Generic Design Assessments
Blog: Blog - Adam Smith Institute
The UK nuclear strategy today is a muddle. After a brief description, this blog suggests what it should be. The Government claims that it ensures "that any new nuclear power station built in Great Britain meets high standards for:safetysecurityenvironmental protectionwaste management"The process is called the Generic Design Assessment (GDA) which is claimed to take four years, though insiders say the average is nearer six. But that is less than half the time taken to approve the National Grid Electricity System Operator (ESO) plans to connect up to 86 gigawatts (GW) of offshore wind by 2035. If one is building just one nuclear power plant every 20 years, these bureaucratic delays may not matter much, but the new small modular reactor (SMR) technology means, if we are to achieve net zero by 2050, building five a year over the next 20 years. No chance of that unless the bureaucrats mend their ways.The three stages of the GDA are: "Step 1 (initiation) which can take around 12 months. At its conclusion, ONR will publish a GDA Step 1 Statement setting out its regulatory position at this point. Step 2 (fundamental assessment) can take around 12 months, after which a GDA Step 2 Statement sets out ONR's regulatory position at this point. Step 3 (detailed assessment) can take around 24 months. ONR will then publish one or both GDA Step 3 Statement and Decision Document."If a developer requires a GDA for the building of an SMR at one site only in the UK then the process can be halted after the successful conclusion of Step 2 followed by a Site Licence application. If the deployment of a particular SMR design is envisaged for multiple sites in the UK then Step 3 needs to be successfully concluded.Although the four issues are supposed to be considered as an integrated whole, they can really be considered as two: safety determined by the Office for Nuclear Regulation and site approval determined by the Environment Agency. Good progress was made on the former issue by the nuclear regulators in the US, Canada and UK agreeing that safety clearance in any one of the three would be regarded as approval by all three. The two-way (and hopefully now three-way) transatlantic agreement to harmonise the licensing process has thus far only tackled the administrative landscape work that occupies much of the first 2 years. There are few safety aspects to Stage 1 which could be reduced to 4 months in all 3 countries if a standardised and transportable approach is agreed. Note that activities covered by the various steps and/or stages of the UK, US and Canadian systems do not match precisely at this stage, and some harmonisation is still required. Stage 3 is where the actual design is critically scrutinised and an iterative dialogue with design teams is established. This work can take around 12-18 months depending upon the lineage and degree of novelty in the reactor design. If the applicant has a well-resourced design team that can respond quickly to regulator questions and design modifications, it can proceed quickly. In the case of the AP1000, there were still almost 1,000 issues raised by the ONR that remained (and still remain) unresolved. This is where delays can become significant and where developers may be to blame. Trying to design and license a new reactor on a shoestring budget can result in developers complaining to their shareholders that the regulator is being unreasonable. Institutional shareholders then lobby politicians to influence regulators to be more accommodating of the newcomers.Terrapower has realised the importance of having an FOAK (First Of A Kind) site under development and adequate financial resources to quickly design, build and commission their first reactor. They are now putting pressure on the US NRC to bring the same urgency to their licensing application. There is one other point worth noting. The history of nuclear safety thus far, together with the accidents that have happened, have largely surrounded the use of water/steam as a coolant and transporter of heat energy from the reactor to where it is converted into mechanical and electrical energy. But when it comes to HTGRs and MSRs there is no high temperature and pressurised water in or around the reactor. These advanced SMRs have much simpler and safer systems and this needs to be reflected in their design processing time.How should it work? We need to distinguish, as the current system does not, between the traditional gigawatt reactors such as proposed for Sizewell C and small or advanced modular reactors (SMRs) which are rapidly coming onto the market. But more importantly we need to distinguish between an FOAK reactor that has never been licensed before and an NOAK (Nth Of A Kind) reactor for which design approval has already been awarded in either the US or Canada, or indeed another country that meets the UK's exacting standards for nuclear licensing. With more than 20 SMR designs likely to make it to the licensing and construction stages in the next decade, the US, Canada and the UK can ill-afford to repeat the administrative and technical procedures given the time, staffing and financial constraints that each regulator is experiencing or is likely to experience once these designs are submitted. The US and Canadian regulators are already facing political pressure to speed up SMR design assessments and to be proportional given the size of these reactors when compared to giga scale reactors. The UK is limiting SMRs to a small number deemed most useful for the UK situation. However, adopting this line risks excluding exciting new technologies and applications from both the UK nuclear industry and the nation's effort to decarbonise industry and power generation.Of the possible new nuclear plants going through the mill in 2010, it was deemed necessary for their sile locations to be approved by 51 separate quangos. It was all a complete waste of time as in March 2011, Energy Secretary Huhne says Britain may back away from the use of nuclear energy because of safety fears and a potential rise in costs after the Fukushima disaster. No new nuclear plants were approved after Sizewell B and Hinkley Point B in 1967. In other words, the process had become disapproval, not approval.Future strategy: NOAK for NuclearThe UK would be much better off if we abandoned trying to be first of a kind (FOAK) but aimed for Nth (NOAK). The advantages would be:The GDA process could be immensely streamlined as the environmental issues could mostly be taken from predecessor countries and, if Canada and/or the US had already given approval, the safety aspects could be eliminated altogether. Instead of starting from scratch, new GDAs should only have to look at environmental issues that differ from GDAs already approved. For example, if the radioactive elements of SMRs were sited underground, approval of one should automatically apply to another similarly sited. Hinkley Point C is a glaring example of the risks and costs of trying to be first. The French utility EDF, which is developing Hinkley Point C, has reported a cost increase to £33 billion (30% higher than initial estimates) and potential delays in its start date.Construction and operating skills would have been developed elsewhere first. A Science Direct paper took NOAK build time to be 6 years versus 10 for FOAK.The market would have had time to grow, be competitive and both capital and running costs reduce."The average capital cost of SMR NOAK plants currently under development worldwide is approximately $5,130 per kilowatt electric (kWe), which is estimated to be 30% less than FOAK plants. However, it's essential to note that specific costs can vary based on factors such as design, location, and technology."From these comparisons, we should conclude that both for large Giga Watt reactors and SMRs, the UK should abandon its enthusiasm for being the world leader. NOAK is quicker, cheaper and more reliable. Nuclear skills can be the better learned from other countries as well as developing our own. The car industry has proved not just to be a UK success story, but a NOAK success story too. Nuclear should learn from that.