Despite the long-standing use of the term SMR, there is still no internationally standardised definition for it. An IAEA definition describes SMRs as a group of small power reactors which, compared to today's nuclear power plants, have a lower output ranging from less than (up to) 10 MWe (microreactors) up to a typical output of 300 MWe. Conventional reactors, however, have an output of over 1000 MWe.

The functionality of this reactor group is very diverse: in a number of concepts, it corresponds to the functionality of today's light water reactors. These types of SMR are, therefore, subject to lower development risks, and the developers can draw on operating experience.

Other types of SMRs are based on novel concepts with little or no previous industrial experience. The latter can be categorised as high-temperature reactors, reactors with a fast neutron spectrum or molten salt reactors.

The current development of SMRs is largely state-funded and is taking place to a large extent in the USA, Canada and the United Kingdom. Provided the right conditions are met, SMRs can not only be built in those countries, but also be sold to others.

Industrial and geopolitical motives as well as military interests play a role in the field of SMRs. The majority of countries pursuing SMR development activities maintain nuclear weapons programmes and build nuclear submarines and/or already have a large "civilian" nuclear programme.

In addition to regular power supply, decentralised power supply for industry and households as well as heat for district heating, seawater desalination and industrial processes are mentioned; concepts for military use, such as mobile microreactors, are also being pursued. In Russia, floating nuclear power plants (Akademik Lomonossow, KLT-40S) are being used to supply remote regions.

In addition to traditional nuclear energy countries, there is growing interest in SMRs from countries with a lack of expertise and infrastructure in nuclear technology, such as Saudi Arabia and Jordan.

If SMRs are also suggested as a solution in the context of combating climate change and the associated reduction in greenhouse gas emissions for global electricity supply, the electricity production they achieve is relevant. Today's new nuclear power plants have electrical outputs in the range of 1,000-1,600 MWe. The SMR concepts considered in the report commissioned by BASE (see info box on this page), however, envisage planned electrical outputs of 1.5-300 MWe.

This means that a 3-1000 times larger number of units would be required to provide the same electrical output. Instead of today's 400 reactors with high output, this would mean the construction of several thousand to ten thousand SMR units.

This goal is a long way off. In addition, the planning process largely neglects various risks associated with multiplying the number of plants, in particular issues relating to transport, dismantling and interim and final storage.

SMRs promise shorter production times and lower production costs thanks to their modularity. Individual components or even the entire SMR should be (mass) produced industrially and transported to the selected locations for installation as required. Similar to a modular system, a single reactor with a low output or a larger plant consisting of several small reactor modules can be constructed from the components (modules) at the site in a short time.

Due to the low electrical output, the specific construction costs are higher than for large nuclear power plants, as there are no more scale effects. The report commissioned by BASE (see info box in the upper half of this page) calculates production costs taking into account scale, mass and learning effects from the nuclear industry: according to this report, an average of three thousand SMRs would have to be built for SMR production to become viable. It is therefore unlikely that the structural cost disadvantage of low-capacity reactors can be compensated for by learning or mass effects.

As with large-capacity nuclear power plants, the provision of SMRs is predominantly state-run or secured by demand (end customers, military). Although spin-offs are also developing from state-funded, large-scale research institutions, and there are also newly founded start-ups, their business models are still based on long-term state funding. It is, therefore, not conceivable that SMR concepts will be able to develop organisational models other than those that have been used in the field of nuclear technology for around 70 years.

Another key reason for the development of SMR concepts is the expectation of shorter time horizons, in particular shorter construction times, and possibly also less complicated dismantling. Looking at plants currently under construction or in operation, this assumption does not appear to be empirically substantiated: planning, development and construction times generally exceed the original time horizons many times over. Experience with historical SMRs indicates that the operating times of non-water-cooled SMR projects are short, and that dismantling them is a lengthy process.

Special application scenarios such as modularity, new manufacturing processes, materials and technological solutions for safety functions often require new regulatory approaches. The planned global spread of SMRs will, therefore, raise entirely new questions for the responsible licensing and supervisory authorities.

To date, there are no SMR-specific national or international safety standards. As many SMR developers are aiming for worldwide use of their SMR concepts, an international standardisation of the requirements would become necessary. This is currently not conceivable, especially for established nuclear energy countries.

On the whole, SMRs could potentially achieve safety-related advantages over high-capacity nuclear power plants, as they have a lower radioactive inventory per reactor and strive for a higher level of safety through deliberate simplification and increased use of passive systems.

Due to their smaller size, developers promise a lower safety risk for the reactors. However, the high number of reactors needed to provide significant amounts of electrical power as well as their planned global utilisation will increase the risk many times over.

Many SMR concepts also aim to minimise safety requirements, for example with regard to the diversity of safety systems. Some SMR concepts even call for the abandonment of current requirements, for example in the area of plant-internal emergency protection. Others completely forego external emergency response planning. These safety concepts, which are also pursued for the sake of cost efficiency, will also increase the risks.

Various non-water-cooled SMR concepts envisage the use of higher uranium enrichments or the utilisation of plutonium fuel and reprocessing technology. This has a negative impact on proliferation resistance - i.e. the need to prevent access to or the technology to produce nuclear weapons-grade material.

Another, often-cited key difference between SMR concepts and today's power reactors is the use of systems that have a long service life and would be delivered as a closed system. Sealing them could simplify monitoring and minimise transports. Furthermore, due to the high burn-up, the fissile material will also become unattractive after some time.

Yet, the high quantity of fissile material required at the start of reactor operation will have a disadvantageous effect. An additional aspect concerns the possibilities of fissile material monitoring by the International Atomic Energy Agency. Many of the standard methods for fissile material monitoring are not directly suited to the special features of SMR concepts, and this would pose new challenges.