As we have just seen throughout this book, industrial nuclear accidents can occur at any stage of the nuclear fuel cycle. They are particularly dangerous in nuclear reactors, as several serious accidents have shown. No nuclear industry is spared, nor is any country. Similarly, radiological, industrial and medical accidents do not spare any nation.
For a long time, the nuclear industry, imitating military nuclear power, has sought to hide accidents and minimize their environmental and health consequences. As a result, our knowledge of the Simi Valley, Windscale and many other accidents is very limited. The turning point was taken with the Three Mile Island accident and especially with the Chernobyl disaster. The fact that the USSR no longer existed a few years after the accident, and in particular the appropriation of part of the research in radioecology and radiation protection by academics and independent researchers, explain the spread of more and more varied knowledge.
The studies on the Chernobyl and Fukushima accidents are considerably more numerous than on all other accidents. The bibliographical references in this book bear witness to this. Studies carried out by university teams and independent nuclear research organizations ensure results without suspicion of conflicts of interest.
Both accidents are serious and have major environmental, health and socio-economic consequences. It is always difficult to compare two accidents, especially when one is older, such as Chernobyl, even though serious remedial measures still have to be implemented, and the other is at an even more uncontrolled stage. However, it is possible to make some provisional observations and the two major accidents classified at level 7 of the INES, generated different consequences [CHA 15].
First, they occurred in different circumstances. Chernobyl was caused by a sudden and uncontrolled increase in the nuclear reaction (multiplied by one hundred), resulting in the explosion of the reactor core, the destruction of the building and the fire of the reactor graphite. Fukushima was the result of a loss of power supplies and cooling sources leading to the degradation of nuclear fuel and subsequent core meltdown of three reactors, followed by decompression of containment vessels and hydrogen explosions. The accidents resulted in very significant damage at the two nuclear sites (one reactor destroyed in Chernobyl and four severely damaged reactors in Fukushima).
Unlike Chernobyl, atmospheric releases of low-volatility (e.g. strontium, ruthenium) and refractory (e.g. plutonium, zirconium, cerium) radionuclides have been limited to Fukushima. Follow-ups have detected low concentrations of 89.90Sr, 110mAg, 140La/140Ba, 95Nb and 238Pu in the samples from the most contaminated areas around Fukushima Daiichi, as well as ephemeral traces of 239Np measured only in Iitate on April 11, 2011.
The majority of authors agree that the Chernobyl accident has quantitatively higher environmental releases than the Fukushima accident. For example, Steinhauser et al. [STE 14] estimated that in almost every respect, the consequences of the Chernobyl accident far exceed those of the Fukushima accident. In both accidents, most of the radioactivity released was down to volatile radionuclides (noble gases, iodine, cesium, tellurium). However, the amount of refractory elements (including actinides) emitted during the Chernobyl accident was about four orders of magnitude higher than that of the Fukushima accident. For Chernobyl, a total release of 5,300 PBq (excluding noble gases) was established as the most commonly cited source term. For Fukushima, Steinhauser et al. [STE 14] estimated a total source term of 520 (340–800) PBq. According to Achim et al. [ACH 14], 10% of the radioactivity emitted into the atmosphere at Chernobyl was caused by the release of 131I and 137Cs.
Emissions into the atmosphere from Chernobyl increased in altitude and therefore spread more quickly worldwide. In contrast, those in Fukushima remained more localized over Japan and a large part (80%) dispersed over the Pacific Ocean near the Japanese coast. Estimates of the quantities of activity released to the environment vary depending on the source. For cesium 137, the estimates of the quantities released at the Fukushima plant shortly after the accident were about one-third of those released at Chernobyl. But gradually the quantities released at Fukushima were estimated to be increasing and would eventually be relatively close to Chernobyl. On the contrary, the quantities of plutonium and strontium released are lower in Fukushima than in Chernobyl and the opposite is true for rare gases such as xenon 133.
A “leopard spot” pattern of depositions were found throughout Europe following the Chernobyl accident, while atmospheric deposition only reached up to 250 km from Fukushima.
