Why we should say NO to
nuclear power

2. Because it is too expensive

The first point to note is that there can be no hard data about the actual cost of delivering electricity from hypothetical nuclear power stations that haven’t been built.

I recall working in the United Kingdom in the 1970s when the government was considering future electrical generation capacity. There was a brisk debate about the relative economic performance of coal-fired power stations and nuclear. Each side was able to produce convincing calculations showing it would produce cheaper power! While a nuclear power station is more complex and expensive to build than its coal-fired equivalent, it needs less fuel so the running cost is lower. How much lower depends on future fuel prices for coal and uranium, which cannot be known. So each side of the debate chose assumptions that gave it the desired answer.

The overall economics also depends on how the higher capital cost is offset against the lower running cost. Economists use a technique called discounted cash flow. It discounts future costs and revenues to bring them to a “net present value”, allowing the lifetime performance to be calculated.

But this computation depends critically on the discount rate, the factor used to reduce the future costs and revenues. A high discount rate makes future fuel costs less important and so favours coal, while a low discount rate makes future costs more important and favours nuclear. Proponents of each approach choose the assumptions that make their favoured technology look best.

Ironically, the running costs of renewables like wind and solar are even lower than for nuclear, because they don’t use fuel at all. So the assumptions chosen by pro-nuclear advocates, to make that technology look attractive against coal or gas, also improve the economics of renewable supply systems.

There is a further complication. The running cost of a nuclear power station is almost independent of the amount of electricity generated, but the revenues from sales are critically determined by how much power is produced. The power supplied by a power station depends on its reliability, since it can’t produce power while off line for maintenance or repairs.

It also depends on other factors that have nothing to do with its reliability, such as the demand for electricity, how the load varies during the day and strategic decisions made by controllers about which power station to use from those available.

If a power station is on line most of the time, it can deliver as much as 90 per cent of its theoretical capacity over time. Some reliable stations in well-managed systems achieve that sort of figure. But if the economic calculation assumes the power station will deliver 90 per cent of its theoretical capacity and it actually achieves only about half of that, as with some reactors in the United Kingdom, the power is twice as expensive.

So statements about the future economic performance of nuclear power stations should always be taken with generous pinches of salt, even if based on proven designs in systems with a long operating experience like those in the United Kingdom or France. You should always examine the assumptions, since they determine the credibility of the estimates.

Given all that, the conclusions of the Switkowski report about the likely economics of nuclear power in Australia would not have you rushing for your cheque book. Let me quote from two paragraphs:

Nuclear power is likely to be between 20 to 50 per cent more costly to produce than power from a new coal-fired plant at current fossil-fuel prices in Australia. This gap may close in the decades ahead, but nuclear power, and renewable energy sources, are only likely to become competitive in Australia in a system where the costs of greenhouse gas emissions are explicitly recognised. Even then, private investment in the first-built nuclear reactors may require some form of government support or directive.

... While carbon pricing could make nuclear power cost competitive on average, the first plants may need additional measures to kick-start the industry.

So the view of the pro-nuclear task force was that nuclear power would not be economically attractive even though putting a price on carbon emissions would make electricity from coal or gas more expensive.

The most optimistic estimate in the report was that nuclear power would cost 20 per cent more to produce than the present mix, but even this figure was based on very ambitious assumptions about government subsidies. More realistic calculations in the same report put the likely price of nuclear electricity at least 50 per cent higher.

These estimates don’t include any provision for insurance, for the very good reason that the commercial insurance industry is not prepared to cover the risks of the nuclear industry. That means the public would essentially be the insurer for the nuclear power industry and would pick up the tab if anything went wrong.

The detailed analysis in the Switkowski report simply emphasises the economic improbability of nuclear power. It cites American studies of the likely cost of new power stations, estimating wholesale electricity prices in the range of A$75 to A$105 a megawatt-hour. The comparative Australian prices it gave were A$30 to A$40 for coal, A$35 to A$55 for gas turbines, about A$55 for wind power or small hydro, and A$70 to A$120 for solar thermal or biomass. As I was writing, proponents of geothermal energy were claiming costs of about A$60.

