IV. Africa’s Power Supply Crisis: Unraveling the Paradoxes

International Monetary Fund. African Dept.
Published Date:
April 2008
  • ShareShare
Show Summary Details

Africa’s Power Sector in International Perspective

Sub-Saharan Africa faces major infrastructure challenges, the most severe of which are arguably those in the power sector. Not only is the region’s energy infrastructure meager compared with other regions but electricity service is costly and unreliable. Indeed, in recent years more than 30 of the 48 countries in the region have suffered acute energy crises. This chapter presents preliminary findings from the Africa Infrastructure Country Diagnostic (see Box 4.1 for further details), which aims to unravel the paradoxes of sub-Saharan Africa’s troubled power sector.

Box 4.1.Introducing the Africa Infrastructure Country Diagnostic

Comparatively little is known about Africa’s infrastructure sectors, with sparse coverage of information in most standard international databases. The Africa Infrastructure Country Diagnostic (AICD) aims to reverse this situation by creating a comprehensive infrastructure database for the continent and an associated body of analytical work. AICD is a two-year multistakeholder knowledge program—currently nearing completion—that is sponsored by the Infrastructure Consortium for Africa as well as the African Union, the New Partnership for Africa’s Development (NEPAD) and Regional Economic Communities (for example, EAC, WAEMU, and the Southern African Development Community). The project covers all major economic infrastructure—energy, information and communication technologies, irrigation, transport, and water and sanitation—in 24 countries that together account for 85 percent of the gross domestic product, population, and infrastructure aid flows of Sub-Saharan Africa. The scope of data collection and analysis for each sector and country include public expenditure, investment needs, and sector performance. The underlying data and associated studies will be made available to the public through an interactive Web site. It is expected that AICD will provide a baseline against which future improvements in infrastructure services can be measured, as well as a more solid empirical foundation for prioritizing investments and designing policy reforms in the infrastructure sectors in Africa. Phase II of the AICD project will extend the coverage to additional African countries.

The entire generation capacity of the 48 countries of sub-Saharan Africa, at 63 gigawatts (GW), is comparable to that of Spain. If South Africa is excluded, sub-Saharan African generation capacity falls to 28 GW, about the same as Argentina’s. Normalizing for population, and subtracting South Africa, the installed capacity of sub-Saharan Africa is only one-third of South Asia’s, and about a tenth of that of other developing regions (Figure 4.1a). Moreover, the region’s generating capacity has been stagnant for many years; its growth rates are barely half those in other developing regions (Yepes, Pierce, and Foster, 2008). To make matters worse, as much as one-fourth of sub-Saharan Africa’s plant is currently not in operating condition.

Figure 4.1.Evolution of Power Infrastructure in Sub-Saharan Africa Relative to Other Regions

Rates of electrification are correspondingly low. Some 24 percent of the sub-Saharan African population has access to electricity versus 40 percent in other low-income countries, and electrification is proceeding more slowly than in other low-income countries (Figure 4.1b). Electricity consumption in the region is a fraction of consumption in other regions (Figure 4.2) and, excluding South Africa, is only about 124 kilowatt hours (kwh) a year, less than one-tenth of China’s. Although electricity tariffs in some sub-Saharan African countries have been kept low, the cross-country average tariff is rather high at US$0.13 per kwh—about double those in other parts of the developing world and almost as high as in OECD countries. Nevertheless, the prices fail to cover costs.

Figure 4.2.Electricity Prices and Consumption in Sub-Saharan Africa Relative to Other Regions

As a result of such low power consumption, the contribution of sub-Saharan Africa’s power sector to global carbon dioxide emissions is no more than 520 million tons a year, with South Africa being by far the major contributor. In all other sub-Saharan African countries, the bulk of greenhouse gas emissions come from land use and deforestation. While power consumption in the region will need to grow substantially to meet unsatisfied demands, a significant share of the increment could be met from hydropower, thereby mitigating the climate change impact. For example, in Southern Africa alone, it has been estimated that more regional trade could reduce incremental carbon emissions by 40 million tons a year.

Unreliable supply adds to the cost. African manufacturing enterprises report power outages on an average of 56 days a year,1 costing firms 5–6 percent of revenues. That is why many firms operate their own diesel generators, at a cost of about US$0.40/kwh. In the informal sector, where firms rarely have the capital for backstop generation, lost revenues from power outages can be as high as 20 percent.

Deficient power infrastructure dampens economic growth and weakens competitiveness, for example, through the detrimental effect on productivity. Escribano, Guasch, and Pena (2008) estimate the impact of infrastructure on firm productivity relative to other variables and also decompose the contribution of various components of infrastructure. They find that in most sub-Saharan African countries infrastructure accounts for 30–60 percent of the adverse impact on firm productivity, well ahead of factors like red tape and corruption. Moreover, in half the countries analyzed, power accounted for 40–80 percent of the infrastructure effect. In another study (Calderon, 2008), simulations based on panel data show that if the quantity and quality of power infrastructure in all sub-Saharan African countries were improved to that of a better performer (Mauritius), long-term per capita growth rates would be 2 percentage points higher.

The scarcity of power in sub-Saharan Africa also affects both the delivery of social services and the quality of life. Without electricity clinics cannot safely deliver babies at night or refrigerate essential vaccines. Lack of illumination restricts the ability of children to study at night and fosters crime in periurban areas.

