Cover Image
close this bookWater Management in Africa and the Middle East: Challenges (IDRC, 1996)
close this folderPart I - Concepts
View the documentDemand-side Management, Conservation, and Efficiency in the Use of Africa’s Water Resources
View the documentAllocation of Water Resources in Africa: Potential for Moving Water in and out of Agriculture
View the documentWomen, Men, and Water-Resource Management in Africa

Demand-side Management, Conservation, and Efficiency in the Use of Africa’s Water Resources

Geoffrey Stiles

Energy Conservation Advisor, SADC Energy Management Project, Harare, Zimbabwe

Introduction

This paper provides a conceptual framework for improved demand management of water resources in Africa based on the experience of demand-side management (DSM) of energy resources in Canada and the United States. (The difference between demand management and DSM is explained in the section “Demand-side management for water: a basic paradigm.”) For examples of efficient water use and for an overall view of water management in Africa, the paper focuses on sub-Saharan Africa. North Africa is discussed only in relation to the experience of irrigation on the Nile. The paper also concentrates on the use of water for human consumption and excludes discussion of water’s role as an ecological entity (for example, its use for habitat preservation).

Experience from energy-demand management provides a useful analogy to demand management of water resources, even if incomplete. As Brooks (1994) pointed out, the analogy works despite its imperfections: water is often oversupplied relative to demand, generally underpriced relative to its intrinsic and economic values, and governed by “institutions geared to augment supply rather than to manage demand.”

Indeed, the analogy may be particularly useful for Africa, where a relatively high proportion of energy comes from water sources (that is, hydroelectricity) and many major water projects are dual-purpose (they provide water both for consumptive uses, such as irrigation, and for nonconsumptive uses, such as electricity generation).

There are a number of parallels in the use of water and energy resources that should be noted:

· Water, as with energy, is commonly transmitted from its point of collection (generation) to its point of use, and it is in the transmission process that substantial losses (inefficiencies) are typically incurred. Moreover, for both resources, means of transmission are extremely variable. In the case of water, they range from fairly simple manual technologies for gravity-based irrigation canals to extremely complex capital-intensive transmission by pumps and pipelines. In water, as in energy transmission, the transmitter and the user are faced with important technological choices that may affect overall system efficiency.

· The end-uses of water, as with those of energy, typically require some form of transformation technology. This can range from a simple faucet or showerhead in a domestic reticulation system, to canals and spray irrigators in agriculture, to boilers and purifiers in industry. Although the parallel to energy is incomplete because the transformation is not necessarily thermodynamic (Brooks 1994), it is close: water must often be changed in form, as to steam or hot water, or in force or pressure before it can be used effectively.

· Water, as with energy, is a critical resource and lends itself to the imposition of control and authority. This point was made by Wittfogel (1957) in his classic study of “hydraulic civilization.” Wittfogel was preoccupied primarily with control over the supply of water, but it is equally important to consider how demand might be controlled and what role water providers can take in efforts to improve demand management of water.

My principal source of information on the energy side is the DSM experience of power utilities in North America and Europe. Energy DSM provides special insights into how user needs can influence the resource management process and resource management authorities in particular, a process that can lead to a profound change in financial management and investment patterns. There are also cases from North America where DSM has been specifically applied to the management of water.

Water use and demand management in Africa

Water use in Africa, as in other developing regions, is dominated by agriculture. Indeed, Africa has the highest percentage use by agriculture of any major region and, conversely, the lowest percentage of domestic or industrial uses (Xie et al. 1993). As a rule, countries increase their relative share of domestic and industrial use as they become more urbanized and more “modernized” or as income levels increase (World Bank 1992). Thus, the dominance of agricultural end-uses in Africa indicates that domestic and industrial supply systems are relatively undeveloped, with a few notable exceptions like South Africa, Egypt, and Zimbabwe. The extremely rapid increase in Africa’s urban population has done little to change this, and most rural-urban migrants end up in informal settlements with dangerously inadequate water supplies. There is also little capital available to pay for the further development of industrial-commercial water supplies in burgeoning African cities. Because of this, Africa has been the target of a number of massive and relatively well documented efforts at water management, particularly those involving riverbasin transfers. Africa also provides examples of the use of water-efficient technologies and management systems in all sectors.

Water end-uses

Commercial or market-based agricultural use

It is estimated that 88% of water in Africa is used for agriculture (Xie et al. 1993). The area under irrigation is about 5.3 x 106 ha, of which 2.6106(53%) is for commercial or large-scale agriculture (Wyss 1991). Commercial irrigation schemes include irrigation from pumped groundwater, irrigation from large dams, and various kinds of river transfers to irrigation systems. The balance, mostly traditional or small-scale irrigation, consists of private flood, swamp, surface, and low-lift irrigation (Wyss 1991). Most river-based schemes are used for capital-intensive farming, with production primarily for the market, although a surprisingly large amount is used for small-scale farms in countries such as Sudan (OECD 1984; World Bank 1990).

Although these figures give the impression that irrigation is relatively common in African agriculture, only 5% of total crop land is, in fact, irrigated (Grainger 1990; MacLean and Voss, this volume), compared with 29% for Asia. This is to some extent a function of the peculiar geography of Africa, which includes inaccessible aquifers in West Africa, making groundwater irrigation extremely expensive, and unreliable surface-water flows elsewhere, making other kinds of irrigation highly risky (Grainger 1990). Rivers in many parts of the continent run dry for long periods of the year because they originate in drought-affected highlands. The impact of these erratic flows is increased by devegetation and the resultant low capacity of the soils to absorb water. It is in these relatively marginal and seasonally semi-arid conditions that a majority of the continent’s farmers, both commercial and traditional, attempt to make a livelihood.

Despite these adverse conditions, there is great potential for increasing irrigation in sub-Saharan Africa (MacLean and Voss, this volume). The irrigation potential of this subregion is estimated at 20-33 x 106 ha (Elahl and Khushalani 1991; Bryant 1994), although much of this is land that could only be irrigated after substantial capital investments. The additional irrigation that might be achieved, given market constraints and limited capital, is probably less than half this amount.

By far the largest amount of large-scale commercial irrigation for farming is in Sudan, where a number of projects have had a massive impact on agricultural production. Good examples are the Gezira and Rahad schemes, which, together, cover several thousand square kilometres. The overall environmental and economic impacts of such schemes are extremely contentious, and, despite major increases in production of commercial crops, the irrigation systems feeding the main growing areas have been subject to erratic management and poor maintenance. As a result, crop yields actually declined in the years following initial development (Elahl and Khushalani 1991). Ironically, much of this farming is small scale, with farms typically less than 8 ha in size and operated by families and small businesses. The capital expenditure for these schemes has been vast, of the order of 320 million United States dollars (USD) for the most recent scheme at Rahad. Many experts, as well as the farmers, question the wisdom of these investments, given the difficulties that have ensued.

The major concentration of irrigated commercial farming activity in sub-Saharan Africa is in the south and east, particularly in South Africa, Zimbabwe, Namibia, and Kenya. Southern African countries have introduced extensive irrigation, primarily from groundwater sources, to an essentially arid and semi-arid environment. Average rainfall in South Africa, for example, ranges from 1 200in coastal KwaZulu-Natal, to 400 mm in the Orange Free State, to 100-200in the northern Cape Province. The average for the country as a whole is 500 mm. Evaporation rates for open water systems in these areas are 1 500-2 500 mm/year (up to 4 000 mm/year in Namibia). Reliance on rainfall alone, therefore, is insufficient to conduct this kind of farming (Davies and Day 1986). As a result, complex, capital-intensive systems of riverbasin irrigation, rainwater dam, and groundwater (borehole) irrigation have been developed. These systems are fairly energy intensive, relying on electrical (in some cases, wind-driven) pumping systems to bring the water from source to end-use. Riverbasin and rainwater dam irrigation is, of course, subject to high losses through evaporation and leakage; groundwater (borehole) irrigation is typically associated with open spraying, which also leads to significant evaporation losses (Davies and Day 1986).

Riverbasin irrigation is widely practiced in South Africa, especially in the Vaal and Orange river systems. Much effort has been devoted to the rationalization of these systems. The latest is the massive Lesotho Highlands Water Project, which will theoretically result in increased efficiency on the supply side through improved storage at source, in the high plateau of Lesotho, where much of the Orange River catchment is located.

The Nile, the major example of riverbasin irrigation in Africa, has been particularly heavily used by Sudan (Adams 1992). However, the people around the Nile still depend largely on traditional floodplain irrigation, rather than on commercially piped or channeled irrigation. Comparable use of river-based irrigation can be found in connection with major dam developments like Akosombo Dam in Ghana, Kossou Dam in Cd’Ivoire, and Aswan High Dam in Egypt.

