Previous Page Table of Contents Next Page

3. Available water-lifting technologies


The main types of water-lifting device relevant for small-scale irrigation (and potable water supply or stock-watering applications of a similar scale) are those powered by:

These six categories are briefly described under Main types of water-lifting device below, and in more detail in Annex H; some categories contain a wide spectrum of different types of water-lifting device. Each device has a particular range of applications for which it is potentially suitable.

An application is usually described in terms of discharge and useful head, where discharge indicates how much water is lifted per second or per hour, and useful head indicates how high the water is lifted. (The term ‘head' means pressure expressed in the equivalent depth or height of a water column: for instance, a pressure of one atmosphere is equivalent to about 10 metres (m) head of water.) Useful head may be subdivided into static lift (the vertical height of the device's delivery point above the water source's free surface) and device delivery head (the pressure in a pipe at the delivery point, capable of driving water further up a hill or along a pipe against friction). Device delivery head is zero for many devices where the water simply drops out of the device outlet into a channel or reservoir.

The suction lift of a pump (the vertical height from the water source's free surface to the pump inlet) is limited, because the water always has to be pushed up to the pump inlet by atmospheric pressure: in practice the suction lift cannot be more than about 7 m.

If the water source is more than about 7 m below the delivery point, the pump itself has to be below the delivery point, often below ground level. Sometimes the whole device is mounted on a shelf partway down a well, and sometimes the pump is mounted some way down a well or borehole while its engine is at surface level. Where the suction lift is more than some limit (depending on the pump type) special arrangements may be needed to prime the pump, i.e. to get water into it at the start of a period of pumping. On a motorized pump, this priming may be done by a small auxiliary hand pump. Pumps that are located below the source's water surface, or that can create sufficient suction by pumping air to draw the water in by themselves on start-up, are called self-priming pumps.

Water-lifting devices use energy in one form or another to lift and/or pressurize water. The amount of useful water-lifting work done is the product of the volume of water lifted and the useful head through which it is lifted (the level difference between the water source's free surface and the delivery point, plus any pressure head at the delivery point). Pumps usually have to do more work than this, because of friction losses in pipework and fittings. The overall energy efficiency of a water-lifting device (engine or other power source, pump, and pipework considered as a whole) is the ratio of useful water-lifting work produced to the energy put into the device in chemical, mechanical or electrical form. Pumps by themselves may be up to 80 percent efficient; complete devices are often less than 10 percent efficient. All this is explained more fully and precisely in Annex H, which sets out the principles on which the efficiency of devices can be compared, and describes the efficiencies and limitations of the more interesting types.


This section summarizes the information in Annex H, where the precise meanings of the various terms are explained.

a) Human-powered water-lifting devices

There are many different types of human-powered water-lifting device. Their efficiency is seldom measured or seriously investigated, even by the developers and makers of the devices, possibly because the input energy is not easily measured and does not have a financial value. This is unfortunate, since human muscle-power has scarcity value (and thus economic value) as with any other kind of power, and is a critical resource for some farm families. Human mechanical work capacity is estimated at about 250 watt/hour (W/h) per day, and humans convert energy to mechanical form with efficiency about 10 percent on average. As far as can be determined, the efficiency of human-powered water-lifting devices is variable, even among the modern types. Some people are unwittingly wasting a great deal of their time and effort lifting water with inefficient devices when more efficient are available.

The simplest traditional type is the carrying of buckets, gourds or other containers by hand. Water is lifted and, at the same time, transported horizontally and sometimes distributed to crops. When it serves all these purposes, it may be an appropriate technology for some situations. However, it is never a very efficient way of lifting water. Slightly more efficient is the shadouf, a long-established device in many parts of Asia and northern Africa. Other devices are known in Asia and Africa, such as swing gourds and swing baskets. They can be reasonably efficient and very cheap, although their output per operator is low.

