Radioactive Waste Management in Sweden: Decision-Making in a Context of Scientific Controversy

Abstract

In early 2022 the Swedish government took a decision to allow the construction of a repository for spent nuclear fuel at the Forsmark nuclear power plant. This chapter describes the long process that preceded the decision and started in the 1970s. Both the siting of the repository and the case made for its long-term safety has seen controversies. In particular, there has been a science-based debate on the long-term safety of using copper as a material for encapsulation of the spent nuclear fuel. The government’s decision relied on claims from the industry and the nuclear regulator that the Swedish KBS multi-barrier concept could still be safe enough. Whether this view will survive the coming decade until the repository will be ready for operation remains to be seen.

10.1 Introduction

Sweden has a long military and civil nuclear history, with all the challenges for radioactive waste management (RWM) that this envisages. Despite this, Sweden is one of the countries that have made most progress in developing and implementing technical systems for the management and disposal of radioactive waste, with a relatively advanced governance system since the 1980s.

When in January 2022 the Swedish government decided to allow the construction of a repository for spent nuclear fuel (SNF), the country joined only Finland in having appeared to have found both a site and a method for final management of long-lived high-level nuclear waste (HLW). But the decision was controversial, and it was a challenge to take it as there was severe scientific criticism of the use of copper as a disposal canister material. The canister is the most important barrier in the safety case and there were claims by highly renowned scientists that the copper would only last for some hundreds of years instead of the required 100,000 years.

This controversy stretched the Swedish governance system for RWM to its limit. Only by claiming that long-term safety would always be good enough as a clay buffer and the bedrock provide two complementary if uncertain barriers, could the decision be taken. The decision may still be found to be in conflict with the implementation of Swedish environmental legislation, which requires the precautionary principle to be met and that decisions cannot be taken based on insufficient knowledge about long-term safety.

This chapter describes the long process that has led to the decision to allow the construction of a repository for SNF, and the controversies that have arisen and remain. The most important controversy has been the copper canister corrosion issue, which has been central to the discussions of long-term safety of the repository since 2007, as well as to the repository licence review process from 2011 until the January 2022 decision.

Although the final decision was taken by the government, a number of other actors were involved in the process, including regulatory bodies, the environmental court system, the government’s scientific advisory board, the scientific research community, the nuclear waste communities, and the environmental movement. The role of different actors has changed over time, and as the discourse became more technical certain actors, including independent scientists and the environmental movement, became influential. Also of interest is that as the government decision grew closer, the repository issue became politicised, something that had been avoided in Swedish politics since the 1970s.

To facilitate the understanding of the developments in recent years, Sect. 10.2 gives an historic background to Swedish nuclear power and nuclear waste management. This is followed in Sect. 10.3 by a description of the Swedish governance and legal framework for RWM. The governance system has some transparency problems that are discussed.

The main focus of this Swedish case study is covered in Sect. 10.4, with a detailed description of the controversies and the decision-making process for the repository for SNF.

In the final section some concluding observations are presented.

10.2 Historic Background

Sweden became a nuclear country quite early and therefore has a long nuclear history. After the Second World War, the military interest in nuclear weapons started a process where a combined military and civil nuclear program developed in the 1950s. The civil program was originally a cover for the military effort and included uranium mining and a plan for a reprocessing plant. A small heavy-water moderated reactor using natural uranium as fuel to produce heat for district heating and electricity was built underground in Ågesta, a suburb south of Stockholm. The reactor also produced plutonium of nuclear weapons quality. The military project was abandoned by the late 1960s when Sweden joined the Nuclear Non-Proliferation Treaty that entered into force in 1970.

As the military interest decreased and the commercial and economic interests increased, the domestic heavy-water reactor programme was converted into a major light-water reactor programme, and 12 nuclear power reactors became operational between 1972 and 1985 at four nuclear power plants (NPPs)—Barsebäck, Ringhals, Oskarshamn and Forsmark.

With the large and fast expansion of nuclear power, a nuclear debate started in Sweden. The opposition to nuclear power allowed the Center Party to lead an opposition coalition in 1976 to break the 40-year political rule of the Social Democratic Party. The anti-nuclear Center Party tried to prevent the start-up of new reactors by the adoption of new legislation. The “Villkorslagen” (Stipulation Act) was formulated so that the industry had to prove that a safe method and a safe site for disposing of HLW existed, before an operational licence could be given.

In parallel to the political developments on nuclear policy, a growing public anti-nuclear movement wanted a referendum on the future of nuclear power, and this was held in 1980. One clear “No” option was put against two different “Yes, for a while” options. When the “No” option did not receive a majority, the result of the referendum was politically interpreted to mean that nuclear power was to be phased out by 2010.

With time, this interpretation was abandoned, but through the years, and more rapidly in recent years, Sweden has reduced its nuclear capacity from 12 to currently 6 reactors supplying about 30% of the total Swedish electricity production of 166 TWh (Energiföretagen, 2021).

Since the political turmoil and wide public debate on nuclear power and nuclear waste management in the 1970s, from the early 1980s onwards there was almost no political interest or public debate in Sweden on nuclear waste issues. The siting process for the SNF repository caused some local political discussions in the mid 1980s, but no actor on the national political level seemed to find nuclear issues interesting enough to stimulate a national debate about nuclear waste. An underlying political understanding developed that the issue of RWM was not to be politicised. Nuclear waste was seen as an issue that had to be resolved because the waste was being produced and something had to be done with it. Also, it was generally felt that the Swedish system for developing a repository for SNF was working well.