In many cases, development of dams has been driven primarily by energy needs, rather than by irrigation needs. Three of the largest hydroelectric developments in Africa, namely, Kariba in Zimbabwe-Zambia, Kafue Gorge in Zambia, and Cabora Bossa in Mozambique, have resulted in little commercial irrigation because of the low population densities and relatively poor agricultural soils in the areas adjacent to the impoundments. There is, however, substantial informal use of waters from Lake Kariba on the Zambian side. Although only about 6000 ha at present, the area could be expanded four- or fivefold if plans to initiate commercial wheat farming upstream of Victoria Falls are implemented (Kasimona and Mkwaya 1995). The proposed use of the Zambezi’s waters above Lake Kariba for municipal, agricultural, and industrial water requirements is now a major area of contention (see “Industrial use”).

Traditional or communal agricultural use

Although dominated by commercial farming, agricultural use of water in Africa also includes a large number and range of indigenous or traditional forms of water distribution for cropping:

· seasonal diversions of river flooding along the Nile and in smaller rivers such as the Pongola in northeastern South Africa;

· smaller-scale and less-commercialized floodplain irrigation along rivers like the Hadejia and Jama in Nigeria and the Zambezi in the Lozi tribal areas of Zambia; and

· small-scale irrigation systems based on river flow or on small rainwater dams in many countries, including Nigeria, Madagascar, Tanzania, Senegal, Mali, Sierra Leone, Sudan, Burundi, Chad, and Somalia.

Although small-scale irrigation is usually thought to be minor in extent and impact, it actually constitutes almost 50% of irrigation in Africa by land area, or 2.38106out of an estimated total of 5.02106 ha (FAO 1986). Such forms of irrigation are indigenous in two senses: many predate colonialism, and most are locally controlled.

Adams (1992) argued that these traditional systems of water catchment and distribution are typically as efficient as and commonly more conservative than large-scale systems and therefore deserve far more attention from planners and financial agencies. Whether this is so depends on a range of variables: the maintenance provided to the system, the types of crops being irrigated, and the methods used to carry water to the crops. Adams’ use of efficiency is probably closest to what Xie et al. (1993, p. 5) called “conveyance efficiency,” that is, efficiency in carrying water from the source to the distribution system. Whether such traditional systems are inherently more efficient in terms of the local provision and end-use of water (or water distribution and application, to use the terms of Xie et al.) is a moot point, especially in view of the controversy over the extent to which irrigation is a consumptive use of water (Seckler 1993).

Municipal use

Municipal uses of water include supply to the domestic sector, as well as to commercial buildings and to washing facilities. Municipal use represents the second largest use (though a distant second) in Africa, at 7% (Xie et al. 1993). However, this use ranges from 81% in Equatorial Guinea, 72% in Gabon, and 63% in Zambia to less than 1% in Madagascar and Sudan (Xie et al. 1993). Generally speaking, the higher percentages are probably a function of conditions (such as war in Mozambique or decommercialization of agriculture in Zambia) that have reduced agriculture’s dominant role. The lowest percentages, on the other hand, are a function of the very dominant role of irrigation agriculture in countries such as Sudan.

Urban water-reticulation systems in most African cities were originally developed by colonial governments and sized to meet the needs of a much smaller population. Extending water supply to rapidly expanding urban and peri-urban communities populated largely by rural migrants and workers has proven a daunting challenge. As an example, domestic water demand in South Africa is expected to grow by about 2% per year over the next 30 years: if this rate of growth is realized, demand in 2020 would be three times that of 1970. This would also increase the share for domestic users: by 2010, more than 17% of South African water use would be domestic, compared with 9.3% now (South African Department of Water Affairs 1986). This trend will, if anything, be strengthened by the recent decision of the new government to provide substantial low-interest loans and subsidies for low-income housing, which will inevitably result in a substantial increase in serviced lots and, correspondingly, in requirements for centrally supplied water.

As a rule, municipal water in African cities, as in European and North American ones, is supplied from open reservoirs fed by dammed rivers or by direct rainwater accumulation. In arid and semi-arid areas, there are also numerous examples of groundwater- or borehole-based municipal supplies, although these usually augment reservoir supplies. The most striking example of the use of boreholes to supply a major city is Windhoek in Namibia, which has also recycled sewage for many years as a water-conservation measure.

Municipal water supply in Africa illustrates perfectly the dilemma created when there is competition for use of an underpriced resource. Few African countries have been able to deal with the rising demand for safe, reliable municipal water supplies because they have already allocated much of the river-based and groundwater supplies in surrounding regions to agriculture and, in a few cases, to industry. Because these bulk uses are often underpriced relative to the long-run costs of system development or even to the short-run costs of operation, governments do not recover sufficient revenue to pay for the very large investments required for expanding domestic and commercial-sector reticulation systems. As a result, several countries are experiencing substantial urban water-supply crises.

The problem of domestic water supply in rural areas is entirely different. In West Africa, there are numerous examples of rural domestic water being supplied from irrigation systems (rivers or lakes). In South, Central, and East Africa, however, most rural villages depend either on boreholes or, much more commonly, on water from nearby rivers or lakes. Tanzania is an exception: a number of small, traditional irrigation systems in areas such as Kilimanjaro and Lake Natron supply both domestic and agricultural needs (Adams 1992).

Industrial use

Industrial use of water is still a relatively minor factor in Africa. Estimates place total African industrial use at 5%, the lowest among the six major regional groupings of nations. However, this small average masks substantial variation, from 25% in Zaire and 22% in Lesotho to virtually nil in Sudan, Somalia, Burundi, and Madagascar. Further, industrial use is likely to grow faster than use in other sectors as African countries undergo major economic changes over the next 10 years. Because industrial water use is generally more efficient than agricultural use, this shift may presage an overall improvement in efficiency, but it is unlikely to signify a reduction in overall water use (Adams 1992).

By far the most significant demand for water from the African industrial sector comes from mining and metallurgy (smelting and refining). Coal mining in Mozambique has been estimated to require up to 4 m³/t or 1 m³/s of water in the mining and washing processes combined (David 1988). Large diamond mines, such as that at Orapa in Botswana, and aluminum smelters, such as that at Richards Bay in South Africa, use substantial volumes of groundwater in the processing of minerals. Use of water at Orapa is large enough to reduce water levels at the major source of this groundwater, the Okavanga Delta.

In some countries, like Zimbabwe, Kenya, Botswana, Namibia, Tanzania, and South Africa, water is used to generate steam, both for electricity and for industry, and to cool coal- or oil-fired generators. Inasmuch as many generators are located in semi-arid areas, most of this requirement comes from groundwater. The South Africans have, however, pioneered in the use of air-cooled generators and have succeeded in substantially reducing demand on local groundwater supplies in arid areas such as the eastern Transvaal (Davies and Day 1986).

Hydroelectric power constitutes a small but important end-use in Africa. A number of major riverbasin developments are tied to hydroelectric dams, including those on the Volta, Zaire (at Inga), Orange, Ruacana, Kafue, Zambezi, and Nile rivers. Several of these projects resulted in substantial back-flooding and reservoir formation and simultaneously reduced the impact on agriculture of traditional seasonal flooding downstream of the dam. Yet, in general, it may be said that the use of water for hydroelectric power is still in its infancy in Africa. The World Bank estimated that new projects, in various stages of planning, could collectively increase hydropower’s contribution to Africa’s energy supplies by a factor of seven, assuming there is increased integration of power transmission systems to justify the increased investment in hydroelectric generating capacity. Although these figures are probably excessively optimistic and may underestimate the cost of delivering electricity from remote sites in politically unstable countries such as Zaire, it is clear that the use of riverbasins for hydroelectric power will increase substantially in sub-Saharan Africa over the next 25 years.

Conclusions

Because water use in Africa is so heavily dominated by agriculture, it is appropriate that concerns over demand management focus on this sector. Opportunities for reducing agricultural demand are considerable, particularly for large-scale commercial farms, and agricultural use of water on such farms is extremely responsive to price.

On the other hand, agriculture’s pivotal role in many African economies also makes it a target of political strategy. This is particularly true for traditional or communal agriculture. African governments, assisted in many cases by donors and multilateral banks, have spent large amounts of scarce capital on dam and irrigation projects that are essentially supply-side initiatives. By world standards, the costs of these initiatives are extremely high, ranging from 1 000 USD/ha for smaller, locally funded projects to 8 000-20 000 USD/ha of reservoir surface for large-scale capital projects (Wyss 1991). Although these initiatives have done little to alleviate water shortages and, in some cases, may actually have worsened them, it seems unlikely that governments would voluntarily place the burden of reducing demand on the user, rather than simply seeking more support for supply, no matter how promising the prospects for correcting the water crisis. Pressure for improvement in DSM in agricultural use must come from outside: from local pressure groups and donors.

