The South African Government’s policy on nuclear power III – Cost and financing issues

The first brief in this series reported the current status of the new nuclear project. The second brief dealt with the nuclear project in the context of energy policy. This brief will deal with construction and financing issues and draw conclusions from the brief series.


Construction cost

In analysing construction cost estimates for new nuclear power plants, there are a number of aspects which need to be kept in mind and which may not be immediately evident from presentations by players in the industry:


  • In any comparative cost analysis of world-wide nuclear power plant construction, there are big disparities and it is difficult to arrive at a meaningful forecast on this basis.  There are large variations in the construction cost in different countries and in addition, serious delays in some of these projects have led to cost overruns to such a degree that the final cost even ends up as a multiple of the originally envisaged cost.  Massive cost overruns are not restricted to the construction of nuclear power plants, but they also occur in conventional power station construction, as is evidenced by Eskom’s own experience with Medupi.  Eskom’s 2015/2016 annual report states that the Board had approved a revised amount of R145 billion for Medupi .  This compares with the amount of R80 billion initially estimated for construction costs in terms of an Eskom media statement issued on the day of the sod-turning ceremony at Medupi on 14 August 2007.  This represents an 81% cost escalation.  As far as the construction delays at Medupi are concerned, Eskom’s 2006/2007 annual report envisaged the first unit at Medupi to come into service in 2011, but this only happened in 2015.


  • Given the international and South African experience with delays and large cost overruns, will the preferred bidders in terms of the RFP be prepared to take any financial responsibility in this regard  -  or are cost overruns to be carried by Eskom (and by extension, the South African consumer) ?  It may even be preferable for Eskom to pay a premium for a fixed price contract, which would normally be more expensive since the contractor effectively underwrites the fact that the contractual amount will not increase.  Such a fixed price contract would at least mitigate the financial risk.


  • Is the cost of financing included or excluded in any construction cost estimate?  In this context, the term “overnight capital cost” is often used, which excludes the cost of financing.  When including interest during construction (and especially if delays are experienced), the all-in cost would exceed the overnight cost by very substantial amounts.


  • How are decommissioning costs dealt with in any proposal ?  They are often included in the operating costs, where they would appear to be relatively modest numbers if expressed on short-term ongoing basis.  However, the aggregate decommissioning costs for a nuclear power station represent very substantial absolute amounts.  If it is not mentioned upfront, it is therefore important to obtain confirmation as to where these costs are housed, to be able to make meaningful comparisons with renewable energy, where decommissioning costs are of a nominal nature.  For a very recent example of the magnitude of the decommissioning cost of nuclear plants, one can look at the German experience, following the declared government policy of abandoning nuclear power.  According to a report in the Frankfurter Allgemeine Zeitung of 10 October 2016, it has now been agreed that the nuclear power companies will be responsible for the dismantling of the power stations and the German Government will look after the waste storage.  In terms of this agreement, the German companies are to transfer a cash amount of 23.3 billion Euros (equivalent of R350 billion).  The estimated additional cost for the companies to dismantle the power stations is 20 billion Euros (equivalent of R300 billion).  The total envisaged cost is therefore R650 billion.  There are 8 nuclear plants are still in operation in Germany with a total capacity of 11 357MW and 27 decommissioned plants with an aggregate capacity of approx. 16 509MW (  To place this in some perspective: purely from a rough order of magnitude comparison of the respective aggregate capacities, these plants equal three times the South African target of 9 600MW new nuclear plants.

In summary, any analysis of construction cost estimates needs to answer the following questions:


  •  Who takes the risk of delays and ensuing cost overruns?  If it is to be Eskom, Eskom’s consumers will be at risk, since additional costs will be recovered from them through Eskom’s tariffs.  Are consumers happy to take this risk, given the global experience with nuclear power delays and cost overruns, as well as Eskom’s own experience with Medupi?

  • Is financing cost included or not?  If not, what is the financing cost?

  • What are the decommissioning costs and how do they fit into the aggregate cost estimate?


Storage of high-level nuclear waste

A permanent solution to dealing with high-level nuclear waste (i.e. spent nuclear fuel) has not yet been found and implemented.  Owners of nuclear power plants therefore have to be prepared and able to manage whatever high-level waste they produce for an indefinite period, on the assumption that any long-term offsite storage solution is not going to be found any time soon.  Mention is often made in literature produced by the nuclear industry that countries are in the process of designing and constructing deep underground engineered facilities for the long-term storage of such waste, but no such facilities are operational yet, 60 years after the construction of the world’s first commercial nuclear power stations. 

Eskom’s website puts it as follows:

So governments have no need to rush their decision about what they will do with HLW (ie. high-level waste) in the long term. They are in a position to weigh all the options on behalf of their citizens.  As a result, very few governments in the world have committed themselves to a final disposal strategy.

Eskom’s website also provides the following on Koeberg’s high-level waste:  when removed from the reactor vessel it is stored in special "fuel pools".  After ten years in a fuel pool it is "cool" enough to be moved into thick-walled casks which can be stored above ground for up to 40 years.  A two-reactor pressure water reactor power station like Koeberg (capacity of 1 860MW) generates approximately 32 tons of spent fuel each year. Over a 40-year lifetime that would add up to 1 280 tons. 


If additional nuclear power stations are to be considered, the risks involved in the storage of high-level waste have to be understood and accepted.


Eskom is to finance the construction costs

The Minister of Energy has now confirmed that the nuclear power project is to be funded by Eskom off its own balance sheet (therefore distinct from Government).   In analysing this new approach, we can take a look at Eskom’s 2015/2016 financial statements: Eskom has non-current debt of R307 billion and an operating cash flow (EBITDA) of R32 billion.    It is obvious from these numbers that Eskom cannot “leverage” its balance sheet any further for debt in a conventional sense and presenting the debt subject in this way is not accurate. 

However, Eskom may still be able to finance the project, but that would be on the basis of the lender taking a view on Eskom’s future cash flows (and not its current balance sheet), in a manner commonly referred to as “project financing”.  In this manner, Eskom’s debt service would be funded by its future tariff revenue (ie. by consumers) and lenders would count on Eskom’s effective monopoly remaining in place for many years to come, in order to ensure continuing cash flow which would fund the debt service obligations.  It would therefore be more accurate to say that Eskom will seek to fund its debt obligations through its future cash flows, as funded by consumers, instead of Eskom “leveraging” its balance sheet. 

In addition, even if no out-and-out guarantee is provided by Government in this regard, reliance would be placed by lenders on what is seen as an implicit Government guarantee, as it is assumed that Government will not permit Eskom to default on its financial obligations. 


General conclusions

The next step is to await publication of the Integrated Resource Plan (IRP) by Government  -  the IRP is to provide an overall assessment of future electricity demand and how that demand is to be met.    In order to ensure a meaningful measure of public participation on an issue of great importance to the South African economy, it will be important for South Africans to participate in the evaluation of the IRP, specifically as to the reliability and credibility of its assumptions and conclusions. Even if Government has recently taken a more measured approach by delaying the issuing of a Request for Proposal (RFP) for new nuclear power stations until the IRP’s publication, a proper analysis will have to be made on the IRP’s conclusions and more specifically, on the need for more nuclear power and its affordability, when compared to the alternatives.  This is necessary to avoid Government pursuing an option which may not be in the financial and broader economic interests of the country.


Anton van Dalsen
Legal Counsellor