F. Lempérière and X. de Savignac, HydroCoop, France
The major roles to be played by some of Africa’s large river systems in meeting water and energy needs throughout this century are discussed here. It can be seen that by the end of the century, low cost hydro generation or storage could meet most of the continent’s energy needs.
By the middle of this century, the African population will probably have increased from 1 to 2 billion. The average income per capita, at present less than 10 per cent of the corresponding income in industrial countries, should have increased at least fourfold (assuming a rate of 3 to 4 per cent per year). The total income may thus have increased by a factor of more than 8. This will require much more water for irrigation, and an increase in electricity generation corresponding to the total income increase. Generation at present is 750 TWh/year, as a large proportion of Africans have no electricity; within 40 years it should be more than 5000 TWh/year, which means an increase of 2500 kWh/year for 2 billion people.
Potential hydropower development and the need of water are analysed here for the whole of Africa, with a particular focus on the four most important rivers: the Congo, Zambezi, Niger and Nile. Possible solutions for specific problems are suggested for Nile and Niger (see Fig.1).
1. Hydropower in Africa: general comments
By mid-century, hydropower could account for 15 per cent of African electricity generation; it may be even more useful for energy storage linked with wind and solar energy, and this additional target could impact the design of many hydroelectric schemes currently designed only for generation.
1.1. Hydropower generation
Fig. 1. The African countries (from H&D World Map showing major schemes under construction.)
A key advantage of African hydropower is the low cost per kWh of many schemes; it may be much less than US¢ 5 (2010 value), which can be compared with the likely cost closer to US¢10 for most other sources in the future. But the potential of low-cost African hydropower is limited to 1000 TWh/year. About10 per cent is already developed, and a reasonable target for mid-century seems 70 per cent of the potential; that is, 200 GW for 700 TWh/year. About 90 per cent of the potential is in Central Africa, where only 5 per cent has been exploited so far (see Fig. 2).
Fig. 2. The major river systems of Africa.
Hydropower in Africa requires less resettlement than in most Asia, because the density of population is lower. But a large part of the investment is in foreign currencies and many countries may require external financing or help; this may cause delays and sometimes questionable extra costs linked with excessive care imposed for environmental impacts. For this reason, most Asian dams are built without foreign financing.
Hydropower schemes will be able to remain in operation throughout one century. Their design should therefore not be based only on the conditions prevailing over the next 30 years. This applies to sedimentation problems, and also to future possible uses of the reservoirs for energy storage.
1.2. Hydropower for energy storage
Fig. 3. Likely electricity resources for Africa during this century.
About 80 per cent of African electricity is from fossil fuels at present; by mid-century this share is likely to be reduced to 20 to 30 per cent, as will also be the case in other parts of the world, partly for reasons of diminishing resources, as Africa does not have much fossil fuel.
The contribution of nuclear electricity will be 10 to 15 per cent of world supply, and much less in Africa. More than 50 per cent of electricity generation in Africa will therefore probably be from wind and solar power (see Fig. 3). Wind farms may be cost effective along the northwest or southwest shore, in the Sahara or Egypt, but most of Africa has limited cost-effective wind resources. The key electricity source will therefore probably be solar energy, as there are many quite desert areas with 3000 to 4000 hours of sun per year, with a probable cost after 2030 of less than US¢10/kWh. The annual peak demand is during the sunny months, and during rainy season more hydropower may balance the reduction in solar energy. Solar energy can be implemented essentially by local engineering and labour within a short time. The total area required to generate 3000 TWh/year of solar energy will be 30 000 km2 mid century, about one thousandth of the area of Africa, and this would mostly be in desert areas. Major further development would thus be possible (see Fig. 3). A part will be through small domestic capacities but most energy will be by large solar farms.
The drawback of wind and solar energies is their intermittent supply capability, which must be balanced by flexible backup, such as fossil energies (to be avoided) or by hydropower reservoirs if designed for energy storage. The following may be possible, for example:
• Associating the energy from solar farms during day with hydropower energy during the night: this may require an increase in the hydropower capacity used during a shorter time throughout the year.
• Linking two reservoirs as pumped-storage schemes along the rivers, equipping them with reversible pump turbines.
