Streamlining nuclear regulation is essential if targets are to be met

Nuclear regulation is widely seen as strangling the delivery of new projects. Timescales are long and costs are high. Hundreds of permits are needed, and there is a constant threat of judicial review when permits are finally granted, adding delays and costs. These problems are not unique to the UK, but they are more prevalent in the west than in Asia, and are one of the reasons that Russia, China and even India are taking the lead in nuclear development.

Last week I attended the Nuclear Hackathon, an event organised by DESNZ to bring together stakeholders from across government, regulators and industry to try to address the challenges of delivering new nuclear projects, with a particular focus on nuclear regulation. The Government has a target of 24 GW of new nuclear generation capacity by 2050, something which will be difficult if not impossible to deliver without significant changes to the permitting processes.

Why is delivering new nuclear so difficult?

The UK is generally bad at large infrastructure projects, and when you add nuclear into the mix, it’s easy to understand why things become materially more difficult. A year ago, Bim Afolami, Conservative MP for Hitchin and Harpenden summarised the problem in the Telegraph:

“First, we need to tackle the alarming growth in poorly made and conflicting regulation which permeates every area of economic life, strangling economic activity for small businesses. Second, we must address the continued expansion of judicial review, which is slowing or stopping critical decisions and pushing up costs for everybody. Only when we have started to make progress on these aspects of government will we be able to turn our policy intentions into results on the ground”
– Bim Afolami, Conservative MP for Hitchin and Harpen

He went on the say that in 2019, the National Audit Office found that there were roughly 90 regulatory bodies in the UK, regulating about 45% of the private sector, imposing monitoring and compliance costs, estimated at over £100 billion per year. That’s not to say that regulation is redundant and unreasonable, but it’s hard not to agree that the regulatory burden has started to get out of hand, and some pruning is needed. There has been a distinct lack of accountability on the part of regulators, as neither the Government nor Parliament have required them to justify themselves.

This is not a new complaint, even on these pages. I have bemoaned the lack of oversight of Ofgem, which like many other regulators, is accountable to Parliament, but since Parliament is not a sentient thing it’s not straightforward for it to hold regulators to account. Parliamentary committees have to decide to take action, and with so many regulators, and also new legislation and all kinds of other things to scrutinise, it’s no wonder that Parliamentary oversight is often lacking until things go dramatically wrong.

When asked about the CMA, Jeremy Hunt, the Chancellor said “I would not want to undermine [the CMA’s independence] at all, but I do think it’s important all our regulators understand their wider responsibilities for economic growth.” This was echoed in relation to net zero ambitions during this week’s hackathon, but more on that later.

The other issue identified by Afolami was the expansion of judicial review (“JR”). Professor Richard Ekins of Oxford University, a foremost expert in the area, notes that “judicial review has been much extended across the last few decades”. The fear that it will be overruled by judges is one of the most significant factors that government must take into account when it wishes to act. This is particularly an issue with major infrastructure projects.

The Nationally Significant Infrastructure Project planning regime was established through the Planning Act 2008 to provide more certainty on the construction of nationally significant projects, and to begin with it worked well with no successful JR applications in its first decade. But the National Infrastructure Commission has noted that since 2012 consenting times have increased by 65% from 2.6 to 4.2 years on average, and the rate of JR jumped from a long term average of 10% to 68%.

The data show that except for immigration cases, the numbers of JR applications have remained stubbornly high in the past few years despite efforts by the Government to implement reforms. This article by law firm Linklaters describes some of the recent drivers of the increase in JRs but there are also longer-term factors in play:

First, governments are having to regulate in new areas, which means more scope for challenge. Second, the challenge of delivering net zero will inevitably create winners and losers – who does and does not receive subsidies, where new infrastructure such as windfarms and pylons are built and so on, all create the conditions for dissatisfaction. Third, is the question of who pays when both state and domestic finances are tight, and an aging population introduces issues of inter-generational fairness. And finally, is the fact that judicial reviews generally do what they are intended to do, which is ensure that the courts interpret and apply the laws that Parliament enacts.