Probably the best-known group of modern human-powered devices is the wide range of treadle pumps. One type was developed in Bangladesh in the 1980s. It is very widely used there for relatively small suction lifts. Other types have been developed entirely in Africa. Most treadle pumps comprise two vertical-axis cylinders whose pistons are operated by two levers, called treadles, which one or two operators work with their feet. A mechanism causes the two pistons to operate alternately. The principle can be applied to a wide range of applications by varying the cylinder dimensions and the lever geometry: total heads range up to about 15 m, and discharges up to about 3 litres per second (l/s) (11 cubic metres per hour (m3/h)). Useful power outputs can range up to about 100 W. Some treadle pumps are efficient and relatively cheap, though some are surprisingly inefficient. Brochures and information sheets about particular treadle pumps apparently never state the input power needed or the efficiency, so it is effectively impossible for potential users to find out how efficient the different models are.

Hand pumps are familiar and widely used for small-scale water supply purposes, but are seldom used for irrigation. There is, however, a ground-level two-cylinder two-person hand pump, which can lift enough water for small-scale irrigation. It appears not to be as efficient as a good treadle pump but may be appropriate for people, who for cultural or other reasons, are not able or willing to use a treadle pump. A rower pump is another kind of low-lift hand pump; a few prototypes have been tested in Africa but are not used on any significant scale.

The rope and washer pump has an endless rope that draws a succession of round discs upward through a plastic pipe to lift water. In some models, developed and tested in Zimbabwe, the rope was driven by hand cranks, and the cost was kept low by using recycled materials such as old car tyres. This may be one of the more efficient water-lifting devices, partly because it does not involve a reciprocating motion or valves, but it is very seldom used for irrigation.

One of the great merits of treadle and other human-powered water-lifting devices is their low cost relative to the motorized types. Many development efforts are aimed at achieving cheaper versions. Various models are marketed at about US$50 - 80, while the more specialized, used for deep wells or high discharges, cost around US$160.

b) Animal-powered water-lifting devices

Animal power has been used for lifting water since ancient times, especially in Asia, though these devices are now almost entirely superseded by motorized pumpsets. Various ingenious devices have been developed in West Africa, especially in the French-speaking countries, but very few are still in service. Energy has a price, of course, in financial or economic terms: one animal's food is variously estimated to cost US$0.30 to US$0.50 per day at 1980 prices, or the crop residues of a hectare of irrigated land.

It appears from the sparse data quoted in Annex H that the animal-powered devices have relatively low efficiencies. They are also quite expensive and mechanically complex, even if made of traditional materials such as wood and natural fibres. In any continent farmers, who have access to animal power, are usually able to afford the more convenient and efficient modern alternatives; conversely, farmers who cannot afford modern motorized or renewable-energy devices usually cannot afford animal power. The Programme will need to review all known devices, but a large-scale development of animal-powered water-lifting in Africa is unlikely to emerge.

c) Motorized pumpsets

There are several types of motorized pumpsets available in West Africa that burn fossil fuels: mostly gasoline or diesel, but sometimes kerosene (see the Glossary for other names for these fuels, in French and English). Information about the pumpsets is fragmented and incomplete, and they are often poorly matched to their applications. Recent tests done by ANPIP in Niger covered useful heads of 4, 6 and 10 m at discharges of 2, 3 and 4 l/s. The pumpsets tested were working at between 3 to 40 percent of their rated power, which illustrates the degree of mismatching of equipment to application that is common on all continents. The one diesel pumpset tested used about four times as much fuel as an efficient pumpset would, while most of the gasoline pumpsets were using more than seven times more fuel than an ideally matched gasoline pumpset, and the best of them more than four times. These very low overall pumpset efficiencies are because the engines are running on part load.

Research in India, The Netherlands and Niger has shown that the fuel consumption of many pumpsets as used on small farms can be halved quite easily. In India, this was achieved on existing water-cooled diesel pumpsets by three simple measures. More recent research in Niger and Holland has studied gasoline pumpsets: even when an engine is running at a low fraction of its full-throttle power the fuel consumption can sometimes be halved by careful attention to the idling adjustment. These recent findings are isolated as research projects cover only a few sample devices. Therefore, information on efficiencies and fuel consumptions of motorized pumpsets is hard to find and there is no standard format for its presentation, or measurement.

