Feature: Botswana’s Options For Future Power Generation Introduction

This article was written by Dr Keith Jefferis with the active support of KANTRAS. Keith is a former academic and Deputy Governor of the Bank of Botswana. Keith works in Botswana, the Southern African region and further afield, and specializes in research, economic commentary and policy-related analysis in a variety of fields. 

During 2020, Botswana will need to make some crucial decisions regarding investment in new electricity generation capacity. Although the new capacity is not required immediately, energy generation investments are large and have a long lead time, and hence investment decisions need to be made soon. At one level the key decision can be presented in simple terms: does Botswana intend to continue relying on coal as its main source of electricity, or should a switch to large-scale solar power generation be made now? In other words, will the next big power investment be in coal or solar power generation? The Government is also developing an Integrated Resource Plan (IRP) for power generation, which will provide a masterplan for electricity provision over the next 20-30 years.

While the decision can be presented in simple terms, the issues to be considered in reaching a decision are complex. They include the economic and financial feasibility of the different options; the projected growth of electricity consumption and patterns of demand; a wide range of technical issues regarding power generation, emissions and electricity storage; other potential sources of supply and regional power markets; as well as the roles of different institutions in the power generation and distribution mix.

Also important are broader trends and concerns about global warming and climate change. This is now one of the world’s most pressing geo-political issues, and actions taken by Botswana have to be seen in this context as well as in terms of the commitments that Botswana has made as part of the international community. Partly because of the urgency of climate change issues, the environment within which power generation decisions are being taken is changing rapidly, from all perspectives, including technical, financial, economic, policy and political. Anticipating where these changes will lead is important in making decisions that will prove to be appropriate over the next 10-20 years.

Making appropriate power generation choices means balancing three, sometimes competing, objectives. These are:

1. Minimising the costs of electricity

2. Maximising the security and reliability of supply

3. Reducing carbon dioxide (CO2)/greenhouse gas (GHG) emissions

Figure 1: Balancing power supply objectives

Sometimes these objectives conflict. For instance, solar power is cheap and has zero GHG emissions, but does not provide security of supply. Adding electricity storage to provide security of supply makes the cost much higher. Coal provides more security of supply, at medium cost, but has high GHG emissions. This article discusses these issues and trade-offs in more depth.

Current power generation

Botswana has historically relied upon coal-fired power generation to provide the majority of its domestic electricity supplies. From 1989 until 2011 this was based on the 132 megawatt (mW) Morupule A power station, operated by the Botswana Power Corporation (BPC). In 2014 the much larger 600mW Morupule B power station was commissioned. Both Morupule A and B have been beset by technical problems. Partly as a result, these supplies have never been sufficient to meet total domestic demand, and have been supplemented by power imports from South Africa (which also generates most of its electricity from coal). There are also small amounts of power imported from Mozambique, Zambia and Namibia. At times, imports from South Africa have been constrained, and Botswana had to resort to (very expensive[1]) supplementary diesel generation capacity, and load shedding when total supplies from all sources were insufficient. There are relatively insignificant contributions from small-scale solar power generation. After being closed in 2011, Morupule A power station has been rehabilitated and is being recommissioned.

BPC currently has a monopoly over large-scale power generation, importation, transmission, and customer (end user) supplies. Energy policy is the primary responsibility of the Ministry of Minerals, Green Technology and Energy Security, with the Ministry of Environment, Natural Resources Conservation and Tourism playing a subsidiary role. The Botswana Energy Regulatory Authority (BERA) is the sector regulator. In principle, the law allows for the provision of electricity by Independent Power Producers (IPPs), but no such deals have yet been made by BPC.

Electricity consumption and demand

In general, electricity demand and consumption rises along with economic growth, with a general expectation that, as a country develops, the growth of demand for electricity will be somewhat higher than economic growth as, for instance, the rate of household electrification increases. In Botswana, electricity consumption increased on average by 4.6% a year over the two decades from 1998 to 2018, very similar to the average real GDP growth rate over the same period. However, the pattern of growth has changed over time, with relatively fast growth (7.5% a year) in electricity consumption from 1998-2008, but much slower growth (1.7% a year) from 2008-2018. The slower recent growth may be for a number of reasons, including supply constraints and load-shedding in some years, and the closure of the BCL mine – previously the country’s largest single consumer of electricity - in late 2016. There has also been an increase in self-provision by consumers, e.g. using diesel generators and rooftop solar panels. In 2018, total consumption of electricity through the national grid was 3,919 gigawatt hours (GWh), of which 70% was generated domestically and 30% was imported.