Journalist Isabel Oakeshott illustrates the problem pretty well with the account of the JR application made in respect of her attempts to replace a dilapidated shed with a new beach hut, and pointed the finger at no-win-no-fee lawyers and environmental zealots abusing the process either to rake in fees or to block projects of which they generically disapprove, typically by identifying some “procedural error by planning officers; presenting council bosses with an enormous dossier highlighting the mistake; and praying that they do not have the stomach or resources for a fight”. This is particularly true with nuclear projects where anti-nuclear campaigners weaponise the JR process to derail specific nuclear applications time after time.

The Government has already tried to address the issue. In 2020 it commissioned a review chaired by Lord Edward Faulks QC to look into the use of judicial review, which did not recommend any significant changes and warned that the Government should think long and hard before curtailing the powers of the judiciary. But the even just the threat of judicial review is distorting the planning process for new nuclear projects as discussed below, so it is likely that some form of limits will need to be applied.

Regulators obsessing about minutiae and forgetting about the bigger picture

Another problem that emerged during the Hackathon was the way that regulators focus on the minutiae of a specific site rather than considering the big picture. A great example of this is the requirement to fully specify the exact type of generators that will provide backup power at an early stage of the permitting process, with various environmental requirements being attached. But what are these backup generators and when would they be used? In a nuclear power station, these generators ensure that the cooling pumps will continue to operate in the event of a loss of power to the site.

The importance of these generators was highlighted in the Fukushima incident when the backup generators were swamped by the tsunami as they had not been moved to higher ground as required by the regulator. Although the reactors were all shut down either before the earthquake or in its immediate aftermath, the failure of the generators after they were inundated, meant that it was not possible to cool the reactors for days afterwards, ultimately leading to the hydrogen explosions which destroyed part of the reactor buildings. It was only through the diligence of the plant staff, with the support of the auxiliary fire service that pumps were brought to site to spray sea water onto the reactors to control the situation.

In other words, the backup generators are required to keep the reactors cool and prevent any potential release of radiation in the event that there is a disruption to the site power supply. Under these critical conditions, does anyone really care about the emissions that might be created by these generators? Other than under testing, these generators are expected to only be used rarely, but when they are needed, they would be absolutely critical, and in the prevention of a radiation release, just about any other types of emissions would be considered acceptable in the moment. So while of course the generators should not be unduly polluting, the key consideration is that they will work when needed and can run on easily available fuel. Getting into the details of the exact specification at an early stage seems excessive, particularly when the time from early permitting to operation can be decades long, and the exact model of generator would be likely to change in this time anyway.

Another example was raised in relation to biodiversity. Delegates complained about the difficulty of navigating habitats and biodiversity permitting which is spread across many different bodies and is very time consuming to address. They proposed bringing all of these requirements under a single body to avoid overlap and ensure there was a holistic and joined-up approach from regulators. There was pushback from environmental regulators in the room who denied there was overlap, but developers said they were often asked to submit broadly the same information to different regulators in different formats.

Another delegate made the excellent point that when looking at the bigger picture, it might be acceptable to sacrifice biodiversity at nuclear sites since the high energy density of nuclear power means that much less land is needed for the same amount of electricity. The alternative might be a much wider impact on biodiversity as more land is needed for other technologies, particularly renewables which have very low energy density. This argument also applies to the wider net zero target – in order to achieve this at a national level, it may be necessary to make compromises in relation to other aspects of the regulatory landscape.

There are also trade-offs to be made between harm to people versus harm to plants and animals: people are directly harmed both by blackouts, particularly if they occur after dark, and by expensive energy which leads to deaths related to an inability to properly heat homes in winter. It  might therefore be reasonable to accept a higher environmental impact on a limited area of land in order to avoid such direct harms to people (noting that lack of biodiversity can create indirect harms, but that these are a lower priority than direct harms).

Another example of the flaws in this approach includes the attitude towards the co-location of small modular reactors at industrial sites. Currently this is not possible since nuclear reactors can only be situated on sites identified as potential nuclear sites in the Nuclear Policy Statement EN-6, but a new version of this planning document is due, and will likely address this constraint. But there remain concerns among developers that new nuclear sites are considered by environmental regulators to be “green-field” when they are not actual green fields since there is already an industrial plant on the land which already impacts the local environment. This should lower the bar in the environmental impact assessments, but developers were not confident that would be the case.

