SMRs need to be radical

A strong argument against small nuclear power (SMRs) is unfavorable economies of scale. Building bigger is usually cheaper on a per unit produced basis. This is why ships, planes, tractors, casinos, hotels, as well as power plants, and wind turbines tend to become larger over time.

Containerships 1956-2019

source: OECD International Transport Forum

To avoid a cost increase by deliberately building smaller nuclear power plants, we must be sure mitigations are in place. Among these countermeasures are volume production, modular design, faster build time with reduced stick build, passive safety, and more. As economy of scale usually scales by the power of 0.6 , and economy of volume scales by the power of 0.2-0.3, must have strong additional cost reduction drivers. 

The jury is still out on this, as very few projects have matured into reality (mainly in Russia and China).

In the meantime, we can ask ourselves some hard questions:

Firstly, do nuclear power plants produce cheaper electricity as they scale their capacity? It should be no shock, but they usually don’t. In recent decades, new builds in the West have been sparse and are usually plagued by timeline and budget overruns being one-off mega projects (that they are). Here, we can not look at economies of scale as the same plant (and capacity) costs more every time it is built. To see economies of scale in action, we must look back to the golden age of nuclear energy: before 1976 in the USA. As a striking example, we can see how Westinghouse reduced cost per kW capacity from an initial $9000 to less than $2000 by scaling up their PWR water reactors. It is worth noting that the main reduction comes from scaling from 60 to 500 MW capacity. Scaling beyond  700 MW did not show a significant reduction (we might suggest that this is a case of running to stay in place due to increased regulatory costs that plagued the industry). 

Data from: Lovering 2016

The UEA Barakah project, built by Korean KEPCO,  has been justly hailed as a win for large nuclear, with overnight costs below 4000$/kW. Assuming this a golden standard for cost, we scale the Westinghouse curve from the above graph to meet this point. This scaled curve gives a crude “guesstimation” of the cost of building "smaller-capacity Korean PWR plants". This graph adds PWR SMR projects from two maturity levels: 1. Already built or final stages: Rosatom and KAREM. 2. Several years away from the first build:  Nuward, Rolls Royce, KAERI, Last energy. And lastly, Nuscale, with its initial estimation several years ago and current estimation (that halted the first project for UAMPS).

Data Sources: public company statements

Although it is expected that first-of-kind builds will be much more expensive, it is still informative to see that mature projects fit nicely into the scaled curve.

The unavoidable conclusion is that

SMRs need to make dramatic design changes to be competitive in the electricity market.