What is the real cost of new build?

April 9th, 2015, Published in Articles: Energize


Although new build projects are coming on line this year, attention needs to be given to projects which need to be initiated in the near future, and decisions taken on the mix of new build. One of the issues clouding decisions on this issue is the cost of various options. This article takes a comparative look at the possible new build costs with the intention of bringing some light onto the issue.

Unfortunately the question of cost has not been looked at objectively, and the cost of certain technologies has been considered in isolation, without taking the cost of implementing equivalent capacity using other technologies into account. Opinions are further confused by the different costs which are held up for comparison, like LCOE, grid parity, cost/watt cost to generate etc. These disparate costs are usually used together in the same argument, and isolated costs find application as an argument of last resort, for example, to provide a reason for rejecting a technology. It is often argued that a certain technology would be an ideal solution to one of the problems facing generation but the construction costs are unaffordable. The argument of installed price is used instead of the actual generation cost per unit, while at the same time the generation cost per unit of energy is used to justify or support another technology. While the installed cost of coal, nuclear and gas is held up as a reason not to proceed with these technologies, the real installed cost of other technologies is ignored.

Table 1: Nameplate capacity requirements vs capacity factor.
Capacity factor Capacity adjusted sizing (MW) Capacity adjusted (MW)
100 1000 1,05
95 1053 1,05
90 1111 1,11
60 1666 1,67
40 2500 2,50
35 2857 2,86
30 3333 3,33
25 4000 4,00

To clarify some of the confusion, or sow some more doubt, this article focuses purely on the installed cost of a specified annual energy generating capacity, with figures being taken from published sources, and details of how the calculated costs have been reached will not be discussed. The article does not deal with LCOE or other hypothetical costs, as these depend on which system of accounting is used, and can anyway only be determined accurately after a power plant has been decommissioned.

Table 2: Cost of nameplate capacity.
Technology Cost/nameplate MW capacity Source
Coal R31,4-million (estimated Medupi)
Solar PV R30-million (REIPPP) [2]
Wind R25-million Published
Nuclear low R40-million Published data
Nuclear high R70-million Published data
Nuclear price cap R65-million IRP 2010

Capacity adjusted sizing

For the purpose of comparison and to highlight real costs “capacity adjusted size (CAS)” is defined as the size of plant, in nameplate MW, required to produce a fixed amount of energyper year. Capacity adjusted MW (CAM) is the nameplate capacity in MW required to produce the same energy as 1 MW capacity at 100% capacity factor.

In this case, the reference unit will be 1000 MW at 100% capacity factor, giving an annual production of 8760 GWh of energy. The CAS will depend on the capacity factor of the power plant. Table 1 shows the nameplate capacity required to achieve the reference energy generation for a range of capacity factors.

The 60% row represents the possible usage profile for load-following or mid term generation.

The impact is that with a capacity factor of 25%, the plant would require four times the amount of nameplate capacity of a plant at 100% capacity factor, to produce the same amount of energy per annum.

Consider the costs of providing this capacity using different technologies. Table 2 shows the estimated installed cost per nameplate capacity of different technologies. Figures are obtained from recently constructed projects in SA where possible. The installed price of nuclear is an indeterminate factor, with reported figures in the range $6700 to $4900/kW for completed projects but with new contracted prices of $3500/kW also being reported. An upper and lower range has thus been given, as well as the capped price of R65-million/MW used in the IRP 2010.

Using capacity adjusted figures the cost of plant capable of generating 8760 GWh/year is shown in Table 3.

From this we can deduce a “capacity adjusted MW cost” for the various technologies as shown in Table 4. This would the cost of an equivalent MW of capacity taking capacity factor into account.

This shows that the cost of using solar PV would exceed that of nuclear for the same energy generating capacity at all capacity factors, wind would be comparable with nuclear as a mid-term range. Coal remains the cheapest source at any capacity factor.

Table 3: Comparative installed cost of 8760 GWh/year plant.
Capacity factor Coal Solar PV Wind Nuclear upper Nuclear cap Nuclear low
% R-million R-million R-million R-million R-million R-million
100 34 100 N/A N/A 70 000 65 000 40 000
95 35 895 N/A N/A 73 684 68 421 42 105
90 37 889 N/A N/A 77 778 72 222 44 444
60 56 833 N/A N/A 116 667 108 333 66 667
40 N/A N/A 62 500 N/A N/A N/A
35 N/A N/A 71 429 N/A N/A N/A
30 N/A 100 000 83 333 N/A N/A N/A
25 N/A 120 000 100 000 N/A N/A N/A


Table 4: Capacity adjusted cost/MW installed.
Capacity factor Capacity adjusted cost/MW coal Capacity adjusted cost/MW solar PV Capacity adjusted cost/MW wind Capacity adjusted cost/MW nuclear upper Capacity adjusted cost/MW nuclear cap price
% R-million R-million R-million R-million R-million
100 34 70 65
95 36 73 68
90 38 77 72
60 57 116 108
40 63
35 71
30 100 83
25 120 100


