Nuclear power - unaffordable, or lowest cost energy available?

Nuclear expert Dr Anthonie Cilliers has calculated estimates for the cost of building nuclear power in SA and believes the high costs being punted may not be accurate.

THE persistent mentions in the media by energy experts that nuclear electricity is expensive and unaffordable has become all too frequent.

These claims can be nothing more than speculations without South Africa being able to issue request for proposals (RFPs), and in most cases I have found it baseless and echoing an ongoing narrative.

Recently we have also heard conflicting statements made by politicians and experts on the affordability of nuclear which in my view confused the public as much as it did the media. Pressure groups then like to cry foul and jump on the statement they prefer, ridiculing the other.

As an academic, I prefer to look at numbers and calculations. The result may be surprising, but as we will see, both statements could be right.

Countries like China, Russia and India are building nuclear power plants at an enormous pace at the moment, more countries are signing contracts to join them, whilst in the US, France and the UK we find costs escalating and construction even being stopped.

Of course, depending on the side of the fence you are on, you ignore the ones that don’t fit the narrative.

The secret, as I have found, is in the funding model. I have also realised that the energy debate has very little to do with nuclear, or renewable energy, or even coal. The energy debate is a proxy debate about ideologies. One will always result in expensive energy (even the “cheap” ones) as a commodity, the other will use energy as a driver for economic growth.

To unpack this, I will use a few real-world examples (I find models to be dishonest on the best of days, designed to fit a certain narrative – but that is a personal view). 

Real world funding models

According to World Nuclear News, the World Nuclear Association and Reuters: Russia and Egypt signed an intergovernmental agreement in November 2015 to collaborate in the construction and operation of a nuclear power plant equipped with four 1 200 MWe units.

The agreement includes provision of a Russian state-backed loan of $25bn for the $30bn project. The Russian state loan will cover 85% of the plant's construction costs, with Egypt to raise the remainder from private investors.

The project is to be completed within 12 years and Egypt will start repayment of the loan at an interest rate of 3% from October 2029.

In May 2016, the government confirmed that the loan was for $25bn, to cover 85% of the cost of four 1 200 MWe units, with repayments to start upon commissioning. The general contract discussions for construction is now concluded and should be signed in 2017. Local content for the first unit is expected to be 20%, increasing for subsequent units.

Two desalination facilities (large enough to supply the entire water requirements of the whole Western Cape) are also planned for these plants. 

This sounds rather expensive, and not knowing what exactly is covered in the agreement, the questions you might ask:

  1. Why did Egypt accept this agreement? They even seem rather proud of it.
  2. How did Egypt get a 3% loan interest when they have a worse credit rating than South Africa?

The answer to the second question is rather simple – this is the normal rates obtained when export credit agreements are reached, and corresponds to the rates Eskom has received in similar agreements according to Eskom’s 2017 financial report.

The answer to the first question is the subject of the calculations in this article.

Model used in this article

The model used in this article has been verified against IRP2016-Draft data and accurately reproduces the same Levelised Cost of Electricity (LCOE) values using the following parameters at a discount rate at the internal rate of return (IRR) of 8.2%, this corresponds to a private financing model as used by the IPP office:

rated capacity

The draft IRP2016 uses an overnight cost of R55 260 per installed kW, realising a Levelised Cost of Electricity (LCOE) of R0.97 per kWh. When using the same parameters for the overnight cost of the UAE/Korea – Barakah-1 nuclear plant, we obtain a LCOE of R0.75 (including interest during construction, fuel, operations and maintenance cost).

It should be noted that this does not include transmission and grid connection costs (for any energy sources). My understanding is that in updated IRPs this should be included.

For projects of this magnitude, we often see the financing model utilised being public rather than private funding, or, as suggested by this article, a combination of private/public funding model.

The above described Egypt/Russia agreement is such a private/public funding agreement:

Vendor state funding: 85% at 3% interest.
Private funding: 15% at 8.2% internal rate of return.

If this funding model is applied to the UAE/Korea – Barakah-1 nuclear plant, we obtain a LCOE of R0.21 per kWh over the 60-year economic life of the plant. Servicing the debt over 22 years, we see a cost of R0.45 per kWh for the first 22 years and R0.10 (in 2017 Rands) thereafter. Already we start seeing the difference between the models.

The object of this article is to investigate the effect of this private/public financing model on the Egypt/Russia agreement.

