Canada’s involvement in the Manhattan Project in the 1940s and its subsequent renunciation of the bomb foreshadowed early on that Canada’s nuclear identity would include inconsistencies. [1] Over the years, these inconsistencies have contributed to apparent contradictions in Canadian actions with respect to nuclear technology. Canada’s nuclear paradox refers to the tension between Canada’s reputation as a champion of nonproliferation on the one hand and factors that undermine its nonproliferation credentials on the other.

Canada has earned its reputation as a champion of nonproliferation. Since the start of the atomic era, Canada has been an active participant in, and often the driving force of, a long list of nonproliferation achievements. This involvement includes membership in and strong support for the International Atomic Energy Agency (IAEA) and the Nuclear Nonproliferation Treaty (NPT); a fully safeguarded domestic nuclear industry with an impeccable record in materials accounting and control; and membership in the nuclear export control groups, the Zangger Committee and Nuclear Suppliers Group (NSG). [2] Canada actively promotes nuclear disarmament initiatives as signaled by: its integral role in drafting and advocating the Comprehensive Test Ban Treaty (CTBT); its support for the 1987 Missile Technology Control Regime (MTCR); and its ongoing efforts to achieve a Fissile Material Cut-Off Treaty (FMCT). [3] Canada is also one of the more active participants at the Conference on Disarmament (CD) and proactively seeks out new disarmament initiatives. Ottawa has supported and often sponsored every major nonproliferation initiative since 1945. Its nonproliferation credentials are widely regarded as among the best.

Several factors, however, have worked against Canada’s otherwise excellent nonproliferation credentials. Canada’s renunciation of the bomb was an easy choice given early indications that Canada could rely on the U.S. nuclear arsenal as a security umbrella, a benefit later embodied in the North Atlantic Treaty Organization (NATO). [4]Until 1984, the United States deployed dozens of nuclear warheads in Canada mounted on Canadian delivery vehicles that it trained Canadian forces to use. [5] Canada also produces about one-third of the world’s uranium and has exported it to both nuclear weapon states and non-nuclear weapon states, including directly supplying the U.S. nuclear weapons program. [6] Additionally, the premier of Saskatchewan, home to Canada’s largest uranium reserves, currently wants a uranium enrichment capability for the province because of its economic benefit, potentially setting a precedent for the spread of sensitive fuel cycle technologies. [7] Canada also remains non-committal on the 2005 US-India nuclear deal. [8] This article focuses on a further element of the Canadian nonproliferation paradox, one that is currently gaining greater attention than it has for years – Canada’s sales of nuclear reactors.

The International Energy Agency (IEA) predicts that nuclear power use could double by 2030. [9] The projected increase in demand has sparked a “reactor race” in which the world’s leading nuclear suppliers in France, Russia, the United States, and Canada are competing for new markets for nuclear power. Given this possible expansion, the relationship between Canada’s nuclear reactor exports and nonproliferation efforts will be of paramount importance to Canada’s otherwise strong reputation for nonproliferation.

Reactor Sales at Odds
Canada has long been a supplier of nuclear technology and materials, most notably the Canada Deuterium Uranium (CANDU) nuclear power reactor. CANDU reactors have been operating in Canada since 1962 and abroad since 1974. [10] Currently, these reactors operate in Canada (22), Argentina (1), Romania (2), South Korea (4), China (2), India (2) and Pakistan (1). [11] The crown corporation Atomic Energy of Canada Limited (AECL) sells the reactors and assists with their operation and management in Canada and abroad. [12]

CANDU power reactors are a proliferation challenge for two reasons. First, they use natural uranium as a fuel source, making them proficient producers of spent fuel that a state can reprocess to produce plutonium for a nuclear device. [13] Second, CANDU reactors do not need to be shut down to refuel, making them more difficult to safeguard against the diversion of material than light water reactors (LWRs), the most common type of reactor in the world’s nuclear reactor fleet. [14]

The problem with Canada’s nuclear reactor exports is that the AECL has succeeded only in selling them to states that have policies incompatible with Canada’s nonproliferation efforts. These sales occur because the economic benefits of CANDU reactor sales tend to trump Canadian nonproliferation policy. The AECL and trade officials cite the economic benefits of selling a reactor abroad, while nonproliferation experts are left responding to the attached proliferation concerns with little more than ramped up nonproliferation rhetoric. [15] This is the case now, as indicated by the AECL pushing reactor sales in Jordan and Ukraine, but it has also been the case historically.

