Analysis: How much will new nuclear electricity cost South Africa?

A new study by EE Publishers looks at the initial capital cost as well as the levelised cost of electricity (LCOE) generated by the proposed 9,6 GW new-nuclear build in South Africa.

This study estimates an initial overnight capital cost (including owner’s development costs, but excluding interest during construction) of the 9,6 GW new-nuclear build at US $50-billion (R776-billion at a rate of exchange of $1 = R14).

The levelised cost of electricity (at base date May 2016) is calculated to be R1,30 per kWh under the stated assumptions, rising to R1,52 per kWh if Koeberg’s fuel, operating and maintenance costs per kWh are used in the calculation.

This compares to the LCOE for new-coal generation in South Africa of R1,05 per kWh and R1,19 per kWh for Medupi and Kusile respectively, and the LCOE for new wind and new solar PV in South Africa of R0,69 per kWh and R0,87 per kWh respectively (all figures adjusted to May 2016).

Context

As with any large infrastructure development project, it is important to understand clearly upfront the total capital cost in order to determine whether it is affordable, how it should be financed, and how the capital cost may impact the financing of other projects that may be contemplated by the shareholder.

Similarly it is important to know the full ongoing costs over the economic lifetime of the project, including interest and capital repayments to the lenders, return on investment required by the owners, and all other fixed and variable operating and maintenance costs.

It is similar in principle to building a new house. How much will it cost? Where will I get the money from? How much will I need to put down myself? How much will I need to borrow from the bank? At what interest rate and repayment terms? Will my income be sufficient to cover not only the monthly capital repayments and interest, but also ongoing maintenance and running costs, such as water, electricity, rates and taxes, etc.? How will the down-payment and monthly repayments affect my other plans? Am I taking on too much debt?

Levelised cost of electricity

In the field of power generation, a planning tool known as the levelised cost of electricity (LCOE) over the lifetime of the plant is used to understand and compare the cost of electricity from a power plant, whether it is coal-fired, nuclear, hydro, gas, wind or solar.

The LCOE of a generation plant is the average cost per kWh unit of electricity delivered over its lifetime, which will recover the full costs, including the initial investment, cost of capital (including dividends and interest), fuel, and all other fixed and variable operating and maintenance costs.

The LCOE is useful when comparing and making investment decisions for power generation projects. While the concept is relatively simple, the calculation of LCOE is somewhat technical, involving such parameters and concepts as the vendor capital costs, owner’s development costs, overnight capital costs, weighted average cost of capital, interest during construction, plant nameplate and net capacity, capacity factor, plant construction time, plant lifetime, fuel costs, other fixed and variable operating costs, and more.

In stating the LCOE calculated, it is essential to also clearly state the economic and technical assumptions used, without which the LCOE presented has little meaning, and without which it cannot be compared with the LCOE of other projects or technologies.

Relevance

An updated initial capital cost estimation and LCOE calculation for the proposed 9,6 GW new-nuclear build in South Africa is particularly relevant at this time in view of the fierce debate raging between Department of Energy, National Treasury, Eskom, economists and independent analysts on the optimum future energy mix in South Africa.

Decisions in respect of nuclear, coal, gas and renewables are imminent, and yet the assumptions and LCOE for nuclear power (and the other generation technologies) presented in the national Integrated Resource Plan for Electricity IRP2010-2030 and the 2013 Draft IRP Update Report are significantly out of date.

Purpose of the study

The purpose of the study by EE Publishers is to update the base LCOE presented in the 2013 Draft Update Report for a 9,6 GW new-nuclear build in South Africa.

For comparative purposes, the EE Publishers study calculates the base LCOE as at May 2016 using the same methodology and general assumptions at the 2013 Draft IRP Update Report, but updating the initial overnight capital cost, and taking into account three years of inflation and significant variation in the rate-of-exchange.

The base LCOE is then recalculated under various alternative assumptions, and the results tabulated to enable readers to easily determine the LCOE of the proposed new-nuclear build resulting from a range of alternative and perhaps more realistic assumptions (see below). The tables also indicate the sensitivity of the LCOE calculated to the various key assumptions.

