6.1. Power Generation

Africa has some of the best renewable energy potential in the world. These include wind, solar and significant remaining hydropower potential. Good geothermal potential is geographically limited to the Rift Valley. Bioenergy includes bagasse processing in sugar cane producing countries. The potential for solar is good virtually everywhere but other resources are confined to specific parts of the continent.

Costs vary by project and by site. In case of remote siting, transmission systems must be added. Costs for renewables tend to come down over time as learning effects occur with increasing installed capacity.

Costs for African projects tend to be higher than in other countries due to the need to import equipment, transportation costs and import levies. Where elements can be constructed locally, costs are reduced. This is the case for items such as dams for hydropower projects, foundations and towers for wind turbines.

Equipment from OECD countries tends to be more expensive than equipment imported from China and India. However, quality also varies and so does the energy yield, as well as operation and maintenance costs. Given the scarcity of operation and maintenance skills, there is a trade-off.

Important economies of scale exist. Larger plants tend to be cheaper per unit of capacity. Typically, a size increase of one order of magnitude reduces the unit capital cost by half.

Costs for new renewable energy equipment tend to decrease rapidly. For each doubling of installed capacity, costs tend to come down by a fixed percentage. Much progress has been achieved in recent years in the reduction of the capital cost for solar. These considerations mean that cost projections are not straightforward.

A typical measure for cost comparison is the levelised cost of electricity (LCOE). This excludes any subsidies or taxes and treats all electricity as being equally valuable, be it supplied to meet base load or peak load. It does not account for the cost of grid integration, such as backup or storage capacity in the case of intermittent renewables. At the same time, external effects are excluded. This is therefore only one indicator of electricity cost. Cost will, in reality, be higher.

The cost of financing is also a key factor. The calculation in Table 2 assumes a 12% real cost of financing. With high political risk and high inflation, the financing cost may be twice as high or even higher. However, concessional loans and risk guarantees can reduce the cost of capital. This is very pertinent for capital-intensive renewable energy technologies.

The cost of working capital is significant and can drive overall costs significantly higher. This is particularly true for large projects that require substantial infrastructure works, for example hydropower dams. The longer the time between the initial start-up of the project and completion, the higher the working capital cost.

Solar PV and solar CSP have been corrected for dust, heat impacts on performance and for gradual degradation as the equipment ages (Performance Ratio PR in Table 2).


1 USD 6,000/kW for 100 MW Shams-1 plant in Abu Dhabi

2 USD 3,750/kW for 100 MW Cape Town wind park project in South Africa, USD 2,175/kW for 300 MW Turkana wind park project in Kenya. This includes all project cost (AFD, 2011).

3 USD 7,000/kW for 185 MW Olkaria project expansion in Kenya (AFD, 2011).

5 Given a 20% capacity factor, a 1 kW panel produces 1,740 kWh. If half of the electricity is stored, evenly divided over the days of the year, 2.4 kWh of daily storage is needed. If battery decharging is limited to 25%, 10 kWh of battery storage capacity is needed. For deep-cycling lead acid batteries this capacity costs USD 1500.

6 The kW price refers to standard testing conditions of 1000 W/m2. In practice average irradiation in Africa ranges from 5-6 kWh/day. Therefore the test condition with a 20% capacity factor would have an annual irradiation of 1,752 kWh. In reality the irradiation is 1,825-2,190 kWh. Therefore the panel in Africa will yield in practice 5-20% more electricity than under the test conditions.

As illustrated in Table 2, large hydropower is a clear winner in terms of production cost. It is followed by biomass co-combustion. However, this excludes the cost for grid connection, transmission and distribution. If these are added, the decentralised solutions seem cost-competitive.

All listed options have costs that are below those for diesel generators. However, for most options, the costs are somewhat higher than for coal- or gas-fired power plants. For fossil-fuelled plants, the fuel costs are highly variable and depend on market developments. Such risk does not exist for renewables plants. Also, renewa-bles plants tend to be smaller, which reduces transmission cost and allows a more gradual expansion avoiding major supply variations. Decentralised generation reduces the risk of massive power outages.

But LCOE will not be the only factor that determines successful uptake of renewable electricity. Consumer ability to pay will also be a key factor. According to the World Bank:

"Surveys held in Mali concluded that the willingness to pay for electricity in rural areas averaged EUR 11.1/month (about USD 15), ranging from EUR 8.2 to EUR 16.7 (about USD 11 to USD 22.5) (Mostert, 2008). In Senegal, most rural households already spend USD 2-24 per month on kerosene and dry cell batteries to meet their lighting and small power needs, and hence are likely to be willing and able to pay for electricity use (de Gouvello et al., 2007). In Guinea, rural surveys obtaining data on avoided costs found that the willingness to pay for basic electricity services was about USD 1.6/month (Mostert, 2008), which would cover the cost of 12 kWh per month at the average tariff of the Sub-Saharan region (USD 0.13/kWh)." (WB, 2010).

Any form of electricity is a major expense under such conditions. A logical response is an emphasis on least cost solutions. However, this can be misleading. As stated earlier, LCOE is not always the right measure for economic evaluation of projects, and it is recommended that other economic valuation methods be explored.

4 Assumes 15% annuity plus 5% O&M. Excludes inflation and taxes/subsidies