After the long political calm on nuclear waste issues, there was an abrupt change around 2020 when nuclear power again became a divisive political issue in Swedish politics. With an upcoming decision on the SNF repository, nuclear waste issues were drawn into a heated political debate.

10.2.1 Early Work Towards a Repository for Spent Nuclear Fuel

As the nuclear waste issue became of increasing political importance in the 1970s, in 1972 the government started a process to investigate the possibilities for permanent disposal of HLW. The result was the AKA (Använt Kärnbränsle och radioaktivt Avfall) report, published in 1976, included a roadmap for the radioactive waste facilities that were later developed.

In 1973, the nuclear industry created a company called SKBF that was to take responsibility for Swedish RWM and disposal. The company was later renamed Swedish Nuclear Fuel and Waste Management Company (SKB), and is owned by the nuclear utilities. After the publication of the AKA report and as a result of the political developments that led to “Villkorslagen”, in 1976 SKBF started a project called “Kärnbränslesäkerhet (KBS)”: The project published three reports in 1977, 1978 and 1983 presenting different versions of what is still called the KBS method for disposal of radioactive waste. The KBS project and the reports allowed the nuclear industry to show that they had fulfilled the conditions of “Villkorslagen”. But according to “Villkorslagen” the industry also had to show that a site could be found that would be safe when using the KBS method. This was quite problematic, but the situation was finally politically “solved” by stating that theoretically it was likely that such a site existed.

The further development of the KBS method and the siting of a repository for SNF, through to the government decision in early 2022 to allow the construction of a repository, is described in Sect. 10.4.

10.3 Governance and Legal Framework: Two Parallel Tracks for Licensing

An application for a license for a repository for radioactive waste in Sweden is processed in two parallel decision-making judicial processes, according to the Nuclear Activities Act (1983)¹ and the Environmental Code (1998).² There are also clear legal obligations in the decision-making governance system for access to information, access to public participation and consultation, and access to justice.

10.3.1 The Nuclear Activities Act

The 1983 Nuclear Activities Act put all the responsibility for management and disposal of waste onto the nuclear industry. To keep the operational licenses for the nuclear reactors, the industry needs to present so-called “Fud reports” every three years, with the latest translated into English in September 2019 (SKB, 2019). The Fud reports include reporting on and future planning of research development and demonstration (RD&D), and a strategic plan for the future of Swedish RWM and the decommissioning of nuclear facilities.

Each RD&D programme is reviewed by the Swedish Radiation Safety Authority (SSM), which includes a broad societal consultation process. The regulator then makes a report to the government with recommendations. A separate review is provided by an independent scientific advisory board to the government, the Swedish Council for Nuclear Waste. The government then takes a decision on the Fud-report and can give conditions for further RD&D work.

Historically, the Swedish regulator for radiation safety (historically the Swedish Nuclear Power Inspectorate, SKI, and since 2008 SSM) has been relatively positive concerning the SKB work plans as presented in the Fud reports, and has in its review to the government recommended support of the plans. There has been a tendency to refer any perceived problems back to the industry, which has the legal responsibility for resolving them. This means that there has been a risk that problems just “disappear”, as it might not be in the industry´s interest to find or examine problems that can hinder its work or plans.

In summary, the legal Fud programme process has been a relatively weak steering process for industry plans or future RD&D work. With a political disinterest in RWM issues and a generally industry-supporting regulator (whether SKI or SSM), most government decisions on the Fud reports have had no effect on the work of SKB.

10.3.2 The Environmental Code

The Swedish legal framework for decision-making on issues that have an environmental impact was greatly improved in 1998 when the Environmental Code became part of the Swedish judicial system. With the new legislation, a special environmental court system was set up to take licensing decisions on activities that could cause harm to humankind or the environment. The legislation, among other things, regulates the content of the environmental impact statement as well as the consultation process necessary for its development (in Chap. 6). The European legislation on environmental impact assessments and strategic environmental assessments are implemented in the Environmental Code.

A special Chap. 2 of the Environmental Code specifies the criteria for an activity to be allowed. The environmental impact assessment has to include rigorous descriptions of alternative siting and methods. The precautionary principle has to be used, and there is a condition that enough knowledge about the activity and its possible influences on the environment must be had.

Of importance for the parallel decision-making of facilities according to both the Nuclear Activities Act and the Environmental Code is that licensing decisions taken according to the nuclear legislation must follow Chaps. 2 and 6 of the environmental legislation.

Decisions of the Land and Environmental Courts can be appealed to a Land and Environmental Court of Appeal, and the decisions of that court can be appealed to the Supreme Court.

10.3.3 Access to Information, Public Participation, and Justice

Parts of the Environmental Code implement the 1998 Aarhus Convention’s second and third pillars of access to public participation and access to justice (UNECE, 1998). The legislation mandates that the implementer of activities that have an environmental impact, including nuclear activities, must carry out and document a process of public consultation while developing the environmental impact statement for a new or changed activity or facility. Part of the decision-making process is that the Land and Environment Court must approve the consultation process and that issues raised have been properly taken into due account. Environmental organisations are given special importance in the legislation and have the right to appeal all decisions taken. MKG, the Swedish NGO Office for Nuclear Waste Review, is a non-governmental environmental organisation established in 2004 to work specifically with nuclear waste issues.³

To facilitate the participation of civil society in the decision-making for the repository for SNF, the government has provided resources for environmental NGOs. Between 2005 and 2016 it was possible for environmental NGOs to seek funding from the Nuclear Waste Fund. About EUR 300,000 per year were made available, and since 2017 a similar sum is available through the state budget.