• Using a hydropower reservoir as the lower reservoir of a pumped-storage scheme, with the upper reservoir being constructed out of the river.
• Creating a pumped-storage scheme using sea water as the lower reservoir, with an artificial basin being built high above onshore, preferably on a cliff. This solution may be attractive close to mountainous areas, or in desert areas where there is no fresh water. It may also be very attractive in many parts of Africa where most people live within few hundred kilometres of the sea.
By mid-century, for a total intermittent energy in Africa of 3000 TWh/year, or 350 GW average electricity supply, the theoretical pumped-storage capacity should be close to 350 GW, although it may be reduced by the use of some conventional hydro or fossil fuel generation. However, the need for pumped storage may reach 200 GW by 2050, equivalent to the hydropower capacity of 200 GW by this time (see Fig. 4)
The extra cost for storage will be about US¢2/kWh of intermittent energy.
Fig. 4. Likely hydro capacity for Africa during this century.
2. Water resource in Africa
The total runoff of African rivers is some 2500 × 109 m3/year, that is, at present more than 2000 m3 per capita, but 90 per cent is in the central part of the region, between the Sahara desert and the Zambezi river. Runoff per capita is less than 1500 m3/year in northern and southern Africa, and much less in many places where the rainfall volume is also very small. The long distance transportation of water is difficult and expensive. The water resources in northern and southern Africa therefore presents a difficult problem, and irrigation may be a higher priority than hydropower with respect to the use of available river water.
3. The Congo
For a catchment area which is considerably greater than 3 × 106 km2, corresponding essentially to the Democratic Republic of Congo, the average flow of the Congo reaches 40 000 m3/s, that is, 1200 × 109 m3/year, equivalent to half of the flow of all African rivers. As there is also a high level of precipitation in this area, the need for river water for irrigation will remain very limited in Congo and its neighbouring countries.
The navigation facilities could be improved, but discussion on that is beyond the scope of this paper.
The economically feasible hydro potential of Congo is considerably more than 300 TWh/year. The Inga scheme alone has a potential of more than 200 TWh/year at a cost close to US¢1/kWh, by far the cheapest cost in the world for electricity, but at present its exploitation has been negligible. The population, at present close to 70 million people, will probably double within 50 years, by which time all this potential could possibly be used locally. It would be possible to sell, between 2020 and 2060, most of the generation which would be beyond the requirements of the Congo. On average, 200 TWh/year sold at several cents per kWh would yield more than US$5 billion/year. Producing 10 to 15 per cent of world’s aluminium could use 50 TWh/year. Another possibility would be to export to countries such as South Africa and Nigeria 10 to 20 per cent of their electricity needs, which would be more than 100 TWh/year. The link to Nigeria could be offshore and direct. The total cost of power could be less than US¢4/kWh, including the cost and losses associated with transmission.
Within the next 50 years, hydropower may be very useful for the development of Congo. Other electricity sources, mainly solar, may be necessary before the end of the century, probably associated with hydropower.
4. The Zambezi
The annual runoff of the Zambezi is close to 100 ×109 m3. It is essentially used for three countries with populations currently totalling 50 million people: Mozambique, Zambia and Zimbabwe. The river water requirements for irrigation will probably remain at less than 10 billion m3/year. Navigation is not foreseen.
Hydropower is an important source of low-cost energy and will be locally the most important energy source throughout the next decades. The potential of the Zambezi and tributaries is 100 TWh/year, which may be essentially developed as some 10 large schemes along the Zambezi. Half the potential is in Mozambique.
There is currently 25 TWh/year of hydropower, of which 10 TWh is transmitted from Cahora Bassa to South Africa along a1400 km line. Some 80 per cent of the present electricity of the three countries is derived from hydropower, which seems the most cost-effective source for further needs. But the population increase and of the per capita income will probably require development of all the hydroelectric potential within 20 or 30 years, and other sources will also be needed.
There is some coal in Zimbabwe and Mozambique, but the main additional energy source will probably be solar energy. By mid-century it could be as important as hydropower. The design of future hydro schemes should thus take into account the need for storage linked with this intermittent energy. This may require extra capacity of powerplants to be used mainly during the night or the provision of reversible units between two reservoirs along the Zambezi (as along the Douro river in Portugal).
