Nuclear Energy

The argument that is most widely made in support of nuclear energy is that there is no alternative. In a sense, that is true. But there is a fallacy here: if we are considering options A, B, C and D, the fact that options A, B and C do not exist cannot be interpreted as proof that D exists.

The big three energy providers are oil, gas and coal; nuclear, which supplies electricity to the grid equivalent to about 2½ percent of the world’s final demand for energy, is very small in comparison. And those three, as detailed in Energy Prospects, are in trouble, so the incentive to turn to nuclear energy as the only remaining alternative is strong. The catch is that nuclear energy is in trouble too.^N105

At first sight, the nuclear energy industry is a model of clean and orderly electricity-generation set in an idyllic environment. The favoured portrait shows a nuclear plant with its graceful cooling towers as a backdrop to a peaceful reed-lined lake, home to happy-looking ducks. But the industry has a rather unfair advantage in these matters. Most of the mess it makes is far away, both in space and time; some of it is invisible; and nuclear fuel itself is a compact item, which most people have never seen. Relative to coal-fired power stations, with their heaps of coal and ash, nuclear reactors seem to be close to magic (Energy Prospects > Coal).

But the reality of the nuclear cycle is rather different. It involves a sequence of many stages, almost all of which produce large quantities of radioactive waste, and use large quantities of energy with carbon emissions to match. It starts—like coal—at the mine, and if you read on, and look at the table, you will find out what happens next. Roughly. It is something of a disgrace that the story cannot be made more robust, more immune from critics armed with contrary numbers, than it is. The problem is that, with one exception, no one has ever carried out a comprehensive analysis of the whole nuclear energy chain, attaching fully-researched estimates of the energy used at each stage. That one exception is the nuclear engineer, Jan Willem Storm van Leeuwen, who (in partnership with the late Dr. Philip Smith, until his death in 2005) has produced a complete life-cycle assessment of nuclear energy.

LIFE CYCLE
Nuclear Energy

It is criticised on all sides—both by the nuclear industry and by informed critics of nuclear energy—but, as one critic recognises,

Storm is the only person who has tried to put the whole nuclear chain together. His work is the best analysis we have at the moment.^N107

As for his critics among the uranium miners, who insist that there is still plenty of uranium available,

You have to remember that mining is optimism. It’s a heroic and risky business. Optimism is what keeps them going; they just have to believe the stuff is there.^N108

The following paragraphs use Storm van Leeuwen’s numbers. Don’t take them too precisely, because by the time you read this, they will have been revised. But bear in mind that the overall story is clear and decisive—nuclear can only be a minor player for the rest of its short life and, before long, it will stop being a net supplier of energy; instead, it will be needing to use more energy than it can supply.^N109

So, back to the mine. A typical reactor needs 200 tonnes of natural uranium for a year’s output. To keep a reactor in fuel for a year, therefore, it is necessary to mine around 140,000 tonnes of ore. How come? Well, uranium ore is contained in rocks at ore concentrations of between 0.2 percent and 0.01 percent (richer ores are rare and largely exhausted). Let’s assume a generous average of 0.15 percent. At this ore grade, the milling process which extracts the uranium oxide from the surrounding rock is 95 percent efficient—and 200 × (100/0.15) (100/95) ≈ 140,000. At the proposed opencast Olympic Dam mine in Western Australia, the ore grade is 0.03 percent, and at this grade the milling process is 86 percent efficient, so some 800,000 tonnes of ore has to be mined for one year’s supply of fuel for one reactor. Moreover, the ore bed itself is 350 metres down, so a large ‘overburden’ has to be removed before the mining itself can start.^N110

Extraction of the uranium oxide consists of crushing and grinding, leaching with sulphuric acid, washing and neutralising with salt, and then further treatment with (amongst other chemicals and solvents) ammonia, lime and nitric acid. The mill tailings must then be removed, and the ore has to be washed, clarified, filtered, further refined by precipitation, and then cleaned and dried to produce uranium oxide, or yellowcake.^N111