Sweden has a long tradition of open access to official documents of the government, going back hundreds of years. With some exceptions for commercial or security secrecy, and regarding interaction with foreign governments, all documents and even e-mails and messages that concern official business must be registered and made available upon request. Also, the contents of important phone calls have to be noted and registered. This means that the activities of the government, of the nuclear safety regulator and local communities can be followed publicly, together with documents and other information.

However, the nuclear waste company SKB is a private company, which is outside the remit of the legislation and therefore not obligated to disclose any information. This means that SKB, which has the sole legal responsibility for research and development on RWM and repository technology, can keep its work secret; SKB can hide any problems because the documentation or research results never have to be disclosed. In practice, this means that the company publishes only results that support its safety case.

The availability of public information on governmental activities, coupled with lack of access to information from the implementer of RWM, means that the Aarhus Convention’s first pillar of access to information is relatively weakly implemented on RWM issues in Sweden.

10.3.4 The Two Parallel Tracks for Licensing: A Complex Process

Before a licence application is prepared, the nuclear waste company SKB must carry out a consultation process to prepare the environmental impact assessment document. This document is necessary for both applications, but the application to the Land and Environmental Court does not need to go into the same detail on radiation safety issues. After the two applications are submitted by SKB, the nuclear regulator SSM carries out a review according to the Nuclear Activities Act. The Land and Environment Court (hereafter the Environmental Court), carries out a review according to the Environmental Code. The main focus of the SSM review is the safety analysis document in the application, while the Environmental Court reviews general environmental issues but can decide to what extent radiation safety issues may be of relevance. This parallel track for licensing continues throughout the decision-making process and causes complications. The process is shown in Fig. 10.1, and also described in Bjällås & Persson (2011).

Fig. 10.1

(Source: MKG)

Process for licensing under the Environmental Code and the Nuclear Activities Act.

Both reviews start with an analysis of whether the application is sufficiently complete. Both SSM and the Environmental Court carry out a broad external review process. SSM can provide opinions in the Environmental Court review. The Environmental Court process is more open than the SSM process, and SKB is asked to comment on opinions, and in turn interested parties can comment on the opinions of SKB. This can take several iterations and SKB can introduce complementary material at will.

After the applications are considered complete enough in both reviews, they are formally announced by SSM and the Environmental Court. This means that the review on issues formally starts. Again, both SSM and the Environmental Court carry out broad external reviews. SSM continues the regulatory review according to the Nuclear Activities Act on its own after receiving the external comments on issues.

In a more open process, where anyone interested can take part in the review on issues, the Environmental Court passes opinions back and forth between SKB and other interested parties, in the same way as in the earlier review on completeness.⁴ In this part of the process, SSM delivers its opinion on the application according to the Environmental Code to the Environmental Court. This is an important step as it also gives the first indication of what the final verdict of the regulator will be on radiation safety issues and the safety case.

When the Environmental Court finds that enough information is available to proceed to the next step it announces the main meeting of the Court. In this public oral meeting anyone can take part, and SSM is an important participant. The main meeting ends the Environmental Court’s review.

The final decision to allow and license facilities involving radioactive waste is taken by the Swedish government. The next step of the licensing process is therefore that SSM and the Environmental Court give their opinions on the application to the government. The government reviews the opinions. According to the Environmental Code, in the review the government can take comments on the opinion of the Environmental Court.⁵ The government also has to get a positive decision from the local community that has can veto the repository, even this late in the process. The government finally decides whether to give permissibility to the repository according to the Environmental Code, and a licence permit according to the Nuclear Activities Act. The government can add conditions to its decisions.

The government decisions can be appealed to the constitutional Supreme Administrative Court, that only considers whether the law has been followed, i.e., it restricts its decisions to issues that have a clear legal connotation.

After a government decision to give a license according to the Nuclear Activities Act, the regulator continues a stepwise decision-making process of examining revised safety cases to allow construction, pilot operation and full operation of the repository. After a government permissibility decision according to the Environmental Code, it is the Environment Court that gives the final license. This is also the Environmental Court’s last opportunity to influence the repository project. The Environmental Court will again make a review and ask for opinions, but the review in this part of the process is only of conditions that can be put on the license.

The Environmental Court finally makes a license decision that can be appealed to two higher court levels, based on details in the conditions, but the government decision forms the basis for a license and binds the courts, which must follow the government’s permissibility decision.

When SKB has both final licenses for a repository, construction can begin. The present Swedish legal system therefore means that nuclear projects have to have licenses according to both the nuclear and environmental legislation. There can be conflicts between these processes, even though experience and praxis has been developed to allow coordination.

10.4 The Long Decision-Making Process for a Repository for Spent Nuclear Fuel

As described in the Sect. 10.3, the Swedish judicial process for allowing construction of a repository for SNF is complicated and time-consuming. It has taken from 2011 to 2022 just to process SKB’s license application. But the start of the project was actually begun in the 1970s, i.e., 50 years ago.