The evacuation area where radioactivity was about 37 MBq.km−2 has a radius of 30 km, an area of 2,830 km2 and includes 135,000 inhabitants in Chernobyl. The same area has a radius of 20 km with an area of 628 km2 and 85,000 inhabitants in Fukushima [RAM 11]. The IRSN [IRS 11b] estimated that, on the 66th day after the accidents, the surface area of soils contaminated by cesium 134 and 137 to more than 555 kBq.km−2 was respectively 10,700 km2 and 874 km2 in Chernobyl and Fukushima (8.5% in Chernobyl) and affected 270,000 and 70,000 inhabitants respectively (26% in Chernobyl). The major problems in Chernobyl are the contamination of the environment and the health status of local populations. The major problems in Fukushima are the radioactive waters that contaminated the ocean until April 2011 and the population’s suffering [CHA 15].
There is also a significant difference between the two accidents. The Chernobyl accident is a unique accident, having taken place over a short period of time, while the Fukushima accident is multiple involving several reactors and pools and took place over a much longer period of time, still not being definitively completed in fact.
However, for some consequences, the Fukushima accident seems more serious than Chernobyl. This is particularly the case for the volume of radioactive waste. Just remember that in Fukushima, three reactors melted and three coriums have to be processed over several decades. This is not to mention the solid and liquid waste still present on the site or in the immediate environment, nor the uncertain fate of three pools containing very large quantities of uranium rods. In addition, the environmental consequences of the Fukushima accident for wildlife appear to be more severe than those of Chernobyl. Unlike Chernobyl, where the terrestrial environment was particularly affected, in Fukushima, as a result of the meteorological conditions and the position of the plant, it was the marine environment that was heavily contaminated, both at the start of the accident by atmospheric fallout, and in the medium and long term by run-off and underground infiltration from the site where many leaks are still present. At soil and forest level, Pumpanen et al. [PUM 16] compared 137Cs concentrations in conifer needles and soils between 1995 and 2013 at two sites affected by the two Chernobyl and Fukushima accidents, one in southern Finland (Hyytiälä) and the other in Fukushima. The vertical distribution of deposited 137Cs in soils was similar in Hyytiälä and Fukushima with the highest concentrations of this radionuclide in the upper layers of the soil. These concentrations have decreased exponentially over time. However, the behavior of cesium in the soil was different between surface and depth in both 1995 and 2013. This shows that 137Cs is still mobile in the upper layers of the soil 27 years after the Chernobyl accident. On the contrary, this radionuclide is more stable in the aerial parts of trees than in the ground and, despite its mobility and active role in metabolism, 137Cs remains in the trees for decades. This means that there is not much exchange of 137Cs between the tissues of the heartwood and those of the bark of the stem. The fate of 137Cs in the physical environment therefore changes over time. Based on an initial analysis of approximately 3,000 articles, Gralla et al. [GRA 14] identified 121 publications that measured cesium 137 levels in food, forage and wood. They found that the global mapping of the impacts of nuclear accidents on the provision of ecosystem services requires a significant increase in peer reviews, including the evaluation of existing gray literature. Assessments should follow a consistent protocol, providing consistent information on the location of sampling and the identification of ecosystem services. In addition, accident studies are necessarily transdisciplinary [GRA 15].
Mousseau and Møller [MOU 14a] summarized observations of effects on wildlife in Chernobyl and Fukushima. Recent advances in genetic and ecological studies of wildlife populations have demonstrated significant genetic, physiological, developmental and physical effects resulting from exposure to radioactive contaminants. The few genetic studies conducted in Chernobyl generally show high rates of genetic damage and high mutation rates. All the main taxonomic groups studied (birds, bees, butterflies, locusts, dragonflies, spiders, mammals) showed a small population size in the highly radioactive parts of the Chernobyl exclusion zone. In Fukushima, population censuses of birds, butterflies and cicadas suggested that abundances were also negatively affected by exposure to radioactive contaminants, while populations of other groups (dragonflies, grasshoppers, bees, spiders) did not decrease significantly, at least during the first summer following the disaster. To date, there is insufficient information to assess the effects on the life history of the various zoological groups, with the exception of insects and birds. The differences observed between Fukushima and Chernobyl may be down to the fact that the maximum exposure was felt at different seasons (early spring and late winter). In addition, the accumulation of multigenerational mutations in populations is higher in Chernobyl than in Fukushima. There has been considerable variation between taxa in their apparent sensitivity to ionizing radiation and this partly reflects life history, physiology, behavior and evolutionary history.