Figure 2: Electricity Consumption, 1998-2018

Of more relevance than total consumption is peak demand, i.e. the maximum load that the system will have to deliver at any point in time. Demand for electricity varies considerably over time - throughout the day, throughout the week, and throughout the year. In Botswana, where the largest source of demand is households, the maximum power demand is roughly 6pm-8pm and 6am-8am in the winter (June-July), at around 550mW. In theory, Morupule B power station can deliver 540mW (net), and hence just about meet Botswana’s peak demand, although in practice it has performed far below its design capacity and has been unable to meet domestic demand; hence the continued reliance on imported power.

Morupule B is currently undergoing a major repair and refurbishment exercise, and if this is successful, it should – along with the refurbished and rehabilitated Morupule A – be able to meet current peak demand. However, if peak demand continues to grow at 3-5% a year, the combined maximum Morupule A and B capacity will be exhausted within the next 3-5 years. There is also uncertainty over how much power Morupule A and B will be able to deliver in practice, even when refurbished. Given that there is a lead time of up to five years between an investment decision and the commissioning of new power generation capacity, a decision on where the next capacity will come from is required as early as possible in 2020.

Power generation options

As noted above, Botswana has traditionally relied upon coal-fired power stations to generate electricity. This partly reflects available technology when investment decisions were made (in the 1970s for Morupule A and the 2000s for Morupule B), along with Botswana’s readily available coal deposits. These are estimated at over 200 billion tonnes – although proven reserves (to conventional standards of geological certainty) are much lower than this. Much of the coal is close to the surface and hence easy to mine though open-pit or shallow underground mines. Its quality is, however, generally average to poor, with a relatively low thermal capacity and high ash and sulphur contents. Overall, however, for mine-mouth power generation, Botswana’s coal is a relatively cheap resource in financial terms (although not necessarily in economic terms once the cost of externalities such as pollution and CO2 emissions is taken into account).

Coal-fired power generation has other perceived advantages. In principle, the technology is well-established and reliable – although the problems experienced with Morupule B with unreliable performance, frequent breakdowns and delivery of power well below design capacity shows that this perception may be misplaced. Coal can provide stable base-load power and has some flexibility to increase and decrease electricity output to meet variations in demand. Botswana’s coal-fired power stations should be able to provide electricity at a cost of USD 0.08-0.10 per kilowatt hour (kWh)[2]. However, the poor experience with Morupule B suggests that the real cost of electricity from this source may be higher than this.

Solar – Photovoltaic (PV) and Concentrating Solar Power (CSP)

Solar energy is another method of power generation that is reckoned by many to have high potential in Botswana. In general, the country has amongst the highest levels of direct normal (solar) irradiation (DNI) of any location in the world, with over 3,000 kWh/m2 per annum. So, while many people argue that Botswana should use its ample reserves of (cheap) coal for power generation, the same argument applies to its ample supplies of (free) sunshine. Clearly there is much more to the choice of power generation technology than the cost of the inputs (for instance, capital costs are much more important that input costs in determining the overall costs of power from different sources), but the point is that access to those inputs does not provide any useful guidance as to whether coal should be preferred over solar, or vice versa. Other issues are much more significant.

Figure 3: World Solar Irradiation

There are two main types of solar power generation technology. Concentrating Solar Power (CSP) uses mirrors or lenses to concentrate the sun’s energy and heat up a suitable liquid medium, such as molten salt. This liquid is then used for thermal power generation through a steam turbine, in much the same way as coal is burned to provide heat to create steam. Solar PV uses the sun’s energy to generate electricity directly through a chemical reaction in solar panels.

Of course, one of the main issues with solar power using either of these two technologies is its intermittency – power is only generated directly when the sun is shining, and hence no power is generated at night and less is generated when the weather is cloudy. This contrasts with coal, which can be used to generate electricity regardless of the weather and at any time of day or night.

Hence, for solar power to provide dispatchable electricity (i.e., in response to demand at any time of day or night), some form of storage is required. The need for storage is one of the weaknesses of solar energy, as storage technology is not yet as well developed as power generation technology – although this is changing very fast. CSP can be combined with thermal storage, whereby the heat generated during is transferred to the molten salt, which has a high heat retention capacity, which can then be used to produce steam for power generation at a later point in time. CSP with storage can now deliver electricity on a 24-hour basis. Reliable solar PV requires battery storage, which is still expensive. The cost of direct solar PV electricity generation has been falling very fast, and it is now a competitive way of meeting daytime electricity needs. However, it is not currently competitive for utility-scale grid supplies of dispatchable power, or meeting peak power requirements, due to battery storage costs. Nevertheless, solar PV with some battery storage capacity is already a cost-effective way of providing off-grid electricity solutions, and the cost of large-scale battery storage solutions is falling fast.