When the KBS concept for a repository for disposal of SNF was developed in Sweden in the mid-1970s by the nuclear waste company SKB, there were three main project reports, in 1977 (SKBF, 1977), 1978 (SKBF, 1978), and 1983 (SKBF, 1983).

The KBS-1 report (SKBF, 1977) was focused on a repository for reprocessed HLW, as at this time reprocessing was the plan for Swedish SNF. The vitrified glass waste was to be encapsulated in a steel container 3 mm thick and then placed in a titanium canister 6 mm thick. Between the steel and the titanium would be 10 cm of lead. The canister was to be surrounded by a buffer of sand and clay. The disposal depth was to be about 500 m. From this concept only the depth of 500 m survived into the final KBS concept used today.

The KBS-2 report (SKBF, 1978) came only a year later when it became unclear whether reprocessing was to be used in Sweden for all the SNF, and provided a similar concept for a repository for direct disposal of unreprocessed SNF. The spent fuel rods were to be taken out of the fuel elements and placed in a copper canister of 20 cm thickness and the whole canister was then to be filled with lead. The canister was to be surrounded by a buffer of blocks of clay. The depth of the repository was to be about 500 m. From this concept the choice of copper as a canister material and the clay buffer survived into the final KBS concept.

The system was further developed and optimised, and in 1983 the KBS-3 concept was presented in a third report (SKBF, 1983), with a 10 cm thick copper canister containing complete fuel elements instead of separated fuel rods, but still with the canister filled with lead. With the final KBS project report, the concept was almost finalised, with copper canisters containing the SNF to be deposited in holes in the floor of tunnels about 500 m underground in granite bedrock (see Fig. 10.2). All tunnels and shafts are also to be filled with clay.

Fig. 10.2

(Source: MKG)

The KBS method.

By the mid-1990s the cylindrical canisters were finalised to be 5 m high and 1 m in diameter and made from only 5 cm thick copper. Inside the copper canister is a cast iron insert (instead of lead) to hold the spent fuel elements in place and provide higher strength to the encapsulation (SKB, 1999).

There is always flowing groundwater in the granite bedrock, and even though the copper canister is supposed to be relatively immune to corrosion, it is surrounded by a clay called “bentonite clay” that will swell when subjected to water. The swollen clay buffer is to provide tight protection of the copper canister from the groundwater. The deposition tunnels and other parts of the repository system will also be filled with bentonite clay so that the whole bedrock system is to become as tight as the bedrock itself to the flow of water.

The long-term safety case for the KBS-3 concept thus relies on two artificial engineered barriers—a copper canister, a bentonite clay buffer—and a semi-natural barrier of the bedrock with tunnels filled with clay. In practice, the tunnels in the bedrock are disregarded in the safety case as it is assumed that no water can flow through them. In the safety case analyses, a rock that has few cracks and fissures therefore becomes a better barrier than a rock with many.

Copper was finally chosen as canister material instead of titanium because theoretically copper is thermodynamically immune to corrosion by water. The immunity of copper to corrosion by water later became an important scientific controversy that became a part of the license review for the repository for SNF.

The reason for originally decreasing the copper thickness from 20 to 5 cm was mainly improved safety analysis modelling, that allowed less and less importance to be put on the copper canister as a barrier. Still, from the 1970s to the early 1990s the bedrock was assumed to be an important barrier in the safety case, meaning that it was important to find the best Swedish bedrock for a repository. This led to a complicated siting process. However, when the siting process made it necessary to be able to site a repository in almost any rock formation in Sweden, the importance of the rock barrier of the bedrock was toned down in the safety case. More emphasis was put on the function of the artificial copper and clay barriers. More information on the siting process is given in Sect. 10.4.1.

10.4.1 Siting of a Spent Fuel Repository—A Long Road to Acceptance

Originally, the bedrock was seen as the most important barrier in the KBS concept. The siting process for a repository for SNF was started in the mid-1970s by the nuclear waste company SKB, with exploratory drilling all over Sweden to find the best bedrock. The arrival of SKB was often met with local resistance, where citizen groups were spontaneously created to oppose the drilling. The opposition to SKB became better organised with time, and a network of local opposition groups called Avfallskedjan (The Waste Network) was formed in order to provide common support. There were almost 20 local groups, and several of them survived long after there was any threat to the local rock becoming a repository.

In 1985, the resistance led to the stop of the siting process. At a demonstration at Almunge, near the city of Uppsala north of Stockholm, police with dogs broke up a demonstration against drilling in a local rock formation. The closeness to the major cities allowed the conflict to become major TV news, and the sight of elderly people being dragged by dogs caused concern, not in the least politically. The Minister of Environment declared that the SKB strategy of selecting a site in conflict with the local community had to stop.

This meant that SKB had to restart the siting work, and a voluntary process was initiated, with all the communities in Sweden contacted to see if they were interested. Two communities in the north, Storuman and Malå saw that a repository could provide jobs and volunteered. In both communities there was political support for the siting projects, but when local referenda were held in 1995 and 1997, the inhabitants of the communities said “No”.⁶ The remaining conflict within the communities after the referenda took time to heal, and these examples did not encourage other communities to follow.