The current approach to habitat regulation essentially assumes that if habitats are protected on a site by site, project by project basis, national level biodiversity is protected. This is not necessarily an appropriate approach to take because it ignores the wider national interest.

There was wide support at the Hackathon for requiring regulators to give consideration to wider net zero goals in their assessment of new nuclear projects. To this could be added the requirement for maintaining energy security. The Government recently amended Ofgem’s remit to require it to consider net zero in its regulation, but this was not done quickly as there was a need for public consultation first. It would be inefficient to repeat this process for all other regulators. Instead, the Government could legislate to require all regulators and similar bodies to give consideration to net zero goals, energy security and other strategic national priorities which the Government may designate from time to time in their decision-making.

Regulatory inflation is strangling new nuclear projects

We heard during the Hackathon from speakers at from Sizewell C (“SWC”), who expressed gratitude that Office for Nuclear Regulation (“ONR”) has allowed them to build an exact copy of Hinkley Point C (“HPC”) at the site, but despite this, and the fact there have been no relevant legislative changes since the HPC permits were secured, the SWC process attracted materially more regulatory questions than HPC. In the first round of the Development Consent Order (“DCO”) application 1,327 questions were asked in relation to SWC by the Examining Authority (plus 221 questions from other parties, so 1,548 in total), whereas at HPC there were only 8 questions (an additional 9 from other parties so 17 in total).

Law firm BDB Pitmans, pointed out that the first nuclear power DCO application was for HPC in October 2011 with just 17 questions in the first round, whereas the second nuclear DCO for Wylfa Newydd in June 2018 attracted 1,006 questions – around 300 more than the previous record for any DCO application. And overall for SWC, the questions were divided into six separate documents, with a covering document containing abbreviations and definitions. One was devoted to biodiversity and habitats alone. In total there were 2,229 questions – more than double the number for Wylfa and 130x more than for HPC.

BDB Pitmans went on to point out that not all questions are equal or require equal amounts of effort to address. They can range from pointing out typos to, in the case of SWC “please will the Applicant take into account any comments made by the Wylfa ExA [in its report] when preparing the next drafts of the DCO and the Explanatory Memorandum and explain why it proposes or rejects them”.

The regulatory pathway above describes the process up to initial permitting, but securing these permits is by no means the end of the story. After this, operating permits are needed as well as construction permits – around 160 of them for SWC. The majority of these are very simple such as land drainage permits, but others can be more complex for instance covering the supply of power to the construction site, or water management. These are not nuclear in nature but apply to many infrastructure projects, and are issued by several different regulatory bodies. In fact, the sheer number of regulators with whom developers must engage is intimidating.

One might ask why, if there have been no relevant regulatory changes, the amount of documentation needed to support this huge range of permits has expanded so much. One reason is a huge fear of judicial review. As noted above, the JR process has been weaponised by campaigners and lobby groups who disapprove of wider policy objectives aided by law firms who have found a way of earning easy money since they are able to recover their costs if they “win” either a JR eventually finding in their favour, or, more commonly, the government body or regulatory authority capitulating.

One way such bodies have sought to protect themselves against the JR threat has been to increase the information obligations of developers who are now required to submit thousands of pages of additional information to support their applications. This raises the bar to any potential JR applicants as it makes the preparation of an application more time consuming and therefore more expensive. However it imposes a huge burden on developers and on the regulators themselves who then have to prepare and respond to the additional material. This makes the entire process longer and more costly.

The solution to this problem is for the Government to legislate to reduce the amenability of planning processes for strategically significant projects to judicial review. This can be done in a number of ways:

Standing Requirements: The Government could introduce stricter standing requirements to limit who can bring judicial review cases. For instance, they could require that only parties directly affected by the planning decision or those with a demonstrable interest in the matter have standing to bring an application for a JR. This would help prevent campaign groups with no direct stake from initiating legal challenges.

