphically as the cost of hydraulic energy
versus the peak daily demand for hydraulic energy (see Figs. 164 to 168).

323
To eliminate one parameter, the output was calculated for each option, not in terms of
volume of water pumped, but in terms of hydraulic energy output; this effectively
combines the volume of water pumped and the head, since units of m3.m, or cubic
metre-metres were used. However it should be realised that this is only valid for
comparing similar systems; you cannot realistically compare systems operating at
radically different heads such as a 100m borehole pump with a 10m head surfacesuction pump purely on the basis of m3.m. To convert a figure in m3.m to flow at a
specific head it is only necessary to divide by the head in question; e.g. 10m3.m could be
2m3 pumped through 5m head. To convert a unit cost of 5cents/m3.m to obtain a cost per
unit of water, it is necessary to multiply by the head in question; eg. that energy cost at
2m head represents a water cost of 10c/m3.

Fig. 164 Human and animal power: output cost versus scale (note log-log scale)

324

Fig. 165 Diesel and kerosene pumping sets (note log-log scale)

325

Fig. 166 Windpumps at various mean windspeeds (note log-log scale)

326

Fig. 167 Solar pumps at various mean insolation levels (note log-log scale)

327

Fig. 168 Hydrams (note log-log scale)

The final results are presented in terms of output cost versus the hydraulic energy
demand. This is because the unit costs of different options, and hence the economic
rankings, are sensitive to the size of system used. Therefore the choice of technology
will differ depending on the scale of operation; systems which are economic for larger
scale operations are often uneconomic on a small scale, and vice-versa.
Figs. 164 to 168 show the results for the different options analysed. In some cases, such
as solar and wind powered pumping systems, the variability of the energy resource was
allowed for by recalculating the band of results three times, i.e. for a mean of 10, 15 and
20MJ/m2 per day (2.8, 4.2 and 5.6kWh.m2 per day) of solar irradiation and, similarly for
three mean wind-speeds of 2.5, 3.0 and 4.0m/s. The lower levels chosen are
deliberately selected because they are sub-marginal conditions, while the middle level
was judged to be marginal rather than attractive for the technologies concerned; so the

328
results of all except the 20MJ/m2 per day (for solar) and the 4.0m/s (for wind) examples
would not be expected to show these technologies particularly favourably.
A problem with Figs. 164 to 168 is that they had to be plotted on a log-log scale,
because of the large range of power and cost considered, or either a very large sheet of
paper would be needed to show the results, or the results at the lower end, which are of
great interest to many people, would have been compressed to insignificance. The
trouble with log-log scales is that the eye interprets distances linearly, so they can be
misleading if simply inspected. This makes it difficult to compare the various options
shown in Figs. 164 to 168. Therefore, Fig. 169 has been provided as a simplified
composite of these results, using mean values (between the highs and lows) of the other
graphs (to avoid too much of a confusion of curves) and moreover it was plotted against
linear axes over a necessarily smaller size range, up to only 1 000m3.m/day. This range
is of most interest as the relative rankings do not change much once an energy demand
of about 1 000m3.m/day is exceeded.

Fig. 169 Mean values of the results of Figs. 164-168 plotted with linear scales on the
same axes for daily hydraulic requirements of 1000 m3m. 'Jumps' in the curve are points
at which it becomes necessary to use another unit to meet the demand.

329

Fig. 170 Expected range of unit energy costs for three levels of demand, if 100,1000 and
10,000 m3m/day for different types of prime-mover

330
Another, perhaps more easily interpreted presentation of these results is given in Fig.
170, where histograms of the cost spread for each system at daily energy demands
equivalent to 100, 1 000 and 10 000 m3.m are given, and compared linearly rather than
logarithmically for ease of comparison. (To put this in perspective, we are a considering
say, 20, 200 and 2 000m3 per day at 5m head, or half those amounts at 10m head, etc.).
This set of histograms also reintroduces the "optimistic" to "pessimistic" spread for each
technology, which was lost in the previous comparison of Fig. 169. It is important not to
lose sight of the possible range of costs applicable to any given technology, especially
as in some cases the band of possible costs is very wide even on the basis of quite
plausible assumptions in all cases.
It is interesting to see, for example, how a 10kW diesel system is by far the most
expensive at the smallest demand level, where wind and solar pumps are at least
competitive, but when 10 000 m3.m are needed the situation is completely reversed.
i.

