Nuclear Waste Production

Introduction

In nuclear reactors energy is released by fission of the uranium isotope U-235. In this process two fission fragments are formed as well as 2.5 new neutrons on average. From these neutrons, one is needed to sustain the fission chain reaction, while the other neutrons are absorbed or will leak away. Part of the excess neutrons will be captured by U-238 to form Pu-239 (plutonium), which is a fissile nuclide like U-235. Subsequent neutron capture processes produce heavier plutonium isotopes as well as higher actinides. Because these latter elements are produced in much smaller quantities then plutonium, they are called minor actinides and consist mainly of neptunium (Np), americium (Am) and curium (Cm).

The material balance in the fuel cycle of a large uranium-fueled reactor of 1 GWe is shown in the table below. From the 20,000 tons of uranium ore needed, about 24 tons of enriched uranium can be manufactured, which can be used to produce 7,000 million kWh of electricity. The spent fuel contains 23 tons of uranium with a U-235 contents of about 0.9% (a little bit higher than that of natural uranium), 240 kg of plutonium, and about 720 kg of fission products.

Mining 20.000 tons of 1% uranium ore
Milling 230 tons of uranium oxide concentrate (with 195 tU)
Conversion 288 tons of UF6 (with 195 tU)
Enrichment 35 tons of UF6 (with 24 t enriched U)
Fuel fabrication 27 tonnes of UO2 (with 24 t enriched U)
Reactor operation 6,000 million kWh of electricity
Spent fuel 27 tons (23 tU, 240 kg Pu, 720 kg FP)

After irradiation in a nuclear reactor, the fissile plutonium contents of the spent fuel equals about that of the fissile uranium (about 0.9%). Furthermore some even plutonium isotopes (30% of the total plutonium) are formed, which are not fissile in thermal reactors, but which could be fissioned in so-called fast reactors. The spent fuel contains about 4% fission products (by weight), which implies that 4% of the actinides have been fissioned. Roughly spoken about 35% of all electricity produced in the reactor comes from plutonium, even though there was no plutonium in the fuel initially.

Fission Products

Most uranium atoms that are fissioned do not split in two equal parts, but in a heavy and a light nuclide. Occasionally ternary fission may be observed producing three fragments. The distribution of fission products as a function of mass number is called the "camel curve" and shows two peaks for mass numbers around 95 and 140.

There is very little difference in the fission product distribution between a thermal-neutron induced fission and a fast-neutron induced fission. Larger differences can be observed between one actinide isotopes and the other. Comparing the fission product distributions of U-235 and Pu-239, for example, the latter is shifted in the direction of heavier fragments. This seems logical as the mass of the initial plutonium isotope is larger. Between the two fissile plutonium isotopes Pu-239 and Pu-241, differences are much less. Compared with the camel curve that can usually be found in text books, the real distribution shows a more erratic behavior.

Even more erratic distributions can be found for the fission product distribution in spent fuel. Because the fission fragments are neutron rich, they usually decay via beta- emission. This implies that the mass of the fragment remains the same during decay, but that the atomic number (equal to the number of protons in the nucleus) increases. Most fission fragments decay via several steps before a stable nucleus configuration is reached. During these steps, a nuclide with a large neutron capture cross section may be formed, such that the neutron capture rate dominates the decay rate of that nuclide. Consequently, relatively few fission fragments with that particular mass number will be found in the spent fuel. These dips in the distribution can be found due to Cd-113, Xe-135, Sm-149, and Gd-157.

It is mainly the fission products Cs-137 and Sr-90 and the actinide nuclide Am-241 that dominate the activity of the spent fuel for the first 300 years. Could one separate these two fission products from the spent fuel and give them a special treatment, like above ground storage for 300 years, the decay power production of spent fuel would be drastically reduced. Many research efforts concentrate on the separation of cesium and strontium as the storage capability of underground disposal sites are mainly determined by the thermal load of the waste.

A better parameter for the potential radiological hazard to people is the radiotoxicity of a nuclide. This parameter accounts for the decay mode of each nuclide (gamma, beta or alpha decay), the biological half life in a human body, and more. The radiotoxicity of spent fuel 10 years after discharge of the fuel from the reactor is equally due to the fission product and the actinides. However, the first contribution is decreasing more rapidly with time then the latter. The dominating half life time of fission products is about 30 years due to Cs-137 and Sr-90, while most actinides in the spent fuel have a much longer half life, even up to several hundreds of thousands of years.

Although the radiotoxicity of the spent fuel decreases rapidly with time, the remaining radiotoxicity is very long-lived: up to 16 millions of years for I-129 (iodine). There are five isotopes that dominate the radiotoxicity due to fission products after 1000 years of storage, namely Tc-99 (technetium), I-129, (iodine), Cs-135 (cesium), Zr-93 (zirconium), Sn-126 (tin) and Se-79 (selenium). For a large nuclear reactor with power of 1 GWe, the production rates of these nuclides are 21, 4.6, 12, 20, and 0.5 kg per annum, respectively. Because of the small neutron absorption cross sections, the isotopes Sn-126 and Zr-93 cannot be transmuted to non-radiotoxic nuclides. Furthermore Zr-93 cannot easily be separated from the other zirconium isotopes produced in spent fuel. This latter argument also holds for Cs-135, which is intimately mixed with the isotope Cs-137. Effectively only the Tc-99 and I-129 isotopes can be transmuted in nuclear reactors.

