Increasing attention is being given internationally to the risks associated with unburnable carbon and stranded assets (UNEP 2015, IEA, 2014; Citi, 2015). Stranded assets are assets that “have suffered from unanticipated or premature write-downs, devaluations or conversion to liabilities” (Caldecott et al, 2014), arising from environmental regulation or other regulatory shifts, as well as market forces and other causes, including the falling costs of competing technologies and structural market decline. Citibank (2015) has estimated that the value of stranded assets is as high as $100 trillion to 2050.
These are significant (mis)investments globally and in South Africa, with potentially substantial impacts on companies and investors. There are, however, other systemic impacts of stranding assets, especially for an energy-intensive economy such as South Africa’s with a large coal-based capital stock and growing energy demand. These include the risks facing Eskom, Sasol or independent coal power producers, or the risks faced by companies and investors in private and state-owned generators (UNEP, 2015). But the risks also extend beyond, to include national risks to energy security (including supply, affordability and acceptability), macro-economic impacts on the demand side (for energy-intensive industry and poor consumers) and welfare risks as a developing country with high levels of poverty and inequality.

The UNFCCC’s Paris Agreement will require commitment even from developing countries to reduce their greenhouse gas emissions, and continued investment in high-emitting infrastructure will create costly risks for South Africa in the future. The energy sector, which in South Africa accounted for 80% of emissions in 2010 (DEA 2014), will need to meet decarbonisation targets in the medium to long term. Limiting warming to 2°C over the course of the century will require the complete phase-out of coal-based electricity generation without CCS by 2050 (Johnson et al, 2013). Investing in new coal-fired assets in the short-term may well prove costly in the longer-term, as the risk associated with not recouping those investments due to policy shifts or technology changes grows higher, especially for plants built after Medupi (Bazilian et al, 2011).

Research Objectives

This paper aims to understand the implications of South Africa reducing emissions as part of a global agreement to limit temperature rise to below 2oC and to examine the potential risks of stranded assets in the energy sector. Can South Africa meet carbon constraints without stranding assets? Given the structure of the energy sector and existing infrastructure, what are the potential effects of the country stranding energy assets to meet mitigation targets? What are the cost implications of investing in power plants that are later underutilised? What is the impact of ignoring non-electricity emissions on the costs of transition?

Key Findings and Policy Recommendations

The Department of Environmental Affairs is not currently allocating South Africa’s carbon budget (i.e. its long-term climate mitigation commitments) between different sectors. The country may also have to reduce its emissions further in the long-term, as the world moves towards emission reductions consistent with 2oC and net negative carbon emissions after 2050. A scenario with a 14Gt constraint is broadly consistent with the mid-PPD of SA’s climate policy, but since global country contributions are not yet consistent with limiting warming to below 20C, many countries, South Africa included, may need to reduce emissions further.
Having examined more stringent scenarios, we highlight two major risks currently facing South Africa. Firstly, that electricity planning will assume a higher share of the national carbon budget for the sector to 2050 and will invest in fossil power accordingly; in later years, this capacity will either have to be stranded wholesale and retired prematurely, or become stranded capacity with plants run at low average load factors. As in Johnson et al, (2013) we find that less stringent near-term climate policy results in longer-term stranded capacity. This risk was highlighted in the IRP update (DoE, 2013: 25); the study was extended to 2050 partly to overcome the risk, because “by excluding the period after 2030 there is a risk of building coal-fired generation in the period leading up to 2030 on the assumption that the carbon emission caps would continue at the same level, but this would lead to a constraint in reducing the emissions or under-utilisation of generation capacity if the cap needed to be reduced over time as indicated by the government’s peak-plateau-decline (PPD) objective.” Given that the electricity sector is a lower-cost option for decarbonisation, we argue that the IRP 2010 emissions constraints (which allow for new coal-fired capacity to be built), should be revisited.
The figure below shows how the constraints we have imposed differ markedly from the emissions constraints in the IRP update. The IRP update assumed that the electricity sector would retain a proportional share of national emissions space to 2050; however, this is not the least cost mitigation option.