Conclusions to be drawn from economic analysis

The economics for most options are particularly size-sensitive, so that what is correct at
100m3.m/day is not generally true for a hydraulic requirement of 10 000m3.m/day.
The only low unit-output cost options which apply almost right across the entire size
range of interest are:
a. mains electricity; providing it is already close to the field so only minor connection
costs are incurred
b. hydro-powered devices (hydrams or turbine pumps); but depending on suitable
site conditions
c. windpumps; but only for locations with high mean wind speeds (i.e. greater than
4m/s)
d. animal power is cost-competitive for energy demands exceeding about
500m3.m/day; but it does not generally seem a realistic new option where animal
traction has not traditionally been used.
e. human power is cost-competitive in very small-scale applications (under
100m3.m/day); but only if a very low "opportunity cost" is assigned to human
labour, and this conflicts with many development goals.
Where land-holdings are so small that the demand is less than 100m3.m/day, then
human power is relatively inexpensive and animal power appears to be competitive.
Solar and windpumps are both potentially competitive and so are spark ignition engines,
providing they are reasonably efficiently sized and operated. Diesel power is not
generally cost-effective for such small energy demands. Although the renewables in
some cases appear competitive at this small demand level, the absolute costs of water
are still rather high and it is important to ensure that irrigation will in fact produce a
profitable yield in relation to the high water costs involved. It may be better to try to
consolidate a single larger water system shared between several such small land
holdings where such an option is feasible.
At the medium size range analyzed, namely 1 000m3.m/day, all the options are generally
more cost-effective than they are at 100m3.m/day energy demand, and there is an
overlap between most options; animal power, windpower (with V greater than 4m/s),

331
water power, i.e. engines (only if efficiently operated) and mains electricity appear
marginally the best options.
At the large size range of 10 000m3.m/day, diesel power comes into its own, and unless
mains electricity or water power is available, diesel will probably be the best option.
Therefore, in summary, mains electricity (providing no connection costs are involved) or
water power are most economical. Windpower is next most attractive if windspeeds are
high, (but it is decidedly unattractive with low or uncertain winds). Solar power is
generally expensive but has the potential to fill a useful gap in the 100-1000m3.m/day
demand level range once the cost of solar-powered systems falls a little more. Engines
have a very wide band of uncertainty relating to their unit costs at small energy demand
levels, ranging from competitive to unacceptable. Spark ignition engines are more
attractive in the small to medium range of 100-1 000m3.m/day while diesel engines
become more competitive at energy demands exceeding around 1 000m3.m/day.
Biomass-fuelled spark ignition engines will generally cost more to run than kerosene or
gasoline fuelled engines (where fuel is at world prices), but will be worth considering
where petroleum fuels are either not available or have a high opportunity cost; obviously
a suitable low-cost biomass fuel resource needs to be readily available.

332

5.2

PRACTICAL CONSIDERATIONS

It is worth elaborating on some of the practical, in addition to economic, considerations

that relate to the different types of water lifting system.