By the end of the century, 100 million people may need 500 TWh/year, of which 300 or 400 TWh could be solar. The corresponding capacity of pumped-storage plants could reach 20 to 30 GW, as much as all hydro generating plants. Hydropower reservoirs may then also be used as the lower reservoirs of pumped- storage plants, with artificial basins built high above.
Directly or indirectly, the Zambezi river’s hydropower will be essential throughout the century to meet local energy needs from renewable sources.
5. The Niger
The Cahora Bassa dam on the Zambezi river in Mozambique.
The Niger is 4000 km long; the problems and solutions are very different for three stretches of the river.
• In the downstream 1000 km (in Nigeria), precipitation is more than 1 m per year and the runoff from local rains reaches 200 × 109 m3/year. The water resources are therefore not critical. The hydropower potential of 30 TWh/year may be developed soon, but will meet a small part of the needs of 300 million people by mid-century. And solar energy may be combined there with electricity production from gas, involving less need for energy storage before 2050.
• The upstream 1500 km in Guinea and mainly in western Mali have an average yearly runoff reaching 40 × 109 m3 in Mopti, flowing essentially within six months; 40 per cent evaporates downstream in the very flat Internal Delta (30 000 km2) between Mopti and Timbuktu. A storage of 1 × 109 m3 at Selingue retains some 100 m3/s in the Niger river system at Bamako during the dry season. Increasing the storage at Selingue up to 5 × 106 m3, which is being studied at present, will supply Bamako with more than 300 m3/s in the dry season: this will be essential for irrigation in Mali, and will be favourable for both hydropower and navigation between Bamako and Gao.
The hydropower potential of 5 TWh/year could meet most of the needs of Mali within the next 15 years, at low cost, especially by the construction of run-of-river plants along the Niger in Mali, and by the upgrading of Selingue. But the main future electricity source in Mali will be probably solar energy, which will be less expensive than using fossil fuel after 2020. This may be associated with hydropower, and will require, by mid-century, major pumped-storage units, for instance at Selingue and in various places in the Internal Delta such as Korientze. The future electricity demand in most of Mali could thus be met entirely by renewable resources at a very acceptable cost.
• The central stretch of the river, with a length of around 1500 km (mainly eastern Mali and Niger) receives less than 0.5 m of rainfall in three months, and the key water resource for 30 million people in the future will be 25 × 109 m3/year of the Niger flowing within few months from the Internal Delta. Some dams are under design or construction for storing several billion m3, and these will be essential for irrigation and development. It may also be possible in dry season to obtain several billion m3/year from the Internal Delta losses, with attention being paid to environmental aspects.
The hydro potential will be developed, but it is very limited. Most future electricity will probably therefore be from wind and solar power, associated with pumped storage in the Internal Delta or close to new reservoirs.
6. The Nile
The Merowe dam on the Nile in Sudan.
The Nile is very long and more complex than the other rivers; its catchment area includes three countries (Egypt, Sudan and Uganda) and a part of six others; that means, a population of 250 million people now and more than 400 million by 2050. The runoff per capita, which is 12 000 m3/year for the Congo and 2000 m3/year for the Zambezi, is less than 500 m3/year
for the Nile. The hydro potential per capita, which is more than 3000 kWh/year for the Congo and 2000 kWh/year for the Zambezi is less than 1000 kWh/year for the Nile (less than 500 kWh/year if Ethiopia is excluded).
The Nile basin can be regarded as having two parts, each with about the same population:
• The upstream rainy part, and particularly Ethiopia, Uganda and South Sudan supplying virtually all runoff and having more than 80 per cent of the hydropower potential.
• The downstream dry part, essentially Egypt and northern Sudan, with an average rainfall of 20 cm/year and a Nile runoff flowing almost totally from the upstream rainy countries.
6.1. Nile hydropower
The cost-effective hydropower potential is 300 TWh/year; 50 TWh/year has been developed, providing 25 per cent of the total present electricity supply of 200 TWh/year. The present and future data are very different for each country:
• Ethiopia has a hydropower potential of 200 TWh/year for a present population of 80 million; it currently uses 5 TWh/year of electricity, but hydro plants which will produce an additional 50 TWh/year are at various stages of construction, design or planning; hydropower will be the key basis for the development and some power may also be sold to neighbouring countries. This hydropower may be able to meet the total electricity demand of Ethiopia up to the middle of this century.