The situation for industrial and domestic uses is more positive. These sectors continue to have a very small impact on African water balances, although their role is growing steadily. Indeed, in a few countries, notably South Africa, Namibia, Zimbabwe, Sudan, Ethiopia, and Egypt, the crisis point has already been reached in the larger urban areas. This crisis has been brought about by a combination of increased industrial uses and the pressure from burgeoning urban and peri-urban townships for water access. Bulawayo, Zimbabwe’s second city, is an example. The city’s continuing water shortage over the past several years has severely affected both domestic and industrial use and has lately produced a shift of population and industry away from the town. Solutions to the crisis centre almost exclusively on water-supply megaprojects, and virtually no efforts have been made to deal with the question of demand, except through emergency quotas and penalties for overuse.

Nevertheless, crises of this sort have had a salutary effect on planning in some countries: the crises bring water to the forefront of public debate and help focus discussion on water needs, or at least on perceived needs. Though these crises may be solved by patchwork supply increments, there is some hope that realization of the immense costs of long-term supply-based solutions will force reexamination of the need for demand-based solutions.

Africa’s water-management crisis

Africa’s water-management crisis is well documented. At least 10 African countries, virtually all in North Africa and the Sahel, were officially classified by the World Resources Institute (1992) as “water-stressed.” It is expected that at least six more will be added to this list by 2025 (Bryant 1994). Moreover, these figures exclude South Africa, where the sources of stress are both overuse and environmental limitations, and those other countries of southern and Central Africa that have recently experienced severe drought conditions (Zimbabwe, Zambia, Botswana, Mozambique, Malawi, and Namibia) and are usually classified as semi-arid. Taken together, well over half of the countries in Africa have significant water-access problems stemming from a combination of overuse, poor management, and adverse climatic conditions. Indeed, some authors argue that per capita water supplies in Africa have actually declined by as much as 50% since 1950 (Bryant 1994).

Much of this crisis centres on growing demand for agricultural water. Most African countries depend heavily on agriculture as a source of income for rural populations, and many have also developed substantial export markets for products such as tea, coffee, sugar, cocoa, rubber, tobacco, and flowers crops that require reliable and abundant water supplies and earn substantial foreign currency.

The crisis of African water management cannot, therefore, be solved simply by reallocation of uses among sectors, that is, from agricultural to domestic or industrial uses. Nor can it be solved by increasing supplies: in most water-stressed countries, the major readily accessible sources of water supply have long been identified and, for the most part, developed. There are plans to develop a few others soon. The sole remaining solution is to limit demand without seriously degrading the quality of life of Africa’s burgeoning populations. The prognosis is not good. As Davies and Day (1986, p. 120) pessimistically argued for South Africa, failure to deal urgently with the issue of water demand and efficiency could have dire consequences:

It is generally assumed that the ‘best case’ we can realistically expect, a combination of the greatest possible use of surface resources, the slowest population growth and negligible exploitation, will bring the drought on for good in 2020.

The experience of developed countries versus that of developing countries

Searching for solutions to this crisis inevitably brings us to compare the recent experience of developed countries with that of developing countries. However, the comparison is not necessarily a helpful or instructive one: developed countries have mismanaged their water resources just as much as developing ones, and in many arid areas of the developed world, the water-supply situation has also reached a crisis point.

The advantages that developed countries have over developing ones are, of course, their stronger revenue base and better access to efficient technologies. Most western countries have well-developed agricultural and domestic water-supply systems supported by a combination of user charges and general tax revenues. Cost-reflective pricing is still a rarity, particularly for agricultural users, although it is attracting increasing attention. As Frederick (1993, p. 22) showed for the member countries of the Organization for Economic Co-operation and Development, “markets and prices have rarely been the principal or even important mechanisms for allocating water resources.” In North America, agricultural (irrigation) tariffs remain well below cost-recovery levels, despite major conservation efforts.

Several studies have identified the need for a demand-based approach to water management in the industrial countries. These studies describe numerous efforts, particularly by the United States, to increase the efficiency of water delivery and use. For Canada, Brooks and Peters (1988) suggested a water-demand management strategy that would first identify sectors where water supply was becoming a problem, then try to remedy the barriers to demand management, and finally derive a sector-by-sector program combining regulatory and pricing measures with marketing and technology-transfer measures. Their study showed that the identification of barriers to demand management is particularly important:

· Economic barriers involve the undervaluing of water by consumers, underpricing by suppliers, and failure to charge self-supplied users for water rights.

· Market failures include lack of adequate information on conservation opportunities, lack of access to affordable capital, unnecessarily restrictive investment criteria, general indifference to the cost of water, and separation of the costs of conservation from its benefits.

· Organizational barriers are characterized by a preoccupation of authorities with supply and with engineering solutions and by concern over the financial risks of demand management.

These barriers are virtually universal and could be applied with minor adjustments to developing countries in Africa. However, there is an overriding difference between experiences in northern industrial countries and those of countries in Africa. The former countries are involved in minor adjustments to relatively prosperous market economies in which there is an established pattern of treating most resources as commodities. The latter countries, on the other hand, are in various stages of transition to a market economy in which resources, traditionally regarded as public or communal, are only beginning to be treated as commodities. In the imperfect and transitional market economies of many African countries, the economic and market barriers are truly substantial and diverse. Nevertheless, many of the same problems exist: lack of information, lack of capital, low or indifferent pricing, traditions of usufruct, and high financial risks.

The experience of Africa

The African continent has generated a number of significant endeavours in water-demand management, although the bulk of effort has been concentrated on the improvement of conveyance, rather than on distribution or end-use efficiencies. The following sections discuss four areas where demand management either has been practiced in the past or is being considered for the future:

· management of riverbasins;

· management of municipal water supplies;

· management of irrigation sytems; and

· use of pricing as a demand-management tool.

MANAGEMENT OF RIVERBASINS - Riverbasins should be a major target of demand-management activities in Africa. As Adams (1992) noted, virtually every river in Africa has now been dammed for either agricultural (irrigation) uses or electricity generation or, more often, both. The Zambezi, the Kafue, the Senegal, the Nile, the Vaal, and the Niger and its tributaries have multiple dams, and there are plans for more. From a planning perspective, Africa has a surprising number of cases where riverbasins affect more than one nation. Of the 13 cases in the world where five or more nations form part of a river or river-lake basin, Africa has 5: the Niger, Nile, Zaire, and Zambezi rivers and Lake Chad (Gleick 1989).

As a general rule, damming of major rivers in Africa has taken place with minimal international (or regional) cooperation and, hence, with little attention to either downstream or upstream impacts. Nevertheless, although demand management has mostly been neglected, sevearal efforts have been made to reverse this historic tendency and institute multinational planning systems to improve the efficiency and minimize the negative impacts of water use in riverbasins. Examples are the Zambezi Action Plan (ZACPLAN) in the Zambezi basin area, the Volta River Authority, the Niger Development Board, and the Nile River Water Agreements.

The consolidation of water-supply management in riverbasin authorities naturally involves a political element, just as it does with agricultural water uses in Africa. As Scudder (1989) pointed out, economics invariably takes second place to political considerations in riverbasin projects, with the result that supply concerns tend to take precedence over demand concerns. The ZACPLAN omission of demand considerations perfectly illustrates this: neither the donors nor the governments involved in this ambitious scheme wish to consider the intractable problem of whether the demands for Zambezi water use are legitimate and have been accurately quantified, and neither the donors nor the governments have considerated mechanisms for limiting this demand.

MANAGEMENT OF MUNICIPAL WATER SUPPLIES - In some respects, domestic water supplies in Africa represent a more tangible crisis than is found in any other sector. The reason for this is simple: although agricultural and industrial uses may be under more objective threat, scarcity of water for households affects the population, particularly the middle-income population living in cities, more directly. Disproportionate attention is therefore given to the water crisis in municipal areas.

Solutions to the problem of domestic water supply in arid and semi-arid or otherwise water-stressed areas have been many and varied. On the supply side, the solutions include the following:

· expansion of conventional supplies through increased damming of rivers and streams;

· development of boreholes on a large scale, typically in combination with damming; and

· interbasin transfers.

Supply-side solutions of this sort are still being considered and actively planned in many African urban centres. The burgeoning peri-urban populations resulting from rural migration have created a need for substantial new supplies. This, in turn, has prompted urban planners to look for new sources. For example, Bulawayo in Zimbabwe has already utilized all of the available river-based sources; six major dams serve the 350 000 residents of this city. The alternative, using boreholes drawn from aquifers, has been exploited as an emergency option, but this is not sufficient to meet the expected needs of the city. As a result, effort has shifted to megaprojects, such as the proposed Zambezi-Bulawayo pipeline.

