Pressuring light water?

26 June 2018

John Lindberg looks at the impacts of path dependencies on nuclear innovation. Is the dominance of pressurised water reactors hampering industry growth prospects?

BORN OUT OF HIROSHIMA AND Nagasaki, the peaceful use of the atom quickly attempted to differentiate itself in the public mind from its violent background. Results were mixed. Nevertheless, the first decades were filled with considerable optimism about the revolutionary potential of atomic energy on humankind’s future. The combination of fallout from nuclear tests, Three Mile Island and Chernobyl severely dented the optimism and nuclear proponents have ever since adopted an apologetic and defensive posture.

Over the last few decades nuclear energy has tried to regain its momentum and find a way to increase public acceptability, with climate change seemingly offering a chance to rebrand and reframe. This has provided a potential opening for nuclear power, both in terms of rapid decarbonisation and prevention of pollution deaths. However, the efforts of the last few decades have had, at best, a limited impact with global nuclear generation falling due to closures outpacing new build, as well as the ease of deploying fossil fuels. Innovation has been seen as the cure, but despite an increasing focus on small modular reactors, Gen IV and nuclear innovation generally, there has to date been very little focus on the barriers to innovation, which are well-known in other sectors of the economy. Path dependencies, both institutional and technological, affect the innovation agenda, as explored in the article below.

Path dependencies in theory

Path dependency seeks to explain pathways and barriers for technology innovation, based on the notion that the use of certain technologies is based on temporally remote events, historical preferences and small historical events such as minor accidents or political issues. Institutional commitments are an essential but dangerous facet of complex infrastructural innovation that tend to remain entrenched for long periods of time, aided by institutional inertia. Heavily regulated and politically sensitive industries, like nuclear, are especially driven by the institutional environments from which they emerged, and within which they operate. They arise from factors such as increasing return and the sunk cost fallacy1, 2.

Early influences have large effects. The end state cannot be predicted in early phases and is defined by increasing path commitment which can lead to lock-in of certain technologies. Four features of technologies and their social environment tend to have an effect that grows over time:

1. Large setup or fixed costs. In these environments, companies and institutions have a strong incentive to identify and develop a single technology. The perception that continuous commitment is needed because investments have already been made, as per the ‘sunk cost fallacy’, strongly influence path dependency.

2. Learning effects. Knowledge gained from operating technologies or systems generate higher returns from continuing use, as repetition often leads to cost reductions and more effective operation. Such trends can clearly be seen in regards to nuclear, where operating experience has increased load factors across different reactor families.

3. Coordination effects. These arise when individuals or companies benefit more if others adopt the same technology and its support infrastructure. In turn, this increases use of the technology and investment in it.

4. Adaptive expectations. If there are strong negative consequences associated with selecting the ‘wrong’ technology, there might be a perceived need to ensure that the ‘right’ technology is adopted.

The dominance of light water

In the nuclear industry now a clear equilibrium state exists in favour of light water reactors (LWRs). This is clear when one looks at the deployment of reactor technologies worldwide. The first decade of civilian nuclear power was a period of technological uncertainty with no reactor design being hegemonic, but LWR concepts (especially PWRs) have become completely dominant in the nuclear market (see Figure 1). This would not have been clear in the first years of civilian nuclear power, where gas- cooled reactors (GCRs) seemed poised to dominate. 

When installed capacity is considered this is even clearer, stemming from the larger size, especially of PWRs. A similar pattern can also be observed by examining the plants under construction worldwide (Table 1) which reinforces the PWRs’ hegemonic position in the post- Chernobyl reactor market. LWR was not an obvious choice for broad commercialisation over gas-cooled or heavy water-moderated designs, and the choice was driven by largely technology-external factors, as shown in Cowan’s (1990) landmark study Nuclear Power Reactors: A Study in Technological Lock-in 3.

When fission reactors were developed governments played a leading role, as they largely defined research and development programmes. Plutonium production for the nuclear weapons programmes were at the fore, and other applications were secondary. Private sector involvement was initially small, and was only allowed to grow as governments decided to differentiate civilian and military nuclear power – indeed, most companies were involved with both. This is also seen in the early deployment phase of nuclear reactors worldwide, which is dominated by GCRs and PHWRs. These designs, with online refuelling, were suitable for producing plutonium for nuclear weapons and thus acted to ‘kick-start’ the development of nuclear reactors for other, non-military purposes.

The complete dominance of LWRs is the remnant of military needs for propulsion and energy politics, rather than optimisation for fuel efficiency or waste management. The early focus on GCRs and PHWRs changed as the military focus shifted towards propulsion. Two different concepts were investigated for propulsion purposes: PWRs and sodium-cooled fast reactors (SFBRs). The influence of Rear-Admiral Hyman Rickover on the US naval propulsion development programme cannot be overstated. He strongly advocated PWRs, and the decision was made easier by the issues faced by the SFBR-operated USS Seawolf. Despite offering superior characteristics, the SFBR did not match the reliability of the PWRs, so the latter became the preferred reactor for submarines and aircraft carriers. This ‘battle of preferences’ would come to influence nuclear power globally until this day.

Once LWRs became the preferred design, the mechanisms described earlier ensured that it gradually became entrenched. However, LWRs were not an inherently superior choice of reactor technology and they exhibit traits, such as fuel usage inefficiencies and poor waste management characteristics, that arise due to technological path dependency.

