I would suggest a consensus for a multinational framework for SMRs, as a formal statement of principle. This would be a tier below an internationally binding convention and could draw from models of the International Maritime Organization, International Civil Aviation Organization and the Naval Ship Code.

There are templates outside of the nuclear sector, for global fleet design and safety licensing, design change management and design authority, and requirements management, that have proven to be successful and could be employed (Soderholm, 2013). These include maritime regulation mechanisms for nuclear transport, and aircraft licensing. Aircraft licensing in particular could be a useful template to regard for standard SMR plants deployed in multiple countries (Goodman and Raetzke, 2013).

20.3 Conclusion

Keeping in mind the global consequences and effects of both disasters and advances, we should look at the risks of a lack of shared cooperative mechanisms for nuclear licensing, safety, monitoring and fuel. Developing countries and the rest of the world share so many systemic risks (economic, health, political, environmental) that new principles of sharing and cooperation, responsibility and mutual obligation will have to supersede narrower concerns such as sovereignty. Arguably, this is already taking place within the nuclear realm, under the NPT (nuclear non-proliferation treaty), and the concept would simply be extended, as far as SMRs are concerned, to the arenas of climate change and technology transfer. This idea deserves further study. SMRs, standardized, safety-enhanced, small-scale and flexible nuclear power technology, would be suitable for such new cooperative arrangements.

There will be a change of global understandings of risk, due to the raised awareness about climate change (because of acute observable effects and incidents). This will affect the economics of SMRs. The trade-offs and embedded risk biases will become more transparent in crisis. As a result, it is anticipated that emissions — reduction and climate-change strategies and policies will gain importance; and that nuclear safety monitoring, regulatory activity, and supply chain will be increasingly internationalized.

Above all, in the very idea of the ‘developing’ country, there is a continuum, the idea of progress. What is development for? We are aiming for the attainment of basic human functional capabilities by all, no matter where, with no exceptions (Nussbaum, 2000), for the ‘unfolding of powers that human beings bring into the world’ (Nussbaum, 2011).

To reiterate, SMRs provide the opportunity to go back to basic first principles (just as technology designers have) and consider what supporting institutions, infrastructure, and rationales for their use, are actually fit for purpose in a world of increasingly shared benefits and detriments.

We have to start somewhere.

[1] AP1000® is a trademark or registered trademark of Westinghouse Electric Company LLC, its subsidiaries and/or its affiliates in the United States of America and may be registered in other countries throughout the world. All rights reserved. Unauthorized use is strictly prohibited.

[2] Identification of relevant attributes for evaluation and selection, looking at the specific country taken into consideration.

2. Definition of measurement and evaluation process of each attribute (quantitative or qualitative, monetary or not, etc); each NPP design has to be evaluated on each attribute.

3. Definition of attribute’s hierarchical structure as required by a fuzzy analytical hierarchy process (AHP) application.

4. Expert elicitation to get attribute weights; each expert has to fill in a questionnaire of pairwise comparisons between attributes or groups of them. Fuzzy AHP permits judgements through linguistic variables (Yang and Chen, 2004).

5. Pairwise comparison matrices from different decision makers are aggregated through the geometric mean method presented in Kuo et al. (2002). Buckley’s method (Buckley, 1985) is then applied to the hierarchical structure and to get final attributes weights; these are fuzzy sets, so a decoding process is needed to obtain crisp values, the most common being the centroid method (Opricovic and Tzeng, 2004).

6. The TOPSIS method is applied for the final integration, looking at the five steps in Opricovic and Tzeng (2004).