Next comes “enrichment”. The isotope that starts the process off in the reactor is uranium-235, but there is only a trace of it—0.7 percent—in natural uranium, and this has to be brought up to 3.5 percent. The yellowcake is reacted with fluorine compounds to produce uranium hexafluoride gas, which is then placed in centrifuges. For every tonne of uranium hexafluoride enriched—and then fabricated into ceramic pallets for use in the reactor—some five tonnes are depleted. Depleted uranium hexafluoride is toxic, radioactive and explodes on contact with water; it abruptly turns from a solid to a gas if it warms up (to 56.5°C) and, as a gas, it belongs to a group of chemicals (the halogenated compounds) whose impact on the climate ranges up to some 20,000 times as much as the same mass of carbon dioxide. For final disposal, it must be converted into uranium oxide or into metallic uranium, before being placed in secure containers and buried deep underground in a geologically-stable repository.^N112

Nuclear fission itself does not produce carbon dioxide, but most of the processes surrounding that event are energy-consuming. The reactor itself has to be built, and then comprehensively refurbished at least once in its lifetime, before eventually being decommissioned, cleaned out, broken into pieces, packed in containers and buried. For each year of operation, 200 tonnes of fuel becomes 200 tonnes of high-level waste. That waste is very radioactive: stand close to it for a moment or two and you are dead. As they become more tightly-packed, the rods need to be separated by boron panels to stop active components of the waste—such as plutonium-239—forming a critical mass. It needs to be kept under water in cooling ponds for a minimum of 10-20 years before being placed in containers. At present, temporary containers are used. Permanent storage requires casks made of steel, lead and electrolytic (very high quality) copper, which are then buried in deep repositories lined with bentonite clay. And the cooling ponds in which fresh high-level waste is stored require a constant flow of water; if they dry out, the spent fuel rods can be expected to catch fire. A power supply and stable institutions are needed to keep the water circulating for the necessary decades.^N113

The processes listed in the table as “front end” are those involved in producing electricity for the grid; “back end” processes constitute the clearing up afterwards. At present, almost none of the waste produced by the nuclear industry has been treated and buried, but this must eventually be done. The total energy cost of the front end and the back end is indicated in the table by the carbon emissions which they produce. As it happens, the energy costs for the front end and the back end of the process are approximately the same, at around 55 grams of carbon dioxide per kilowatt hour of electricity delivered to the grid. At an ore grade of 0.15 percent, the energy needed by the front end is equal to about 25 percent of the gross output of energy to the grid. As is the energy needed by the back end. In other words, the net power available for consumption—after allowing for those front end and back end energy costs—is just 50 percent of the gross power produced.^N114

Now, all this depends crucially on the availability of high quality ore. The problem is that, as the industry matures, it has to turn to poorer grades, and there is a theoretical limit at which so much ore has to be mined and milled, and so much energy is needed for the task, that the energy balance turns negative: more energy has to be put in to the process than comes out of it. The theoretical limit for the grade of ore from which you can get a positive energy balance—assuming that everything goes according to plan—is (for hard rock) 0.02 percent, or (for soft rock) 0.01 percent.^N115

But that really is only a “theoretical” limit, because things do not generally go according to plan. Here we can draw on a crucially important measure of energy wealth: “energy return on energy invested” (EROEI). The EROEI of the entire economy of the United States in 2008 has been estimated at about 40:1—in other words, about 2.5 percent of the economy’s total output is devoted to gathering the energy it needs. If EROEI drops to 10:1, some 10 percent of the economy must be devoted solely to energy-gathering; at 1:1, it is 100 percent: energy limits, when they occur, are decisive. So it is important not to fudge the real value of EROEI, and to make a clear working distinction between the “theoretical return on energy invested” (TROEI) and the “practical return on energy invested” (PROEI). If (for instance) the ore deposit is very deep, or if there is constant flooding, or if there is no water (so that the process depends on desalinated seawater), or if diesel oil becomes scarce, then, in practice, the return on energy invested can be worse than 1:1. It can turn negative. For an energy project to be worthwhile, it needs to produce a PROEI of at least 20:1.^N116

