THE finance minister's recent budget discussed a significant focus on nuclear energy, particularly the use of Small Modular Reactors (SMRs) and the need to amend India's nuclear liability law: The Civil Liability for Nuclear Damage Act, 2010. Though this has been a demand of the US nuclear industry for a long time, it has been on the back burner as the US nuclear industry has iteself been in a crisis for quite some time. Westinghouse emerged from Chapter 11 bankruptcy protection only in 2018 and commissioned the Vogtle nuclear plant in Georgia, US, in 2023. It took them nearly 15 years. The two Westinghouse AP 1000 units cost Georgia Power nearly $37 billion, more than twice that of the initial estimate. A similar plant in South Carolina was abandoned after the utility company spent more than $9 billion on construction costs. Though a few older players, including Westinghouse and General Electric's nuclear business (rechristened as GE Vernova), are rebranding themselves as suppliers of SMRs, no such plant has yet been built in the US. Even Niti Ayog's SMR Report (The Role of Small Modular Reactors in Energy Transition) says that the only functioning SMRs are in Russia and China. However, both these are much smaller in size and are experimental in character.

The recent interest in SMRs in the US was based on the belief that the AI revolution will demand huge, energy-guzzling data centres. Though the SMRs have high capital costs, running at constant load, they will have lower unit cost and carbon footprints. DeepSeek however has provided an alternate model for AI with much lower capital costs. The large data centre model of AI promoted by the US biggies – OpenAI, Mircosoft, Google, Amazon – may therefore not be the best way to go.

So why this sudden interest in SMRs in India and the need to modify our Nuclear Liability Act? We already have the indigenous capability to design and build nuclear plants in BARC and Nuclear Energy Corporation, and an engineering industry consisting of BHEL, L&T, BHVP, etc., which can supply the equipment. Why this sudden revival of interest in inviting Westinghouse and GE again and willingness to bear both the high costs and the risks of any nuclear plant? Particularly as the US itself does not have any appetite for the same; at least in the utility sector, which is back to promoting gas and oil with Trump's slogan of Drill, Baby Drill!

This takes us to the larger geostrategic issues and what the Trump administration wants to do. Clearly, the US has decided that as the world's leading military and financial power, it will extract "tribute" from the rest of the world. This means the so-called Rule Based Order of G7 and the trade regime under WTO are to dead. The US has military competitors, Russia and China, and economically China, with which the US will negotiate. But from others, he expects tribute. Though India is too big to be treated like Mexico or Canada, it is still expected to give concessions. The willingness to buy the rebranded costly US SMRs and US military aircraft, weapons, and LNG all appear to be demand for tribute in different forms.

We are going to focus here on the import of nuclear plants, more specifically, of the US SMRs. The US record on nuclear energy has been one of huge cost and time overruns. We will also examine whether the US SMRs are technologically only scaled-down versions of the plants that Westinghouse and GE were building earlier that have no takers today. Or do they have any novel technologies? In this context, we will also look at what India did to build its nuclear energy capabilities and why it appears to be willing to sacrifice these to appease the US.

In the US, unlike most countries, including India, both the owners of the nuclear plants and the suppliers/builders are very large private players. While General Electric and Westinghouse supplied the plant and equipment, the utilities are private energy companies combining all three activities: generation, transmission, and distribution. Incidentally, after the Enron disaster in California, the deregulation or dismantling of these vertically integrated electricity utilities has slowed down significantly in the US.

Let us return to the import of US SMRs, for which India has now agreed to amend the Civil Liability for Nuclear Damage Act (CLNDA). This has been a long-standing demand of the US, so that the US suppliers have no liability. Only the owners of the nuclear plant, namely the Nuclear Power Corporation, would be solely liable for any accidents, even if the nuclear accident was due to faulty equipment supplied by US companies. Unlike the US, no other international supplier, be it the Russians or the French, has made nuclear liability a precondition for supplying nuclear plants and equipment.

The SMRs in the US do not appear to have advanced or novel designs of reactors but only reduce the size of the reactors. Even earlier, when Westinghouse's Vogtle and South Carolina VC Summer plants were being supplied, though they were 1100 MW in size, they were sold as modular designs. This meant that the reactor vessel would be fabricated in modules in the Westinghouse manufacturing plant, erected and put together only at the site. The SMRs have the same approach but are only smaller, hopefully allowing for quicker construction. That reducing the size of the plant can lead to lower cost per unit of production goes against the basic engineering economics of what is called economies of scale. On the contrary, scaling up the size of the plant leads to lower capital costs per MW. This was why nuclear plant suppliers scaled up the size of their plants then, reaching 1100-1600 MW in size. Interestingly, the opposite – reducing the size of the plant – is now being touted as an argument for lowering the cost without explaining how such a reduction leads to either lower cost per installed Mega Watt or per unit of electricity produced.

A recent study by the Institute for Energy Economics and Financial Analysis (IEEFA) has analysed the claims of the SMRs for lower cost of the plants. I quote, “The rhetoric from small modular reactor (SMR) advocates is loud and persistent: This time will be different because the cost overruns and schedule delays that have plagued large reactor construction projects will not be repeated with the new designs. But the few SMRs that have been built (or have been started) paint a different picture – one that looks startlingly similar to the past. Significant construction delays are still the norm, and costs have continued to climb.” Interestingly, the study above also gives details of the costs of the SMR plants mentioned by the Niti Aayog as examples of successful SMRs, the one in Russia, the other in China, and the third under construction in Argentina. The cost escalation has been 3-4 times for the Russian Floating SMR and the Chinese Shidao Bay plant as well, while Argentina's CAREM 5, which is yet to be commissioned, costs are already 7 times the original estimate! Another SMR project, between NuScale Power and Utah Associated Municipal Power Systems, has also agreed to terminate its contract in November 2023, citing high costs and offtake uncertainty. If the technology remains the same, it is difficult to see how the costs per unit of electricity can come down by simply reducing the size of the plant. In other words, as the IEEFA Report title says, SMRs: Still Too Expensive, Too Slow and Too Risky.

The cost of solar and wind, the two energy sources expanding the fastest globally, have seen significant drops both in the cost of installed capacity per MW and per unit of electricity produced. Simply put, we already have a revolution in renewables, which has seen both its capital costs and its cost per unit of electricity drop well below any other grid-connected energy source, be it fossil or nuclear. The capital cost per MW makes nuclear power the most expensive option today. Any economic argument for nuclear power seems particularly problematic, as in renewables, we have not only achieved economies of scale but can only improve as we invest in it further. It is no longer only an environmental imperative but also the most economic option. Yes, we will still need storage, with pumped hydro for the grid looking like the best option. Renewables are no longer only for climate change activists; they make hard, commercial sense.

Why, then, are we barking up the wrong tree? Why are we promising to buy small modular nuclear reactors from US suppliers who cannot sell them successfully in their own country? Concessions to the US may be necessary in India's geopolitics, but should that lead to another disaster for the country's electricity sector, and a far bigger one than the Enron fiasco?