• Egypt has a population of 85 million, and is using 140 TWh/year. The present contribution of hydropower is 15 TWh/year, mostly from local potential; the power from Aswan dam which contributed 50 per cent of the Egyptian electricity in 1975 represents only 5 per cent now. As the main electricity source is fossil fuel (gas and oil) the need for pumped storage is currently low. By mid-century, it may be more cost-effective to export gas and to use wind or solar energy to cover most of the demand; pumped storage may then be necessary, possibly using the Aswan lake and the sea as lower reservoirs.
Other countries within the Nile basin have a total population of 80 million people, a present electricity consumption of less than 20 TWh/year and a hydro potential of 80 TWh/year, mostly in northern and southern Sudan and Uganda. Those countries will probably develop most of their hydro potential within 30 years, and they may also import some hydropower from Ethiopia. After 2030 they will probably use a lot of solar energy, which may well be combined with hydropower reservoirs and pumped storage.
By mid-century, hydropower will therefore represent quite a small proportion of electricity generation in most of the countries along the Nile, but pumped storage may favour the development of a substantial amount of solar energy.
6.2. Nile Water
Sharing 72 × 109 m3/year of the presently available Nile water between 400 billion people would be a major problem because the Nile is the essential water source for half the population (Egypt and northern Sudan) and because other countries will wish in the future to use some of the Nile water generated from local rains, specially for irrigation. The associated legal or political rights are not covered in this paper which is looking only at the possibility of increasing the water availability.
The total runoff of the Nile and tributaries is in fact 140 × 109 m3/year but half is lost for natural or human reasons. Virtually all this runoff reaches the Sudan and includes on average:
• An eastern flow of 83 × 109m3 (essentially from Ethiopia) through Sobat (16 × 109 m3), Blue Nile (55 × 109 m3) and Atbara (12 × 109 m3). About 4 × 109 m3 is lost from the Sobat in the East Swamps of the Southern Sudan; virtually all flows between July and November. Seasonal storage of more than 30 × 109 m3 will be possible in future at Ethiopian dams and could reduce the need for storage in Aswan Lake and at Merowe.
• A southern flow of 42 × 109 m3 throughout the year from the White Nile upstream of Sudan; 20 × 109 m3 is lost by evaporation, in the Central Swamps (the Sudd). • A flow of up to 15 × 109 m3 between August and December from the South Sudan tributaries (mainly the Jur and Lol) if virtually lost in the Western Swamps.
About 40 × 109 m3/year is thus lost by evaporation in the swamps of southern Sudan. Major evaporation losses also occur at Aswan Lake up to 15 × 109 m3/year; and some billion evaporate in Sudan at the reservoirs of Merowe, Roseires, Djebel Aulia, and others.
The annual precipitation depth at Victoria Lake exceeds the evaporation level by more than 0.3 m, but evaporation is higher than the rainfall level by more than 2 m in northern Sudan and at Aswan. The water storage at the 67 000 km2 Lake Victoria is thus more attractive than the Aswan storage, but is poorly used at present, because 80 per cent of the extra flows, more than 20 × 109 m3/year, reaching the Sudd are lost there by evaporation.
The main possibilities of increasing the Nile’s water availability are thus in southern Sudan and by the management of main reservoirs.
6.3. Saving water from the Sudd
The Sudd, which means the central swamps of Southern Sudan, is an extremely flat region and its soil is impervious. Swamps are thus created by the flows of the upstream White Nile, that is, 15 to 65 × 109 m3 yearly and the rain over the swamps, that is, 5 to 15 × 109 m3 yearly. The swamps area varies thus significantly according to the years and the seasons, from virtually zero to 10 000 km2 in very dry years, such as 1921 and from 20 000 to 35 000 km2 in very wet years, such as 1964 and 1988, with disastrous impacts on the local population. The annual flow downstream of the swamps is usually between 10 and 20 × 109 m3/year.