Africa’s Acute Power Problems

Africa’s overstretched electricity systems have become exceedingly vulnerable to supply shocks, resulting in widespread outages and load shedding (Figure 4.3). With economic growth in the past decade raising demand for electricity, the lackluster expansion of generation and transmission facilities has stripped away any cushion from excess capacity that may have existed previously. In recent years, when droughts reduced power in the hydro-dependent countries of East Africa, prolonged blackouts became commonplace. In countries like South Africa, plant outages for maintenance—in a context of low reserve margins—have had serious consequences (Box 4.2). Countries whose power infrastructure has been damaged by conflict have also suffered severe shortages. And high petroleum prices have created enormous cost pressure in countries like those of West Africa that depend on imported oil products for generation.

Figure 4.3.Countries Affected by Acute Power Sector Crises in 2007

Box 4.2.Regional and Economic Effects of South Africa’s Power Supply Crisis

South Africa has long been a sizable producer of low-cost electricity, reflecting its abundant coal reserves. It is by far the region’s biggest producer and consumer of electricity, accounting for over half of electricity production in sub-Saharan Africa. Electricity prices for both households and industry are exceptionally low; this has been an important factor in the development of South Africa’s energy-intensive mining and mineral processing sectors.

South Africa’s electricity supply has remained stagnant in recent years but demand has continued to increase, resulting in power shortages. Attempts to encourage greater investment in generation by the private sector proved unsuccessful (partly because South Africa’s low electricity tariffs were unattractive to independent power producers) and at the same time resulted in delays in investment by the state-owned electricity provider, Eskom. As a result, the spare capacity (“reserve margin”) in the system to cope with peaks in demand has declined, leaving the country prone to periodic rounds of rolling power cuts, sometimes with very little warning.

This has resulted in gridlock on the roads as traffic lights fail; millions of rand lost because businesses cannot operate; and houses regularly without power for up to 12 hours. Electricity supply to large industrial users was also reduced in January, resulting in a temporary shutdown of production in the mining sector, causing global prices for gold and platinum to spike. South Africa exports about 5 percent of its electricity production to neighboring countries—such as Botswana, Namibia, and Swaziland—that import at least half of their electricity needs from South Africa. These countries have been affected by a similar regime of rolling blackouts; moreover, some South African opposition and union groups have called for a complete halt on power exports.

The government’s response to the crisis was a series of measures to manage demand in the short and medium term until new capacity comes on stream. This will involve power rationing, modeled on Brazil’s response to its energy crisis in 2001, with a view to reducing demand for electricity by 12.5 percent. Large mines have already been rationed to 90 percent of their normal electricity supply. Eskom also plans to increase its generating capacity by some 50 percent over the next 9–10 years. Electricity prices are likely to increase substantially over the next several years to help finance investment (and reduce demand). Nevertheless, the supply-demand balance is likely to remain tight for the next few years.

Source: Clément and Shanaka (2008).

An increasingly common response to the crisis has been short-term leases for emergency power generation by a handful of global operators (Table 4.1). Though this capacity can be put in place within a few weeks, it is expensive. The costs of small-scale diesel units, for example, are typically about US$0.35/kwh. The equipment is typically leased for up to two years, after which it reverts to the private provider. An estimated 700 megawatts (MW) of emergency generation are currently operating in sub-Saharan Africa; this represents more than 20 percent of installed capacity. The total price tag ranges from 0.5 percent of GDP in Gabon to 4.3 percent in Sierra Leone.

The recent energy crises are symptoms of a deeper malaise, the causes of which need to be understood and addressed. Four paradoxes shed light on the complex challenges that need to be faced: abundant energy but little power; high prices but even higher costs; widespread but ineffective reform; and high expenditure yet inadequate financing.

Table 4.1.Emergency Power Generation in Sub-Saharan Africa
CountryDateContract DurationEmergency CapacityPercent Total Installed CapacityEstimated Annual Cost as % of GDP
Angola20062 years15018.11.04
Ghana20071 year805.41.90
Kenya20061 year1008.31.45
Madagascar2004Several years5035.72.79
Rwanda20052 years1548.41.84
Senegal20052 years4016.51.37
Sierra Leone20071 year20133.34.25
Tanzania20062 years18020.40.96
Uganda20062 years10041.73.29

Paradox 1: Abundant energy but little power

Ironically, sub-Saharan Africa is richly endowed with both renewable and exhaustible energy resources. At present, for instance, it exploits only 8 percent of its gross hydropower potential of 3.3 million gigawatt-hours (gwh) annually. The countries on the Gulf of Guinea hold 4.9 percent of the world’s proven oil reserves (some 60 billion barrels) and 7.8 percent of proven natural gas reserves (some 14 trillion cubic feet); if converted to electricity, the natural gas currently flared during oil production could itself meet a substantial share of Africa’s power needs. Southern Africa is rich in coal; Botswana, South Africa, and Zimbabwe hold 5.6 percent of the world’s proven reserves (more than 50 billion tons). There is also significant geothermal potential in the Rift Valley.

However, the continent’s energy resources tend to be concentrated in a handful of countries where physical and political barriers to trade make it difficult for them to access centers of power demand—and their economies are too small to for them to develop their own resources. For example, the Democratic Republic of Congo (DRC) alone accounts for about 40 percent of sub-Saharan Africa’s hydroelectric potential, and Ethiopia accounts for another 20 percent. But both are far from the economic centers in southern, western, and northern Africa, and the multi-billion-dollar investments needed to exploit hydro potential are too big for their economies.