On the demand side, solutions include the following:

· establishing seasonal quotas and usage restrictions;

· rationing by time of day or area;

· recycling domestic and, if feasible, industrial wastewater, including sewage; and

· setting price penalties.

These actions apply primarily to urban water use. Water use in rural areas outside organized municipal authorities is, of course, less, both per capita and absolutely, and is less likely to be subjected to restrictions, although many rural areas in Africa are certainly water stressed.

Some of these solutions are practiced to a certain degree. For example, the development of quotas is restricted primarily to emergency situations, such as those experienced by municipal suppliers in relatively developed cities like Johannesburg, Harare, and Nairobi. Time-of-day rationing also occurs, primarily in emergencies similar to that caused by the 1991-92 drought in Central and southern Africa. However, such rationing is becoming fairly commonplace in many cities in areas of the continent that have seasonal rain: because local reservoirs and dams are only rarely recharged to their full extent, they require substantial restrictions on use toward the end of the dry season. Very few examples of price penalties are found in Africa.

MANAGEMENT OF IRRIGATION SYSTEMS - Efforts to manage African irrigation systems for greater efficiency are numerous in both large-scale commercial and small-scale traditional forms. Recent reviews of this subject by the World Bank show conclusively that management of large-scale commercial irrigation schemes is problematic in much of sub-Saharan Africa. On the whole, small- and medium-scale schemes that are locally managed and employ traditional methods tend to fare better in terms of both economic efficiency (that is, rate of return) and water-use efficiency. Lele and Subramanian (1991, p. 60), for example, suggested that, in contrast to large-scale schemes in Africa,

small-scale irrigation has had largely positive experiences and this offers considerable untapped potential for expansion involving both public and private investment.... Where the private sector is involved, an effort should be made to decentralize irrigation systems and make their clients directly accountable for them.

The difficulty with this argument, as the authors themselves admit, is that few governments have been willing to forego investment in large-scale schemes simply because they are inefficient. Large-scale systems are more amenable to central-government control and, therefore, more attractive to the entrenched urban bureaucracies of many African states. Moreover, involvement of the private sector is not really a pertinent issue in many countries. Small, locally managed irrigation systems are more often “public” than “private” in terms of effective ownership and management philosophy. Nonetheless, the localization of control still confers significant efficiency benefits. It is important not to conclude, as the World Bank has done in several studies, that private ownership of the resource or private-sector management of its distribution will necessarily bring about increased efficiency (Wyss 1991).

Adams (1992) presented a very cogent argument for the inherent efficiency of small-scale, traditional irrigation systems. He showed, for example, that small-scale furrow irrigation in Tanzania by the Chagga of the Kilimanjaro region and by the Sonjo of Lake Natron is very carefully managed and that, because of this, the Tanzanian government has been able to use traditional irrigation as a basis for further technical improvements, rather than replacing it with modern large-scale systems. Similar examples of productive traditional irrigation systems may be found in the rainwater-harvesting techniques of northern Kenya, the use of terraces for erosion control in Burkina Faso, and the Molapo Development Project of the Okavango Delta in Botswana (Adams 1992). Adams cautioned that intervention in such indigenous or traditional irrigation and water-management systems has drawbacks: for example, altered perceptions of ownership and responsibility, increased economic uncertainty for smallholders, and increased dependence on outside support and technology. The solution to this is not to leave indigenous systems alone on the premise that they work better without interference but, rather, to insist that “integrated resource management” be followed and that information about new technologies and training in their use be provided as part of the development package (Adams 1992).

Much the same argument could be applied, of course, to virtually all aspects of irrigation development and rehabilitation. Locally oriented solutions are preferable because of the high cost and economic risk associated with externally imposed solutions and because of the “hidden hand” of environmental damage and possible loss of user control over water management. On the other hand, uncritical acceptance of local solutions is little better: a large number of small-scale irrigation systems can be just as environmentally damaging in the aggregate as a few large-scale ones and may be less efficient in their application than they could be with appropriately scaled technological improvements and proper information and training.

The private-public argument deserves closer scrutiny. In southern Africa, particularly, the confrontation between private (commercial) and public (subsistence) agriculture has been a major political and economic management issue. Because most of the countries of this region are arid or semi-arid, the issues of access to and management of water resources are a central part of the debate. Commercial agriculture has thrived in countries like Zimbabwe, Namibia, and South Africa in part because of substantial public investment in irrigation infrastructure and support systems. This investment includes government-funded programs for damming of major rivers, provision of free technical advice and water monitoring, and subsidized loans for boreholes and private dams. Indirect support has been provided in the form of price subsidies to farmers for the products of irrigated agriculture, along with price controls on the raw materials they require to develop this agriculture. The result has been an economy with built-in distortions favouring commercial (mainly white, male) farmers over communal or traditional (mainly black, female) ones. The latter do receive some of the benefits of agricultural extension work, but investment, particularly in irrigation, has been minimal, except recently in some communal areas of Zimbabwe and Zambia (Kasimona and Makwaya 1995).

The pertinent issue for this paper is whether disproportionate investment in the modern or commercial sector has increased or reduced the efficiency of agricultural water use in the countries concerned. The answer is not simple. For example, the commercial sector’s use of large-scale spray irrigation for tobacco, maize, wheat, and other cash crops has resulted in higher evaporative losses than occur with, say, drip irrigation. On the other hand, evaporative losses are low compared with losses from flood irrigation or even canal irrigation. In addition, some field crops, such as wheat, cannot be effectively irrigated by any other method. Still, the investment and operational costs of spray irrigation are very high, and the successful operation of these systems requires careful and informed management. Davies and Day (1986), for example, argued that efficiency levels for spray irrigation are of the order of 30% and that the low efficiency of these systems is compounded in South Africa by poor maintenance, erosion damage, and salination effects. Others have shown that the on-farm efficiency of spray irrigation in Israel and other parts of the Middle East can be as high as 75-80% and in hot, dry areas with maximum evaporation loss can be as high as 60% (Schwarz 1991).

As these discussions suggest, investment in spray irrigation, drip irrigation, or improvements to irrigation-water distribution (for example, by lining canals) are capital intensive. They almost always require some degree of public investment, whether the eventual beneficiary is a private commercial farmer or an indigenous communal farmer. In southern Africa, governments have supported major public-investment campaigns to help the commercial farming sector be competitive internationally. These governments have invested in irrigation and other water-supply measures as a central part of this economic philosophy. Correspondingly, these governments either have failed to invest or have invested on a much smaller scale in water-management schemes to support the communal farmer, and political confrontation, economic mismanagement, and adverse ecological impacts have ensued. In the Save River catchment area of southeastern Zimbabwe, for example, traditional dependence on floodwater irrigation, coupled with massive government-induced increases in the local indigenous (smallholder) farmer population, has resulted in severe degradation of the entire river system. The river now flows (when it flows at all) well below the soil surface and is effectively inaccessible. In northeastern Zimbabwe, near the Nyanga Mountains, development of large irrigation dams has pushed traditional farmers (for whom the irrigation was intended in the first place) into marginal areas because there is insufficient cultivable area near the new dams. Nevertheless, small-scale commercial farming, using a combination of groundwater (boreholes), summer-flow storage dams, and off-river flow for irrigation, has been relatively successful and efficient in Zimbabwe, as documented by ter Vrugt (1990).

Other examples of efficient irrigation based either on technological improvements or on better management and operational control can be found in Africa. Xie et al. (1993) summarized the essential factors in efficient management:

· ensuring a reliable water supply to the end-use point;
· assessing soil characteristics and plant requirements;
· improving management skills;
· improving maintenance;
· using surface and groundwater conjunctively;
· disseminating information on efficient technologies and techniques; and
· implementing demand management through legislative and administrative intervention.

To this list, Grainger (1990), who has studied the effect of improved water management on the desertification process, would add four items:

· improving project design to ensure that drainage is provided for and farmer needs are being met by the design supply;

· expanding farmer involvement in the design and implementation phases;

· promoting more attention to rehabilitation over new construction; and

· increasing use of small-scale approaches that are within the competency and scale of small farmers and that are effectively decentralized in development and management.

USE OF PRICING AS A DEMAND-MANAGEMENT TOOL - The use of pricing as a management tool is relatively underdeveloped in Africa. Although most irrigation systems pass the cost of water on to the consumer in some form or other, as do most domestic and industrial water-provision systems, the tendency is to underprice this resource or to use price to control access but not the level of actual use. Correct pricing of water resources, therefore, presents a significant opportunity for DSM.