Table 5: Cost of adding storage to renewable.
Capacity factor % Nameplate cost/MW with 30% storage (R-million) Cost with 60% storage for      nameplate              MW              (R-million) Capacity adjusted MW cost including 30% storage Capacity adjusted MW cost including 60% storage
40 38,64 47,28 96,60 118,20
35 37,56 45,12 107,31 128,91
30 36,48 42,96 121,60 143,20
25 35,40 40,80 141,60 163,20

Cost of storage

These calculations ignore the variable nature of renewable energy, so the comparison is not strictly valid. For low penetration of renewables this is not a problem, but once higher levels are achieved the use the energy generated is limited. Both on-site and centralised storage are seen as the solution to this problem. The amount of storage required to achieve dispatchability of RE has not been calculated, but upper and lower limits of 30 and 60%  of the daily production sound reasonable.

The cost of storage depends on the technology used, and the cheapest present day cost of established technologies ranges from R3-million to R6-million per MWh [3]. The lowest cost will be assumed here.

Table 5 gives the calculated possible range of costs for adding storage to renewable technologies – taking the nameplate MW cost for PV as well as the capacity adjusted MW cost.

Comparison: Future build nuclear vs renewables

Leaving coal and other carbon based fuels out of consideration, how does nuclear as a carbon free fuel stack up against renewables? Consider for instance the proposed nuclear fleet of 9600 MW and assume that 2400 MW would be used as mid-term, with the balance as base load, Assuming that the capped price for nuclear is the lowest and considering costs at two higher prices and assuming 30% capacity factor for renewables gives the values shown in Table 6.

Table 6a: Cost of nuclear new build project.
System Low end cost Mid cost High end cost
95% capacity R65-million/MW R70-million/MW R75-million/MW
60% capacity R102,92-million/MW R110,83-million/MW R118,75-million/MW
Total project cost for nuclear R715-billion R770-billion R825-billion


Table 6b: Cost of renewables new build.
Storage capacity 0% 30% 60%
Project cost for renewables plus storage (R-billion) 960 1153,44 1372,80

To provide the same energy as the proposed nuclear fleet of 9600 MW using renewables would cost R1100-billion to R1380-billion, a figure which has been stated as impossible to afford, even at a price higher than the IRP 2010 revision cap of R65-billion, using renewables could cost up to R500-billion more than nuclear.

The difference here is that renewable energy will be purchased via the REIPPP program, funded from private sources, so the build cost don’t accrue to Eskom or the government, and build cost, i.e. actual costs, are therefore of no concern and will accrue to someone else. Yet all this will be coming from the the private sector. If the private sector is going to be willing to invest that amount of money in renewables then surely the same investment could be made in nuclear. If new nuclear energy were to be obtained using an REIPPPP style bidding process would this alter the picture, and remove some of the objections?

As a final example consider the new build scenario sketched in the policy adjusted IRP2010, and the often quoted statement that renewables will comprise 42% of the new build program, based on the policy adjusted IRP2010.

Table 7: An analysis of energy generation capacity and costs.
Technology Coal Nuclear Renewables
Nameplate capacity (MW) 6300 9600 18 000
Total cost (R-million ) 214 830 624 000 684 000
Annual energy generation (GWh/y) 41 391 78 840 39 420
% of total annual generation 19,33 36,82 18,41

A quick look reveals that new renewables will cost approximately the same as new nuclear, while only delivering 50% of the annual energy generation, The figure of 42% refers to nameplate capacity and in fact renewables will only contribute approximately 19% to the new annual energy generation capacity,

Although the cost figures used here are estimates and could be questioned, the article serves to illustrate how easy it is to use selective cost information to support or oppose a particular technology. Realistic evaluation should compare all cost factors on an equivalent basis.


[1]    T Durning: “Nuclear and renewables shared goal and comparative cost”, The energy collective, http://theenergycollective,com/tracey-durning/2154591/i-went-hunting-clarity-nuclears-comparative-costs-and-heres-what-i-found
[2]    Jasper solar photovoltaic power plant, South Africa, www.power-technology.com/projects/jasper-solar-photovoltaic-power-plant
[3]    Sandia national laboratories: DOE/EPRI 2013 Electricity storage handbook in collaboration with NRECA, www,sandia,gov/ess/publications/SAND2013-5131.pdf

Send your comments to: energize@ee.co.za

Related Articles

  • South African Government COVID-19 Corona Virus Resource Portal
  • Ministerial determinations propose 13813 MW of new-build by IPPs, none by Eskom
  • Crunch time for South Africa’s national nuclear company, Necsa
  • Dealing with the elephant in the room that is Eskom…
  • Interview with Minerals & Energy Minister Gwede Mantashe