5.2.1

Status or Availability of the Technology

Some technologies are more "available" than others. Table 5.4 indicates technologies in
general use, or with "future potential", plus some that are obsolete; a few qualify for
more than one of these categories. In some cases, such as small Stirling engines, it is
believed that not one commercial manufacturer currently offers a viable product even
though in this case they were widely used in the past and there seems no technoeconomic reason why they would not be attractive today. Similarly, the Chinese Turbine
Pump, which is widely used in southern China, appears to be most attractive
economically (and has numerous operational advantages) but is not currently produced
outside China. Therefore, for some time to come, such technologies may only generally
be considered as a potential option in most countries.
A more general problem with all new or unfamiliar technologies, even if they are
nationally commercially available, is to obtain the necessary information and advice in
order to:
a. procure the correct size and specification of system
b. install it correctly
c. operate and maintain it effectively
It is probably best if all but the more adventurous (and wealthy) of small farmers play
safe and stick to familiar and "available" technologies where help, advice and spares are
readily available and risks are minimized. However, if everyone took this advice, new
and perhaps eventually better technologies would never become available, so therefore
it is worth suggesting that it is necessary for government, international aid agencies and
institutions, with a commitment to the future development of small-scale agriculture, to
take risks in this area on behalf of their local farmers and to test and demonstrate any
technologies that appear promising in the local irrigation context.
Problems must be expected with pilot projects. It is therefore vital to measure, monitor
and record the behaviour of any innovative systems that are tried. Even if no problems
occur, unless such pilot projects are properly monitored it will not be possible to come to
any conclusion as to whether the new technology being tried is competitive with what it
is supposed to replace.
Actual performance monitoring is important, but so are qualitative comments on
operational aspects, such as maintenance or installation difficulties, or shortcomings as
perceived by the user. Feedback on these aspects needs to be absorbed by the
manufacturers and developers of this kind of equipment, so that the necessary
improvements can be set in hand, otherwise development will be delayed.

333
5.2.2.

Capital Cost versus Recurrent Costs

As explained previously, low recurrent costs tend to have to be traded for high capital
costs. High capital costs represent a real barrier for small farmers to take up new
technology even if the unit output costs are competitive. Worse, low capital costs are
often an incentive to install inefficient systems (eg. small kerosene pumps sets with
inadequate distribution pipe). Where there is a good case for farmers to be encouraged
to use a high capital cost technology (even to move from spark-ignition to diesel
engines), then it will generally be necessary for appropriate credit facilities to be made
available as an incentive.
Renewable energy systems, with their high capital costs and low recurrent costs may be
of particular interest to institutions having access to grant or aid finance for capital items,
because they do offer a means for investing in low recurrent costs. Many rural
institutions face major problems with meeting the recurrent costs of running conventional
pumping systems, so in some situations it may make sense to introduce high capital cost
equipment simply to reduce the maintenance and fuel budgets.

5.2.3

Operational Convenience

This factor varies considerably with different types of pumping system. For example, a
windpump will be highly dependent on adequate wind to operate, so if high risk crops
are grown where the provision of water on demand is vital to the survival of the crop,
then a large (and consequently expensive) storage tank will be necessary to ensure
water is always available. Alternatively less risky and probably less valuable crops could
be grown with only a small storage tank or even in some cases with no storage at all.
Therefore the flexibility of the device, or its controlability, must be taken account of as
they affect such fundamental decisions as the choice of crop to grow under irrigation.
Other factors relate to aspects such as size and portability. Small engines and small
solar pumps may be quite portable, which means they can be moved around to irrigate
with only short, but effective distribution pipes, while a windpump, a larger engine, or a
hydram will inevitably have to be fixed. However small portable items are also vulnerable
to theft in some regions, which makes the relatively large and fixed installation less at
risk in that context.
Few options can rival the operational flexibility of an i.e. engine powered system in terms
of rapid start up, portability, provision of power on demand, etc., but of course one of the
reasons for looking at the other options is that the i.e. engine generally suffers the major
drawback of needing petroleum fuels. So the operational shortcomings of many of the
alternatives need to be weighed against the fuel needs and the likely future availability
and cost of fuel.

5.2.4

Skill Requirements for Installation, Operation and Maintenance

Two key factors apply here; the absolute skills required and the level of familiarity with
the equipment. Commonplace, but complex machines like diesel engines can often
appear to be simpler to handle than much simpler but less familiar technologies such as
solar pumps (from the maintenance point of view). Due allowance for the need to learn
about a new technology must therefore be made before dismissing it as too