The most common pricing strategy in water management, as in energy management, is to cross-subsidize among various consumer categories. As a general rule, such subsidies take the form of overpricing domestic water use on the premise that domestic users can more easily conserve than agricultural or industrial consumers and are more likely to be able to afford such an increase without serious stress. The equity implications of this could be considerable, but, to the extent that low-income domestic consumers get their water from central water sources (communal pumps and pipe outlets) and, therefore, do not pay directly for the resource, adverse equity effects are mitigated.

In Zimbabwe, agricultural water use is effectively subsidized through a “blended price,” which ameliorates the cost to the consumer of new irrigation developments. Although blending ensures that new entrants to the market do not bear an unreasonable cost burden, it tends to depreciate the value of water and thus reduces interest in water conservation (Helming 1993). This is a cross-subsidy in the sense that other customer classes are paying directly or indirectly for the underpricing of water to farmers.

Experiences from elsewhere in Africa suggest a similar picture. The true cost of irrigation development is rarely passed on to the farmer, partly, at least, because so many irrigation projects are funded by donors. Where the project is based on loans rather than grants, as in World Bank projects, there is a tendency to enforce cost recovery as a matter of principle, and requirements of this sort are often embodied in loan covenants. However, such cost-reflective pricing will only succeed where the producers are able to recover fair prices for the sale of resultant crops. The price increase to water users must therefore be balanced by a removal of price-setting on food intended to protect consumers (Barghouti and Subramanian 1991).

Hydroelectric generation is a significant element in a number of African riverbasin schemes, and, indeed, prospective revenue from electricity sales is a major factor in decisions to invest in such schemes in the first place. Commonly, however, the water used for generating hydroelectricity is regulated by one organization but a different organization is responsible for, and derives revenue from, electricity generation. Generating electricity from water represents an opportunity cost (other uses of that water, such as for industry, recreation, or irrigation, having been reduced), so it is appropriate to charge a rent to the generating authority for the use of this resource. Ideally, this would create a revenue stream for the water authority that could be used for further development and maintenance of the hydrological system, provided, of course, the money is not diverted to general revenue by government! In addition to improving economic efficiency, charging a rent for the resource also provides an incentive to conserve, inasmuch as the utility is charged only for water actually used in generation.

A study is under way on the Zambezi River system to develop just such a tariff to fund the Zambezi River Authority (ZRA). If approved, the tariff would be paid to ZRA by the national utility of each country - the Zimbabwe Electricity Supply Authority (ZESA) and the Zambia Electricity Supply Company (ZESCO) - on the basis of electricity actually generated from the Kariba and Victoria Falls power stations each month. This would replace the current system whereby the two utilities are given an annual allocation of water based on their theoretical requirements but no rent is levied and there is no penalty for exceeding the allocation (P. Robinson, Harare, Zimbabwe, personal communication, 1994).

Many similar opportunities to use pricing as a mechanism for water-demand management exist in Africa. However, pricing is rarely a solution in itself. For example, most small-scale farmers have not had any significant influence on the type of irrigation or other water-delivery systems they use. Thus, raising prices to these farmers, who may be using an inherently inefficient or inappropriate delivery system, will not in itself improve conservation of water and may, in fact, lead to economic distress. On the other hand, Africa is full of examples where authorities failed to provide sufficient revenue to pay for the maintenance and periodic renovation of irrigation systems, with the result that the systems quickly fall into disrepair, with attendant losses in efficiency. It is clear, therefore, that, for irrigation agriculture to operate efficiently, there must be some form of cost recovery, whether through the pricing mechanism itself or by some other means.

Pricing reform means effectively that water is treated as a commodity rather than as a communal good or right. Although this argument has more merit in the context of market economies and large-scale commercial projects, where there are substantial investment and operational costs to be recovered, it can also be applied to small-scale, traditional systems. In Kenya, for example, the efforts to augment and rationalize small-scale irrigation systems by making them the cornerstone of market agriculture are a case in point. Water rights, though traditionally supported, are now seen as something that must be earned through proper maintenance and efficient operation of the system (Adams 1992).

In much of Africa, the driving force behind the shift toward market-based approaches is the economic-reform program initiated by the International Monetary Fund or the World Bank and supported by many bilateral donors. Commercialization or privatization of government and parastatal organizations has become commonplace, if not always successful. Water providers (riverbasin, irrigation, and municipal water authorities) have largely fallen outside this initiative, possibly because they are traditionally dominated by public interests in the industrial countries as well. However, there is pressure for water authorities to operate on a profitable basis to ensure that they have sufficient revenue for ongoing management and operational costs. This, in turn, calls for a move toward cost-reflective tariffs for water, just as it does for energy. According to the World Bank (1992), the real costs of water have typically doubled for every new irrigation or municipal water project in Africa, by itself a strong argument for continual upward adjustment of prices.

Demand-side management for water: a basic paradigm

Defining demand management and demand-side management

As applied to water resources in particular, demand management is a term that usually refers to a broad range of techniques and processes based on the end-use requirements of water consumers, rather than on the supply requirements of water providers. Demand-side management, by contrast, is a term used to describe a somewhat narrower set of activities and principles that are typically initiated by the resource provider (utility) itself as a part of its corporate-planning and capital-investment process. In the energy context, DSM is closely associated with efforts to alter utility-load patterns. DSM is also associated with specific kinds of planning, often referred to as “integrated resource planning” or “least-cost planning,” in which demand reduction and supply increase are given equal weight by a utility when making investment choices.

By definition, DSM requires a better understanding of consumer requirements. As a result, power utilities implementing DSM are more responsive and more likely to make socially appropriate investment decisions than would traditional supply-oriented utilities. DSM allows an “integrated look at technology options, customers’ needs and utility considerations” and recognizes that “customer needs” (demand) are not fixed and can be manipulated by both economic and pricing incentives (Gellings and McMenamin 1993, p. 1).

In its most conservative form, DSM refers only to the utility’s efforts to achieve a specific “load shape” that allows it to meet capacity needs at minimal cost. In this definition, the “time pattern and magnitude of the utility’s load” are the only criteria (Gellings and McMenamin 1993, p. 2). However, the application of DSM also includes activities such as customer-initiated efficiency investments, customer generation, promotion of new applications for electricity, information programs intended to bring about strategic conservation in different sectors, and a variety of marketing strategies designed to influence the share that electricity holds of the energy “pie” (Gellings and McMenamin 1993).

The most significant implication of DSM is its impact on resource planning. The convention in energy, as in water, has been to estimate demand based on historical data, using established elasticities of income and price to predict the rate at which demand will increase or, in some cases, employing simple trend analysis. This practice invariably results in generous assumptions about demand growth; supply requirements are chronically overestimated, as has been the case, for example, at Ontario Hydro, Canada’s largest utility, which now has a substantial overcapacity and has been forced to close all or part of several generating stations.

By contrast, DSM is characterized by planning that takes account of all possible investments on both the demand side and the supply side (integrated resource planning). As a result, DSM planners are able to use least-cost planning criteria, selecting investments on a more objective basis, regardless of whether they entail new generation capacity, additional transmission capacity, or consumer incentives to purchase efficient end-use technologies.

Advantages and disadvantages of demand-side management

DSM programs have, for the most part, been implemented by utilities in various industrial countries. There are several reasons for this:

· Power supplies in the industrial countries are run on tight margins, either by private power suppliers or by public power suppliers operating on commercial criteria. Investment in new supplies is therefore closely scrutinized, and decisions must reflect the opportunity cost of foregoing alternative investments.

· The price of electricity in most industrial countries is relatively high and generally approximates the long-run marginal cost of new supply options, which makes efficiency investments more attractive than they would be in low-price regimes.

· Customers in industrial countries are sensitized to price and react quickly and fairly predictably to price increases. They also have the capacity to deal with price increases by selectively reducing demand (for example, by purchasing more efficient appliances or other end-use technologies). Their sensitivity and responsiveness are, of course, partly a function of higher income levels and more sophisticated consumer markets.

Utilities initiating DSM in the industrial countries usually have a fairly precise notion of the supply cost of conservation options, that is, the discounted cost of an investment in efficient end-use technologies for a particular customer or class of customers. Comparing this with the discounted cost of supplying the same amount of energy provides a more realistic planning base. A utility typically does this by developing conservation-supply curves, which compare the cost per kilowatt-hour saved for different conservation options. These can then be objectively compared with the cost per kilowatt-hour of different supply options. For example, the conservation-supply cost to the utility of a technology such as compact fluorescent lights might be 2.5 cents/kW×h of energy saved over the life cycle of the product, against a delivered cost of 4-5 cents/kW×h of energy supplied from a conventional power station. If the utility can mount a promotional campaign or offer a “prescriptive” subsidy on the purchase of such products at a total cost that is less than the avoided cost of the conventional station or other supply options, then this represents a realistic alternative investment, one on which the utility could earn an attractive rate of return. Even though this investment might only defer, rather than displace, the supply option, it can still produce substantial savings to the utility by delaying the need to embark on expensive capital investments.