SKB then moved the search for a site for the repository to communities that already had nuclear facilities or neighbouring communities. By the late 1990s the siting process was focused on two nuclear communities, Oskarshamn and Östhammar, where the Oskarshamn and Forsmark NPPs are situated. At the Oskarshamn NPP there was also already the central intermediate storage site for SNF, Clab, and at the Forsmark NPP there was already the SFR repository for short-lived radioactive waste. At the Oskarshamn NPP, SKB had in 1995 also established an underground hard rock laboratory down to 500 m depth underneath the island of Äspö, where large-scale experiments on copper and clay and on groundwater flows could be carried out.⁷

Both communities were accustomed to nuclear power, nuclear waste facilities and SKB, which undertook large information campaigns to build up trust. This strategy has worked, and a large majority of the population in each community has a positive view of the company. Since 2003, SKB takes yearly polls to measure the popularity of the planned radioactive waste facilities in the communities, and as an example, the support in Östhammar community for a repository for SNF has been between 75 and 84%.

SKB decided to make detailed site investigations between 2002 and 2008 in areas adjacent to each of the NPPs. As discussed above, emphasis in the safety case had shifted in the 1990s from the importance of a tight bedrock to the ability of the artificial barriers of the copper canister and the bentonite buffer to contain the radioactivity in the long-term. This meant that the nuclear waste company could suggest that almost any bedrock could be used for a KBS repository, as the safety case models could show that the artificial barriers of the copper canister and the clay buffer would provide safety for the hundreds of thousands of years needed.

In 2009, a site just south of the Forsmark NPP in Östhammar community was chosen for a repository in favour of an Oskarshamn site. Interestingly, the reason was that the Forsmark bedrock had fewer cracks than in Oskarshamn, so the safety analysis showed that Forsmark was a little safer. The site for an encapsulation plant (Clink) was to be co-localised with the existing SNF storage facility, Clab, at the Oskarshamn NPP site. The copper canisters were to be transported by ship from Oskarshamn to Forsmark.

Before the decision was announced, most assumptions were that SKB would choose a site just south-west of the Oskarshamn NPP for the repository. This would put it near the encapsulation plant, and the Oskarshamn community appeared to be more politically and administratively well-organised to support a repository decision. But SKB appeared to realise that having a better rock could be more important.

The change in focus over time regarding the importance of the rock barrier in relation to the artificial engineered copper and clay buffer barriers is very important. As is worth repeating, the KBS concept is supposed to rely on three independent barriers for long-term safety. In reality, however, the barriers are not independent of one other, and each may have its problems. The question of to what extent the robustness of the whole system, in a holistic approach, allows the weaknesses of individual barriers to be ignored was to finally become the central issue in the licensing review for the repository for SNF. The long-term-integrity of the copper canister became increasingly questioned throughout the review.

10.4.2 The Copper Corrosion Controversy

Before describing the review of the licence application for the SNF repository, it is vital to understand more about the copper corrosion controversy that has been so important in the decision-making process. The basis for using copper as a canister material is that theoretically, according to classic thermodynamics, copper is as immune as gold to corrosion in a repository environment. The reason for this is that there is no dissolved oxygen in the groundwater at repository depth that can corrode copper. As a comparison, oxygen in the air corrodes copper roofs, turning them green, but without oxygen in the repository, or in the water in the repository, the copper surface will theoretically not be affected and only minor corrosion from other processes will take place (King et al., 2010). In addition, any oxygen in the repository during deposition of the copper canisters will be consumed after the sealing of the holes by bacteria and very fast chemical reactions, with the process taking only a few months.

The problem is that the scientific hypothesis that water without dissolved oxygen (anoxic water) does not corrode copper may be false. This was discovered experimentally by a researcher at the Royal Institute of Technology (KTH) in the mid-1980s (Hultquist, 1986), but the results were ignored by SKB and the Swedish regulator at that time. As the repository consultation process started, the researcher together with colleagues published new experimental results in 2007 and then onwards with new studies (e.g., Szakálos et al., 2007; Hultquist et al., 2009; Hultquist et al., 2013; Hultquist et al., 2015). SKB has strongly contested the research, but the results have been repeated by other researchers (e.g., Becker & Hermansson, 2011; Cleveland et al., 2014). It is now also theoretically understood that water molecules can corrode a copper surface (e.g., Macdonald & Sharifi-Asl, 2011).

As a result of the controversy, there is now an ongoing scientific paradigm shift to the fact that water can directly corrode copper even when there is no oxygen. This raises the question of how fast the reaction can take place at the temperatures and in the complicated water chemistry of the repository. The researchers from KTH claim that some copper canisters can start to collapse after only a few hundred years in the hot and chemically complex repository environment. The issue is not only how fast a general corrosive process occurs but also if pitting corrosion is possible, whereby corrosion can continue relatively fast where it has commenced; much like in a car where rust tends to corrode right through steel once it starts to occur ar a certain spot. Of importance is also that when copper reacts with a water molecule there is a release of hydrogen that can build up inside the copper and make it brittle. Also, the repository gets hot, close to 100 °C, and corrosion proceeds much faster at higher temperatures.

SKB has strongly questioned all criticism and points to its own research, which maintains that of the approximately 6000 copper canisters to be put in the Swedish repository, less than one will start to leak in a million years.