Moreover, in most sub-Saharan African countries energy markets are too small to take advantage of efficiencies from large-scale electricity production. With today’s technology, full economies of scale in thermal power generation kick in at about 400 MW; in only 14 countries in sub-Saharan Africa do national power systems meet this threshold. In another 14, power systems have only 100 MW of capacity. With relatively little cross-border trade, many sub-Saharan African countries use technically inefficient forms of generation (Figure 4.4). In eastern and western Africa, about one-third of installed capacity is diesel-based generators. These countries have few domestic energy resources of their own, even though there are sufficient hydro and gas resources in neighboring countries to support much lower-cost forms of generation.

Figure 4.4.Drivers of Operating Costs for Sub-Saharan African Power Systems

The consequences of this technically inefficient pattern of power generation become evident when average operating costs of different types of power systems are compared (Figure 4.4b). The average for predominantly diesel-based power systems is as much as US$0.20/kwh more expensive than the cost of hydro-based systems. Similarly, the operating cost penalty for countries with national power systems of less than 200 MW installed capacity can run up to US$0.30/kwh relative to countries with power systems above 500 MW. An additional penalty for landlocked and island states relative to coastal nations is the much higher cost of importing fossil fuels.

That is why regional power pools have been formed in Central (CAPP), East (EAPP), Southern (SAPP), and West (WAPP) Africa. The pools are at very different stages of development, both technically and institutionally. The political process is most advanced in the WAPP, supported by political agreements at the head of state level through the ECOWAS. The pools, particularly the WAPP and SAPP, have facilitated significant cross-border exchanges of power. A number of countries, such as Botswana and Niger, rely on imported power; others, such as Nigeria and Mozambique, are major exporters of power. However, none of the pools is yet at the point where the arrangements are fully competitive.

Paradox 2: high prices but even higher costs

The variation in electricity charges across sub-Saharan African countries is huge, and spans some of the cheapest power in the world (at less than US$0.05/kwh in hydro-based systems and in South Africa based on cheap coal) to some of the most expensive power in the world (at over US$0.30/kwh in countries with diesel-based systems and landlocked or island locations, such as Chad and Madagascar). Nevertheless, across countries the average charges today look high by international standards and have increased recently, reflecting higher oil prices and tightening supply conditions worldwide. Average revenue has risen from US$0.07/kwh in 2001 to US$0.13/kwh in 2005. In countries reliant on diesel-based power generation systems, average revenues have risen from US$0.08 to US$0.17/kwh.

Yet the average revenue in sub-Saharan African countries reliant on diesel-based generation still falls significantly short of the average operating costs of US$0.27/kwh (Figure 4.5), even though average revenue in these countries has risen dramatically during the last five years, from US$0.08 to US$0.17/kwh.

Figure 4.5.Electricity Costs and Revenues by Type of Power System (US$/kwh)

Despite such comparatively high average revenues, the vast majority of sub-Saharan African countries are doing little more than covering average operating costs (Figure 4.6). The correlation between average revenue and average operating cost is as high as 90 percent, indicating that operating cost recovery is usually the driving principle behind power pricing. Nevertheless, once average operating costs exceed US$0.20/kwh, there is a tendency to price below the 45 degree line in Figure 4.6a. The implication is that past capital costs of power sector development have historically been almost entirely subsidized by the state.

Figure 4.6.Average Power Sector Revenue and Various Cost Benchmarks

Nevertheless, a comparison of current average revenues and average operating costs misrepresents the long-term cost recovery situation for two critical reasons. First, because of major inefficiencies in revenue collection, the average revenue collected per unit of electricity sold is substantially lower than the average effective tariff that is being charged. Second, owing to major inefficiencies in generation technology and the growing trend toward regional trade, the average incremental cost of power in sub-Saharan Africa for many countries is somewhat lower than the average historic cost of power production (including both operating and capital costs). Thus, a truer picture of the long-term cost recovery situation is gained by comparing the average effective tariff with the average incremental cost as in Figure 4.6b. The result shows that for many (though certainly not all) countries, even the current tariff would be adequate for cost recovery purposes if only revenues could be fully collected, and the power system could transition toward a more efficient structure of production.

The presence of large historical capital subsidies to the sector raises questions about the distributional incidence of these. In a recent study, Wodon and others (2008) use evidence from household surveys to establish the distribution of such subsidies in 18 sub-Saharan African countries. In all the countries, power sector subsidies are found to be highly regressive. Across the bottom half of the income distribution, barely 10 percent of households have access to electricity; because the poor are almost entirely excluded from service, they cannot possibly benefit directly from subsidies (Figure 4.7).

Figure 4.7.Electricity Service Coverage in Sub-Saharan Africa

In urban areas of low-income countries, about a third of households lack access to electricity. At least half of this unserved urban population areas lives close to an electricity grid, suggesting that demand-side barriers—such as connection charges or household tenure—are contributing to restrict access (Banerjee and others, 2008).

In rural areas of low-income countries, only 12 percent of the population on average has access to electricity, and for at least 17 countries the figure is below 5 percent. The dispersed nature of the rural population means that grid extension does not always prove economical, although a handful of countries—most notably Ghana and South Africa—have had successful large-scale grid-based electrification programs, based on technically and financially strong utilities and careful policies to address affordability.