Tempting Fate With Every Sale
The so-called peaceful nuclear explosion (PNE) conducted by India in 1974 was of direct consequence to Canada because of the major role Canadian technology and assistance played in the Indian nuclear program. Canada gave India its first nuclear reactor, a research reactor known as CIRUS, in 1956 within the context of the Colombo Plan to foster development. [16] Ottawa also sold India two power reactors in 1963 and 1967 as a part of the AECL’s early efforts to sell CANDU technology internationally. [17] These exports preceded the entry into force of the NPT in 1970, so the safeguards regime was still in its infancy. The CIRUS research reactor was not placed under what are now known as safeguards because of opposition from India, and the reactor ultimately provided the plutonium for India’s 1974 nuclear test. Not only did Canada export the technology India used to divert fissile material, but along with the United States, Canada helped train over a thousand Indian atomic scientists and engineers, providing the Indian government with at least some of the technical expertise needed to establish a weapons program. [18]

In response to the Indian “PNE,” Canada formally declared that any state to which it sells its nuclear technology must be party to the NPT and accept safeguards as a condition of supply. [19] Canada also strengthened its bilateral safeguards substantially, though there were some slippages to follow (see below). The experience with India also led Canada to become one of the loudest voices in favor of stringent export controls in the newly-formed Nuclear Suppliers Group (NSG) in 1975. [20] On the surface, Canadian export policy clamped down on proliferation. Despite the Indian experience and the government-imposed restrictions on reactor sales that followed, however, the Canadian nuclear industry continued to pursue contracts with some customers that presented clear proliferation risks.

The AECL, for example, initiated reactor sales with South Korea and Argentina in 1973. Supposedly learning its lesson from the Indian experience, Canada concluded bilateral safeguards agreements in 1975 and 1976 respectively for each state. It was widely known, however, that in 1974, South Korea had lobbied the French to purchase a reprocessing facility, which would have given it the ability to produce plutonium that could be used in a nuclear weapon. [21] Argentina was also a poor choice. Buenos Aires had not signed the NPT, was not a party to the 1967 Treaty of Tlatelolco banning nuclear weapons in Latin America, and had never made a commitment not to develop nuclear weapons. [22] To make matters worse, Argentina maintained a policy of nuclear autonomy and was in an intense regional rivalry with Brazil, leading both countries to consider a nuclear weapon option. [23] Although South Korea and Argentina were high proliferation risks, the AECL successfully negotiated reactor sales to both.

In addition to India, South Korea and Argentina, the AECL has sold reactors to Pakistan, Romania, and China. The reactor sale to Pakistan in 1956 came despite Pakistan’s regional rivalry with India, which made it a likely candidate to seek nuclear weapons, a goal that it ultimately realized in 1998. When the 1978 sale to Romania was completed, Romania was a member of the Warsaw Pact, and its decision to purchase Canadian over Soviet reactor technology raised fears that it had ulterior motives to acquire an independent nuclear weapon capability. [24] Finally, controversy broke out over a 1996 sale to China because, not only was China a nuclear weapon state, but China had supplied nuclear technology to Pakistan and perhaps others to aid their nuclear weapon ambitions. [25]The record shows, then, that of the six states to which Canada sold nuclear reactors, two went on to detonate a nuclear device, two had at least some nuclear weapons aspirations, one deliberately aided another state in its nuclear weapon program, and one was a member of an opposing alliance.

The AECL and Canadian government were not indifferent to the security implications of CANDU sales. Canada’s inability to compete in major nuclear power markets like Western Europe, where proliferation concerns were less, forced the AECL to look elsewhere. Lorne Gray, the President of AECL at the time of the Argentine sale, for example, admitted that Argentina was not a desirable customer, but that the AECL had little choice but to take what contracts to export CANDU reactors it could; LWR was the technology of choice in Europe, and Canada’s small domestic market could not generate enough orders to make the AECL profitable. [26] The bilateral safeguards agreement with Argentina was weak because the AECL had no other potential sales at the time, Germany was competing for the sale, and Canada could not leverage U.S. influence as it did in the South Korean deal. [27] The economics of the global nuclear industry dictated that if Canada wanted to sell CANDU reactors, it had no choice but to sell them to less desirable states that other nuclear suppliers could afford to avoid.