In Niger the retail price of kerosene is 40 percent lower than that of gasoline, presumably the result of tax policy. As a result, considerable efforts have been made to investigate the use of kerosene as a fuel for small pumpsets. Many gasoline engines can easily run on kerosene once they are warm, but they have to be started on gasoline, so they have to be modified to have two fuel tanks and a fuel switching valve. There may be strong financial incentives for farmers to seek such engines, even though they cost somewhat more than normal gasoline pumpsets. From an economic (national) viewpoint, as opposed to the local financial, there may be little or no advantage in using kerosene instead of gasoline. Dual-fuel engines represent a dubious technical compromise in response to a distorted price situation.

Further research is already being planned to improve the efficiency of available pumpsets in West Africa, in a further phase of the ANPIP project in Niger. One initiative is the production and marketing of smaller engines, better matched to the common West African applications than those marketed up until now. There is also an initiative to use pumps that are more efficient.

In addition to ordinary diesel and gasoline-driven pumpsets, there are small numbers that use electricity for the transmission of power from engine to pump, via a generator and motor. These are usually for applications with a static lift of more than 7 m so that the pump needs to be below ground level. Technically this is a complicated and expensive compromise, but it may have advantages for a few users. Other motorized technologies include the Venturi or jet pump and the airlift pump, but low efficiencies make these unattractive for most applications. ANPIP has been associated with the development and testing of an engine-driven axial pump capable of delivering up to 18 l/s at very low heads, or 10 l/s at about 2.5 m suction head.

The purchase price in West Africa of a Japanese-made gasoline motorized pumpset, of about 1.5 to 4 kW design output, is usually in the range US$300 - 600, and a diesel is around US$990. Indian-made pumpsets tend to cost around US$180, while those made in China are considerably cheaper at around US$110. All these pumpsets are often used in applications for which they are seriously overpowered, resulting in unnecessarily high running costs. Not surprisingly, in view of the above-mentioned fuel consumption comparisons, there are various reports of big reductions in pumping cost achieved by improved matching of device to application. The tests carried out for ANPIP in Niger were accompanied by economic comparisons from the farmer's financial viewpoint. Those reported at the end of Abric et al. 2000, and those related in the Niger country report in Annex E, must be interpreted with caution because they contain a number of sweeping assumptions that may distort some of the comparisons. In particular, the useful lives of all engines are assumed to be the same.

d) Devices powered by electric motors connected to remote electricity supplies

In places where there is a public supply of electric power near farmland, it can be economically, financially and technically appropriate to drive small pumps by means of electric motors using the remote power supply. In West Africa, however, there are not many such places apart from parts of the Senegal valley. This type of water-lifting technology is little used for small-scale irrigation. It is, of course, more common for domestic water supply. Where such devices are used, the usual problems of matching the device to the application still arise, but they are not as significant as for the engine-driven pumpsets because electric motors are available for almost any power output. Submersible axial pumps are easy to use when the power is a remote electric supply, which makes this category especially attractive where groundwater is deeper than 7m.

The technology of any pumpset involving electric-power transmission is significantly more complicated than that of the ordinary engine-driven pumps. Safety considerations are important. For these reasons, as well as cost and the limited availability of public electricity supplies, this category of water-lifting device is not considered likely to have much of an impact on small-scale irrigation in West Africa.

e) Renewable energy devices: sunshine, wind or water power

These energy sources do not have the long-term and loss-free energy storage inherent in fossil fuels. The energy supply is therefore usually unreliable, while the equipment needed to capture and apply a useful amount of power to a pump is expensive.

Solar power is widely used for applications requiring relatively small power inputs at remote locations: telecommunications and small isolated potable water supplies are typical examples. Despite many years of intensive research attempting to develop cheap and robust solar energy gathering devices, they remain expensive relative to their power output. Both they and the associated equipment for bringing the energy to a pump or other load are quite delicate and sensitive. Experience of their use in remote locations for pumping potable water has been mixed, with pumpsets often out of operation for long periods awaiting repair or spare parts. Solar-powered devices must be kept on the list of potential technologies, hoping for future improvements in cost and robustness. In the mean time, they are not cost-effective for any but a few low-power and specialized applications, certainly not for water-lifting for small-scale irrigation.