Fortunately, the dose rates to the population after 100,000 years of storage due to leakage of nuclides from the geological repository are mainly due Tc-99 and I-129. Nuclear waste that is stored in salt domes underground may leak from the repository when the salt dome rises and comes into contact with the groundwater. The resulting subrosion rate (dissolution rate of the salt in the groundwater) is estimated to be 0.1 to 1 mm per annum. For a salt layer of 100 m, this means that it will take 100,000 to one million of years before the groundwater comes into contact with the vitrified waste in the repository. After the nuclides come into contact with the groundwater, they may be dissolved in the water and be transported to the surface. This process depends on the groundwater flow rate (for an overburden thickness of several hundreds of meters, this may take 50,000 to 5 millions of years), and the adsorption rate of the nuclides in the underground. Unfortunately, both Tc-99 and I-129 are not retarded due to adsorption in the underground, which makes them the dominant contributors to the expected dose rate to the population in a salt subrosion scenario.

Once the nuclides are released to the environment, it is assumed that they all will be ingested by human beings. This can occur via various routes like water that is used for drinking water, irrigation or other purposes. For a repository containing the vitrified waste due to a reactor park of 100 GWea (200 times the annual production in the Netherlands), the maximum expected dose rates to the population reach values of 3 micro-Sv per annum, while the average value does not exceed 0.03 micro-Sv. These dose rates can be neglected compared to the natural background dose rate in the Netherlands, which measures about 3 milli-Sv per annum.

Several reactor types can be considered for the transmutation of Tc-99, like fast reactors, heavy water moderated reactors like CANDU or PWRs (either with UO2 fuel or MOX fuel). It turns out that the transmutation half lives in a fast reactor vary between 15 and 20 years, while those in a CANDU are about 25 years. In a PWR the transmutation half lives exceed even 40 years. In any case, for all reactor types the technetium inventory would amount up to several tons, which makes transmutation impractical and very expensive.

Other solutions that can be envisaged are immobilizing nuclear waste, this means storing the technetium in a synthetic rock such that the release rate of technetium to the groundwater is reduced, or retarding the transport of technetium to the surface (e.g. by storing technetium in a molecular form that is easily adsorbed by the underground rock layers).

Spent fuel contains significant amounts of rare-earth elements that are applicable as automobile exhaust catalysts and chemical catalysts (palladium, rhodium), and in the electrical (palladium, rhodium, ruthenium), electrochemical (ruthenium), glass (rhodium) industries. Furthermore these rare metals could be used in future as electrode fuel cell catalysts and in hydrogen energy systems. The concentration of these metals, especially those of Tc, Ru, Rh and Pd, are about a million times larger in spent fuel compared with the concentrations in the earth's crust. Also in PWR spent fuel with a three times lower burnup value than that of fast reactor spent fuel, the amounts exceed one kg per ton HM.

Actinides

Radiotoxicity due to actinides is dominated by plutonium and americium. The recycling of these two actinides seems first priority in order to reduce the radiotoxicity of the nuclear waste. The radiotoxicity of plutonium equals that of the natural uranium needed to manufacture the fuel only after 200,000 years. For americium this is around 5,000 years of storage. Although the radiotoxicity of natural uranium seems a safe limit to consider, one could argue that, because of differences in composition between nuclear waste and natural uranium, other limits are better. Also the radiotoxicity of spent fuel might not be a good measure to rank the impact of the spent fuel on the environment. Leakage dose rates, for example, are dominated by Np-237 and its decay products, which contribute very little to the radiotoxicity of the waste.

After several 100,000 of years, the radiotoxicity due to plutonium equals that of uranium in the spent fuel, while the radiotoxicity of americium equals that of neptunium. Curium gives a significant contribution (> 1%) only during the first century of storage. The actinides to which the radiotoxicity of nuclear waste can be attributed depend on the time lag between discharge and spent fuel reprocessing. Because Pu-241 decays with a half life of 14 years to Am-241, the contribution due to americium increases while that of the plutonium decreases as the time of reprocessing the spent fuel is delayed (although the latter is partly compensated by decay of Cm-242 to Pu-238, which increases the radiotoxicity due to plutonium).

The table below shows the yearly production of actinides in a typical PWR (a French N4 with power rating of 1450 MWe), and in the European Union (EU). The attitude of people towards plutonium varies from "Plutonium Hell No" to "Hello Plutonium". In order words, in some countries like the US plutonium separation is seen as a proliferation problem, while in other countries like Japan and some European countries, plutonium is considered a fuel worthwhile to extract from spent fuel.

