For SMRs to remain competitive there will need to be heavy state-side subsidizes for consumers, as the initial cost of energy produced from them is considerably above wholesale auction prices.
https://rethinkresearch.biz/articles/smr-boon-forecast-by-nei-rethink-doubts-it/—SMR boon forecast by NEI, Rethink doubts it, By Connor Watts, 24 Aug 22, Some 19 utilities surveyed by the US Nuclear Energy Institute (NEI) says they see potential for up to 90GW of small modular reactors (SMR) within the United States by 2050. This would work out to around 300 reactors producing 731.3 TWh of energy, most of which would come online before 2050. Those numbers look pretty good until you look at current generation figures
During 2021 the US consumed around 790 TWh of electricity from nuclear power plants and 899 TWh from coal plants. The EIA projects that 16GW of nuclear plants will be decommissioned, alongside 69GW of coal plants up to 2050. This means that as much as 95% of potential SMR energy production could be going directly to the replacement of existing facilities.
This leaves only 5GW of remaining demand potential for SMR energy production up to 2050 outside of decommissioning existing facilities, we think even this may be optimistic.
Traditional nuclear plants take upwards of a decade to build and often run massively overbudget, making the energy it does produce once built more expensive as costs are attempted to be recouped.
SMR producers promise to do away with these drawbacks through standardizing their designs to enable factory production. Minimizing cost while shortening production times, theoretically addressing the main weaknesses of traditional reactors.
In practice it’s a little more complicated.
Considering the complexity and risks involved with nuclear power generation, commercial production and design of SMRs remains a slow and meticulous process. This has left many SMR sites still in the planning or design phase years after their announcements, almost competing in deployment time with traditional nuclear plants.
Once a new SMR gets deployed after its design and development period, it will need to be monitored for a few years to inspect for defects and inefficiencies within the design to prevent any mishaps. This is likely to add yet more time to an already long production horizon, adding costs as new units cannot go into production.
The cost savings achieved through modularity and standardization are borne through mass-production and deployment. In a way this is the “gigafactory” approach for nuclear. Considering long initial production times, this will contribute significantly to a short-term increase in the price of nuclear electricity, minimizing its applications where it remains competitive.
SMRs are supposed to come to market at $60 per MWh LCOE – but already wind and solar are considerably below that, and what level will they be at after SMRs arrived on the scene in volume, by say 2030?
For SMRs to remain competitive there will need to be heavy state-side subsidizes for consumers, as the initial cost of energy produced from them is considerably above wholesale auction prices.
Another issue with nuclear power generation is water usage. Earlier this month nuclear plants in France had a rule concerning water discharging waived as heatwaves boiled most of Europe.
Typically, the reactors would reduce their output to minimize damage from discharging hot water into the nearby ecosystems, but this rule has been waived until the 11th of September to ensure energy supplies in the short term. SMRs will also need to use local water supplies as a coolant, which makes them ineffective in a drought.
This can be mitigated through the use of alternative coolants such as liquid metal, gas and molten salt, but many SMR designs currently work similarly to traditional nuclear.
To use the time horizon for SMRs makes them look economically unfavorable, and while these 19 utilities may genuinely feel they like the idea of more nuclear, their controlling state utilities commission may well have something to say about whether they ever actually get installed.