The only changes made to the IRP parameters in these calculations is the inclusion of financing information as deduced from the Egypt/Russia nuclear agreement.

For readers not interested in the long calculations, the results are (please still read the end, I promise, it is worth it):

Egypt/Russia agreement with private funding model:

LCOE over 60 years economic life: R1.57 per kWh. This corresponds with published models often quoted utilising the private financing model.

Egypt/Russia agreement with 85% vendor state backed loan at 3% and private funding at IRR of 8.2%:

LCOE over 60 years economic life: R0.39 per kWh servicing the debt and equity over 22 years at R0.89 per kWh; R0.10 per kWh thereafter.

Again, the difference in the model is staggering to say the least.

So, let’s delve into the calculations…

What a typical (similar) nuclear power agreement could look like in South Africa.

Four 1 200 MWe = 4 800 MWe in Egypt at $30bn (including interest during construction). South Africa could be looking at eight of these to provide 9 600 MWe at $60bn (without assuming a learning rate as the projects progress).

At a conservative R14.25 to the US dollar, a 9 600 MWe project value would be R855bn. This results in a contract value of R89 062.50 per installed kW for the 9 600 MWe nuclear (substantially larger than the UAE/Korea costs). One has to remember that this is not merely capital costs, as it includes interest during construction (IDC) - that makes up a significant portion of the costs. *

* Due to ambiguous media reporting it cannot be verified that IDC is included at all times.  If additional IDC at the vendor loan interest rate is added for the construction time of each unit it amounts to an additional R101 522 392 087.86 ($7 124 378 392) worst case. Costs with this additional value will be indicated throughout. It is, however, our view that the quoted number is the value of the debt at the time of repayment start.

If a vendor state backed loan of 85% could be agreed upon at a 3% interest rate (as in Egypt), and the rest locally, the split is:

Vendor state backed loan: R712 500 000 000.00 ($50bn).
* additional IDC R64 208 554 468.88 ($4.5bn).

Private funding at 8.2% IRR: R121 500 000 000 ($10bn).
* additional IDC R37 313 837 619 ($2.6bn).

We have to remember that nuclear plants produce electricity all day long with an annual capacity factor of 90% (as used in the IRP) internationally. This means that 90% of the installed capacity is available over a year.

On this basis, the 9 600 MWe nuclear will produce 75 686 400 MWh of electricity per year. Of course, we also need to take the Operation, Maintenance and Fuel cost into account, for this we use the information available from the UAE/Korea - Barakah-1 contract value (operating, maintenance and fuel of new nuclear plants tend to be very similar). We get:

O&M for 9 600 MWe nuclear (including fuel) = R8 142 857 142.86 ($571 428 571) per year in 2017 rands.

To put these figures into context we now need to calculate what the electricity would cost to ensure that the capital, fuel and operations & maintenance are be covered.

Based on the Egypt/Russia agreement, we use a figure of 3% interest on 85% of the contract value, paying it of over 22 years, whilst using 8.2% IRR on the remaining 15% of the contract value and paying it of over the 60 years economic life of the plant as well as covering the operation and maintenance costs. We obtain the following:

Vendor state backed loan: R0.59 ($0.042) per kWh for 22 years.
* with additional IDC: R0.64 ($0.045) per kWh for 22 years.

Locally raised funding: R0.17 ($0.012) per kWh for 60 years.
* with additional IDC: R0.21 ($0.015) per kWh for 60 years.

O&M (including fuel): R0.10 ($0./007) per kWh for 60 years in 2017 rands.

R0.85 ($0.06) per kWh for the first 22 years.
*with additional IDC: R0.95 ($0.067) per kWh for the first 22 years.

R0.30 ($0.019) per kWh for the remaining 38 years.
* with additional IDC: R0.29 ($0.021) per kWh for the remaining 38 years.

Since the plant economic lifetime is 60 years, we determine this average breakeven tariff over the 60 years as:

Nuclear breakeven tariff:  R0.48 ($0.034) per kWh.
* with additional IDC: R0.54 ($0.038) per kWh.

Variation of repayment periods depending on the source of capital is always possible. Servicing the entire debt and equity over 22 years (to reduce interest paid), repayment break-even tariff becomes:

Vendor state backed loan: R0.59 ($0.042) per kWh for 22 years.
Locally raised funding: R0.20 ($0.014) per kWh for 22 years.
O&M (including fuel): R0.10 ($0.007) per kWh for 60 years in 2017 Rands.


