Too close to home: the planning law limits on population around a nuclear plant
This article was published in our nuclear briefing, Nucleus.
It is no secret that it is difficult to get planning permission for a nuclear power station in the UK. One of the reasons is our mathematical test on the population around a reactor: the bemoaned semi-urban density population criterion.
The determination process
The Planning Act 2008 classifies a nuclear plant with capacity of more than 50 megawatts as a nationally significant infrastructure project. Applications for such projects in England and Wales fall to be determined by the Secretary of State for Energy Security and Net Zero. The Act obliges the Secretary of State to have regard to national policy statements on nuclear energy generation when deciding whether to grant or refuse an application. The latest policy statement (EN-7) came into force on 18 December 2025. In a move away from its 2011 predecessor (EN-6) its focus is no longer limited to eight specified sites.
The new policy statement encourages a broader range of sites to be brought forward in recognition of advancements in modern reactor designs. EN-7 sets out a framework of factors which should tilt the Secretary of State towards a 'yes' decision where satisfied. The factors include geophysical attributes, impact on biodiversity, and impact on water availability. The move to a flexible framework of factors is welcomed by the nuclear industry. Less welcomed by the industry is the retention in EN-7 of the Office for Nuclear Regulation's semi-urban density population criterion.
The maths test, not quite as easy as Pi
The semi-urban density population criterion applies a complex mathematical formula to the population around a potential site. It is a pass or fail test. In this respect it stands in contrast to the other factors listed in EN-7.
It works by totting up local population figures in a series of one-kilometre radial bands from a potential site out to 30 kilometres. The number of people in each band are separated into 30° segments like slices of a pie with the reactor at the centre. Next, weighting factors are applied to come up with a figure for each segment and each cumulative radial band. The weighting factors are derived from the atmospheric dispersion of radioiodine in a hypothetical nuclear incident. They reflect the average dose that a person would receive in each radial band with no protective measures. The effect is that a smaller, densely packed settlement near the centre of the pie has a worse impact on score than a larger, more evenly dispersed settlement towards the crust.
The final numbers are processed against fixed benchmarks and converted into 'site population factor' statistics, all of which must be less than one. The first benchmark, in unweighted terms, is 5,000 people per square kilometre when you look at each 30° segment. The second benchmark, in unweighted terms, is 1,000 people per square kilometre when you look at each cumulative radial band. In other words, the 5,000 people/km2 test must be satisfied for twelve out of twelve slices of pie. The 1,000 people/km2 test must be satisfied for 29 hypothetically smaller pies, not just the pie as a whole. These benchmarks are what the ONR considers an acceptable risk to the public. Determining what constitutes an acceptable risk to the public is an inherently political and philosophical art. It is nevertheless a judgement that governments make on matters of medicine, transport, housing and welfare every day.
Bad science
The benefit of a mathematical formula with fixed benchmarks is that developers know where they stand. And a comparison of the harm to be suffered from a nuclear incident in different locations is not a bad thing. However, the UK Nuclear Regulatory Taskforce submitted to Parliament this year that the criterion is based on outdated technology and bad science. This is not hard to believe for a policy that was introduced less than a decade after the UK’s first commercial reactor was deployed and a full two decades before typewriters were replaced by office computers.
The methodology used in the criterion has not been substantially updated using decades of operational and safety data that has been collected since the 1960s. It does not care about reactor size. More importantly, the criterion fails to take account of safety improvements in modern reactor designs.
Shrinkage of reactor output in many modern designs depicts the criterion as a blunt instrument. Whilst it is an oversimplification to say size equals risk, a smaller reactor contains a smaller amount of fuel. In these very simple terms, there is a greater risk posed by a 1.6 gigawatt PWR than an SMR which is smaller in output by a factor of twenty. The criterion makes zero allowance for this reduction of risk.
Modern reactor designs employ a range of independent safety systems. Many of these are passive systems, meaning a reactor in distress will shut down in a safe and controlled manner without human intervention. These systems can rely on physical processes such as natural circulation or gravity to perform safety functions without power or operator action. For example, gravity-drop control rods are held up by electromagnets and will drop into a reactor in the event of an unexpected loss of power. The use of molten salt as coolant in some designs is another example. When molten salt overheats, it expands and naturally absorbs more heat from a reactor core, slowing the chain reaction.
The retention of an over-zealous local population criterion is therefore a missed opportunity. It curtails the benefit of reactor shrinkage and reactor safety improvements. The solutions are to revisit the numerical basis of the criterion (and end up with something probably even more complex) or to move to a qualitative test. A third solution is to do away with safe local population density as a planning consideration in EN-7 altogether.
The REPPIR alternative
To the extent that population density should remain a planning consideration, it could harness an exercise an operator must already carry out to comply with the Radiation (Emergency Preparedness and Public Information) Regulations 2019 (REPPIR). These regulations require an operator to carry out a hazard evaluation and consequence assessment of potential radiation emergencies. The consequence assessment looks at the incidents likely to occur with that particular design in that particular location with that particular set of design and operational mitigations that are in place.
The consequence assessment also takes a real world look at the harm. For instance, in the majority of Great Britain, we have a prevailing south-westerly wind. This means an escaped cloud of radioactive particles is most likely to blow towards the north-east. A site located close to an infant school or a retirement home will have more particularly vulnerable people in harm's way than is revealed by simple population numbers. By contrast to the REPPIR exercise, the maths of the semi-urban density population criterion takes no account of these nuances.
The main duties of an operator under REPPIR are to carry out the hazard evaluation exercise, discuss it with the local authority, and implement an appropriate emergency preparedness plan. REPPIR behaves like a planning requirement in that these duties are a pre-condition to use of a site for ionising radiation. However, it is unlike a planning requirement in that it is focused on mitigating the impact of that use rather than questioning whether the use should happen in the first place. Nor does it give the public a say in the process.
If the REPPIR exercise were to replace the semi-urban density population criterion in the planning process, it would give the Secretary of State a better-tailored framework to consider the number of people in harm's way from a nuclear incident, the gravity of the risk, and the mitigations. Many will feel that the risk of a nuclear incident to the local population should remain a planning consideration so that the public has a voice on it. Evaluating nuclear risk should not be the exclusive preserve of the ONR.
The future
The critics of the semi-urban density population criterion say they do not want to compromise nuclear safety, they simply want the test to get with the times. Unless revisited in some way, the criterion will inhibit the UK government's pledge to quadruple nuclear generation to 24 gigawatts by 2050. We live on a densely populated island, and the new power stations will have to go somewhere.