It may appear strange that the scientific controversy of how copper behaves in a repository environment has not been resolved long ago. To the initiated observer it is clear that one problem is that SKB and its sister organisations in Finland and Canada have never carried out experiments in the laboratory or in hard rock laboratories with the purpose of scientifically examining this issue in detail. And when there have been experiments that could give important information, they have not been used to the full extent. An example of this is the LOT experiment operated by SKB in the Äspö hard rock laboratory discussed in Sect. 10.4.5. Another way that SKB has tried to discredit independent research showing copper problems is to claim that the experimenters have been careless, and that oxygen has leaked into the apparatus, even though this has never been shown to be the case.

Sometimes the implementers lose control of the science. At the end of 2017, after the main meeting of the Environmental Court which is discussed in the next Sect. 10.4.3, the results of 18 years of copper corrosion in an oxygen-free repository environment were published, from the international collaboration experiment Full-scale Engineered Barrier EXperiment (FEBEX) in the Grimsel hard rock laboratory in Switzerland. In the experiment, clay and metals were tested in a tunnel that was heated to simulate disposal of HLW. The report showed considerable and unexpected copper corrosion, also pitting corrosion, which is very serious as it means a larger risk of a hole being created through a copper surface (Wersin & Kober, 2017). SKB has claimed that the corrosion must be due to oxygen leaking into the experiment, but this has not been shown.

10.4.3 Consultation, License Application and Review

From 2002 until 2010, SKB carried out a lengthy and thorough consultation process for developing the environmental impact statement for the SNF repository and the encapsulation plant. In March 2011, SKB submitted a license application package according to the Nuclear Activities Act and the Environmental Code (Sect. 10.3.4 and Fig. 10.1) for a SNF repository system using the KBS method at the Forsmark NPP, and an encapsulation plant at the Oskarshamn NPP.

The application review was started by both the regulator, the Swedish Radiation Safety Authority (SSM) according to the nuclear legislation, and the Land and Environmental Court according to the environmental legislation. The initial review for completeness of the application took some time and SKB had to add extra documentation several times. This phase of the review was finished in 2015, and the application was announced as complete by both SSM and the Environmental Court in January 2016.

During 2016 and 2017 the application was reviewed on issues. Many issues were raised, including problems with the site chosen, better alternatives for siting and choice of method (i.e., very deep boreholes), issues concerning the safety case (canister integrity, clay erosion, hydrogeology and seismology), as well as problems with intentional intrusion scenarios and challenges of transferring information about the repository into the future.⁸

Perhaps most important during the license review process was the issue of possible problems with the long-term integrity of the copper canister. This issue was raised by some actors, including researchers at the Royal Institute of Technology (KTH) in Stockholm and environmental NGOs. The copper corrosion controversy goes back to the 1980s, but became very lively from 2007 onwards with the publication of new studies by researchers at KTH as described in Sect. 10.4.2. The controversy grew to be more and more important during the review process all the way to the end.

An important event in the license review was in June 2016 when SSM presented the nuclear regulator’s opinion on the radiation safety issues to the Environmental Court. The regulator had made a large number of demands for additional information from SKB during the review for completeness of the application. SSM had also carried out its own research on the copper corrosion controversy (e.g., Becker & Hermansson, 2011; Macdonald & Sharifi-Asl, 2011; Hultquist et al., 2013). But during the spring of 2016, SSM had to make the first open decision on long-term safety. In a statement to the Environmental Court it declared was that there was a potential that the repository would be safe enough to fulfil the licensing criteria of the Environmental Code (SSM, 2016). This decision bound the regulator to saying “Yes” also later in the process.

In the autumn of 2017, the main meeting of the Environmental Court was held as the final part of the review process. The nuclear safety regulator SSM told the Court that some remaining issues, i.e., the copper corrosion issue, could be dealt with after a government decision, but that the regulator’s position was still that there was potential for the repository to be safe enough to fulfil the licensing criteria of the Environmental Code. The Environmental Court questioned SSM’s position in the meeting. According to both the Environmental Code and the Nuclear Activities Act, the repository had to be shown to be safe before a government decision. The question was: Did SSM mean that there was a potential that the repository could be safe, or that it had been shown that the repository would be safe?

At the main Environmental Court hearing, leading corrosion scientists from KTH again strongly questioned the SKB position on copper corrosion and the long-term integrity of the copper canister. The Environmental Court even arranged an extra day for the discussion of copper canister issues.

During the hearing that took in total four weeks it was disclosed through leaks of internal SSM documents to media that the regulator had internal conflicts on the copper canister issue in the spring of 2016, just before the regulator first told the Environmental Court that the repository would likely be safe (Lundell, 2017a; Lundell, 2017b). The documents showed that the leading SSM copper corrosion expert had been opposed to the regulator saying “Yes” to the Environmental Court, as it was not shown that copper would be a sufficiently good canister material. Documents also showed that there were SSM scenarios on copper corrosion processes that showed regulatory limits for radioactive releases from the repository could be exceeded.

On January 23, 2018, the Environmental Court made its recommendation,⁹ that the government say “No” to the application, primarily due to the uncertainties regarding the long-term safety of the planned repository due to possible copper canister problems. These issues would have to be resolved before a government decision could be taken.