Materiality Threshold: Implementing a materiality threshold could help filter out JR applications based on minor procedural errors or technicalities. The threshold could specify that only decisions with significant implications or errors that substantially affect the outcome of the planning process are eligible for judicial review. This would discourage frivolous challenges and focus legal scrutiny on substantive issues.

Time Limits: Setting time limits for initiating judicial review proceedings could reduce the risk of delays and uncertainty in the planning process. By imposing strict deadlines for filing applications, the Government could ensure that challenges are brought promptly and efficiently, reducing the potential for abuse of the process through prolonged legal wrangling.

Costs Reform: Reforming the costs regime associated with judicial review cases could deter speculative or unmeritorious challenges. The Government could consider measures such as capping recoverable costs, requiring losing parties to bear a larger share of legal expenses, or introducing financial penalties for frivolous applications. These changes would increase the financial risks for those considering judicial review, discouraging abuse of the process.

Expedited Procedures: Introducing expedited procedures for dealing with JR cases involving strategic planning decisions could help minimise delays and disruptions to the development of new nuclear projects. This could involve assigning dedicated judges or courts to handle strategic planning-related cases promptly and efficiently, ensuring that legitimate challenges are resolved swiftly without unduly impeding the planning process.

Specialised Tribunals: Creating specialised tribunals or panels to adjudicate strategic planning-related disputes could provide a more tailored and efficient alternative to traditional judicial review proceedings. These tribunals could be staffed by experts with relevant experience in planning law and policy, allowing for more informed and expedient decision-making while reducing the burden on the regular court system.

One interesting development was the recent confirmation by the Appeal Court that decisions made by the Government in its capacity as a landlord are not amenable to judicial review. The Government, through Great British Nuclear, is currently in the process of acquiring the Wylfa Newydd and Oldfield nuclear sites, so that might provide some scope for the Government to use its status as a landlord to limit some of the JR risk. I would be interested in the views of any lawyers reading this on whether the Government owning the sites would provide any benefits in this regard.

The approach to risk drives spurious design modifications

In line with the International Atomic Energy Agency, the UK regulatory process requires that the risks posed by a nuclear plant are demonstrably “As Low As Reasonably Practicable” (“ALARP”), taking into account economic and safety requirements. There is no specific definition of ALARP, but it can be found in legislation and case law in many nations. The concept originally derives from UK legislation, particularly the Health and Safety at Work etc. Act 1974, with the term being enshrined in UK case law since Edwards v National Coal Board in 1949 following the death of a coal miner by a rock fall that might have been prevented if the tunnel roof had been shored up.

The appeal court judges ruled that the National Coal Board did not have to take every possible physical measure to eliminate risk – it only had to provide protection where it was required. The phrase used in the judgement was “reasonably practicable”, establishing the principle that a risk must be significant in relation to the sacrifice (in terms of money, time or trouble) required to avert it – risks must be averted unless there is a gross disproportion between the costs and benefits of doing so.

An example of the application of ALARP in the nuclear industry is the requirement that workers’ exposure to ionising radiation is “as low as reasonably practicable”. But what happens once exposure levels fall way below safe limits. What happens when workers are exposed to more ionising radiation walking to site from cosmic rays than from the nuclear processes on site? Is reducing radiation levels “as low as reasonably practicable” still as useful exercise?

The drive for ALARP also means that designs cannot be left alone. A design can be approved, but what if some new development arises, some new technology or approach that could reduce risk? If the project was brand new, then those elements might reasonably incorporated, but what about a design that has previously been approved elsewhere. A good example of this is EDF’s European Pressurised Water Reactor (“EPR”). Designs were approved in France and Finland, and EDF hoped to build the same design in the UK. But this was not possible – the regulators got to work and forced through a raft of design changes, which resulted in a materially different reactor. The most egregious example of this regulatory interference was a requirement for EDF to build an entirely redundant analogue control system which could replace the digital controls if for some reason the computers failed to work. This is despite there being strong redundancy in the power systems, in other words this requirement was to address a risk that the computers didn’t work even when power was maintained to run them.