Many studies have been carried out over the past century century aimed at reducing this loss of 10 to 50 × 109 m3/year by a 350 km-long canal which would bypass the Sudd: a length of 270 km was even excavated between 1978 and 1983 before the civil war.
6.3.1 New canal design
The increased need for water and for the mitigation of major local flooding favours a detailed review of the canal design with a view to increasing its capacity considerably. A better knowledge of local conditions already led to the design for an upgrade, in 1980, to raise the canal’s water level to 1.5 m above the natural ground level, and to increase the discharge from 250 to 350 m3/s, in other words, 10 × 109 m3/year. It seems possible to increase the capacity up to 1000 m3/s, which would mean 30 × 109 m3 throughout the year. A possible way would be:
• Damming the Nile at the canal offtake with a low gated dam, easily built in the dry on the right bank of the Nile, and by constructing low dykes across the Nile and along the upstream banks.
• Completing and upgrading the existing canal works to create a canal 300 m wide, with a maximum water level 2.5 m above the natural ground. This requires two side dykes (4 m high and with 8/1 slopes) of 200 m2 section, each using materials from the canal area excavation. One dyke already exists along a length of some 270 km, and this could be adapted. The total excavation would be:
350000 m × 2 × 200 m2–270.000 m × 150 m2 ≈ 100 × 109 m3
The unit cost was US$1.5 /m3 for 70 × 106 m3 excavated in 1983. It will thus be close to US$5/m3 for future works, that means, an investment of US$500 for earthmoving and US$500 million for the Nile damming, a lock, and permanent road and crossing along the canal. Part of this total investment, in the range of US$ 1 billion ????, will be balanced by more than 1 TWh/year of free electricity supply from the extra water through existing hydropower plants in Egypt and Sudan.
The canal’s wet section will be
300 m × 2.5 + 2 × 200 = 1150 m2
with a 0.9 m/s flowrate; it will thus allow 1000 m3 /s.
It could be increased at low cost if advisable.
6.3.2 Canal operation and impacts
The dam and canal could be operated leaving in the Nile generally an annual flow of between 20 and 25 × 109 m3 equivalent to half of the years between 1900 and 1960. In very wet years, the flow in the Nile will be 35 × 109 m3 instead of 65 × 109 m3, thus avoiding disastrous flooding.
Seasonal swamps could be maintained as in the past, or optimized according to local requirements. The environmental impacts of this solution would be thus favourable.
The local population would be able to keep traditional seasonal swamps and benefit from:
• flood mitigation;
• irrigation along the canal;
• permanent road from Juba to Malakal; and, • navigation along the canal.
In addition, the southern Sudan could obtain a fair share of the huge benefits associated with the extra water for Egypt and North Sudan, perhaps by a fee proportional to the canal flow.
The average annual flow after 1960 has been about 42 × 109 m3. Leaving 22 billion in the Nile as average would save 80 per cent × 18 ≈ 15 billion/year. The yearly flow in the canal will be between 5 and 30 × 109 m3.
This design is much more attractive than the design of 1983:
• its extra cost is balanced by the extra power; • it provides much more water;
• it mitigates local flooding; and,
• it favours good management of Aswan Lake.
6.4. Water saving in other swamps
About 4 × 109 m3/year from the Sobat is lost in the Eastern Swamps. Most of this could be saved by a dam in Ethiopia and/or a canal in the swamps.
Up to 15 × 109 m3/year is lost in the Western Swamps from local tributaries. A large amount of this could be saved by dams and/or canals similar to the Sudd canal design. Pumping could be used to obtain the necessary head in the canals.
These schemes could save up to 10 × 109 m3/year and be undertaken after the Sudd canal with the experience of its construction and operation. The cost per m3 of water saved could be double the cost for the Sudd canal, but remain very attractive.
6.5. Water saving in the Aswan reservoir
The past management of Lake Victoria, based on the rigid rules established 50 years ago, could be optimized and the capacity of a large Sudd canal would provide the opportunity of associating the management of Lake Victoria and the Aswan reservoir for overall optimization.