Costs for both forms of solar power generation have been falling rapidly. In 2018, electricity from newly commissioned CSP projects (with storage) averaged USD0.185 per kWh. Utility-scale solar PV was cheaper, at USD0.085 per kWh in 2018 (without storage). The rapid reduction in costs for these technologies experienced between 2010 and 2018 is expected to continue, with costs of USD 0.08 and USD 0.05 per kWh for CSP and solar PV respectively by 2022 – making them cheaper than typical coal projects, even before the costs of externalities are taken into account[3].

Other sources –hydro, gas, wind and nuclear

Hydro and wind power: Elsewhere in the world the cheapest forms of renewable energy (RE) are hydro-electric power and wind energy, especially onshore wind. Hydro power is particularly good, not just because of its low cost but because of its flexibility in terms of its ability to deliver power on demand. Unfortunately hydro power is not an option for Botswana, given the country’s topography and lack of large rivers. Nor is wind power an option for large-scale power generation, due to low wind speeds.

Gas: Many countries use natural gas-fired power stations to meet a portion of their electricity needs, especially at peak times as supply can be ramped up quickly. Although gas is also a fossil fuel, it is less damaging than coal in terms of pollution and CO2 emissions. Botswana has natural gas in the form of coal-bed methane (CBM) associated with coal deposits. CBM is a potential fuel source for electricity generation, and small CBM projects are currently being developed. The 90mW generating plant in Orapa has been used to generate power using diesel in the past, but can also use gas. Gas-fired electricity generation has the advantage of being very flexible, due to easy storage and the ability to increase output rapidly to meet peaking demand. However, it is relatively expensive[4].

Nuclear: Although designs for micro- and small-scale modular nuclear power plants exist, their adoption globally has been very limited and their economics uncertain. They also have demanding requirements for water for cooling purposes, which makes them unsuitable for Botswana, and as well there are long-term waste storage issues.

Figure 4: Comparative costs of renewable energies

Global warming and climate change

Any discussion of power generation options needs to relate to climate change and global warming issues. Concern about climate change due to rising global temperatures is now one of the most pressing international policy issues (environmental, political and economic), and it is well established that emissions of CO2 and other gases are the main cause of global warming[5]. Emissions come largely from fossil fuels, whether used in power generation or as fuel for transportation (diesel, petrol, aviation fuel etc.). Some steps have been taken for international collective action to limit harmful emissions, dating back to the 1997 Kyoto Protocol, and more recently under the 2015 COP 21 Paris Agreement[6]. This agreement recognised that CO2 emissions would have to be reduced to “net zero” by 2050 (carbon neutrality) in order to prevent a potentially disastrous increase in global temperatures (more than 1.5-2.0 degrees centigrade above pre-industrial levels).

Most countries made commitments to reduce CO2 emissions at COP 21, although these are voluntary and lack an enforcement mechanism. There are major concerns (e.g. as expressed at the COP 25 Madrid meeting) that many countries are falling behind on their emission reduction commitments, leading to much faster global warming. This would lead to rising sea levels, higher average temperatures and more extreme weather, which will be reflected in drought conditions in and lower rainfall in some areas, with major impacts on the habitability of coastal cities and other areas, on food production, and disease.

Although global warming is still at an early stage, there are already real impacts. At the time of writing, Australia is experiencing another round of deadly bush fires that have been widely attributed to global warming. There are also major concerns about the future viability of producers of fossil fuels, such as oil companies and oil producing nations, given that some known reserves of oil cannot be developed if CO2emission targets are to be met, and hence such companies and countries will have “stranded assets”. Coal producers are particularly vulnerable, as coal is the most polluting of fossil fuels in terms of CO2 emissions per unit of energy produced. As a result, many coal projects are now extremely difficult to finance commercially – including coal mines, coal-fired power stations, and coal transport projects such as railway lines – as many banks are not convinced that there will be a business case for coal over the financing lifetimes of such projects. In addition, with international public opinion turning strongly against coal, banks are also concerned that financing coal projects will have adverse impacts on their business more generally. The pull-back of banks from coal projects applies to both purely commercial banks as well as multilateral development banks such as the World Bank and African Development Bank. As a result, any coal projects that do proceed will need to be financed – directly, or indirectly through guarantees – by government, which has implications for public debt and compliance with statutory limits on government indebtedness.