334
sophisticated. In absolute technical terms there are no water lifting technologies more
technically demanding than the diesel engine; most of the renewables are basically
much simpler even if in a few cases they involve little understood concepts (not many
diesel mechanics understand the first principles of a diesel engine either - it is not
necessary to know this in order to overhaul an engine).
The level of support available from manufacturers or suppliers is most important; most
successful technologies have become widely used because they were successfully
promoted and supported by the commercial interests that market them. Even the
simplest technologies stand little chance of being taken up unless they are effectively
promoted and the early users are also properly supported and helped through any
problems.
Often there are trade-offs between the amount of skill needed and the amount of
maintenance/operational intervention required. For example; solar pumps need very little
maintenance, but what they do need is usually unfamiliar and demands specialized and
at present rare skills, (although given appropriate training the necessary skills can be
rapidly attained); i.e. engine pumps need quite frequent and sometimes technically
sophisticated maintenance functions, but because the technology is so widespread there
are many people capable of performing these; the cruder types of village built
windpumps need a great deal of adjustment and running repairs, but to the people who
are familiar with them, these present little inconvenience or difficulty.
Familiarity or lack of familiarity are perhaps more important factors then the absolute
level of technical skill needed, (after all few repairs can be more demanding then
overhauling a diesel injection pump or reboring an i.e. engine, yet few provincial towns
exist in developing countries where such activities cannot be carried out). Therefore
training is a vital aspect of introducing any new or unfamiliar technologies.

5.2.5

Durability, Reliability and Useful Life

Durability, reliability and a long operational life usually cost money, but they also are
frequently a good investment in terms of minimizing costs. Perceptions on the value of
capital or the choice of discount rate will usually control decisions on these aspects.
However it is best to try and show through economic or financial analysis whether the
benefits from buying high quality equipment are cost-effective. Many who have analyzed
the cost of operating machinery in remote areas, particularly engines, have concluded
that their performance, not only in terms of output but also in terms of reliability and
durability, usually turns out to be sigificantly worse than expected. There is therefore
often merit in erring towards oversizing prime-movers and procuring any special
accessories that make the system more "fail-safe".

5.2.6

Potential for Local Manufacture

This is of more immediate interest to policy makers than farmers, although the benefits
to be gained from local manufacture could of course eventually benefit the latter.
One of the principal reasons to seek alternatives to petroleum fuelled engines for
irrigation is because of the inability of many countries to import sufficient petroleum to
meet present, let alone future, needs. Therefore shortage of foreign exchange to import

335
oil equally implies shortage of foreign exchange to import foreign solar pumps or other
such alternatives. Therefore, any system which lends itself to whole or partial local
manufacture is of potential economic importance in terms of import substitution for oil,
(and probably for engines too). However, the benefits of local manufacture do not end
with import substitution.
Other important results of local manufacture, or part-manufacture, are:
-

creation of local industrial employment

enhancement of industrial skills
improved local availability of spare parts
improved local expertize in the technology

in other words, local manufacture can help to overcome many of the constraints
mentioned previously in supporting the initial diffusion of a new technology, and can at
the same time help to develop the industrial/manufacturing base. The economy of the
country gains twice from local manufacture of irrigation equipment, first from internalising
the manufacture and secondly from the enhanced agricultural production once the
equipment starts to be widely applied.

336

5.3

CONCLUSION

The reader will have realized from this paper that there is a wide variety of options for
combining prime movers and water lifting devices in order to pump irrigation water. In
practical terms the situation will be simplified for most people by having to choose from a
much more limited selection of what is available rather than what is possible. However, it
is hoped that by applying an understanding of the technical and economic principles
described, more cost-effective irrigation may be achieved, and hence more irrigation.
Small scale lift irrigation is not yet normally practiced in many countries, so the incentive
to develop the full range of potentially useful technologies has not yet been given to
industry. It is to be hoped that the increasing need to grow more food will drive irrigation
technology forward and result during the next decade or two in a considerable widening
of the choice of equipment on the market. The author hopes this book will have made a
small contribution to encouraging the understanding and interest needed for this to
happen.

337

REFERENCES
1.

Molenaar, A., Water Lifting Devices for Irrigation, FAO, Rome. 1956

2.

World Bank, Energy in the Developing Countries, Washington, DC. 1983

3.

Stern, P S, Small Scale Irrigation, I T Publications, London. 1979

4.