R0.89 ($0.062) per kWh for the first 22 years.

* with additional IDC: R0.99 ($0.021) per kWh for the first 22 years.

R0.10 ($0.007) per kWh for the remaining 38 years in 2017 Rands.

Again, with the plant economic life of 60 years, we determine this average break-even tariff over the 60 years as:

Nuclear break-even tariff: R0.39 ($0.027) per kWh.

* with additional IDC: R0.42 ($0.03) per kWh.

The result is rather staggering, and at first glance one might wonder how it is possible for the figure to be so low.

The reason in this scenario is rather simple: the combined effect of the extremely low interest rate offered by the vendor state as well as the long economic life of the plant.

This often-overlooked combination makes it virtually impossible for other energy sources to compete at the same level. This is the reason why we see nuclear plants becoming the cheapest operating plants on the grids of many countries after around 20 years. It also emphasises the importance of the financing model.

Nuclear costs versus other energy sources in South Africa.

So, how does these nuclear costs compare to other energy sources? Based on the latest Renewable Independent Power Producers (IPP) procurement prices, we see:

Nuclear (Egypt/Russia): R0.39 per kWh despatchable (R0.89 for 22 years, R0.10 thereafter).

Onshore Wind Bid Window 4: R0.75 per kWh intermittent.

Solar Photovoltaic Bid Window 4: R0.91 per kWh intermittent.

Concentrated Solar Bid Window 3.5: R1.80 per kWh despatchable.

This comparison is graphically depicted in Figure 1.


Again, the result goes against everything we have been told. What is even more worrying, is the results when the IPP bid window trends are compared with the capital cost trends for these energy sources.

According to The South African Renewable Energy IPP Procurement Programme: Review, Lessons Learned & Proposals to Reduce Transaction Costs by Eberhard and Naude, the capital costs of Solar PV and Onshore Wind sources are not coming down in line with the bid window tariffs.

In fact, for Solar PV the capital costs went up from bid window 3 to 4 – indicative to reducing profit margins which is unsustainable. This comparison is depicted in Figure 2. We have to remember that, for bid window 1-3 we are already locked into 20 year contracts of tariffs between R3.60 and R0.87 per kWh.

When looking at figure 1, it is important to note that the costs for nuclear electricity could almost double before it starts competing with the second cheapest option. And yet, while the IPP office has had the opportunity to test the market on renewable energy prices, nuclear is still subjected to speculation on the cost.


Additional costs for nuclear power.

I am sure I would be taken to task if I quote these values and calculations without looking at other costs associated with nuclear power plants. Spent management cost are already included in the cost of fuel, but we need to ensure adequate funding for decommissioning the plant, once the 60-year economic life is over (if we do not extend the plant life to 80 years).

Based on best international practices, the decommissioning of a nuclear plant cost in the region of R5bn per 1 000 MWe installed.

Adding R0.02 per kWh to the price – from day 1, and calculating compound interest over 60 years, results in a decommissioning fund of R2 070 208 985 000, this is equivalent to R53 772 168 583 in 2017 value, or R5 601 267 560 per 1 000 MWe in 2017 value. Of course, the break-even tariff allows for enough ceiling to increase this provision even larger if needed.

Calculating cost of decommissioning of other electricity sources is beyond the scope of this article although this also needs to be done.

What would such a financing agreement cost South Africa and what are the risks?

It is important to discuss what it would cost the South African economy, and what the risks are should we enter into such an agreement with a vendor country.

First of all, the risk of becoming indebted to another country. This is mitigated by the repayment of 22 years.

This is no longer than a typical home loan – something we are all comfortable with, and a short duration in terms of planning by sovereign states, as well as almost a third of the economic life of a nuclear plant.

The Independent Power Producers (IPP) contracts also run for 20 years, so this is a norm. I would also suggest the IPP office getting involved in raising the 15% locally financed component of the contract.

When looking at the local content in the plant, the first 3 200 MWe will make up 20% local content which we expect to rise to 43% for the second 3 200 MWe and 68% for the third 3 200 MWe, a mega project such as this allows enough time and incentive for local suppliers to develop a sustainable supply chain.

The second risk is that of guaranteeing the repayment of the loan, since the project value includes interest during construction, and repayment starts once the plants start commercial operation. The risk is mitigated to being able to sell the electricity to the market. This is a rather “nice” problem to be faced with, especially since the average break-even tariff is only R0.39 per kWh.