Even though Botswana is small in the global context, and its GHG emissions have little significant global impact, it is party to the COP21 Paris Agreement, and has committed contribute to global reductions in GHG emissions. Botswana’s (non-binding) Intended Nationally Determined Contribution (INDC) is “to achieve an overall emissions reduction of 15% by 2030, taking 2010 as the base year. Base year emission estimation is 8307 Gg of CO2 equivalent”. From the background documents to the Government’s COP21 submission, this appears to mean a 15% reduction in CO2 emissions relative to the level that would apply in 2030 without such mitigation measures, rather than an absolute 15% reduction from the base year level (which is impractical). Clearly this implies a shift away from coal-fired power generation towards renewable (i.e. solar) energy, as well as limitations on other potential coal-based projects (such as coal-to-liquids).

Despite intensifying concerns about global warming and climate change, achieving concerted global policy action to achieve reductions in CO2 emissions is proving challenging. From an economic perspective, the optimal approach would be to have an internationally agreed carbon tax paid by all emitters of CO2. This has the advantage of making real to CO2 emitters (polluters) the costs of the damage they are causing – i.e. of internalising costs that are currently manifested as externalities, borne by others but not by those who produce the emissions. Under this “polluter pays” principle, making emitters bear these costs, the intention is that behaviour would change to reduce emissions and hence global warming. A carbon tax would have the additional advantage of providing resources to finance mitigation and adjustment measures.

Botswana’s options

So where does this leave Botswana in terms of power generation decisions? As noted, there is very little current solar power generation, despite the rapid global shift towards renewables. Most current solar power initiatives are private and small-scale, mostly in off-grid locations such as farms and safari camps. However, some steps are being taken by government: tenders have been issued for two 50mW solar generation facilities to supply power to the national network, as well as for 12 smaller-scale (1-5mW) solar projects. There are also plans to allow households and small business to generate their own electricity using rooftop solar installations and sell any surplus back to the BPC grid.

However, there still seems to be an official preference for making another large investment in coal-fired power – the proposed 300mW Morupule B Units 5/6 project has been on the cards for years. It has, however, been delayed due to problems in reaching agreement with the project promoters and financiers – Marubeni of Japan and Posco of South Korea, as well those countries’ export credit organisations – who now appear to have pulled out of the project. Nevertheless, a further 300mW of coal-fired power still appears to be in BPC’s planning strategy[7].

It is difficult to see how this could be justified in commercial or financial terms, given the rapidly declining costs of solar power generation. It is even more puzzling in economic terms. A proper planning decision – as per the formal principles adopted by the Ministry of Finance and Economic Development and as laid out in the Planning Officer’s Manual – specifies that major public investment decisions should undergo an economic appraisal reflecting economic costs and benefits, rather than financial costs only, using shadow prices where necessary to incorporate externalities. Applying these principles to power generation decisions requires that the shadow price of CO2 emissions is included in the economic assessment of any potential investment in coal-fired power generation. A rough calculation using the World Bank recommended shadow price of USD30 per tonne of CO2 emissions, and based on Botswana’s historical emissions, suggests that the cost of CO2emissions would add USD0.035-0.040 per kWh to the cost of coal-fired electricity (compared to a financial cost of USD0.08-0.10 per kWh) – i.e. an increase of 35-50%. This significantly shifts the economic calculation away from coal and towards solar power[8].

Another big investment in coal-fired power generation would also appear to be inconsistent with the (draft) Botswana Climate Change Strategy[9]. This proposes “a robust approach to reduction of GHG” emissions, “particularly in the energy sector”, and states that “government will adopt and enforce carbon taxes” (pp.25-26). This too implies a big shift towards solar power generation. Besides properly reflecting economic costs and incentivising the production of green energy, carbon taxes could also raise substantial revenues for government. To be consistent, a carbon tax would also have to apply to petrol and diesel.

However, even if Botswana decides to move decisively towards solar power and build no more coal-fired power stations, this still leaves unresolved the question of solar power’s intermittency, especially if the lowest cost solar option (PV) is adopted. Ample, cheap electricity can be generated by solar PV in the daytime. But as noted above, peak demand occurs on a winter morning, when it will have been dark for more than 12 hours. There are also issues around the reliability of solar power supplies during cloudy periods, which can last a week or more in summer.

There are several ways of dealing with the intermittency problem, and addressing the challenge of meeting peak electricity demand on a cold, dark winter morning. One is to rely on imports from neighbouring countries – which in practice means utilising coal-fired power generated in South Africa (hence “outsourcing” GHG emissions). Looking further ahead, it may be possible to utilise renewable hydro power from Zambia or the Democratic Republic of Congo (DRC), through the Southern African Power Pool (SAPP). A second option is to use Botswana’s CBM gas to meet peak demand. This is still a fossil fuel, but much less damaging than coal. Power generation from CBM in Botswana is still at an early stage, but could be introduced quickly if only government and BPC were prepared to sign a power purchase agreement (PPA). A combination of solar and gas fired power generation could be a compelling and cheaper option that coal.