PWR (kg/tHM) EU (tons)
Uranium9372500
Plutonium1225
Minor actinides1.33.5

Once plutonium is separated from the spent fuel, it should be recycled as soon as possible to prevent the decay of Pu-241 to Am-241. Recycling plutonium in PWRs is an option considered as a first step towards a full partitioning and transmutation strategy of actinides. If plutonium would be recycled in PWRs with enhanced moderation, which would enable to fully load the reactor core with MOX assemblies, the amount of plutonium is reduced with a factor of three. Although encouraging, the amounts of americium and curium have increased with a factor of four, leading to a marginal reduction of the radiotoxicity of the spent fuel (factor of two). If one would recycle the plutonium afterwards in fast reactors, the radiotoxicity reduces with a factor of five compared to the once-through scenario where PWRs are fueled with UO2 fuel which is disposed of after irradiation. Due to plutonium recycling, more americium and curium is produced that would have to be recycled as well, if one would like to reach higher reduction factors.

A fast neutron spectrum has several advantages for the transmutation of actinides. Not only increases the number of fission neutrons released if the energy of the fission-initiating neutrons increases, also the ratio of fission to capture cross section increases with energy. This means that a fast neutron spectrum to fission actinides produces fewer minor actinides that are formed by neutron capture processes. It turns out that in a fast reactor, one can consume 50 kg of plutonium per TWh of electricity (a consumption of 125 kg per TWhe would be the theoretical maximum).

Furthermore, in fast reactors, the minor actinides can be recycled without extreme impact on fuel cycle operations. It turns out that mixing 2% of americium in the MOX fuel of a fast reactor increases the neutron emission and decay power with 40 and 70%, respectively. Extra shielding is required because of the 76 times higher gamma emission rate (Am-241 emits gamma rays wile the plutonium isotopes do not). For the reprocessing of the spent fuel after irradiation, measures should be taken because of the 4 times higher neutron emission rate and decay power. However, because fuel manufacturing and reprocessing need to be done robotized anyway, these numbers seem no obstacle for the recycling of americium in fast reactors.

For a reprocessing price of 1000 USD/kgHM, reprocessing spent fuel and extracting plutonium for use in LWRs is estimated to be cost-effective only if the uranium price would be 350 USD/kg at least. The historical price of uranium never reached such high values and the current price is only 50 USD/kg. Therefore, from the economical point of view, there seems no need to start reprocessing spent fuel. Future shortage of uranium is not foreseen either, because an increase from 50 USD/kg to 350, would increase the number of uranium ores that can economically be exploited. On the other hand, because the costs of uranium make up only a few percent of the price of electricity generated with nuclear, the increase of the electricity price is modest: only 0.1-0.2 cents per kWhr of electricity generated. Environmental arguments could be stronger then the economical ones.

Full implementation of a partitioning and transmutation strategy for actinides reduces the estimated volume of a geological repository with a factor of 10 to 20, while the peak dose rate would reduce with a factor of 100. The price of geological storage of nuclear waste would have to increase by a factor of 10 to a level of 3000 USD/kgHM to make it all cost effective.

Different View on Nuclear Energy

The use of uranium and the production of nuclear waste can all be seen from a slightly different viewpoint. In each supernova explosion, billions of tons of dust and rocks are spewed out in the interstellar space. Occasionally, these rocks will be captured by stars and form planets. Our planet Earth can thus be considered as a waste product from a stellar fusion reaction. In the beginning, the earth contained many radioactive nuclides among which plutonium and minor actinides. Although most of these radioactive nuclides have all decayed already for a long time ago, there is still some radioactive potassium, thorium and uranium present. In future, also these nuclides will decay to stable nuclides and the energy release, in total about 40 MeV per nuclide, will contribute to the geo-thermal energy reservoir of our earth.

If we fission uranium, we produce fragments that are radioactive only for a few hundreds of years, which is negligible compared with the billions of years it will take to let all uranium and thorium decay. Furthermore, we have reduced the total energy release considerably. The original uranium nuclide emits 40 MeV in total; after one year the fission fragments emit only 0.2 MeV on average. This means we have reduced the radioactive energy release due to uranium with a factor of 200. It should be realized that already after one year most fission fragments (3 out of 4) are stable nuclides.

As written above, if we let the uranium in the earth, it will contribute 40 MeV per atom to the geo-thermal energy reservoir of our earth. This geo-thermal energy would be the maximum that we can extract from the radioactive decay of uranium. If we use the energy that is released by fission of uranium (200 MeV per fission), we increase that amount with a factor of five. In practice this number will even be much higher because we can fully extract the energy from the fission process (200 MeV), while the energy due to decay of uranium (40 MeV) will largely be "spoiled" by nature. Extracting uranium from the ground to fission it in nuclear reactor seems a very effective and environmental friendly way to extract and amplify the potential geo-thermal energy. Seen in this way, we could rename our nuclear power plants to "Geo-thermal Energy Amplifiers".

The slides can be downloaded here