World Bank, World Development Report - 1982, World Bank/OUP, Oxford. 1983

5.

World Bank, World Development Report - 1980, World Bank/OUP, Oxford. 1981

6.

Leech, G., Energy & Food Production, IIED, London. 1975

7.

World Water, The Right to Rice, Liverpool, Oct 1980. 1980

8.

Booher, L.J., Surface Irrigation, Agricultural Development Paper No. 95, FAO, Rome. 1974

9.

Kraatz, D.B., Irrigation Canal Lining, Land and Water Development Series No. 1, FAO,
Rome. 1977
Vermeiren & Jobling, Localized Irrigation, Irrigation & Drainage Paper No. 40, FAO, Rome.
1980
Doorenbos & Pruitt, Crop Water Requirements, Irrigation & Drainage Paper No. 24, FAO,
Rome. 1977
Kraatz, D.B., Socio-Economic Aspects - cost comparison and selection of water lifting
devices, Proc. FAO/DANIDA Workshop on Water Lifting Devices in Asia and the Near
East, held in Bangkok Dec 1979, FAO, Rome.1981
Roberts, W. & Singh, S., A Text Book of Punjab Agriculture, Civil S Military Gazette,
Lahore, Pakistan.1951
Mead, Daniel W., Hydraulic Machinery, McGraw-Hill, New York and London. 1933

10.
11.
12.
13.
14.

15. Fraenkel, P.L., Food from Windmills, IT Publications, London. 1976

16. Collett, J., Hydro Powered Water Lifting Devices for Irrigation, Proc.FAO/DANIDA
Workshop on Water Lifting Devices in Asia and the Near East, held in Bangkok Dec 1979,
FAO, Rome. 1981
17. Watt, S.B., Chinese Chain & Washer Pumps., I T Publications, London. 1977
18. Gao Shoufan, Yin Jianguo, and Cheng Zuxun, Tube-chain waterwheel, proc.UNDP/FAO
China workshop of Nov, FAO, Rome. 1982
19. Dunn, P.D., Appropriate Technology, Macmillan, London. 1978
20. Intermediate Technology Development Group, Water for Rural Communities, Appropriate
Technology Journal, Vol 9, No. 1, IT Publications, London. 1983
21. Hofkes, Manual Pumping of Water for Community Water supply and small scale irrigation,
proc. FAO/DANIDA Workshop on Water Lifting Devices in Asia and the Near East, held in
Bangkok Dec 1979, FAO, Rome. 1981
22. Morgan, letter in Appropriate Technology Journal Vol 9 No 1, IT Publications, London.
1983
23. Wilson, S.S., Pedalling foot power for pumps, World Water, Liverpool, Nov 1983. 1983
24. Schioler, T., private communication. 1985
25. Khan, H.R.,Study of Manual Irrigation Devices in Bangladesh, Proc. AT in Civil