A third option is to shift the time patterns of demand, rather than take it as given. For instance, long-distance water pumping – such as the North-South Carrier - can easily be done in the daytime using cheap electricity from solar PV, with no storage required. Introducing variable time-of-day tariffs – with more expensive power at peak times and cheaper power at off-peak times – would provide an incentive for consumers to adjust their consumption patterns and spread demand more evenly, and match it to periods of high supply.

Fourth, storage technology is developing rapidly and falling in price. Competitive storage options for CSP can already provide dispatchable power at any time of day. Similar trends under way with large-scale battery storage will soon enable solar PV to provide dispatchable power at any time. Once solar can provide the cheapest source of electricity at 7am on a winter morning, it really is game over for coal, and that point may not be far off.

Conclusion

Botswana needs to consider many factors in deciding – quickly – where the next 300MW of power will come from. Choosing coal rather than solar runs the risk of being backward-looking rather than anticipating likely technical, economic and political changes over the next two decades. Many commentators believe that the world is close to a tipping point with regard to the shift from fossil-fuel based energy to renewable energy (RE), with the dramatic reduction in costs of RE technologies such as solar and wind energy, battery storage costs, and the shift towards electric vehicles. Although collective policy actions to deal with global warming – such as carbon taxes – are proving difficult to agree on, economics is doing the job, in terms of changing the relative economics of fossil fuel and RE sources. Once this tipping point is reached, change will happen very rapidly. Botswana needs to be wary of being on the wrong side of history. But government also needs to be decisive and move much more quickly: bidding rounds for solar energy have been cancelled twice, while the current invitations for IPPs to provide electricity from solar and CBM sources are moving at a glacial pace. Because of this, combined with sub-economic electricity tariffs, the current environment is not conducive for private investment in power generation, despite this being a declared aim of government policy.

A further point is that electricity has been sold too cheaply in Botswana for many years. For political reasons, BPC has not been permitted to sell electricity at a cost that covers generation, import and distribution cost, hence requiring expensive subsidies. In addition, BPC has not had to pay for the environmental damage that its activities cause. If these issues are to be addressed, the price of power will have to rise. A change in the pricing and taxation structure (higher average prices, carbon taxes, variable time-of-day tariffs, while allowing users to partially offset this by generating and selling power back to BPC) would lead to a much more economically and environmentally rational set of incentives.

From a domestic perspective, a great deal of focus is being placed on the role that innovation can play in Botswana’s economic transformation. Investing in coal provides little or no opportunity for useful innovation, as coal is almost certainly a fuel of the past globally. By contrast, solar power generation is set to expand rapidly; the International Energy Agency (IEA) predicts a 50% expansion of renewable power capacity from 2019 to 2024, of which 60% will be accounted for by solar PV[10]. By committing to solar energy, Botswana could provide opportunities for local innovation, that could in turn be at the cutting edge of important global technological developments. This is in turn consistent with Vision 2036 and its ambition of an information based society harnessing the new technologies.

[1] The cost of diesel emergency power was around USD2.50/kWh

[2] Discussions of the cost of power generation generally refer to the “Levelised Cost of Electricity” (LCOE). The LCOE of a given technology is the ratio of lifetime costs to lifetime electricity generation, discounted back to a base year using a discount rate that reflects the average cost of capital.

[3] All figures taken from the International Renewable Energy Agency (IRENA) publication, Renewable Power Generation Costs in 2018,https://www.irena.org/publications/2019/May/Renewable-power-generation-costs-in-2018

[4] Estimated costs of CBM electricity are in the range of USD1.20-1.60 per kWh

[5] Other emissions also contribute to global warming, notably methane; these are generally measured in terms of tonnes of CO2equivalent (tcoe).

[6] Convention of the Parties (COP), members of the United Nations Framework Convention on Climate Change (UNFCC)

[7] Presentation to Botswana-China Business Forum on Electricity Supply in Botswana, August 2018.

[8] This shadow price is similar to the price of carbon under the EU’s Emissions Trading Scheme (ETS) at the end of 2019, but lower than the US Government’s guide price of US$50 per tonne.

[9] Ministry of Environment, Natural Resources Conservation and Tourism

[10] https://www.iea.org/reports/renewables-2019