This modification was not required by either the French or Finnish regulators, neither of which has a reputation for a casual approach to risk. And while the requirement can be described in a couple of lines, in practice it meant a complete re-design of the control room to accommodate the extensive cabling needed to support the redundant system. I understand this was the single biggest factor driving the cost over-runs, although there were various other issues including external ones such as covid and the supply chain inflation triggered by the war in Ukraine.

The inherent problem of ALARP is that it encourages regulators to constantly re-evaluate whether a design can be improved, and while that might sound good in principle, in practice it makes it very hard to get things done, and is a huge inhibitor to the replicability needed to capture economies of scale. (While building many reactors of a particular type is an efficient approach, it also carries risk as evidenced by the recent issues with stress corrosion cracking in the French fleet – as the reactors age, they may begin to fail in similar ways, creating a wider risk to energy security, however there are ways of mitigating this risk for example by constructing a portfolio of different designs and ensuring an appropriate replacement cycle.)

Another problem with ALARP is that in some cases it transfers risks elsewhere. This has been seen in the chemical industry where in order to make one process safer, a more risky upstream process may be necessary, either in the production of a feedstock or piece of equipment. In the power sector, we are facing serious security of supply risks, so the permitting process for new nuclear power stations should also consider what would happen if the plant wasn’t built. Would this increase risks to energy security which in themselves might result in harm – it is difficult to imagine winter blackouts in the UK not resulting in fatalities, through road accidents and falls in the home among the elderly to name just a few.

“…after the Flixborough explosion the manufacturing process was replaced by a less hazardous one, using a raw material manufactured elsewhere by an equally hazardous process. Similarly, when extra equipment was proposed to decrease the amount of benzene vapour discharged to plant environments it was alleged that, on average, we could expect more people to be killed constructing the new equipment than would ever be saved by the reduction in the concentration of benzene vapour,”
– Trevor A. Kletz, Department of Chemical Engineering, Loughborough University

The solution is to modify ALARP. While the original concept already had an element of cost-benefit, that part needs to be broadened to take account of the bigger picture. If a risk is already low, there is no need to reduce it further. If a design has already been accepted, there should be a very high bar to changing it, with regulators having to justify their decisions, with some overseeing body being empowered to hold them to account. The drift towards risk elimination (as evidenced by the management of the AGR graphite cracking issue) must be halted.

Three key reforms for accelerating nuclear development

While the problems described above may seem intractable, there are some reforms that could make a significant impact:

1. Require all regulatory bodies to take keep the national net zero objective in mind when making decisions. This would force them to take account of the big picture in their decision-making: a reduction of biodiversity at one site might be acceptable when set against the wider benefits of nuclear, particularly since its energy density means a lot of energy can be produced from a small area, unlike many of the alternatives. This may well require legislation to avoid having to individually consult on and amend each regulator’s terms of reference separately.
2. Narrow amenability of nationally strategic projects to judicial review. While democratic processes and accountability are important, judicial reviews should not be abused by people who disagree with Government priorities – that is what elections are for. Setting time limits, preventing applicants from making repeated applications with slight amendments to previously rejected applications, placing restrictions on standing, and finding a means of setting materiality thresholds should be found. This would also require primary legislation.
3. Modify ALARP to remove the requirement for endlessly considering if a design continues to represent the lowest reasonably practicable risk. Once a design has been approved by a competent regulator, whether that is in the UK or a trusted country, the bar for requiring changes should be set high and a cost-benefit analysis should be conducted before mandating design changes.

These changes are simple to say, less so to implement, and time is short before Parliament is dissolved ahead of the General Election, which by law must be held before the end of January 2025. But the Government still has a good Parliamentary majority and could make some progress, particularly if it can secure the co-operation of the Labour Party which also supports nuclear power and the net zero ambition which will be difficult to achieve without the security and reliability of nuclear power.

It’s clear that unless things change, the nuclear target will not be met, and likely with it, the net zero ambition. The Government – and everyone else – knows this, hence the Hackathon and related consultation, but the time for acting in this Parliament is running out, so it needs to get on with it. And of course, these changes must also apply to regulation of the existing fleet to avoid un-necessarily premature closures if energy security is to be maintained in this decade, well before the net zero target date arrives.