DSM has become a major part of utility planning in many industrial countries and provides a distinctively customer-oriented focus at a time when many utilities are seen by their customers as frankly predatory and profiteering. However, success has proven difficult to reproduce in developing countries, particularly in Africa, where the provision of electricity is typically a state monopoly and where commercial criteria have, for the most part, not been successfully applied. Several important caveats apply to the use of DSM in developing countries:

· DSM does not work as well where electricity prices are low. Because of political factors, electricity prices in many jurisdictions in Africa have been artificially depressed, although this is changing rapidly. It is, of course, possible to use nonprice incentives, such as information programs, where pricing is still too low to encourage voluntary consumer investment.

· There must be a scarcity of the key resource (energy or water) to compel the national utility to pursue some form of investment strategy aimed at coping with increased demand. Where energy (or water resources) are available and well distributed, there is far less incentive to undertake DSM.

· There must be a substantial investment by the utility in both new planning methodologies and customer-information databases. Hence, the utility must maintain a good cash flow and strong revenue base, which are major obstacles among African utilities at the moment.

· Barriers to acquisition of efficient technologies, including tariff or import barriers, must be removed or minimized to ensure that these technologies can be acquired at a fair and competitive price. Protectionist policies of the sort found in many African states until recently are a disincentive to implementation of consumer-driven DSM.

· Consumers must have access to good market and product information without excessive cost. This implies a strong and well-managed information program and a sound statistical base. These are again difficult to achieve in many African utilities, which often lack the revenue base and human resources to carry out such programs.

Notwithstanding these significant limitations, a number of African utilities have already embarked on DSM programs, and others are closely studying the prospect for DSM. Utilities in relatively wealthy countries, such as ESKOM in South Africa, are, of course, better positioned for this kind of initiative. ESKOM has undertaken fairly detailed assessments of DSM prospects and has, in fact, initiated a limited “small customer” DSM program, despite the fact that it has a 25% overcapacity! Poorer utilities, such as Electricidade de Mobique in Mozambique, ZESCO, and Tanzania Electricity Supply Company, have also embarked on such programs (with donor assistance) because they are faced with significant supply shortages and they badly need to encourage consumers to use electricity more efficiently.

Conditions for success in demand-side water management

By reviewing the pros and cons of energy-DSM programs, one may deduce several conditions for successful water-DSM programs:

· Detailed knowledge of end-uses, or customer consumption patterns, is a precondition of successful DSM programming, as suppliers need to be able to target specific customer classes for incentive or subsidy. Electricity utilities meter consumption for most classes of user, so the basis for a detailed consumer (end-use) database already exists. However, this is not so for water utilities, although the gross requirements of different classes of consumer are generally known and many specific efforts to measure usage have been undertaken. Metering of flows to irrigators, measurement of water passing through and consumed by hydroelectric generators, and (although rare in Africa) metering of domestic and industrial consumers to provide revenue to municipalities are some examples.

· It is necessary to alter official perceptions of the relationship between water pricing and consumption, particularly those that characterize centralized water-management schemes. Water authorities must establish systems to measure water consumption and then charge according to consumption. In fact, it can be argued that the principle of cost-reflective pricing of water should be attached, not just to river- or rainfall-based irrigation or municipal water supplies, but to groundwater (borehole) supply as well, because use of this sort involves depletion of a national resource for the benefit of a few.

· Pricing regimes must reflect costs. In other words, the overall cost of operating, maintaining, and (perhaps) expanding the water-supply system should be included in the average consumer price charged for water and (possibly) in the prices charged to specific end-users or classes. Such short-run marginal-cost pricing, which accurately reflects operating and maintenance costs but not expansion costs, would be superior to the present system of politically expedient pricing and would be easier to develop and administer than long-run marginal-cost pricing.

· Water suppliers must have general knowledge of the types of consumer end-uses and their typical requirements, namely, the amount of water required for specific kinds of activity, such as industrial cooling, production of beverages, washing, or household consumption. This knowledge will enable the suppliers to identify areas of wastage or inefficiency more readily and to gear their DSM programs to the most wasteful consumers.

· There must be clear lines of authority over the storage, catchment, and eventual distribution of water. Authorities, whether central or local, must be accountable to their consumers.

· There must be access to water-efficient technologies or processes. In practice, therefore, these technologies or processes must be available when and where needed and at a competitive cost.

Implementing demand-side management for water

Analyzing customer end-uses

The first and most obvious task in implementing a successful DSM program for water is to analyze customer end-uses. Doing this with accuracy requires either (1) a system of data collection that is built into the revenue collection or administration system or (2) an accurate means of differentiating customer classes and estimating the number of customers in each.

The first requirement presupposes an organized and centralized water-distribution system, such as we might find in large-scale irrigation schemes in Sudan, or a municipal water-supply system, such as we might find in many southern Africa cities. These systems can fairly easily produce data on average consumption per customer over a stipulated period. Such data, however, are useful only to indicate broad shifts in consumption patterns.

The second requirement assumes that there is a clear division by customer type in the distribution system, perhaps a geographical or locational breakdown, so that different domestic areas can be distinguished on the basis of gross consumption and distinguished again from predominantly industrial areas. Although less precise, an analysis based on this kind of information can be done with minimal skills and can identify average consumption per unit by customer class.

Neither method will provide a detailed end-use breakdown. For example, they will not differentiate hot-water use from cold-water use or, in the case of industry, water for boiler makeup from water used in the product itself. In the absence of metering, achieving this would require an end-use survey: a sample survey of water users to identify typical end-use breakdowns in different subsectors and to permit extrapolation to the population as a whole. This is the method favoured by energy planners and, increasingly, by energy utilities. However, it is an expensive method and is likely justified only for large industrial consumers and for a small sample of domestic consumers of different income levels.

Once an end-use breakdown is obtained, it is possible to generate an analysis of water-quality requirements in each sector and subsector. In the case of industrial water use, for example, this would identify which uses are consumptive and which are nonconsumptive, what the requirement is for pretreated water, whether recycled water from certain internal processes is acceptable, and whether there is a need for (or tolerance of) water from thermal sources (such as recycled cooling water from a thermal generator). In the case of agriculture, an end-use breakdown would determine the specific water needs of different crops and whether those crops can tolerate recycled water.

With information on both water quality and end-use, it is possible to generate a minimum-water-requirement program for the area, sector, or subsector, broken down, if possible, by type or quality of water required. Requirements should be expressed as a range to make some allowance for periodic variation. Ideally, the program also includes information on temperature, salinity, or other water-quality considerations. A program for a hypothetical textile plant is shown in Table 1. The basic elements of this approach are outlined in Figure 1.

Comparing the costs of demand-side and supply-side options

The second major step in establishing a DSM program is to develop a conservation-supply curve, which describes the relative life-cycle costs of different conservation and supply options. For water, the comparison might be between, say, the cost per cubic metre of installing a drip-irrigation system and the cost of water delivered by a conventional spray-irrigation scheme or the cost of obtaining a new source of water.

In such an assessment, the supplier compares a range of conservation and supply options to see what the most advantageous investment would be, and at what point (if any) supply options are more cost-effective than conservation options. An exercise of this sort is fraught with many assumptions. To establish what the cost of a given conservation investment might be, for example, one must make assumptions about initial cost, product lifetime, operating costs, price elasticity of demand for the product and of the particular consumer group being targeted, and likely adoption rate of the product without incentives. Similar assumptions are made, of course, when one is planning a supply investment. In the case of water, the choice of conservation technology is limited, as is the experience of consumer reaction to price or cost incentives. Technologies to consider include the following:

· low-pressure pipes, sprinkler systems, and drip systems for irrigation;
· water-recycling applications;
· improved lining materials to reduce seepage from canals;
· flow restrictors for domestic or industrial ablution;
· low-flush toilets;
· seasonal price variations; and
· leak-detection programs.

This analysis must reflect the true cost to local consumers of purchasing and operating a specific technology or application. It must also be updated periodically to reflect changes in local markets, tariff or tax effects, and cost of operation.