The concentration of power services in upper-income echelons might suggest that full cost recovery pricing is feasible. However, the reality is more complex. In low-income sub-Saharan African countries, even households in the highest income quintile have monthly budgets of only US$260 to support families typically comprising five people.

Banerjee and others (2008) estimated the electricity affordability problems of sub-Saharan African households in different pricing scenarios, assuming modest consumption of 50 kwh/month. Bills are considered affordable if they do not absorb more than 5 percent of household budgets. With cost recovery prices of about US$0.25, as is currently the case in high-cost countries, a subsistence monthly bill would be US$12. Except in a relatively small group of middle-income and better-off low-income countries (Cameroon, Cape Verde, Côte d’Ivoire, Republic of Congo, Senegal, and South Africa), a substantial share of the population would be unable to afford cost-recovery tariffs. Today, household spending on electricity service is significantly below this level (Figure 4.8). However, if costs could be reduced to US$0.12/kwh—in line with sub-Saharan Africa’s average incremental cost of power—the monthly bill of US$6 would be affordable to most of the population, except in the lowest-income countries like Burundi, Democratic Republic of Congo, Ethiopia, Malawi, and Uganda.

Figure 4.8.Electricity Service Expenditure in Sub-Saharan Africa

Although residences account for 95 percent of power utility customers in Africa, they contribute only about 50 percent of revenue. Thus, the pricing of power to commercial and industrial consumers is just as important to cost recovery. The average revenue raised from low- and medium-voltage customers seems to be similar, but high-voltage customers tend to pay only about half as much. Globally this relative price differential is not unusual. It reflects the fact that high-voltage customers do not use as much of the distribution network and hence do not create such high costs for the utility. Nevertheless, it suggests that in absolute terms neither residential nor commercial and industrial customers are close to paying full cost-recovery prices. Moreover, a number of sub-Saharan African countries have historically sold power at highly discounted rates to large-scale industrial and mining customers like the aluminum smelting industry in Cameroon and Ghana and the mining industry in Zambia. These arrangements were initially justified as ways of locking in base load demand to support large-scale power projects that went beyond the immediate demands of the country, but they are increasingly questionable as demand has grown to absorb capacity.

Given the problems of sub-Saharan African power systems, cost recovery needs to be discussed along with measures to reduce costs, improve revenue collection, and increase reliability.

Paradox 3: Widespread but ineffective reform

Although they are somewhat behind the reform programs in other regions of the world, sub-Saharan African countries also embarked upon the path of power sector reform orthodoxy, which included reform legislation and sector restructuring to pave the way for competition in generation and private sector participation across the electricity supply chain. As of 2006 more than 80 percent of sub-Saharan African countries had enacted a power sector reform law, 75 percent had experienced private participation in power, about 66 percent had corporatized their state-owned utilities, more than half had established a regulator, and more than a third had independent power producers (Figure 4.9a). Yet few countries have adopted the full range of reform measures (Figure 4.9b).

Figure 4.9.Evaluation of Power Sector Reform

The lack of results has forced a rethinking of whether certain reform principles and programs apply in sub-Saharan Africa. One reform that has not been widely adopted in sub-Saharan Africa is unbundling generation, transmission, and distribution functions to create competition in generation and supply. Besant-Jones (2006) in his global review concluded that restructuring the power sector to advance competition only made sense in countries large enough to support several generators above minimum efficient scale. The power systems in most sub-Saharan African countries are so small that this prescription is largely irrelevant for them. Nevertheless, even in the largest countries, where unbundling could work, there has not been much progress.

There have been nearly 60 medium- to longer-term power sector transactions with the private sector in sub-Saharan Africa, not counting emergency power generation leases (Table 4.2). Almost half of these have been independent projects, with the utility signing Power Purchase Agreements with the private sector to build green field generation plants. With more than US$2 billion of private investment, these have provided nearly 3,000 MW of new capacity—a substantial contribution to available capacity. An independent assessment concluded that these projects can be relatively costly owing to technology choices, procurement problems, and currency devaluation, and are often subject to renegotiation (Gratwick and Eberhard, 2007). A poorly documented issue is the extent to which Power Purchase Agreements are creating contingent liabilities for the state.

The rest of the transactions have been concession, lease, or management contracts, typically for the operation of the entire national power system. These have had a relatively high failure rate; about one-third of the contracts are currently in distress or already cancelled. However, in the more successful transactions performance has improved noticeably. The usual reasons for failure are the lack of financial viability or creditworthiness of the utilities—governments have been unwilling or unable to adjust tariffs to enable cost recovery or pay subsidies to make up the difference—and the lack of access to funding for priority investments to improve efficiency or expand services. The fundamental factors to make private participation work were absent.

Table 4.2.Private Participation Power Sector Transactions in Sub-Saharan Africa
Type of Private ParticipationCountries AffectedNumber of TransactionsNumber of Problem TransactionsTotal Value of Transactions (US$m)
Management or lease contractChad, Gambia, Gabon, Ghana, Guinea-Bissau, Kenya, Lesotho, Madagascar, Malawi, Mali, Namibia, Rwanda, São Tomé and Príncipe, Tanzania, Togo1745
Concession contractCameroon, Comoros, Côte d’Ivoire, Gabon, Guinea, Mali, Mozambique, Nigeria, São Tomé and Príncipe, Senegal, South Africa, Togo, Uganda1251,598
Independent power projectAngola, Burkina Faso, Congo, Côte d’Ivoire, Ethiopia, Ghana, Kenya, Mauritius, Nigeria, Senegal, Tanzania2422,293
DivestitureCape Verde, South Africa, Zambia, Zimbabwe4938
Source: World Bank Private Participation in Infrastructure database (2007).Note: Problem transactions are defined as projects that are now distressed or were prematurely cancelled.
Source: World Bank Private Participation in Infrastructure database (2007).Note: Problem transactions are defined as projects that are now distressed or were prematurely cancelled.