The storage in the Aswan reservoir could reach 157 × 109 m3. The necessary seasonal storage is 30 × 109 m3 and will be probably greatly reduced by the seasonal storage of future hydropower schemes at Ethiopian dams and the Merowe dam. The over-year storage of well over 100 × 109 m3 is thus the key reason for the huge evaporation which has been 15 × 109 m3/year over the last 10 years. Reducing by 50 × 109 m3 the over-year storage would reduce the evaporation by 4 × 109 m3/year and avoid huge losses of water by the Toshka spillway, as occurred recently. The overall saving will be 5 × 109 m3/year. Extra storage in Lake Victoria may be used through the Sudd Canal and would balance the 50 × 109 m3 reduction of over-year storage at Aswan. The cost of this solution would be limited to:
• An earlier powerplant investment (US$100 or 200 million) in Uganda because the storage increase at Lake Victoria may reduce the hydropower supply there over some years.
• A loss of power at Aswan close to 300 GWh/year, because the loss caused by reducing the head is higher than the power from the 5 × 109 m3 of water saved.
6.6. Technical and political agreements
6.6.1 First phase
Reaching an overall agreement between nine countries for sharing the Nile water, as well as the benefits and drawbacks of future dams or canals, may be a long and difficult process. It should preferably not delay the Sudd Canal, which is of great benefit for Egypt and Sudan; they could pay for it with 1 per cent
of their oil revenue during construction, thus avoiding the possible delays associated with foreign financing.
The canal could be completed by 2020 and would have no direct impact on other countries. But it favours a possible review of the rigid and questionable agreement between Uganda and Egypt for the management of Lake Victoria Lake. This could lead to a significant water saving in the Aswan Lake and would help to optimize the management of the Sudd, preventing extreme floods and droughts there.
The total water saving in 2020 could be 15 to 20 × 109 m3/year.
Hydropower dams could be built before 2020 in various countries; most seem to have limited impacts on other countries and such impacts may be favourable.
6.6.2 Second phase
A second phase, after 2020, and based on the experience of the first phase, could include:
• dams and canals to retain water from theWest and East Swamps of southern Sudan;
• possibly a dam on Albert Lake; and,
• a reduction of losses in Sudan’s reservoirs, especially Djebel Aulia.
Relevant savings could reach 10 × 109 m3/year. Corresponding studies are not urgent.
6.7 Water sharing
The key political problem is the sharing of the Nile water between nine countries. This would be easier if sharing 100 × 109 m3/year instead of 72 × 109 m3. This would give the possibility of significant shares for the upstream countries and an increase of the water available for Egypt and North Sudan instead of a reduction.
The Sudd Canal is the key investment: it deserves an early design optimization and an analysis of the benefits and impacts which could be the basis of an agreement between Egypt and Sudan.
By the middle of this century, hydropower could be supplying, at low cost, most of Central Africa’s electricity. It could also be essential for the necessary storage to support intermittent energy sources, mainly solar, to be much developed in North and South Africa after 2020. These possibilities have been presented in the context of the potential of the four main African rivers, in countries representing half of the African population and with the greatest amount of hydropower potential. But similar problems and solutions apply to most other African rivers.
The utilization of the rivers can be optimized for irrigation; specific solutions may be applied for the Nile and Niger. Floods mitigation and navigation improvements are also important related benefits.
Aqua~Media International Ltd, Hydropower & Dams World Atlas & Industry Guide, 2010.
Le Nouvel Observateur, “Atlas économique:, Paris, France;2011
ICOLD, International Symposium on High Aswan Dam, Cairo, Egypt; 1993
Collins, R.O., “The waters of the Nile”; 1990.
Mali Committee on Large Dams, “Suggestions pour la production électrique du Mali”; 2010.
F. Lempérière has been involved in the construction and/or design of 15 hydraulic structures on large rivers including: Cabora Bassa in Zambezi, redesign and excavation for the Jonglei canal in Sudan, the hydropower plant at the Old Aswan Dam and studies for Niger Energy. He is Chairman of the ICOLD Committee on Cost Savings in Dams and of HydroCoop, a not-profit association providing advice on dams and floods.
X. de Savignac has been involved in the construction of large hydraulic structures such as Cabora Bassa (on the Zambezi river). He was also in charge of the excavation of the 350 km long Jonglei Canal, bypassing the White Nile.
Hydro Coop, 4 Cité Duplan, Paris 75116, France.