Discussion on the assumptions used in the calculation of LCOE and their validity

Overnight capital cost per kW of net output

Based on an estimated overnight capital cost of $50-billion (including owners development costs but excluding interest during construction) for a 9,6 GW nuclear fleet with a net output of 6 x 1082 MW using Rosatom VVER 1200 reactors, an overnight cost of $5776 per kW net output is calculated. Therefore this figure is used in the EE Publishers study to calculate the base LCOE. As this would vary depending on the proposals from the different vendor countries, the LCOE in the EE Publishers study has also been presented using alternative assumptions of overnight capital cost per kW of net output, ranging from $4000 to $8000.

Weighted average cost of capital (WACC)

The WACC takes into account return on capital required by the shareholder (government) as well as the cost of borrowing. The WACC used in the 2013 Draft IRP Update report is 8% real, and this has also been used in the EE Publishers study to calculate the base LCOE. Nuclear proponents may argue for a lower WACC, but 8% is the consistent figure used across all technologies in the IRP, and if one were to reduce this for a nuclear build, then there would be little reason not to reduce it for all other technologies too. However, for completeness the LCOE in the EE Publishers study has also been presented using alternative assumptions of WACC, ranging from 4% to 10%.

Plant capacity factor

The average plant capacity factor is a measure of the availability of a power plant for generation at full rated output, taking into account planned and unplanned outages, and output reductions. A capacity factor of 92% is used in the 2013 Draft IRP Update, and this has also been used in the EE Publishers study to calculate the base LCOE. EE Publishers considers this capacity factor to be unrealistic based on international experience and local experience at Koeberg, with a more realistic capacity factor considered to be 88%. Thus the LCOE in the EE Publishers study has also been presented using alternative assumptions of average plant capacity, ranging from 85% to 92%.

Reactor construction time

A reactor construction time of six years is used in the 2013 Draft IRP Update Report, and this has also been used in the EE Publishers study to calculate the base LCOE. Many would argue that this is unrealistically low based on international experience in Europe and the USA, and the new-coal build at Medupi and Kusile, and should be more like 10 years. However, the learning experience through the construction of a fleet of six reactors may tend to reduce this. The LCOE in the EE Publishers study has therefore also been presented using alternative assumptions of reactor construction time, ranging from 5 to 10 years.

Plant economic lifetime

A plant economic lifetime of 60 years is used in the 2013 Draft IRP Update, and this has also been used in the EE Publishers study to calculate the base LCOE. EE Publishers considers this lifetime to be unrealistic without taking into account major plant life-extension costs at mid-life, which has not been done in the 2013 Draft IRP Update Report. A more realistic plant economic life without major life-extension costs is considered by EE Publishers to be 30 or 40 years. Thus the LCOE in the EE Publishers study has also been presented using alternative assumptions of plant economic lifetime, ranging from 30 to 60 years.

Fuel and other fixed and variable operating and maintenance costs

A figure of R0,215 per kWh is used in the 2013 Draft IRP Update Report for fuel and other fixed and variable operating and maintenance costs, and this has also been used in the EE Publishers study to calculate the base LCOE. However, in a recent letter published in Business Day, Eskom’s Group Executive Generation, Mr Matshela Koko, disclosed that the “production costs” of Eskom’s Koeberg nuclear power plant was actually R433/MWh (i.e. R0,433/kWh). The assumption is that he is referring to the fuel and other fixed and variable operating and maintenance costs, whereas the IRP assumes a figure of R0,215/kWh. If this higher and likely more realistic figure is used, the base LCOE calculated increases from R1,30 per kWh to R1,52 per kWh i.e. 117% of the base LCOE.

Conclusions

Using the assumptions of the Integrated Resource Plan for Electricity IRP2010-2030 and the 2013 Draft IRP Update Report, and the other assumptions detailed above, and without taking tax effects, decommissioning, long-term waste disposal, and plant life extension costs into account, an up-to-date base LCOE for the proposed new-nuclear build in South Africa would be R1,30 per kWh.

This base LCOE of R1,30 per kWh would rise to R1,52 per kWh if the actual fuel costs, and the fixed and variable operating and maintenance costs of Eskom’s Koeberg nuclear power plant, were used instead of those assumed in the IRP. DM

Chris Yelland is the investigative editor for EE Publishers.

Photo: A picture made available on 21 June 2016 and taken with wide angle lens shows two cooling towers of the nuclear power plant Grafenrheinfeld, Germany, 31 May 2016.  EPA/DANIEL KARMANN