Perhaps the single most relevant institutional consideration is the governance of the national power utility. It is possible to rate sub-Saharan African power utilities by the extent to which they are managed on sound commercial principles (IMF, 2004a). The rating is based on whether utilities have managerial autonomy with respect to (i) labor policy and (ii) market decisions relating to production and sales; whether utilities are financially viable, measured in terms of (iii) the absence of subsidies; and (iv) tax breaks and the requirement to be (v) profitable and pay (vi) market rates for debt; and whether utilities are accountable, producing (vii) published audited accounts and being (viii) publicly listed on the stock exchange to protect the rights of (ix) minority shareholders.

These good governance practices are not widespread in sub-Saharan African utilities (Figure 4.10), though the majority do report freedom with respect to labor policies, and a sizable minority can make their own market decisions. While most utilities report that they are required to be profitable and pay market rates for debt, in practice the vast majority benefit from sizable subsidies and tax breaks and are not in a position to borrow at all. Only 60 percent publish audited accounts, and stock exchange listing is unheard of. The typical utility in the sample meets only about half the criteria (Figure 4.10b).

Figure 4.10.State-Owned Enterprise Governance Characteristics

Poor governance is reflected in deficient performance. In well-performing utilities around the world, system losses can be as low as 10 percent; two-thirds of sub-Saharan African utilities report losses of more than 20 percent. Similarly, well-run utilities collect close to 100 percent of what is owed them, while 40 percent of sub-Saharan African utilities collect less than 90 percent (Figure 4.11).

Figure 4.11.Frequency Distribution of Power Sector Efficiency Indicators

The inefficiency of sub-Saharan African utilities generates substantial hidden costs. These hidden costs can be quantified (Ebinger, 2006) by comparing the revenues a utility raises against those raised by an ideal utility that prices at full economic cost and keeps distribution and collection losses to best practice levels. In many sub-Saharan African countries, hidden costs can be as high as 2 percent of GDP (Figure 4.12a). About 50 percent of the costs stem from collection losses and another 30 percent from distribution losses (Figure 4.12b). The dividend from improving utility performance is often very high. The contribution of under-pricing to these hidden losses is relatively small, although it varies by country.

Figure 4.12.Hidden Costs of Power Sector Inefficiency

Given the large scale, long lead times, and extensive preparation required to build power infrastructure, careful planning is crucial. However, many sub-Saharan African countries lack ministerial capability for long-term power sector planning. The current power shortages were to a large extent foreseeable, but action was not taken far enough ahead to avert them. Even today—notwithstanding the strong case for power sector development—there is a shortage of bankable electricity generation projects because of bottlenecks in project preparation. To some extent power sector planning has been a casualty of the 1990s reform model that emphasized market-led infrastructure development and allocation of human resources to regulatory rather than planning functions.

Effective power sector planning recognizes critical upstream linkages with fuel supply industries. Security of supply is subject to interrelated infrastructure and incentive problems. For countries with access to natural gas, lack of pipeline capacity and vandalism have been a growing concern. In Nigeria, for example, these problems have reduced the gas available for domestic electricity generation by independent power producers and limited gas trade in the WAPP. Inadequate incentives in gas pricing have also deterred private investment in infrastructure to gather natural gas and in pipelines. For countries that rely on imported diesel fuel, deficient port facilities and transport links add greatly to their costs. The lack of competition and transparency in fuel procurement also exacerbates costs. The OECD Competition Committee has flagged collusive tendering of oil and price fixing as major issues in a number of African countries.

Paradox 4: High expenditure but inadequate finance

Sub-Saharan African countries on average spend 2.7 percent of GDP on power; a significant number spend more than 4 percent (Table 4.3). Typically, more than 90 percent of this spending is channeled through the national state-owned utility; less than 10 percent appears on the central government budget. Operating costs absorb 75 percent of total spending. As a result, public investment in the sector is very low—on average only 0.7 percent of GDP.

The contribution of official development assistance (ODA) to public investment in power has been modest, averaging only US$700 million a year for the last decade. Support has been highly volatile: only a few hundred million dollars a year in the late 1990s but rising toward US$1 billion a year in recent years. Despite the substantial number of private transactions, their value has averaged only about US$300 million a year for the last decade, and once again the flows have been highly volatile because these investments are lumpy. Thus total external capital flows to the power sector in sub-Saharan Africa amount to no more than 0.1 percent of the region’s GDP (Figure 4.13).

Figure 4.13.Long-Term Trends in External Finance for the Sub-Saharan African Power Sector

Sources: OECD (2006); Infrastructure Consortium for Africa (2007); and World Bank Private Participation in Infrastructure Database (2007).

