Analysis and Commentary


What are the steps to building nuclear power stations – by Peter Farley

Analysis and Commentary




In this, the third part of a series about Coalition plans for nuclear power in Australia, Peter Farley asks the question – How would we establish a nuclear power industry?

It is often claimed that many countries are going nuclear and if Australia wants to be a ‘Developed Country’ we should have nuclear power.

If that is the case, how would we go about it?

The first step is to build a regulatory framework. Regulations differ from country to country partly for historical reasons and partly different circumstances.

Then like all codes, the regulations are modified bit by bit in response to new hazards or changes in practice, but they become cluttered with deadwood with slightly conflicting requirements which make compliance difficult and expensive.

Typically, it takes two to three years and US$1-2bn to get a licence to build a new nuclear plant in the US, even if the plant itself is an approved design on an existing site.

We could work with the international agencies to develop a modern set of regulations, but that could easily take 4-5 years.

So, let’s say we agree to just doing a quick copy and paste job after legislation is passed in late 2025 and we start to recruit the necessary staff in 2026.

The UK has 700 regulation staff, France 1,500 and the US 2,700 so by 2030 the regulations would be published, and site selection could proceed.

The French regulator has a budget of €150m/y and French salaries are about 25 percent lower than ours.

We can then estimate that the first thirty years of the Regulator’s life would cost the taxpayer $3-5bn.

The nuclear workforce

At the same time, we need to establish and train the construction workforce.

EDF Energy claims that 22,000 people in Britain are working on the Hinckley Point C power station (pictured). This is nine times the size of the Snowy II workforce and almost three times the size of the entire Australian oil and gas extraction industry.

Most of these people would need security clearance and many would need additional training and certification.

If we say that on average it is 3 months training for two thirds of the staff, that is an initial training load of about 900,000 person days, say 1,000,000 days including training the trainers.

Over seven years, that is a cost of about $200m not including students wages.

However, Britain is building two plants in about 17 years. If we want six plants on line by 2065 we would be building a peak of five at one time so the workforce would reach 45-50,000.

With retirements and departures, the training and security bill will be over $700m over thirty years.

Fabricating nuclear power stations

Now apart from the lack of skills, we don’t have fabrication workshops with twin 500 tonne cranes and appropriate welding and heat treatment equipment.

These cost about $200m to build and equip.

Further, transport constraints and the distance between projects means that new workshops will probably need to be built near each site as would concrete batching plants etc. Some equipment would be shared but $500m in plant costs is not unreasonable.

Then there is the question of build cost. CSIRO used Korean figures, but they are highly questionable as there have been no public updates since 2018 and the company building and operating the nuclear plants is carrying US$150bn in debt, about two years annual sales.

We do have public data on other reactors, although that data usually doesn’t include all the losses made by contractors such as Westinghouse and Siemens who withdrew from projects after billions in losses.

All these plants are built on existing sites with plentiful cold cooling water and robust transmission access and an established nuclear workforce.

Access to cooling water can’t be dismissed. A single reactor cooling tower evaporates enough water for a city of 350,000 people.

Even without the traditional ‘Australian Premium’ for construction projects, believing we could build plants at less than a 10 percent premium over experienced northern hemisphere countries is stretching credibility, so A$42-45bn each in 2024 dollars is likely with another $2-4bn on water and transmission infrastructure per reactor.

Adding up the bill for nuclear power

All up the bill is approaching $270bn over thirty years to build enough nuclear capacity to supply about 50-55,000 GWh/y.

If we build the cheapest plant above and somehow manage to give the Americans twenty-three years start and build them for the same cost, it is still close to A$175bn for less than 50,000 GWh/y.

According to the US Department of Energy, fuel, operations, security, maintenance and other overheads are around A$50-65/MWh.

At current cost of capital, amortising the establishment cost over 60 years, finance and depreciation works out at about $400-500/MWh.

Best case total cost of $450/MWh.

Further, as it is not uncommon to have three or four of six plants offline at once for six weeks or more, we would need to maintain even more gas/coal/hydro than we have now to supply 500-600,000 GWh/y of demand in 2060-70.

Or what of wind, solar?

On the other hand, we are currently building enough wind/solar/storage every year to add 7-10,000 GWh/y, roughly the equivalent of one nuclear power plant.

With enough storage to make wind/solar as reliable as nuclear, if the government offered a guaranteed price of A$120/MWh they would be knocked over in the rush.

Further reading:
Nuclear does not mean reliable power for Australia – by Peter Farley

Peter Farley holds an engineering degree and is a manufacturing leader who built pioneering CNC machine tools for export winning many export and engineering awards. Peter has been studying the electricity sector since his 2012 Election to the Victorian Committee of Engineers Australia.
A realistic time frame for building nuclear- by Peter Farley

Picture: EDF Energy/Hinkley Point C will provide zero-carbon electricity for around six million homes when complete



Share this Story
Analysis and Commentary



Stay Informed


Go to Top