Continuity With Change – Canada and the Nuclear Revival
Today, the global market for nuclear power has shifted significantly. Concerns about climate change, energy security, and the rising cost of fossil fuels has sparked renewed interest in the demand for and supply of “carbon-free” nuclear technology. With 37 states currently declaring their intentions to pursue nuclear energy, nuclear suppliers like France, Russia, and the United States have concluded nuclear cooperation agreements with numerous states in the Middle East, North Africa, and Southeast Asia to pave the way for reactor sales. [28] Unlike the situation in the 1960s and 1970s, all of the states currently interested in nuclear energy are parties to the NPT and are required to have comprehensive safeguards agreements with the IAEA. Further, the 1997 Additional Protocol signals an unprecedented level of transparency of nuclear activities for states that adopt it, although it is not mandatory. The renewed interest in nuclear energy thus coincides with a more robust nonproliferation regime.

The near-universal involvement of states in the nuclear nonproliferation regime has made the reactor sales side of the Canadian nuclear paradox less difficult for the Canadian government and the AECL to justify, even sales to states in unstable regions with volatile governments. Global nonproliferation efforts have created a market in which selling a reactor to nearly any state is acceptable, at least according to formal guidelines such as those agreed by the NSG. Article IV of the NPT states: “All the Parties to the Treaty undertake to facilitate, and have the right to participate in, the fullest possible exchange of equipment, materials and scientific and technological information for the peaceful uses of nuclear energy.” [29] Canada is acting entirely within the spirit of the treaty by providing nuclear technology, as long as it does not assist any state to acquire nuclear weapons. Indeed, it could be argued that Canada has an obligation under the NPT to share its peaceful nuclear technologies. That Canadian technology did not contribute to the weapon programs of Iran, Iraq, North Korea or Libya, has helped avoid further public controversy.

Many states claim that under the NPT they have the right to the full nuclear fuel cycle for peaceful purposes, allowing them to acquire nuclear material. But as North Korea demonstrated, states can also later withdraw from the NPT, renounce their safeguards, and develop nuclear weapons. [30] This loophole is one of the biggest challenges currently facing the global nonproliferation regime. Efforts to close this gap are becoming increasingly necessary given concerns that the argument regarding nuclear “hedging” may be accurate, particularly in the Middle East. (Nuclear hedging is the process by which a state acquires a latent nuclear weapons capability by obtaining civilian nuclear technology that can later be used to develop weapons in the event that its security environment deteriorates.) Many non-nuclear weapon states such as Japan and Egypt have implied that no technical or financial barriers prohibit them from acquiring nuclear weapons, in part because of their nuclear infrastructure. [31] A peaceful nuclear program provides most of the technology and expertise required to build nuclear weapons, so the spread of nuclear energy is inextricably linked to the spread of nuclear weapon capabilities.

The major nuclear exporters have concluded agreements with several Middle Eastern states despite concerns about hedging. Canada appears to be no exception given its ongoing pursuit of a reactor sale to Jordan. [32] The challenge for governments in dealing with nuclear hedging is that states suspected of engaging in it usually have otherwise suitable nonproliferation credentials. Steep competition among nuclear suppliers also means that, if the Canadian government does not allow a reactor sale because of hedging fears, another state is likely to take its place.

Multilateral governance of the fuel cycle is one important way that the nonproliferation regime can reduce the proliferation risk of a revival of civilian nuclear power. The IAEA and nuclear supplier states have put forward several proposals for a guaranteed fuel supply or “fuel leasing,” in which states take back spent fuel after supplying reactor fuel as a disincentive for states to acquire enrichment and reprocessing capabilities. Canada has always supported multilateral approaches to the fuel cycle, as long as they are commercially competitive and participation by suppliers is based on their nonproliferation credentials and legitimate economic justifications. [33] Taking back spent fuel is a more sensitive issue. Canada’s policy on spent fuel is that it will not take it back under any circumstances because of environmental concerns related to storing radioactive waste. [34]

Better Candidates or More of the Same?
The same challenges that bedeviled the Canadian nuclear industry in its formative years still plague it today. With claims of a nuclear “renaissance,” the AECL continues to promote sales of its older CANDU technology, as well as the new ACR-1000 (Advanced CANDU Reactor), while French and Russian LWRs remain the technology of choice in most markets for nuclear power. [35] However, the global surge in demand for nuclear reactors has led to increased interest in CANDU technology from states with better nonproliferation credentials than Canada’s previous customers.