Based on the cost of constructing nuclear power stations in the United States, nuclear power in Australia would still be two to three times the price of power from coal or gas, much more than wind, small hydro or geothermal, and similar to solar thermal or biomass.

The report then claims that the cost of nuclear power in Australia would be very much lower, in the region of A$40 to A$65 a megawatt-hour, “if Australia becomes a late adopter of new generation reactors”. This estimate is breath-taking in its optimism.

It assumes there will be spectacular improvements in a new class of reactors which have yet to be built anywhere, roughly halving the price. Yet no such improvements have ever been achieved in the 50-year history of the industry! The report concedes that building a nuclear reactor in Australia would be more expensive than in the United States because of the lack of construction experience and a regulatory framework. Applying this weighting to actual American figures gives a likely cost in the range A$83 to A$120 a megawatt-hour.

There is also an obvious logical contradiction in this argument. It says we should go ahead with nuclear power now because the cost would be acceptable if we were a “late adopter” of a new generation of reactors. But we can’t be a late adopter if we order now, and we can’t order now if we want to be a late adopter.

If we wait until the promised wonder-reactors are actually built and operated successfully, it will be some time in the 2020s before Australia can place an order. With a further 10 to 15 years construction, it would be some time in the 2040s before the power station delivered its first electricity.

Some other economic arguments used to support nuclear power are even more outrageous.

A research group at the University of Melbourne5 noted in a recent study that the cost of building power stations in the United States is up to US$5,000 per kilowatt of power produced and takes about 10 years. It then says, “Westinghouse claims its advanced PWR reactor, the AP1000, will cost US$1,400 per kilowatt for the first reactor and fall to US$1,000 for subsequent reactors. It also claims these reactors will be ready for electricity production three years after first pouring concrete.” The report concludes that if these targets are met, “it will provide cheaper electricity than any other fossil fuel-based generating facility”.

This is an extraordinary conclusion.

It is true that the economics of nuclear power generation would look a lot better if the cost could be cut by a factor of five and the construction time halved, but they are truly heroic assumptions. It is like conceding that new cars selling for $40,000 and using 12 litres of fuel per 100 kilometres are expensive to drive, but saying if car companies could get the price down to $8,000 and reduce fuel consumption to four litres per 100 km, driving would be cheap! Of course it would be. But you would not allocate government funds in the hope of achieving that sort of outcome. It is purely wishful thinking.

The brutal truth is that the nuclear wave peaked about 30 years ago and has been declining since. As electricity systems around the world liberalised and private companies were forced to weigh up competing options, investors turned their back on nuclear power.

The number of reactors in western Europe and north America peaked about 20 years ago and has been declining since, as closures and cancellations outnumber new orders. In the decade up to 2003, the average annual rate of increase in the world’s nuclear power capacity was 0.6 per cent compared with 1 per cent for oil and coal, 2 per cent for gas, more than 20 per cent for solar and nearly 30 per cent for wind.

Despite the recent hype from the industry, most of the world has rejected nuclear power for alternatives that are cheaper, cleaner, more flexible and less dangerous.

At a conference in Adelaide just a few years ago,6 the editor of Nucleonics Weekly warned of real uncertainty about the likely cost of new designs for nuclear reactors still on the drawing board. He gave two recent examples: the French company AREVA had lost hundreds of millions of dollars on a new reactor still being built in Finland, while the design for two reactors in Taiwan was only 65 per cent complete 11 years after contracts were signed.

He also said that it was not clear any supplier of nuclear reactors would be interested in trying to build one or two in a country with no supporting infrastructure or past experience.

So, firstly, it might not be possible to persuade a commercial operator to build a nuclear power station in Australia. Even if such a company could be found, it is impossible to say what the reactor would cost; even charitable assumptions imply it would be much more expensive than the more attractive renewable supply options. Alternatively, Australia’s experience of building and operating hydro-electricity, wind turbines and solar hot water, provide both the necessary skills as well as hard evidence of what the alternatives will cost.