In recent years, the China Ex-Im Bank has emerged as a major new financier of power infrastructure in sub-Saharan Africa. Between 2001 and 2006 Chinese commitments averaged US$1.7 billion a year, more than ODA and Private Participation in Infrastructure (PPI) combined and equivalent to about 0.2 percent of the region’s GDP. Most of the Chinese financing has gone to six large hydropower projects with a combined generating capacity of over 7,000 MWs. Once completed, these projects will increase sub-Saharan Africa’s installed hydropower capacity by 40 percent. China is also financing 2,500 MW of thermal power, and the India Ex-Im Bank has financed major thermal generation projects in Nigeria and Sudan.

Numerous econometric analyses show that the elasticity of power sector demand with respect to economic growth is close to unity. With GDP growth rates in sub-Saharan Africa averaging above 5 percent a year in recent years, power generation capacity should be growing at a similar rate to keep pace with demand. However, since 1980 annual growth in generating capacity has averaged only 2.9 percent.

A recent study constructs a series of optimization models for each of Africa’s major regional power pools to simulate the expenditures required (Econ Analysis, 2008). The model is flexible enough to consider different assumptions about the extent of regional power trade, the pace of economic growth, the extent of political ambitions for universal access, and the price of inputs like oil and gas.

Table 4.3.Power Sector Expenditure(Percent of GDP)
TotalCentral GovernmentState-Owned EnterprisePublic InvestmentOperating Costs
Lower quartile1.900.041.830.201.70
Upper quartile3.450.363.290.792.65
Source: Briceno-Garmendia and Smits (2008).
Source: Briceno-Garmendia and Smits (2008).

Table 4.4 reports baseline results for a scenario in which full advantage is taken of regional power trade and all countries aim for an access rate of 35 percent by 2015. Each year sub-Saharan Africa would thus need to add about 3,000 MW of new generation capacity and connect about 3 million new households. This scenario costs 6–7 percent of sub-Saharan African GDP, equivalent to US$47 billion annually, split almost evenly between investment and operations and with about two-thirds of the cost coming from power generation needs. Current spending on power averages less than 3 percent of GDP. Investment accounts for more than half that, equivalent to at least 2 percent of GDP, compared with current power sector investment amounting to less than 1 percent of GDP on average. Despite significant expansion in access, the bulk of the expenditure is associated with generation.

The regional averages conceal huge variations between countries. As power trade grows, the burden of investment falls disproportionately on countries with abundant resources. In a handful of cases the annualized expenditure requirement exceeds 10 percent of GDP, mostly for investment in generation for export. The most prominent examples are Ethiopia and the Democratic Republic of Congo, which would each become major exporters of hydropower in their pools. The financing would not necessarily need to be raised from domestic resources but could be underwritten to some degree by importing countries.

Table 4.4.Annualized Power Sector Expenditure Requirements to 2015(Percent of GDP)
TotalInvestmentOperating ExpenditureGenerationTransmission and Distribution
Source: Africa Infrastructure Country Diagnostic database.

Preliminary estimate for the whole of sub-Saharan Africa based on results currently available for EAPP and SAPP.

Source: Africa Infrastructure Country Diagnostic database.

Preliminary estimate for the whole of sub-Saharan Africa based on results currently available for EAPP and SAPP.

There is major potential for expansion of cross-border power trade. For example, in the SAPP alone, the volume traded internationally could rise from the current 45 terrawatt hours (TWh) to 141 TWh a year if trade were exploited to its full economic potential.

Trade necessitates investments in cross-border transmission links but also allows for significant savings from accessing lower-cost power sources. It is therefore possible to calculate the gains from trade as the rate of return on cross-border investments. These have been estimated at 20 percent in Eastern Africa to 167 percent in Southern Africa, in both cases exceeding typical hurdle rates for public investment (Econ Analysis, 2008). While such trade would still only represent about 8 percent of total power demand, in such trading scenarios some smaller countries would depend on imports to meet more than 50 percent of domestic demand.

The savings in the annualized cost of the power sector from trade are relatively small at less than 10 percent, but the gains for individual countries in terms of cheaper power can be substantial. Most countries would reduce the average cost of power by a few cents/kwh—a 20–60 percent saving. For a handful of countries power costs would be reduced by more than 60 percent, or US$0.10/kwh.

The main effect of trade in power is to support development of more large-scale hydropower schemes that would not be viable for a single nation. As a result, the composition of the generation portfolio shifts toward hydropower by 10–15 percentage points relative to the case without expanded trade. The additional hydropower would displace natural gas generation in Eastern Africa and coal generation in Southern Africa. It would also increase the share of power production coming from export countries like Ethiopia and the Democratic Republic of Congo.

Irrespective of trade development, however, the major power consumers—Egypt, Nigeria, and South Africa—continue to be by far the main producing countries in their regional power pools.

The Way Forward

The power sector in Africa is characterized by a set of paradoxes. There are abundant sources of power, significant levels of government funding, and notable efforts at reform. Yet electricity access rates are very low compared with other developing regions, prices are high, and the power supply insufficient and unreliable.

The policy choices that best address these paradoxes are not clear-cut. The traditional model that predominates in the sub-Saharan African power sector—vertically integrated, state-owned monopolist utilities—has yielded disappointing results. Yet reform to increase efficiency and boost competition through private participation has in many cases failed to deliver the expected results: unbundling is limited, failures of transactions and projects have been frequent, and there has been minimal additional investment.