The UK is considering the ACR-1000 as one of four reactor designs to replace its aging nuclear infrastructure. [36] A sale to the UK would demonstrate that CANDU technology is a feasible alternative to the dominant LWRs currently used in Europe. Penetrating European markets could mark a major turnaround for the AECL that might give the Canadian government the flexibility to require better nonproliferation credentials of the AECL’s customers. Much still needs to fall into place, however, for this series of events to unfold, as the AECL still faces steep competition from U.S., French and Russian companies. [37]

The AECL claims that the ACR-1000 is more efficient, safer, and more secure than older CANDU technology, but proliferation resistance credentials are notably absent from its official sales pitch. [38]It will also use “slightly” enriched uranium, meaning that customers will now need to acquire enrichment services from one source or another, and in an era of energy “insecurity,” states may wish to enrich their own uranium. It seems counterintuitive that, given its sales history, the AECL would not try to make CANDU technology more proliferation-resistant. The projected emergence of next generation reactors like the ACR-1000 is an opportunity for Canada to bring the AECL’s reactor sales more in line with the nation’s nonproliferation credentials.

Beyond the UK efforts, the AECL is currently in discussion with Ukraine and Jordan about concluding CANDU sales. Both states have suitable nonproliferation credentials on paper: they are parties to the NPT; neither is suspected of wanting to acquire nuclear weapons in the short-term; and both have a demand for nuclear energy to meet domestic energy needs. Neither state, however, is an entirely desirable candidate for a CANDU reactor. Ukraine was reluctant to give up the nuclear weapons it inherited from the Soviet Union after the Cold War, and since independence its political situation has proven to be somewhat volatile.

A sale to Jordan may be even more problematic because it is located in a turbulent region and is a poor country that could suffer instability sometime in the future. Jordan is suspected of nuclear hedging along with much of the Middle East. Jordanian officials often express qualms about Israeli nuclear weapons, and have more recently conveyed fears that the Iranian nuclear program will have repercussions for Jordan’s security, particularly if the United States becomes increasingly aggressive towards Iran. [39] Jordan, however, is 95 percent dependent on energy imports, suggesting a legitimate need for an energy alternative. But the nuclear option makes little sense for Jordan relative to other sources given the prohibitively high costs the country will have to incur to upgrade its infrastructure if it is to sustain a nuclear program. [40] The incongruity of a nuclear power program in Jordan raises concerns that factors other than energy security may be motivating Jordan’s interest.

Managing the diffusion of nuclear technology to the Middle East so that it is consistent with the goals of the nonproliferation regime is one way Canada can burnish its nonproliferation credentials. [41] Denying the technology to Middle Eastern states, however, is not an attractive solution because states in the region and probably elsewhere will perceive it as discriminatory and inconsistent with the NPT’s “inherent” right of all states to secure nuclear technology for peaceful uses. Another alternative is needed. Deciding on that alternative is one of the major challenges facing the nonproliferation regime in confronting the nuclear revival.

Does Canada Have it Right?
The economic benefits of selling CANDU reactors internationally are clear: they generate extensive revenues; allow Canada to break into new markets; increase employment; and are indispensable to the sustainment of Canada’s domestic nuclear industry. [42] The economic imperative to export them was, and remains, considerable. Exporting nuclear reactors keeps Canada in an exclusive club. The political cost to Canadian leaders of rejecting past CANDU exports because of proliferation risk would have been significant given the size of the nuclear industry in Canada and claims that insufficient exports would lead to its collapse, notwithstanding the qualms of the small but vocal nonproliferation community and the Canadian peace movement. [43]

Whether or not Canada is justified in ensuring the survival of its nuclear industry by continuing to sell nuclear reactors to risky states is an ethical question with no clear right answer. The economic argument breaks down somewhat, however, when one considers that every CANDU reactor sold at home and abroad has been heavily subsidized by the Canadian government, and ultimately the Canadian taxpayer. [44] The benefits of reactor sales go to a subset of Canadian industry rather than to Canadian industry as a whole.

Another question at the heart of the Canadian nuclear paradox is where to draw the line between indiscriminately providing peaceful nuclear technology and taking precautions to prevent proliferation. Supplying peaceful nuclear programs without inadvertently facilitating a later diversion to a weapons program is becoming a bigger part of the Canadian nuclear paradox as new markets for nuclear power emerge in unstable regions. If the AECL is going to continue to market CANDU technology in these regions, Canada needs to find the right balance between lucrative nuclear exports and preventing nuclear weapons proliferation.

Justin Alger – Norman Paterson School of International Affairs, Carleton University, Ottawa