This will enable South Africa to again incentivise electricity use, and incentivise economic growth. This will require affordable electricity tariffs to the customers – a challenge we need to embrace and work towards. The alternative – energy poverty as we have today.

Probably the biggest risk we face when embarking on a project like this, is overruns in the construction period. Existing models suggest 23% overrun in the first 3 200 MWe, reducing to 13% and 8% for the subsequent batches of 3 200 MWe. This would result in break-even tariffs of:

First 3200 MWe with 23% overrun and decommissioning:

Nuclear break-even tariff: R050 ($0,035) per kWh.

* with additional IDC: R0.54 ($0.038) per kWh.

Second 3200 MWe with 13% overrun with decommissioning:

Nuclear breakeven tariff: R0.46 ($0.032) per kWh.

* with additional IDC: R0.49 ($0.034) per kWh.

Third 3200 MWe with 8% overrun with decommissioning:

Nuclear breakeven tariff: R0.44 ($0.031) per kWh.

* with additional IDC: R0.47 ($0.033) per kWh.

All of which remain well within the affordable range of tariff estimates.

Should South Africa rush out and order nuclear power plants?

To the readers that skipped the calculations, welcome back.

The above calculations are nothing more than that, mere calculations that can easily be replicated and analysed or critiqued.

It does, however, stress the fact that speculation on prices can be very dangerous, and to do what is best for South Africa would be to test the market.

Testing the market does not mean to look for least cost only, it would mean to enter into negotiations (non-binding) with vendor countries in order to determine the best financing agreement to be reached.

I am certain that no rational argument can be made against pursuing a cost of R0.39 per kWh, if we can find an agreement such as the UAE/Korea one, we might even look at R0.21 per kWh.

The importance of the financing model cannot be overlooked. Changing the financing model from a public/private partnership model to a private only funding model changes the results completely.

Even more importantly, if South Africa plan the tariffs going into the future assuming a private funding model for nuclear, and we manage to finance it in a similar way to the Egypt/Russia agreement (3% interest on 85% loan), we will see an internal rate of return on the private equity component (15% of the project) of 46.5%.

As a final point, South Africa has excess electricity capacity at the moment. We should not walk with open eyes into the same trap we walked into in the late 90s. Most of our coal plants will be decommissioned in the next 15 years (80% of them in the next decade). Will we avoid load shedding then? 

Next time you hear announcements about nuclear being “unaffordable”, or “cheapest electricity available”, don’t forget, they might all be right, depending on the model. Be wary of those making these statements off as “ridiculous”, or simply “wrong”. We still have some work to do it seems. We also need to ensure any process be as transparent as possible, and free from corruption.

Other African countries currently looking at similar nuclear agreements are: Kenya, Nigeria, Zambia, Ghana, Tanzania, Uganda, Ethiopia.

The South African Network for Nuclear Education, Science and Technology (SAN-NEST) is currently playing a leading role in developing knowledge transfer and developing strategies with these countries to ensure a sustainable nuclear industry in Africa. “Knowledge is power”.

  • Dr. Anthonie Cilliers Pr.Eng is the national coordinator: South African Network for Nuclear Education, Science and Technology.

Join the nuclear conversation. Email Fin24 now.

SUBSCRIBE FOR FREE UPDATE: Get Fin24's top morning business news and opinions in your inbox.

Read Fin24's top stories trending on Twitter:

We live in a world where facts and fiction get blurred
In times of uncertainty you need journalism you can trust. For 14 free days, you can have access to a world of in-depth analyses, investigative journalism, top opinions and a range of features. Journalism strengthens democracy. Invest in the future today. Thereafter you will be billed R75 per month. You can cancel anytime and if you cancel within 14 days you won't be billed. 
Subscribe to News24
Rand - Dollar
Rand - Pound
Rand - Euro
Rand - Aus dollar
Rand - Yen
Brent Crude
Top 40
All Share
Resource 10
Industrial 25
Financial 15
All JSE data delayed by at least 15 minutes Iress logo
Company Snapshot
Editorial feedback and complaints

Contact the public editor with feedback for our journalists, complaints, queries or suggestions about articles on News24.

Government tenders

Find public sector tender opportunities in South Africa here.

Government tenders
This portal provides access to information on all tenders made by all public sector organisations in all spheres of government.
Browse tenders