A difficulty in developing cost data is that many potential efficiency actions involve administrative or management improvements, rather than “technical fixes.” In the case of energy suppliers, this problem is usually dealt with separately through information and training programs. These also represent a cost to the supplier and are intended to offset other supply costs. Perhaps the most important point about analysis of financial benefits of different conservation and supply options is that there must be an organization or agency responsible for providing water supplies, with discretionary authority over the course of future investments in the system and to which the benefits of conservation alternatives would accrue. There is no reason, of course, why this authority might not be a tribal or communal irrigation administrator, as opposed to a large riverbasin or dam authority or a municipal water-management authority. However, the development of DSM programs for water invariably implies an increase in the sophistication and technical competence of the management authority.

Disseminating information and training management

A major barrier to adoption of demand management is the provision of information dissemination and training. In a focused DSM program, therefore, water providers may be of assistance in two additional ways:

· disseminating information on efficient end-use technologies or better water-management strategies; and

· providing training on how to manage water resources more effectively.

DISSEMINATING INFORMATION - Dissemination of correct product information is essential if consumers are to achieve high end-use efficiencies. Many energy-DSM programs focus on information to the virtual exclusion of other kinds of subsidy or assistance and find that, if market conditions are correct (in particular, if the price of the resource is high enough), consumers will purchase efficiency independent of any further incentives. Such an assumption would obviously be inappropriate to the African situation, where consumers may be poor as well as ill informed. However, it is still essential to provide good information, not only on how to choose new technologies for water use, but also on how to manage the resource more efficiently by (in the case of agriculture, for example) applying better soil-conservation techniques, using water-recovery techniques, covering storage areas, or choosing less-water-intensive crops.

A pertinent example from the energy field in Africa is a project sponsored by the International Research Development Centre through the Energy for Development Research Centre in Cape Town. The project targets low-income populations in South Africa’s urban and peri-urban areas. It aims to improve the flow of information about efficient and environmentally friendly domestic energy technologies (for lighting, cooking, space heating, and water heating) through a series of information workshops, efforts to improve appliance standards, and involvement of both the electricity utility and local authorities in the marketing and retailing of these technologies (EDRC 1993).

TRAINING MANAGEMENT - Training is the “flip side” of the requirement for better information. Examples from Egypt and Morocco show that improved on-farm management skills introduced through farmers’ organizations can improve irrigation efficiency by 10-15% overall and productivity by as much as 30% (Xie et al. 1993). Management improvements can be fairly inexpensive if indigenous organizations and community groups are used as the delivery agents. From the viewpoint of water authorities and governments, this is a highly cost-effective investment in improving water-use efficiencies.

Choosing efficient technologies

In energy DSM, a distinction is often made between “prescriptive” and “audit-based” programs. The former involves provision of grants or subsidies to consumers who purchase certain types of energy-conserving equipment, such as energy-efficient motors or lights. The latter involves assessment of a single facility, such as an industrial plant or building, by a qualified consultant. The consultant then recommends a specific set of conservation investments, the costs of which can be shared by the utility and the facility owner. Prescriptive approaches are simpler to manage, although they involve careful market research and precise estimation of savings to the utility at different levels of adoption; audit approaches are more complex but much more precise in targeting potential savings.

Both programs are customer based rather than supplier based, and both involve selection of specific end-use technologies that have previously been studied by the utility and compared with other investments. Audit-based programs are generally preferred by utilities with limited capital because supply options can be deferred by adopting a small number of high-return conservation investments. However, audit-based programs require management sophistication and professional expertise, both of which are lacking in water authorities everywhere, not only in Africa. Prescriptive programs are preferred by utilities with major supply problems, substantial investment capital, and a good consumer-information database. In water management, prescriptive programs are appropriate where the water authority is trying to influence the behaviour of a large number of small customers, hence in the domestic sector and in small-scale irrigation. Such programs would prescribe specific end-use technologies, such as drip-irrigation systems for tobacco farmers or flow restrictors for domestic showers, and set a level of subsidy that will stimulate adoption by the user. In contrast, water audits directed to major industrial or agricultural users can almost certainly be justified financially. Water authorities can turn up significant efficiency opportunities through routine inspections of facilities.

Improving the efficiency of transmission-distribution

Transmission-distribution is a major area of potential improvement for electricity utilities, particularly where, as in much of sub-Saharan Africa, these systems are old and poorly maintained. Losses from these systems can range as high as 25 or even 30%, where 8-10% would be an acceptable figure in normal circumstances of short- and medium-range transmission. Apart from this substantial opportunity for efficiency gain, improvements in transmission-distribution efficiency have the additional benefit that they can be included in regular capital programs and can be managed entirely by the existing organization within the utility: they do not require special market-assessment programs or new-technology testing.

In the water case, the parallel is to replace or repair piping and canal systems in either municipal supply or irrigation. The evidence of efficiency losses in these applications is strong. Davies and Day (1986), for example, noted that known losses through leakage in the Cape Town municipal water-supply system (piping losses) were of the order of 14%. Xie et al. (1993) suggested that, for irrigation systems, there are extreme cases: unlined canals in pervious soils, for example, have transmission losses of 30-35%. With proper soil conditions, appropriate canal or pipe design, and better maintenance, it is possible to achieve transmission efficiencies of more than 95%.

The contrast between developing and industrial countries in terms of distribution losses is significant: whereas the level of distribution efficiency in developing countries overall is estimated at 68%, that of the United States is 78%, suggesting that a 15% improvement over the developing-country figure is theoretically possible. Accurate figures from Africa are not available, but it is estimated that distribution efficiency is even lower than the developing-country average, ranging from 40 to 60% (Xie et al. 1993).

The benefits from improvement in distribution are apparent, both for the end-user and for the provider of water. Actions such as relining canals, repairing and upgrading piping systems, replacing seals in pipeline pumps, minimizing leakage in storage tanks and reservoirs, and, if appropriate, reducing line pressures all require both capital investment and technical knowledge. They are best undertaken by central water-distribution authorities through direct investment, although it is possible that improved distribution efficiencies in irrigation systems could be encouraged by providing subsidies to users for lining local distribution canals or repairing pipes and gates.

Institutionalizing demand-side management programs for water

Institutionalizing demand-side management of water requires that authorities responsible for water distribution have a clear set of responsibilities for maintaining secure and efficient water delivery, that they are responsive to price and quality issues, and that they have a reasonably detailed understanding of their customers’ requirements. As this combination of qualities would be a rarity even in developed countries, it is more realistic to see the goal of institutionalizing DSM at water authorities as a process rather than an event. The status and capabilities of the customers themselves are key to this process. If water consumers cannot tolerate large variations in the price they pay for water (or any price at all) and if they have no opportunities to introduce efficient end-uses themselves, either financially or practically, then they are more likely to support an institutionally driven DSM program.

The final step in the institutionalization process is often the most difficult. DSM programs must be established by the very same organizations and individuals that are traditionally supportive of supply-side management. Existing water-management programs are characterized by a strong interest in developing new supplies and maintaining a given level of supply or quantity, and they are geared to dealing with expanding demand. To shift from this preoccupation to encouraging customers to reduce their demand for water and even to investing in these efforts is a major challenge that will almost certainly have to be supported by government itself, by nongovernmental agencies, and, possibly, by donors and lenders if it is to succeed.

Priorities for demand-side management of water resources

Establishing DSM programs for water resources requires that water-management authorities, as well as the donors and lenders who support water development, set entirely new planning priorities. Because for Africa the mix of organizations and authorities is quite different, the mix of priorities will be different from those of North America or Europe. They should include the following elements in rough order of priority:

1.provement of the maintenance and operational efficiencies of existing delivery systems - This is a major investment opportunity for water suppliers and must be encouraged. In large-scale systems, such investments should be driven by revenue yields: the returns on water sales must be sufficient to make it attractive for the supplier to invest in improving conveyance efficiency directly. In small-scale systems, such as traditional irrigation networks, indirect investment may be more appropriate because suppliers can subsidize the end-users’ maintenance of their own part of the network and thus encourage a mutual concern for efficiency.

2.njunctive management of groundwater and surface-water resources - his is particularly important to avoid depletion of aquifers. However, the converse is also true: by using groundwater resources, it is possible to augment surface supplies and thus defer the cost of new surface reservoirs and other capital-intensive supply systems. This is a strategy that can be realistically applied only at the level of the supplier or local or national government. Conjunctive-use strategies are particularly important in dealing with the crisis of urban water demand that afflicts many African cities, but they can also be applied in agricultural settings.

3.velopment of customer-oriented policies by large-scale suppliers, such as government itself or irrigation and riverbasin authorities - This is a natural development for commercially operated utilities, but it will be more difficult for parastatal organizations, which tend to have a top-down orientation. Such a major shift in orientation can only be accomplished by significant changes in style of management. In other words, it will happen when management sees its services as customer driven.