On the same date, SSM told the government that it could agree to the repository, as some issues, i.e., possible problems with the long-term integrity of the copper canister, could be dealt with later, after a government decision. The regulator also believed that the repository can be safe enough even if the copper canister barrier does not work exactly as postulated as there are other barriers (bentonite clay buffer and bedrock). This regulatory focus on a holistic approach to the robustness of the safety case is important to understand what happened next in the continued government review.

10.4.4 The Government Review of Copper Corrosion Issues

After the January 2018 statements of the Environmental Court and SSM, the government review started.¹⁰ SKB made a submission of complementary information on copper corrosion in April 2019, stating that the Environmental Court had not understood the copper corrosion issues. Very little new information was provided.

The government sent out the complementary information for consultation, and comments from other parties were provided to the government in late 2019. In its answer, SSM stated that its conviction that the repository would be safe enough was said to have been “strengthened” by the new information, despite information in the regulatory review saying there could be problems with the copper canister (SSM, 2019).

At this time the Swedish Council for Nuclear Waste, the government’s scientific advisory body, entered the decision-making process, and said that there may be problems with the copper canister, and also with the cast iron insert, which may show that the KBS concept will not work as intended.

The researchers at KTH persevered in their criticism of using copper as a canister material. They were joined by the SSM corrosion expert who had opposed the regulator saying “Yes” to the court in 2016, who had subsequently left SSM. MKG also stated that it would be wrong to use copper as a canister material.

10.4.5 An Unexpected Development: New Experimental Packages from the LOT Project Retrieved with 20 years of Copper Corrosion

In the middle of the government review an unexpected possibility occurred that had the potential to clarify the issue of whether copper was a good enough canister material for the Swedish SNF repository.

The so-called LOT (Long-term test of buffer material) experiment operated by SKB has been ongoing at 400 m depth in the Äspö Hard Rock Laboratory near the Oskarshamn NPP since around the year 2000. In total there were seven experimental packages with copper and clay in a very good simulation of real repository conditions. Three 1-year packages were retrieved early, but when SKB retrieved one 5-year package in 2006 an unexpected amount of copper corrosion had occurred. However, the reporting on the corrosion that had taken place in these packages was very limited. There were only estimated general corrosion data and no cross-sectional metallographic studies that are necessary in order to understand how much corrosion has taken place and of what type (Karnland et al., 2009; Wersin, 2013).

As results about copper corrosion from the LOT experiment could be important for the understanding and possible resolution of the copper corrosion controversy, MKG had repeatedly demanded that the next package be retrieved and analysed. These demands were not met by SKB, and the regulator SSM has shown little interest in the copper corrosion results from experiment.

Then came the surprise. In the early autumn of 2019, SKB secretly retrieved two 20-year-old experimental packages. This was disclosed by SKB at a meeting organised by the regulator SSM in early October 2019. The disclosure was likely unintentional, and while SKB started to provide more information about the retrieval, the company originally stated that it had no intention of revealing results from the retrieved packages until after the government had taken a decision on the repository.

During the autumn of 2019, MKG worked to get SKB to disclose all relevant corrosion results as soon as possible, so that SSM could check the results by carrying out a scientific quality review. MKG kept the government informed of the developments concerning the LOT packages. Its recommendation was that the government should wait for the corrosion results before taking a decision on the repository results, as if the corrosion was as bad the FEBEX experiment, something could be very wrong with using copper as a canister material.

These efforts were successful, and SKB published a report on the copper corrosion results in October 2020 (Johansson et al., 2020). Although there is a large amount of information on copper corrosion in the report, it was clear that the reporting was scientifically lacking in quality. Even though it was clear from pictures in the report that some surfaces were very corroded, these were not examined. The most heated surfaces where the corrosion was expected to be the largest were not examined in detail, and SKB claimed that other surfaces were examined instead as it was easier to do so. The corrosion on the most heated surfaces was indirectly estimated by measuring the corrosion products in the clay adjacent to the surfaces; a method that gives only a rough estimate. Cross-sectional metallographic studies in the report from less heated areas show signs of pitting corrosion. There is a very high likelihood that such images of the corrosion of the most heated areas would have shown deep pitting corrosion, and that SKB would have had big problems explaining how this had occurred.

During the autumn of 2020, SSM carried out a quality assurance project with the support of the UK consultancy Galson Sciences, and published a report in early 2021 (Hicks et al., 2021). MKG sent four inputs to the regulator to support the review.¹¹

Using the report, SSM made a statement on the LOT results to the government in March 2021. Unfortunately, SSM accepted the SKB reporting of results without much analysis of its own. The researchers at KTH pointed out to the government that if the reporting of the LOT results lacked scientific quality, then “the anoxic copper corrosion rate in Swedish groundwater is catastrophic with respect to the KBS-3 model and this conclusion can be made without further considering the radiation induced corrosion (radiolysis), stress corrosion cracking and hydrogen embrittlement”.¹²

10.4.6 The Government Decision

Despite the copper corrosion controversy Östhammar community in October 2020 told the government that it approved the siting of the repository and would not use its veto. All formal conditions were fulfilled for a government decision. But since 2018 the government had been a Social Democrat and Green Party minority coalition, with a Minister of the Environment from the Green Party who was reluctant to approve any SNF repository.

In the spring of 2021 the political discourse on nuclear power and nuclear waste management became increasingly politicised. Pressure from the pro-nuclear political parties in opposition increased throughout the year. All the parties in the conservative block (Liberal Party, Conservative Party, Christian Democrats and Sweden Democrats) that has been forming in Swedish politics for a few years have nuclear power as central to their climate policies.