Wind power has been used extensively for centuries to lift water, usually for pumped drainage in places with very flat land and persistent winds. Relative to their water-lifting output, both ancient and modern wind-powered devices are large and expensive in comparison with other currently available technologies. They tend to be unreliable, or at least to need a good deal of attention and maintenance. An additional factor is the regional and seasonal availability of strong winds. For much of the time, wind speeds are not very high over most of the cultivable lands of West Africa. Thus, although its scope as an intermediate technology must not be ignored, and some modern developments have improved outputs, the potential use of wind power for water-lifting in West Africa is unlikely to be large.

Water-powered water-lifting devices have been used for centuries. They can be cost-effective where there is a large flowing river or water dropping through a few metres. However, in West Africa this is seldom the case near land where water-lifting is needed. At least, if there were enough water dropping by a significant amount, irrigation would usually be by gravity.

All these renewable energy sources are in principle attractive to resource-poor people because the energy itself comes without financial cost or muscle-work. They should all be included in any inventory of relevant technologies. Nonetheless, the high initial costs, and the limitations mentioned here and in the Annexes, mean they are unlikely to have a positive impact on livelihoods in West Africa in the next few decades. Research into these technologies continues, because of the commercial and economic potential for developed countries. Therefore, it would not be cost-effective to pursue fundamental research into such technologies especially for West African water-lifting applications. All that can be usefully done in this context is to keep technological progress elsewhere under review, so that if any promising technical developments occur they can be identified and field-tested in West Africa.

f) Biological energy sources

Energy can be obtained from biological sources incidental to other activities, in particular from gases produced by the bacterial decomposition of animal and human excreta, sometimes also from plant residues. While, in some circumstances, this may be a cost-effective and sustainable energy source for heating, its use for mechanical power is not expected to be useful for West Africa in the near future. The use of agricultural crops to make fuel for heat engines can be regarded as a biological or renewable energy source. These energy sources should be kept on the list of possibilities for water-lifting in West Africa, though the true economic opportunity cost of the fuel (for instance in its alternative use as fertilizer or foodstuff) must always be taken into account. As with the sources discussed in the previous section, they should be kept under review by observing research carried out in other places and for other sectors.


One of the main aims of the proposed Programme is to provide users and choosers[2] with the information and skills required for choosing the most appropriate water-lifting technology for their particular situations and applications, using these three words in the sense introduced in Chapter 2.

A user is primarily interested in irrigation cost (in terms of both money and effort), and in this context, specifically, in what a certain technology will do and what it will cost over its lifetime. The balance between these factors represents the return or advantage to be gained by use of the technology. Differentiating more finely, the main criteria would normally be:

Obviously, the relative importance of these and other criteria will vary according to the user's situation. However, any initiative to promote informed choice of technology must cover them all. For motorized pumpsets of about 2 to 40 kW size, a useful tool is already available in the form of the spreadsheet PUMPSELECT.xls prepared by the HIPPO Foundation and recently introduced in roving training courses in West Africa (van't Hof 2001c). Although the Foundation is severely hampered, having to rely on sparsely available information on the energy consumption and efficiency of complete pumpsets working in field conditions. Engine manufacturers may publish nominal specific fuel consumption figures but these usually relate only to particular conditions of altitude, load and speed. Pump makers publish characteristic curves, but they refer only to the pump, and experience shows that other parts of the pumpset, especially the foot-valve, may have a drastic effect on energy consumption.

For human-powered devices there does not appear to be any comparable database or selection aid, indeed it would be impossible to compile one because of the lack of complete and mutually comparable descriptions of the devices available. Manufacturers do not publish energy input figures, and even research projects have very rarely measured them.

Currently the above criteria can not be applied, because there is a lack of information for many of them, as well as a lack of understanding of the simple technical principles.

[2] From the subtitle of Fraenkel (1997).

Previous Page Top of Page Next Page