338
Engineering conf., Inst. of Civil Engineers, London. 1980
26. McJunkin, Handpumps for use in Drinking Water Supplies in Developing Countries, IRC,
The Hague. 1977
27. UNDP/World Bank, Global Handpump Project, Technical Paper No. 3., Washington DC..
1982
28. Klassen, The Rower Pump, proc. FAO/DANIDA Workshop on Water Lifting Devices in
Asia and the Near East, held in Bangkok Dec 1979, FAO, Rome. 1981
29. Tata Energy Research Inst., Energy Update, Bombay. 1979
30. Birch D.R. & Rydzewski J.R.,Energy Options for Low Lift Irrigation Pumps in Developing
Countries: The case of Bangladesh & Egypt, ILO, Geneva.1980
31. Hood, O.P., Certain Pumps and Water Lifts used in Irrigation, U S Geological Survey,
Washington DC. 1898
32. Jansen, W.A.M.,Performance Tests of Kerosene Pumpsets Wind Energy Unit, Colombo, Sri
Lanka. 1979
33. Reader, G.T., Hooper C, Stirling Engines, E & F N Spon, London and New York.1983
34. Ross, A., Stirling Cycle Engines, Solar Engines, Phoenix, Az. 1977
35. Fluitman, F., The Socio-Economic Impact of Rural, Electrification in Developing Countries:
A Review of the Evidence, ILO, Geneva.1983
36. Vadot,L., Water Pumping by Windmills, (translation of 'Le pompage de l'eau par eoliennes'),
La Houille Blanche, No.4, pp 496-535, Grenoble, France, Sept. 1957
37. Perry,Experiments with Windmills, US Geological Survey, Washington, DC. 1899
38. Sidney Williams & Co, Comet Catalogue No. 9, Dulwich Hill, NSW, Australia. 1968
39. Eldridge, Frank R., Wind Machines, van Nostrand Reinhold, New York. 1980
40. Fraenkel, P.L., Wind Technology Assessment Study, I T Power Ltd., Reading for
UNDP/World Bank, Washington, DC.. 1983
41. Gao Rushan, Zhu Zhongde, On the Development of Human, Animal, Wind and Water Power
Lifting Devices for Irrigation and Drainage in China, Ministry of Water Conservancy,
Beijing. 1979
42. Schioler, T., Water Lifting Devices for Irrigation - History and General Background,
proc.FAO/DANIDA Workshop on Water Lifting Devices in Asia and the Near East, held in
Bangkok Dec 1979, FAO, Rome.1981
43. Gilmore, E., Barieau, R. E., and Nelson, V., Final Report on the Feasibility of using Wind
Power to Pump Irrigation Water, for the Governor's (of the State of Texas) Energy Advisory
Council, Austin, Texas. 1977
44. Nelson, V., et al.,Wind Power Applications in the US - Irrigation Pumping, Wind
Engineering, Vol 6 No2. 1982
45. Lysen, E., Introduction to Wind Energy, CWD, (formerly SWD), PO Box 85, Amersfoort,
The Netherlands. 1983
46. Pinilla, et al., Wind Energy and Water Pumped: Conversion Efficiency Limits using Single
Acting Lift Pumps, Proc. Conf. BWEA, Cambridge University Press, Cambridge.1984
47. Golding E.W.,The Generation of Electricity by Windpower, E & F N Spon, London (revised
edition). 1976
48. World Meteorological Office, Technical Note No. 175, Meteorological Aspects of the
Utilization of Wind as an Energy Source, Secretariat of the WMO, Geneva.1981