Escaping path dependency

Given the renewed interest in innovation as a possible remedy for the lack of a nuclear renaissance, it is necessary seek strategies to escape the path dependencies that otherwise present formidable barriers to non-LWR concepts. Walker (2000) says: ‘The usual assumption is that technologies die through some automatic process that is triggered in markets by the arrival of superior products and processes. Inferiority becomes self-evident and those displaying it are no longer selected. Even if technologies are not replaced immediately, ‘obsolescence’ will cause their disappearance...’4. However, this cannot be taken for granted because of institutional commitments such as long-term infrastructure investment and government involvement, driven by increasing returns and inertia.

It seems that exogenous shocks are necessary in order for the right conditions of change to occur, as systems tend to create their own equilibrium where individual actors themselves are trapped in the system, limiting the possibility of change from within.

Cowan & Hultén (1996) offers three distinct factors for such potential paths5: crisis in existing technology; regulation; and niche markets.

Crisis in the existing technology

The construction of new nuclear power plants (Generation III designs) across most parts of the western world has been plagued by cost overruns and delays, which has undermined public faith in nuclear. The response to nuclear accidents, the push for tighter radiological regulations and the impact on reactor designs (eg increased redundancy in safety systems) will likely have added to such a socio-political crisis in the existing technology. But while it could be argued that socio- political and regulatory responses have undermined the viability of nuclear energy (either by cancelling nuclear projects or increased regulatory burdens), there has been no push from governments or utilities towards non-LWR concepts. There is little to suggest that nuclear accidents have provoked a technological crisis for LWR concepts, as seen in the construction numbers in Table 1. It is therefore likely that a crisis of confidence in LWRs will result in the end of nuclear activity altogether, rather than a push for different nuclear reactor designs.


Nuclear regulatory frameworks developed and were refined alongside the ascendancy of LWRs. In turn, this will have created intuitional and regulatory path dependencies that must be considered, especially in licensing and regulatory provisions for advanced, non-mainstream LWR concepts. This is especially problematic in countries that have adopted the US Nuclear Regulatory Commission (NRC) approach, which is highly prescriptive and legalistic, and so is vulnerable to (and exhibiting clear signs of) institutional path dependency. Unless regulators allow a flexible licensing process and regulatory framework the institutional and legislative path dependencies will create formidable barriers for ‘challenger’ technologies, especially those relying on non-LWR technology. 

Niche markets

A potential strategy for entering the established nuclear market would be to focus initially on niche market segments. Such niches could range from waste management and off-grid production to industrial or district heat and desalination, with electricity production a secondary benefit.

Niche market applications would create test-beds for concepts and associated infrastructure, which is essential prior to mainstream market entry. This, alongside with commercial-scale proof of concept, will be for a technology breakthrough, which in turn could produce a cost breakthrough.

Note that technologies not only create a physical presence, but also define what is deemed to be a resource. This is clear in the ‘plutonium economy’ concept of using the plutonium created in thermal reactors as fuel via recycling. Whilst further discoveries of uranium deposits worldwide (along with public opposition) helped to remove the urgency for recycling, the hegemony of PWRs reduces the impetus for non-uranium dioxide fuels. It has reinforced the view that plutonium is not an acceptable fuel, thus reducing its value and disincentivising recycling efforts and developments. Niche market applications of offering alternative waste management solutions would be one option to try and redefine what constitutes a resource and what does not. This might be of interest seeing that the current resource regime classifies 96% of the content of spent nuclear fuel as ‘waste’, despite it being usable for energy production in non-LWR concepts.

Future path

The nuclear industry clearly experiences the effects of path dependence and its causes and effects must be understood by regulators, policymakers and developers alike. It does not determine technological or political outcomes, but it acts as a constraining force on the available options. Institutions play a key role in entrenching certain technologies.

For path dependency to be overcome in the nuclear sphere, institutional change is necessary, but achieving it is difficult. Nevertheless, awareness of the challenges posed by path dependency for nuclear innovation will hopefully serve as a starting point for the debate on not only how to deliver the next generation of nuclear, but also deliver sustainable innovation pathways that will learn from the lessons of the past.  

John Lindberg, Incoming radiation/nuclear risk communication PhD student 


1. Arthur, W.B. Competing Technologies, Increasing Returns, and Lock-in by historical events. The Economic Journal, Vol. 99, 1989, pp.116-131.

2. Unruh, G.C. Understanding carbon lock-in. Energy Policy, Vol. 28, 2000, pp.817-830.

3. Cowan, R. Nuclear Power Reactors: A Study in Technological Lock-in. The Journal of Economic History, Vol. 50, Issue 3, 1990, pp.541-567.

4. Walker, W. Entrapment in large technology systems: institutional commitment and power relations. Research Policy, Vol. 29, 2000, pp.833-846.

5. Cowan, R., Hultén S. Escaping lock-in: The case of the electric vehicle. Technological Forecasting and Social Change, Vol. 53, Issue 1, 1996, pp.61-79. 

Nuclear Engineering International has published a special edition for distribution at this year¹s World Nuclear Exhibition.

The focus of the special edition is on how digital solutions, virtual reality, robotics and other emerging technologies have the potential to secure the future of nuclear power.

With existing reactors facing increasing market pressures, digitalisation and automation offer opportunities to improve efficiency and reduce costs.

Robotics, data analytics and new simulation capabilities can aid cleanup of legacy facilities. New technology is also revolutionising the workplace, and can help the nuclear industry to attract a new generation of talent to replace its ageing workforce.

Looking ahead to the next decades,innovative technologies from small modular reactors to Generation IV designs and new fusion concepts, promise to make the next generation of nuclear energy more competitive with other energy sources.

Read your World Nuclear Exhibition Special Edition - New technology encouraging excellence in nuclear energy

WNE Figure 2: Load factors for different reactor types
WNE Figure 1: Number of reactors worldwide by type

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.