Comparison of carbon constraints in the Integrated Resource Plan update and those applied in the current study (‘SAS’ refers to the scenarios where Sasol is assumed to run for its full technical life, which would result in increased pressure on the electricity sector to decarbonise).
Comparison of carbon constraints in the Integrated Resource Plan Update and those applied in the current study (‘SAS’ refers to the scenarios where Sasol is assumed to run for its full technical life, which would result in increased pressure on the electricity sector to decarbonise).

The rate of retirement required to meet the emission constraints exceeded the rate at which the fleet is scheduled to retire, especially when new coal plants are brought on line. Continued investment in coal plants after Medupi and Kusile need to be carefully considered by energy planners since these are likely to be costly (mis)investments for South Africa; Medupi and Kusile already run a high risk of becoming stranded capacity if emissions constraints for the country or the electricity sector are tightened. The 10 GT scenario shows significant stranding of assets as there is no coal-fired generation by 2040.
Further, South Africa’s climate policy does not adequately account for the emissions associated with liquid fuels production. If South Africa is to continue to rely on coal-to-liquids (CTL) for liquid fuels supply to 2050, then higher levels of decarbonisation in other sector will have to take place. If a least-cost mitigation plan is to be adopted for the county, then the Department of Environmental Affairs and National Treasury must (as the implementers of mitigation policy) understand the political and economic trade-offs between these sectors better. While the effects on total investment in the electricity sector (e.g. the 12 and 12 SAS scenarios) may be similar, the electricity price increases earlier when more coal-based electricity capacity is stranded to meet the emissions constraint. In essence, this means that electricity-intensive industry must endure higher electricity prices for Sasol to avoid the stranding of its coal-to-liquids asset and the mines that support it. Alternatively, mitigation in other sectors must be deepened. Continued reliance on emission-intensive industry is risky, both to South Africa’s standing as a global citizen committed to emission reductions but also directly to the economy. Not only will international pressure increase in future years as implementation of the Paris Agreement unfolds internationally, but the coal sector globally is now viewed as being in structural decline. At the same time, industrial policy continues to favour energy- and carbon- intensive industry (DTI, 2015).
The key message is that the stranding of power sector assets is unavoidable in a carbon-constrained world. This will lead to higher electricity prices, as the stranded assets are replaced by lower carbon electricity generation capacity. Building new coal plants will only add to this risk.
A higher electricity price has significant economic impacts on energy-intensive sectors overall, negatively affecting non-ferrous metals in particular. This is likely to result in the shut-down of electricity-intensive sectors or these sectors moving off-grid, both of which are already becoming a reality with current electricity price increases. The socioeconomic implications of this are significant as jobs will continue to be shed, putting more pressure on the fiscus to combat increasing poverty and inequality.
A transition to a low-carbon economy could hedge this risk for South Africa, and provide growth opportunities in the medium to long-term as South Africa recoups benefits through increased competitiveness. This is possible provided the economy shifts away from energy-intensive sectors, especially iron and steel and non-ferrous metals, and coal (upstream and downstream) as these sectors continue to decline in profitability. Without this structural change, the challenge of growing the economy to address high levels of unemployment and poverty would increase substantially if stringent mitigation targets were required in the electricity sector.

Future Work

This paper has used a partially linked energy and economic model to unpack the economic effects of mitigation and the impact of stranding assets on the South African economy. There are key areas of future work that remain. These include:
• Further linking of the energy and economic models, including refineries and CTL, coal mining and demand sectors.
• Examining the coal resource implications of these scenarios, i.e. stranded resources and unburnable reserves and resources.
• Examining the inclusion of the coal baseload independent power producers programme under different future mitigation targets to understand the costs and risks of stranding new coal assets.


A MAPS (Mitigation Action Plans & Scenarios) funded paper. Click here for the link to full working paper and references.