4.intenance of a realistic but generally cost-reflective pricing system - This is the key to effective DSM, but it is not always possible to achieve it instantly or uniformly across customer classes. Nevertheless, most experts seem to agree that African agriculture, particularly commercial agriculture, is excessively cross-subsidized. They also believe that farmers will benefit in the long run (even though they may complain bitterly at first) by paying a larger share of the true cost of their supplies. Pricing must be tempered by the recognition that most African farmers are not involved in commercial agriculture. Similarly, most domestic consumers cannot afford substantial increases in water tariffs. Suppliers may, therefore, have to consider a “lifeline tariff” for these consumers and recover a larger percentage of revenue from those that can pay.

5.velopment of an end-use database - This is critical for water suppliers, whatever their size and kind. Suppliers require detailed knowledge of how their consumers use water: what amounts and quality of water are required for specific end-uses. A user survey can accomplish this at relatively low cost, and the benefits of such an exercise far outweigh its costs. Above all, this kind of information will provide suppliers with a better indication of the end-use efficiency of their systems, as well as a more exact idea of how overall system efficiency can be improved by effecting technical improvements at the consumer level. In the African context, information of this sort is almost entirely lacking at present, except in a few relatively developed countries where metering systems have been used or where a high percentage of the population served falls within the formal economy.

6.ovision of information on affordable water-management technologies - This is a prerequisite for encouraging use of water-efficient technologies. African farmers, as well as domestic users, rarely have such access and, as a result, are not in a position to react rationally to price increases or even to direct conservation efforts. Dissemination of better product information has to come from either government or the supplier. It serves as an appropriate investment by them in increased efficiency of water delivery.

7.ovision of appropriate financial incentives to consumers - This is necessary if consumers are to increase efficiency: where possible, water suppliers should consider the benefits of investing directly in consumer efficiency. This is a key element in energy-DSM programs and will work just as effectively for water-DSM programs, even though the economics may differ because of water’s universality and relatively lower transformation costs. Incentives can range from prescriptive measures, such as providing free flow restrictors to domestic consumers or establishing fixed subsidies for adoption of drip-feed irrigation, to audit-based measures that result in paydowns of efficiency investments for large industrial consumers.

8.creased water recirculation - This is a customer-based activity that must be driven by pricing and information distribution. Suppliers need to identify areas where water from nonconsumptive uses can be diverted to consumers needing additional water at low cost. Just as in energy management obvious savings from recycling (often heated water) are overlooked by consumers because of low prices or lack of technical knowledge, so, too, in water management, recycling may need to be encouraged through either partial subsidies or increased information on recycling opportunities and demonstrations of its application in comparable situations. Recycling of sewage water is quite a different matter and demands strong central control and regulation.

9.provement in the management of water resources - This is particularly important for irrigation systems but may also be applied to municipal and industrial water management. Improved management skills can contribute significantly to efficiency gains and to overall system productivity in the case of agriculture, and it is logical, therefore, to expect authorities to invest in such improvements. Such investments could involve, for example, focused education programs for small farmers, management-assistance activities delivered through agricultural extension programs, and consumer-information programs that stress the benefits of better management.

Recommendations for future research

It is clear that development of more-efficient water-management systems in Africa can be facilitated by application of some of the DSM principles employed in the field of energy. To achieve this, African governments, as well as the private sector, research organizations, and nongovernmental organizations, must act decisively to initiate specific reforms. In addition, they need to support continuing research in the field of DSM through existing university- and government-based research. The following areas are recommended for further research:

· development of appropriate end-use technologies for local conditions in Africa;
· analysis of the effects of pricing on consumer efficiency in different sectors;
· improvement in the accuracy and cost reduction of water-loss measurements;
· development of end-use databases for different customer groups;
· studies on the impact of transmission-distribution or conveyance losses on overall system efficiency; and
· development of conservation-supply curves for different authorities or sectoral applications.

References

Adams, W.M. 1992. Wasting the rain. Earthscan Publications, London, UK.

Barghouti, S.; Subramanian, A. 1991. Policy and management problems. In Barghouti, S.; Le Moigne, G., ed., Irrigation in sub-Saharan Africa. World Bank, Washington, DC. World Bank Technical Paper 123.

Brooks, D. 1994. Economics, ecology and equity: lessons from the energy crisis in managing water shared by Israelis and Palestinians. In Isaac, J.; Shuval, H., ed., Water and peace in the Middle East. Elsevier Science, Amsterdam, The Netherlands. pp. 441-450

Brooks, D.; Peters, R. 1988. Water: the potential for demand management in Canada. Report to the Science Council of Canada, Ottawa, ON.

Bryant, E. 1994. Water: tapping Africa’s most basic resource. African Farmer, 12 (July), 25-28.

David, L.J. 1988. Environmentally sound management of the Zambezi riverbasin. International Journal of Water Resources Management, 4(2).

Davies, B.R.; Day, J.A. 1986. The biology and conservation of South Africa’s vanishing waters. Centre for Extramural Studies, University of Cape Town, Cape Town, South Africa.

EDRC (Energy for Development Research Centre). 1993. Energy efficiency, equity and the environment: improving access to energy services for the urban poor of South Africa. Submission to International Development Research Centre. University of Cape Town, South Africa.

Elahl, M.; Khushalani, B. 1991. Technical issues of irrigation development in sub-Saharan Africa. In Barghouti, S.; Le Moigne, G., ed., Irrigation in sub-Saharan Africa. World Bank, Washington, DC. World Bank Technical Paper 123. pp. 69-82.

FAO (Food and Agriculture Organization). 1986. Irrigation in Africa south of the Sahara. FAO, Rome, Italy. FAO Investment Centre Technical Paper 5.

Frederick, K.A. 1993. Balancing water demands with supplies: the role of management in a world of increasing scarcity. World Bank, Washington, DC. World Bank Technical Paper 189.

Gellings, C.W.; McMenamin, J.S. 1993. DSM: what’s real, what’s not, and how we can tell the difference. Electric Power Research Institute, Palo Alto, CA.

Gleick, P.H. 1989. Climate change and international politics: problems facing developing countries. Ambio, 18(6).

Grainger, A. 1990. The threatening desert: controlling desertification. Earthscan Publications, London, UK.

Helming, S. 1993. Planning for the future: the cost of water. Paper presented at EDI Water Management Workshop, July 1993, Victoria Falls, Zimbabwe.

Kasimona, V.N.; Mkwaya, J.J. 1995. Present planning in Zambia for the future use of Zambezi River waters. In Matiza, T.; Crafter, S.; Dale, P., ed., Water resource use in the Zambezi Basin: proceedings of a workshop held at Kasane, Botswana. International Union for the Conservation of Nature and Natural Resources, Gland, Switzerland. pp. 35-43.

Lele, U.; Subramanian, A. 1991. Sectoral strategy and irrigation development in sub-Saharan Africa: some lessons from experience. In Barghouti, S.; Le Moigne, G., ed. Irrigation in sub-Saharan Africa. World Bank, Washington, DC. World Bank Technical Paper 123. pp. 60-68.

OECD (Organization for Economic Co-operation and Development). 1984. Rahad Irrigation Project performance report. OECD Project Performance Report 5130.

Schwarz, J. 1991. Sector water study in Israel. Presented at an international workshop on water resources management, June 1991, Washington, DC.

Scudder, T. 1989. Riverbasin projects in Africa: conservation vs development. Environment, 31(2).

Seckler, D. 1993. Designing water strategies for the 21st century. Winrock International Center for Economic Policy Studies, Arlington, VA. Discussion Paper 16.

South African Department of Water Affairs. 1986. Management of the water affairs of the Republic of South Africa. South African Department of Water Affairs, Pretoria, South Africa.

ter Vrugt, J. 1990. Private sector irrigation in Zimbabwe. In Barghouti, S.; Le Moigne, G., ed., Irrigation in sub-Saharan Africa. World Bank, Washington, DC. World Bank Technical Paper 123. pp. 88-90.

Wittfogel, K. 1957. Oriental despotism. Pergamon Press, New York, NY.

World Bank. 1990. The Gazira Irrigation Scheme in Sudan. World Bank, Washington, DC. World Bank Technical Paper 120.

World Bank.1992. World development report 1992: development and the environment. World Bank, Washington, DC.

World Resources Institute. 1992. World resources report, 1992-93. Oxford University Press, New York, NY.

Wyss, H. 1991. Prospects for irrigation development in sub-Saharan Africa. In Barghouti, S.; Le Moigne, G., ed., Irrigation in sub-Saharan Africa. World Bank, Washington, DC. World Bank Technical Paper 123. pp. 1-4.

Xie, M.; Kuffner, U.; Le Moigne, G. 1993. Using water efficiently. World Bank, Washington, DC. World Bank Technical Paper 205.