The delivery of Elon Musk’s battery is a triumph of modularity. Musk promised the South Australian Government that he would deliver the battery in 100 days and completed ahead of the deadline. Furthermore, wind and solar projects are regularly being delivered in 2 years.
The lesson is clear – the current investment environment in electricity generation favours modular technologies that can be financed and delivered, on time and on budget, even when, ultimately they are not the optimum outcome. The combination of privatised electricity systems, environmental regulatory hurdles, unions, cost blowouts and time blowouts of mega projects, suggest that factory built and modular technologies have a clear advantage.
Unlike state-sponsored, public interest projects, such as metro trains or desalination plants, electricity generators are forced to compete in privatised markets selling the same sort of electrons. State sponsored projects can adopt the ‘Break-Fix Model’ of project management, or even adopt Hirschman’s Hiding Hand, but these options are not available to generator projects.
The three generation technologies that are most affected are concentrated solar thermal, coal with carbon capture and nuclear. The application of modularity to nuclear is the most interesting of the modularity paradigm.
Nuclear and negative learning
Nuclear is famous for being one of the rare examples of negative learning – up to the early 1970s, nuclear was following a normal learning curve, but the learning rate turned negative, even before Three Mile Island and Chernobyl, and has never recovered.
In the 1960s, GE and Westinghouse believed they had amassed enough in-house expertise in nuclear design that they were prepared to offer design-and-construct turnkey contracts. In part, this was driven by a business strategy of generating momentum and learning-by-doing in order to capture early market share in a potentially profitable industry. But by the 1970s, it became apparent that nuclear vendors had pushed ahead with large scale light water reactors before the full suite of reactor types had been assessed. This left some of the multifaceted social and practical issues of complexity, safety, and waste disposal unresolved. Ultimately, light water was probably the wrong choice for civilian reactors.
The scaling problem
Paradoxically, one of the early drivers of reduced per MW cost was increasing unit size. The near order-of-magnitude scale-up of US reactor size was very rapid; reactor size increased from less than 200 MWe pre-1960’s, up to 200 to 250 MWe mid 1960s, to around 1,100 MWe mid- 1970s. But larger reactors are more challenging to construct for several reasons – a scarcity of large forging facilities presents a bottleneck in reactor pressure vessel production; the very large containment buildings require specialised design, materials and quality control; the logistics and shipping of larger components are more challenging. Increased scale can also adversely contribute to cost escalations that arise from construction delays.
Furthermore, larger units suffer from the problem of geometric scaling – the core power (and radioactive decay power) is generally proportional to the volume of the core, which is related to the cube of the effective core radius. But the heat removal from the vessel is proportional to the vessel surface area, which is roughly proportional to the square of the core radius. Hence, the rate of heat removal, relative to heat generation, declines roughly linearly to the increase in radius of the reactor pressure vessel.
Lessons for modularity
It is clear that the conditions of 1970/80s France and Sweden, that led to rapid and successful state-sponsored nuclear build-outs, are not going to be repeated in the developed world. In countries with state-sponsored electricity mega-projects, such as China, the situation is different. As Quiggin notes,
China today looks a lot like France in 1970. China is ruled by a modernising elite that’s pro capitalist but happy to exercise state control over the economy, and to ignore or crush public opposition.
It is clear that the only path forward for Western nuclear development is factory built, modular units that can be delivered on-time and on-budget. Today, there are 3 Small Modular Reactors (SMRs) in use, 5 under construction, and 10 ready for near-term deployment. The question is whether SMR vendors can bring models to market sufficiently quickly, and at the right price. For example, Terrestrial Energy announced last month that it had successfully completed the first phase of the design review for its Integral Molten Salt Reactor (IMSR), representing the first regulatory opinion by a western nuclear regulator of a commercial advanced reactor power plant design.
The decarbonising effort will be easier with cost-effective and safe nuclear – the perennial question is – can it deliver?