The lesson that emerges is that success in tackling the challenges is not a simple function of the model adopted. The power sector in Africa needs to move to a “mixed economy,” characterized by a range of structures, regulation, and technologies adapted to each country’s context. Successful interventions will tackle several problems simultaneously to put the sector on a positive trajectory of improved sector and utility management, financial viability, new investment, and better customer service. This means recognizing that the power sector has quasi-monopolistic characteristics—particularly in grid-based distribution and to a lesser extent in transmission—and that incumbent utilities will continue to be the largest players in the sector for the foreseeable future. But interventions also need to be innovative and ambitious, recognizing that meeting customer needs means multiple providers, financial viability, and new forms of external financial assistance. Where certain preconditions are in place—including appropriate regulatory frameworks for public-private partnerships, reformed tariff frameworks, and sufficient security of investment for investors—sector reforms can do much to facilitate the entry of strategic private partners.

Consequently, the starting point is sustained and concerted action on three strategic priorities: (i) regional scaling-up of generation capacity, (ii) improving the effectiveness and governance of utilities, and (iii) expanding access through sector-wide engagement. The three are interdependent and must be tackled together. Efforts to boost generation and regional power trade will stumble if the utilities, which will continue to be central actors in the sector, remain inefficient and insolvent. Expanding electricity distribution systems without taking measures to tackle the shortages in generation and to improve transmission capacity would clearly be futile. And focusing exclusively on utility reform would be fruitless unless a start is made on substantial, long-gestation investments in both generation and access to improve quality of service and render the utilities viable. In short, these strategic priorities must progress together. At the same time, the time required to yield results from these actions is such that they need to be complemented by such short-term measures, including demand-side management (for example, the introduction of energy-efficient bulbs) and loss-reduction programs (such as enhanced bill collection and initiatives to tackle electricity theft).

Regional scaling up of generation capacity

The first strategic priority is to tackle the generation capacity deficit head-on. Africa’s considerable hydro, gas, and coal resources remain under exploited. The best way to scale up generation at the lowest unit cost is to develop a new generation of large-scale generation projects. An initial wave of projects could include candidates like Inga III in the Democratic Republic of Congo, which is expected to add about 3,800 MW in capacity, the Temane gaspowered plant in Mozambique (750 MW), Gilbe Gibe III hydropower in Ethiopia (1,800 MW), and further development of generation capacity based on natural gas from Nigeria. However, individual countries do not have the necessary investment capital, or even the electricity demand, to move forward with these large projects. A project finance approach predicated on regional power off-take, in which private sector participation and donor funding are blended, is needed.

Expanded generation capacity is redundant unless the power can be transmitted to users. This is where regional power pools play a critical enabling role. Challenges common to all the pools are rehabilitation and expansion of the cross-border transmission infrastructure to increase the potential for trade, and harmonization of regulations and system operating agreements. Equally important is the formulation of market trading mechanisms so that the additional energy generated from large projects can be priced and hence allocated in an efficient and fair way (for example, via competitive pool arrangements). While the economics of regional large-scale generation projects are convincing, they may give rise to significant political challenges. The gains from trade are much larger for some countries than for others, and considerations of self-sufficiency sometimes have more political weight than access to low-cost power. These factors need to be addressed early in the project development cycle.

Large-scale regional energy schemes have deep financing requirements. Capital expenditure for Inga III, for example, is estimated at US$4–5 billion. This is beyond the capacity of concessional financing from development partners, even after significant increases in aid. Private participation will be pivotal. Yet successful private investments in energy projects have been rare in Africa, and increased private investment will not materialize simply because of large infrastructure financing gaps. The lessons learned from past failures need to be addressed, as private investment will only flow where rewards demonstrably outweigh risks. Large-scale regional generation projects would have several attractions in this respect:

  • Large investments benefit from economies of scale: for a given amount of generation capacity, the total costs (design, engineering, capital items, civil works, safeguards and others) for one large plant are lower than for several smaller plants with the same aggregate capacity. All else being equal, investments in larger projects are therefore likely to be more profitable.

  • The investments would primarily be in standalone generation projects that present fewer risks than investments in vertically integrated utilities whose operational and regulatory risks (organizational inefficiency, lack of financial transparency, geographical distribution of personnel and assets, governance risks, political interference, and contingent risks like an uncertain legal framework) are far harder to price.

  • There is increasing realization that investment in new generation capacity cannot be undertaken in isolation from other efforts in the sector. Capital is more likely to be forthcoming in an environment where other factors—such as the tariff structure, power purchase agreements, and reliable transmission interconnections—have been addressed. Utilities that are better run will eventually be able to move beyond covering operating expenses to invest in system expansion, making the whole sector more viable. Public sector financiers like the World Bank are also becoming more nimble in their deployment of tools to help crowd in the private sector, such as risk-mitigation instruments.

There are early but encouraging signs that scaling up generation capacity through large private-sector-led projects is starting to gather speed. A prominent example is the privately owned 250 MW Bujagali hydro plant in Uganda, supported by World Bank Group guarantees and funded by a private consortium. At the same time, ambitious regional projects undoubtedly present technical, financing, and political risks, and will continue to be complemented by investments at the national level.

Improving the effectiveness and governance of utilities

Shortcomings in how the power sector operates lead directly to many of the suboptimal outcomes detailed in this chapter. Tackling these shortcomings will require improvements in the regulatory and tariff framework at the sector level, as well as better management in utilities.

The lack of strategic policy and planning for the electricity sector at the central government level is a critical weakness. Interventions have been piecemeal rather than integrated; for example, many countries have focused on generation without investing in efficient transmission and delivery of power. A well-articulated plan for the sector will allow governments to move beyond the “firefighting” that has reduced their ability to plan for exogenous shocks, such as drought or high oil prices.