So, how much high-grade uranium remains? Well, depletion estimates are never straightforward, but let us start with one that says there is some 60 years’ supply of uranium rich enough to at least give us a positive theoretical energy return (TROEI). Suppose that, over this period, the number of nuclear reactors that now exist in the world could theoretically continue to produce a net flow of energy to the electricity grid:^N117

• Now deduct the energy that would be needed for the mining, milling and construction (front end) processes described above. That energy—approximately 25 percent of the gross output—is equivalent to 15 years’ supply.^N118

• Now deduct the energy that would be needed for all the cooling, clearing up, container construction, dismantling, and burial (back end) processes needed. That also comes to about 25 percent of the gross energy output—another 15 years’ supply.^N119

• Now remember that, while the industry is here assumed to have 60 years of life ahead of it, it also already has 60 years of life behind it, during which it has been steadily producing waste, none of which has been cleared. That will require energy equivalent to 25 percent of the gross output—yet another 15 years’ supply.

If we add these together, it means that approximately 45 years of a 60-year total supply of uranium would be used up either in the (front end) production process itself, or in the (back end) waste disposal. And that in turn means that our notional and optimistic estimate of 60 years remaining supply leaves us with just 15 years’ worth of electricity delivered to the grid: from the turning point of 2025, the industry will effectively have to direct the whole of its output to the task of clearing up its own waste.

In fact, the idea that there is sixty years’ supply of uranium at the present rate of demand looks optimistic. Already the production of uranium from mining falls well short of present demand of 65,000 tonnes a year; the shortfall (22,000 tonnes) is supplied by drawing down from stockpiles of uranium and dismantled missiles left over from the Cold War, which are expected to reach exhaustion around 2013.^N120 So we have to ask: what if the available supply were significantly less than 60 years? If the supply of ore rich enough to satisfy the requirements of PROEI were half of that—30 years—then the turning point after which the industry would need energy equivalent to the whole of its own gross energy output for the task of clearing up its own wastes is 2010. If there were only fifteen years’ supply of ore at current rates of demand, then the turning point would have been 2000.^N121

If the clear-up is (or has already been) postponed beyond the turning point, then the industry will never be able to generate enough energy to clear up its own wastes, and will have to use other sources of energy which it will never be able to pay back. The alternative is to leave the waste dumps on the landscape indefinitely. But without attention and protection, and a constant supply of energy to circulate the cooling water, the unstable waste would in due course start to leak and catch fire. In many cases it would be flooded in the early years of rising sea levels.

But what if these depletion estimates on uranium supply are overly pessimistic? Suppose that rather than 60 years’ uranium supply left, there were as much as 120 years’ worth, enough to last at the present rate of extraction until 2130. Well, using the same estimates for the energy-cost of dealing with the waste, the turning point at which the industry would need to use the whole of its energy to clear up the waste would be 2055—to be followed by 75 years of working to capacity on the task of waste clearance, with no energy sales to pay for it. It seems unlikely that, even if the industry did find vast and unexpected supplies of uranium, its past and future waste dumps will ever be cleared up.

The impacts of an uncontrolled worldwide legacy of nuclear waste dumps, degrading and catching fire at their leisure, are not known for certain. Would there be a series of localised Chernobyls erupting around the planet indefinitely? If reactors and their spent fuel dumps are flooded as sea levels rise, what would be the implications for coastlines, the oceans and fishing? And how might all this affect our health and well-being? This would be a dainty research project. What we do know is that, if the actual energy costs of preventing this outcome are factored into our understanding of the matter, the nuclear industry cannot any longer be regarded as an available source of energy.^N122

It is possible that more token reactors will be built. It will almost certainly have to be with government funds, since the industry itself is aware of the problems. And by the time the new reactors are ready to start paying back the energy that was used to construct them, the world’s supply of uranium will be so depleted that it is likely that they will never be started up. Meanwhile they will have shifted the priority of energy policy away from solutions that actually work. If they are started up, and then shut down for lack of fuel, at the same time as the electricity grid is down owing to a lack of gas, then there will be no power supply to keep water circulating in the ponds. They will then dry out.

Related entries:

Energy Prospects, Control Overload, Lean Energy.