Ahmed and Aly [AHM 14] estimated and compared the cancer risk resulting from radionuclides released into the atmosphere by the Fukushima and Chernobyl accidents. A screening method using the near-field air dispersion pathway was used. The simulation was performed using 131I, 134Cs, 137Cs and 85Kr. The analysis of the results showed that the cancer risk resulting from the Chernobyl accident is higher than that of Fukushima for different atmospheric stability classes (according to the 1961 Pasquill classification [PAS 61] ranging from extremely unstable to moderately stable) and for various downwind distances. At the same downwind distance, the cancer risk for 131I is higher than for 137Cs, 134Cs and 85Kr. Therefore, it is necessary to take into account the characteristics of atmospheric stability in the development of emergency preparedness and response strategies.
The first consequence of a serious nuclear accident is large-scale soil contamination. The result is an evacuation of populations and an abandonment of agricultural soils and economic activities over a large area. The most tragic disadvantage is that this situation will last for decades and even, for the most radioactively contaminated lands, for several centuries.
In retrospect, now after more than 32 years and 7 years respectively, the major nuclear accidents at the Chernobyl and Fukushima plants provide abundant ecological data. However, ecological studies conducted on fauna and flora chronically exposed to ionizing radiation often yield contradictory conclusions. This is probably caused by the complexity of interspecific relationships within polluted ecosystems, as well as the need for multidisciplinary approaches to interpret observed effects, unbiased but difficult to deploy [IRS 16c].
However, some consequences are obvious, such as the morphological anomalies observed in pine trees in both Chernobyl and Fukushima. The frequency of occurrence of these anomalies should be correlated with the dose absorbed by the trees, but dosimetric estimates are uncertain and do not allow robust dose–response relationships to be established. Any decline in abundance of terrestrial invertebrate populations depends on the species groups studied. Differences in radiosensitivity between species and effects related to chronic exposure over several generations are the main explanations for these differences. A significant number of ecological studies focus on bird communities, including long-term observations in the Chernobyl exclusion zone and in the contaminated territories of Fukushima. The decrease in abundance observed at both sites was correlated with the increase in the ambient dose rate. A dosimetric reconstruction in Fukushima birds showed that the observed effects were consistent with knowledge on the decrease in reproductive capacity related to the increase in absorbed doses [GAR 15]. The abundance of large mammals in the evacuated areas has been the subject of controversy. Some studies observed a decrease, others an increase in abundance. These discrepancies are probably related to observation bias and the fact that individual species have a relatively large range including both highly contaminated areas and possibly others that are much less radioactive.
Møller and Mousseau [MOL 16] looked in the literature for evidence that adaptation to natural radiation or after a nuclear accident exists for some species. They found only a total of 17 suspected cases of adaptation, most based on field experiments with organisms derived only from two or three sampling sites and a single experimental study showing evidence of improved radiation resistance. Finally, these same authors could not find any evidence to support the existence of the phenomenon of low radiation hormesis (a biphasic dose response to an environmental agent characterized by a low-dose stimulation or beneficial effect and a high-dose inhibitory or toxic effect).
In the event of a major accident at a nuclear reactor anywhere in the world, more than 90% of the 137Cs emitted would be transported beyond 50 km and approximately 50% beyond 1,000 km before being deposited. Based on a dispersion model [LEL 12], the risk of contamination by 137Cs particle deposits (> 40kBq.m−2.yr−1) would be greater than 2% per year in the north-eastern United States, Western Europe and Japan. As a result, a considerable number of people would be exposed to contamination, estimated to be greater than 5.106 Bq.m−2.yr−1 in the north-eastern United States, Western Europe, Japan and East Asia. The results are similar for gaseous emissions of iodine 131 [LEL 12]. As already noted in the medical surveillance of survivors of Hiroshima and Nagasaki [AMI 19], there are apparently no specific pathologies related to exposure to ionizing radiation. The effects of radiation are multiple and are responsible for a statistically significant excess of many cancers and pathologies, including those of the cardiovascular system. However, these pathologies may have other causes, which make it very difficult for victims to obtain compensation.