The copper corrosion controversy became more intense in the autumn of 2021. MKG told the government that more scientifically produced results from the LOT experiment should be used to provide vital copper corrosion information necessary before a decision. The Swedish Council for Nuclear Waste stated that more research is needed to understand how the copper canister behaves in a repository environment, but that this could be done after a government decision. The Council also suggested that according to the Environmental Code, the government would be able to give only a construction license, and an operational license could be given separately at a later time. Such a legal construction would, however, be unprecedented and lacked any previous praxis. The Environmental Code states that there should be enough knowledge about the possible environmental effects of a project before a licence is given.

Because of the controversy that had put into doubt whether the copper canister would work as intended in the repository, towards the end of the license review SKB and SSM were forced to emphasise the importance of seeing the KBS concept in a holistic perspective. Both organisations were convinced that there were no copper problems which could be important enough to question the long term-safety of the repository. In their common opinion the safety case with the three barriers (copper canister, clay buffer and the rock/clay in tunnels) that work together proves the repository will be robustly safe. In fact, there are scenarios in the SKB safety analysis that show that even if there are small holes in many of the canisters from the beginning, the regulatory limits will only be exceeded in the very long term if the clay buffer works as intended and the rock of the tunnels is very tight to water flow.

In November and December 2021 there was political parliamentary turmoil regarding the government budget process for 2022. This resulted in the Green Party leaving the government. The new Social Democrat Minister of the Environment almost immediately promised a decision on the SNF repository in January 2022.

On January 27, 2022, the Social Democrat government took the decision to approve the repository. The decision relied almost exclusively on statements from SSM based on the view that in a holistic perspective the three barriers together provide sufficient long-term safety.

MKG and member environmental organisations have appealed the government decision to the Swedish Highest Administrative Court (constitutional court) for judicial review (Naturskyddsföreningen et al., 2022). The organisations primarily want the court to review whether the government has followed the preconditions for a decision according to the Environmental Code (i.e., that the repository has been shown to be safe at the time of the license and that the precautionary principle has been followed) when the government decision so strongly relies only on support from the regulator.

In the appeal, the organisations also question the holistic industry and regulatory approach that the robustness of the whole KBS barrier system means that individual barriers need not function as intended. Their case is that there is strong evidence that the copper canister will not work as foreseen due to corrosion and other processes. If this is so, then the clay buffer will likely be destroyed by the copper corrosion products and will not be tight to water. The clay buffer barrier is not independent of copper canister problems. If the final barrier of the rock/clay in tunnels does not work, there is then a big problem.

In fact, it is not the rock itself that is the weakest part of this final barrier, but rather the clay in the tunnels. If the clay does not completely stop the flow of water as intended, and there is relatively little research to show that it will, then the “rock barrier” is not as strong as envisioned in the safety case. According to the organisations, it is too simple to give each of the three barriers a high protective function in the safety case and claim that as long as one barrier is intact everything is fine. An important issue for the constitutional court to ponder is how big can problems with individual barrier problems be, compared to the “robustness” of the total barrier system, for an environmental project to receive approval.

If the appeal is unsuccessful, the next step in the decision-making process will be the return of the case to the Land and Environment Court for a final licence decision and conditions. This process will not start until mid-2023 and will take several years, with the possibility to appeal any decisions taken.

If everything goes as SKB hopes, there would be a construction start in perhaps five years, and the first copper canister containing spent nuclear fuel would be placed in the repository in fifteen years’ time; unless the science of copper corrosion and/or a better understanding of the weaknesses of clay barriers come back to haunt the decisions taken so far.

10.5 Some Final Observations

There may be some lessons to learn from the Swedish system for governance of RWM, especially from the controversial case of the repository for SNF.

Firstly, the Swedish SNF repository project is a clear example of when over time a project becomes too big to fail. With complete regulatory support from the industry, at times confirmed in weak governmental oversight, there were never any alternatives other than to move forward with the project.

Secondly, it is clearly a problem to have all the responsibility for research and development for radioactive waste facilities attached to a private industrial entity that is not part of the national public access to information system. If the work of private nuclear waste company SKB had been open to public scrutiny it is likely that the copper corrosion controversy would have been resolved much earlier. It would also have been better to have more resources, and a better mandate for the regulator to carry out its own independent research. And to have more general public funding for research on RWM.

Thirdly, the science behind the SNF repository safety case seems to have gone into a mode of confirmation bias early on. It is of course possible that SKB actively avoided doing the necessary research to prove that the copper canister would work as intended. But what has happened can perhaps be explained instead by a tunnel vision based on early basic assumptions that were just thought not needed to be tested, even when there were clear signs that the assumptions might be wrong.

Fourthly, the quality of the decision-making process was improved considerably by the implementation of good consultation processes and systems for access to justice. Even though the final outcome of the decision-making for the SNF repository may have been disappointing to some parties, it is hard to argue that not all voices were heard or were able to influence the process. The quality of the decision-making process was further improved by resourcing local communities and environmental NGOs to be able to participate fully in the process.

Finally, it is of vital importance that all problems that arise during the development and decision-making process for a repository for high-level radioactive waste are fully examined as far as possible. And even more importantly, as early as possible. The challenge is to find a governance system that both encourages and enables this.