Financial viability of incumbent utilities—and hence creditworthiness and access to domestic and international private capital—is important for the overall development of the sector. It demands that utility revenues allow at least the recovery of operating costs and ideally some contribution to capital costs. That means that in many cases tariffs need to be gradually adjusted to levels that will allow these goals to be met, while remaining sensitive to the needs and capacity to pay of poorer households. The corollary of tariff adjustments is the need to significantly reduce operating costs to lessen the financial burden on consumers of efforts to recover costs. Operational efficiency programs are needed to reduce the high rates of technical, nontechnical (electricity theft), and collection losses. These can include capacity building and technical assistance to improve management, business practices, and planning. Priority areas are improved load management (to better match supply with priority customer needs), theft reduction initiatives, and increased revenue collection (through enhanced metering and better-run customer service units). Capital expenditure can also be driven down by using low-cost technology standards, as undertaken in Mali and Guinea. Innovations have included adjusting technical design standards to meet the reduced requirements of low-load systems, maximizing the use of material provided by local communities (such as locally sourced wooden poles), and the use of local employees and supervisors recruited from the community.

Past efforts at improving utility management focused too heavily on technical issues to the exclusion of governance and accountability. Good governance practices in sub-Saharan African utilities are often observed in the breach. Transparency and accountability depend on solid financial management, procurement, and management information systems—for example, requiring the auditing and publication of financial accounts and the use of comprehensive cost-based accounting systems that allow functional unbundling of costs and a clearer sense of cost centers. Oversight and transparency also need to be enhanced by better corporate governance (for example, by reforming how senior managers are appointed, insisting on conflict of interest disclosures, and making staffing practices more transparent and effective). Many of sub-Saharan Africa’s newly established energy regulators can play an important role in this area, even in the absence of private participation in the sector. Reforms to ensure financial and managerial autonomy from state interference in commercial decisions are also needed in some countries.

In practical terms, even with appropriate tariffs, reduced operating costs, and better governance, the combination of expensive, maintenance-intensive equipment and the inability of even moderately wealthy households to pay for the full capital cost of domestic grid extension means that full cost recovery in Africa is not yet possible. Often capital subsidies will still be needed, whether from governments or donors. Governments must therefore be able to articulate, in their strategic policy framework, the economic benefits of subsidies to the sector, as well as the path to eventual full cost recovery. They must also recognize that some households connected to the grid cannot afford even the variable costs of their service; carefully targeted and calibrated lifeline subsidies may be a vital part of power sector strategy.

Expanding access to electricity through sectorwide engagement

The fact that power is often unavailable to lower-income groups means that those who do not have access are not benefiting from government or external financing. From a social, poverty reduction, and political perspective, it is therefore imperative to expand access. Yet financing expansion to lower-income households will further strain the financial viability of the power sector.

Tackling this dilemma will require both significantly higher concessional financing from development partners for access programs, and tariff increases and operating cost reductions. Given the scale of investments needed, a systematic approach to planning and financing new investments is critical. The current project-by-project, ad hoc approach in development partner financing has led to fragmented planning, volatile and uncertain financial flows, and duplication of efforts. Engagement across the sector in multiyear programs of access roll-out supported by multiple development partners as part of a coherent national strategy will channel resources in a more sustained and cost-effective way to the distribution subsector. Coordinated action by development partners will also reduce the unit costs of increasing access, as well as creating new sources of demand that will further make the case for large generation projects at the supranational level.

Since universal household electrification is still decades away in many countries, it is equally important that sector-wide programmatic approaches ensure that the benefits of electrification touch the poorest households, particularly deep in rural areas. While grid extension is often not economical for dispersed populations, off-grid models based on innovative renewable technologies can be cost-effective. For example, low-cost portable solar lanterns are one consumer product that could be accessible and affordable to the rural public, and the “Lighting Africa” initiative is supporting the development of the market. Solar-powered electrification of clinics and schools that provide essential public services to low-income communities is another way of bringing the benefits of investment in electrification directly to these communities.

Finally, it is important to recognize that most of the measures described above are medium term in nature, and cannot be implemented overnight. As strong economic performance continues to escalate power demand, many sub-Saharan African countries will continue to face a very tight demand-supply balance in the coming years. It is therefore critical that longer-term efforts to redress the underlying structural causes of sub-Saharan Africa’s current power supply crisis be complemented by shorter-term measures to soften the economic and social impact of power scarcity. Recent experiences from countries such as Brazil show that well-designed demand-side management measures (for example, a quota system with price signals, combined with a public energy efficiency campaign) can go a considerable way toward trimming peak demand, thereby substantially reducing the extent of power rationing at a relatively low economic and social cost.

Note: This chapter was prepared by the World Bank’s Africa Region Sustainable Development Department. The team consisted of Vivien Foster, Tjaarda Storm van Leeuwen, Cecilia Briceno-Garmendia, Daniel Camos, John Gabriel Goddard, Rob Mills, and Karlis Smits. The research draws on the Africa Infrastructure Country Diagnostic (AICD), a multistakeholder knowledge program supported by the Infrastructure Consortium for Africa. The AICD will include a much more extensive power sector review, to be published later in 2008. This work represents the views of the authors. David Dunn contributed from the IMF.

    Other Resources Citing This Publication