The most serious accidents require the evacuation of a large population. This has had health, sociological and economic consequences that have been largely underestimated in the past. Indeed, evacuated individuals must change their physical and social environment and lifestyle habits. The result is profound psychological imbalance, frequently leading to anxiety, then mental distress and often a drift towards the use of licit and/or illicit drugs. This cost, which is difficult to quantify precisely, is enormous.
The societal impact is not limited to areas directly affected by the nuclear accident. For example, Visschers and Siegrist [VIS 13] conducted two surveys on nuclear power plants for Swissnuclear, one 5 months earlier and another directly after the accident in Fukushima. The longitudinal study was conducted in Switzerland with 790 respondents. Four variables were studied: acceptance, perceived risks, perceived benefits and trust associated with nuclear power plants. These four variables were correlated with each other and determined essentially identical responses before and after the accident. Even after a serious accident, the public can still consider the benefits to be relevant, and trust remains important in determining their perceptions of risks and benefits. However, it should be noted that on May 21, 2017, the Swiss approved by vote a new energy law, which aims to gradually switch from nuclear to renewable energy. Several other countries (Germany, Belgium, Italy) have decided to phase out the use of civil nuclear energy.
The question of the probability of a new nuclear accident is still topical and worrying. Several studies are trying to answer this question. In a sample of five random years, the probability of having a major nuclear accident worldwide (INES level 7) is 14%. The risk of at least one nuclear accident, with an INES level greater than or equal to 4, is 67%. However, these figures are subject to uncertainty because of the vague definition of a nuclear accident [HAD 14]. Taking into account the accidents actually observed to date, the probability of at least one serious accident in the next 10 years is about 63% [RAF 88]. According to calculations by Rose and Sweeting [ROS 16], the overall probability of a core meltdown accident in the next decade, in a world with 443 reactors, is nearly 70%. Owing to statistical uncertainty, however, the probability could range from 28% to 95%. The United States, with 104 reactors, has an approximate 50% probability of experiencing a fusion accident within the next 25 years. After the Fukushima disaster, Rose and Sweeting [ROS 16] analyzed all reactor core meltdown accidents and estimated a failure rate of 1 in 3,704 reactor years. This rate indicates that more than one such accident could occur somewhere in the world in the next decade. According to UNSCEAR [UNS 00a], it should be noted that since 1975, the number of accidents that have occurred, with or without clinical consequences for nuclear workers, has been steadily decreasing (Figure C.1). These various probabilities of nuclear accidents do not fully take into account the efforts required, and currently partially realized, by the regulatory agencies to builders following the Chernobyl and Fukushima disasters.
The cost of a nuclear accident is difficult to estimate and varies greatly depending on whether direct costs or all indirect costs are taken into account. According to Sovacool [SOV 08], nuclear accidents caused 4,100 direct deaths and cost more than US$18 billion (2006 value). However, these expenses only concern the repair of the reactor sensu stricto. The actual costs of a serious nuclear accident are significantly higher. They must take into account environmental costs, such as the loss of ecological services, economic costs, such as those generated by the abandonment of local industries and the loss of agricultural land, for several centuries, health costs related to both physical and mental health, and the costs of building new cities to house and treat displaced populations. All of these costs are spread over several generations and are several hundred billion dollars.
After a severe nuclear accident, large geographical areas are unfit for human activities for several centuries and many generations of human populations are displaced. As the consequences of such an accident are immense, if nuclear energy is to have a future, it is essential and urgent that those in charge of this sector make research efforts to understand nuclear accidents, their triggers and their changes. Efforts must also be made to prevent these accidents. Finally, there is a need to significantly improve the management of the acute crisis, corresponding to the course of the accident, as well as post-accident management.
These research and management efforts will be the subject of the third volume of this series on radioactive risk.