339
49. Leicester, R. J., Study of the World Market for Medium-sized Wind Generators, for
Overseas Development Administration, Crown Agents, London.1981
50. Lipman, N.H., et al., (Editors), British Wind Energy Association, Wind Energy for the
Eighties, Peter Perigrinus, Stevenage, UK and New York.1982
51. ECDC-TCDC, Renewable Sources of Energy, Volume III, Wind Energy, UN Economic and
Social Commission for Asia and the Pacific, Bangkok. 1981
52. Butti, K,. & Perlin, J., A Golden Thread: 2500 Years of Solar Energy and Technology
Marion Boyars Pubs., London. 1981
53. Daniels, F., Direct Use of the Sun's Energy, Yale University Press, and Ballantine Books,
New York. 1974
54. Halcrow/I.T.Power, Small Scale Solar Powered Irrigation Pumping Systems: Technical &
Economic Review, World Bank, Washington DC. 1981
55. Halcrow/I.T.Power, Small Scale Solar Powered Irrigation Pumping Systems: Phase I Project
Report, World Bank, Washington DC. 1981
56. Halcrow/I.T.Power, Small Scale Solar Powered Pumping Systems: The Technology, its
Economics and Advancement - Main Report, World Bank, Washington DC. 1983
57. Kenna, J.P. and Gillett, W.B., I.T.Power/Halcrow, Handbook on Solar Water Pumping,
World Bank, Washington DC and I T Publications, London. 1985
58. Rauschenbach, H.S., Solar Cell Array Design Handbook, van Nostrand Reinhold, New
York. 1980
59. World Meteorological Office, Meteorological Aspects of using Solar Radiation as an Energy
Source, Technical Note 172, WMO Secretariat, Geneva. 1981
60. Rosenblum, Practical Aspects of Photovoltaic Technology, Applications, and Cost, NASA
Lewis and USAID Science and Technology Bureau Office of Energy, Washington DC. 1985
61. Meier, U., Local Experience with Micro-hydro Technology SKAT, St Gall. 1981
62. Alward, et al., Micro-hydro Power: Reviewing an old Concept, NCAT/US Dept. of Energy,
Butte, Montana.1979
63. Summary Report, Proceedings of the FOA/UNDP/China Workshop on Water Lifting
Devices and Water Management, Held in China Nov 1981, FAO Rome.1982
64. Tacke, J.H.P.M., Hydraulic Rams, Delft University of Technology, draft or IRC publication
Alternative Energy Resources for, Rural Water Supply Applications in Developing
Countries, The Hague.1984
65. Nguyen Trong Lac, Norias in Vietnam, proc. FAO/DANIDA Workshop on Water Lifting
Devices in Asia and the Near East, held in Bangkok Dec 1979, FAO, Rome.1981
66. Collett, J., Water Powered Water Lifting Devices for Irrigation, unpublished revision for
FAO of ref. [16], Intermediate Technology Consultants, London. 1980
67. Fraenkel, P., & Musgrove, P, River S Tidal Current Turbines, Proc. Int. Conf. on Future
Energy Concepts., Inst.of Electrical Engineers, London.1979
68. Hall,D.O., World Biomass: An Overview, Proc. Biomass for Energy, UK International Solar
Energy Society, London.1979
69. Kristoferson, L., Bokalders, V., Newham, M., Renewable Energy for Developing Countries:
A Review, The Beijer Inst., Stockholm.1984
70. FAO, Map of Fuelwood Situation in Developing Countries, FAO, Rome. 1981
71. Brown, Lester R., Food or Fuel, Paper No.35, Worldwatch Institute, Washington DC. 1980
72. Earl, D.E., Forest Energy & Economic Development, Oxford University Press, Oxford. 1975
73. Ascough, W.J., Producer Gas for Automotive Use, Proc. Alternative Fuels for I.C.Engines,

340
Reading University and ITIS, Reading and Rugby, England . 1981
74. Foley, G., Barnard, G., and Timberlake, Gasifiers: Fuel for Seige Economies,
Earthscan/IIED, London.1983
75. van Swaaij, W.P.M., Gasification: the Process and the Technology, Proc. Energy for
Biomass, Brighton.1980
76. Damour, V & M-S, Development of Small Scale Gasifiers for Irrigation in India, Paris.1984
77. Coombs, J., Production of Alcohol Fuels, Proc. Alternative Fuels for I.C.Engines, Reading
Univ. and ITIS, Reading and Rugby.1981
78. Radley, R.W., et al, Vegetable Oils as Fuels for Diesel Engines, Proc.Alternative Fuels for
I.C.Engines, Reading University and ITIS, Reading and Rugby .1981
79. Hall, D.O., Vegetable Oils for Diesel Engines, Private Paper, Kings College, London
University.1981
80. Goddard, K., Liquid Fuels from Biomass, Chartered Mechanical Engineer, London, Jan
1979. 1979
81. Hall, D.O., Renewable Resources (Hydrocarbons), Outlook on Agriculture, Vol 10, No. 5.
1980
82. Pyle, D.L., Biogas & Thermochemical Products, Proc. Alternative Fuels for I.C.Engines,
Reading University and ITIS, Reading and Rugby. 1981
83. van Buren, A., A Chinese Biogas Manual, I T Publications, London. 1979
84. Meynell, P-J., Methane: Planning a Digester, Prism Press, Dorchester. 1982
85. UN ESCAP, Renewable Sources of Energy: Volume II, Biogas Bangkok. 1981
86. ITDG Mission to China, Internal Report, Intermediate Technology Development Group,
London. 1981
87. Hofkes, E.N., et al., "Renewable Energy Sources for Rural Water Supply in Developing
Countries", International Reference Centre for Community Water Supply and Sanitation, the
Hague.1986

341

Back Cover