The Nuclear myth: Small modular nuclear reactors are not good complements to renewable energies

Supporters of small modular and advanced nuclear reactors often hail the reactors as the panacea to fighting climate change, but small modular nuclear reactors are more expensive than renewable energies such as solar and have low learning rates.

Whenever anyone dares talk about the advances in renewable energy, someone invariably pipes up saying words to the effect: “What about nuclear?”  There is a more sophisticated version of the above. Instead, they say, “what about small modular nuclear  and advanced nuclear reactors?” They present their arguments as rational, based on cold facts. The reality is different. Nuclear may be a viable energy source under certain unusual circumstances, but as a complement to renewables, at least in most places, forget it.
The error that some climate change advocates make is to argue that while fighting climate change will be expensive; it will be even more expensive in the long run not to fight it. In other words, the opportunity costs of not tackling climate change is high.  But in fact, this does not have to be the case. It is quite possible and perfectly consistent to think climate change poses an existential forest to humanity, but defeating it may eventually make us better off. And when I say better off, I don’t mean better than if we let climate change run a-mock; I mean better off even if anthropogenic climate change is false. To be clear,  I think we are probably underestimating the potential cataclysmic implications of climate change. I just happen to have faith in technology and believe that providing we are not collectively stupid (which we are more than capable of being), climate change can be defeated and reversed at a low cost.
And two words explain my optimism— learning rate. Learning rate describes how the cost of a technology falls with every doubling in its production. One example of a learning rates is Wright’s Law — named after an aeronautical engineer Theodore Wright who  “observed that every doubling of production of U.S. aircraft brought down prices by 13 per cent.” / Moore’s Law is another example of a learning rate.
Historically, learning rates are not well understood by the markets or policymakers. It seems to be human nature to underestimate the implications of learning rates. That is why early predictions about computers were so far off — for example, the Popular Mechanics 1949 prediction: “Where a calculator like ENIAC today is equipped with 18,000 vacuum tubes and weighs 30 tons, computers in the future may have only 1,000 vacuum tubes and perhaps weigh only 1.5 tons.”
For a new technology that could one day be superior to existing technologies, it often needs a nudge. For example, the development of solar power was partly due to the original NASA Space Programme.
War can also lead to innovations that the markets would not have typically funded. The mad behaviour that leads to bubbles can be associated with rapid technological advances, and fighting climate change may be a catalyst for new, ultimately superior, technology.

Nuclear versus renewables — The renewable revolution and learning rate miracle

Consider solar. Solar power is now the cheapest form of electricity generation in history — according to this report.  It turns out that in 2006 annual U.S. solar power shipments measured in a peak kilowatts was 320,208 and by 2019 had increased to 16,372,314 — that is a 50-fold increase in 13 years. Meanwhile, the peak cost fell from $3.50 per peak watt to around 40 cents over the same period.

In other words, output increased by 5,000 per cent; cost fell 87 per cent. Given that solar accounted for just 2.3 per cent of U.S. utility-scale electricity generation in 2020, it is not unreasonable to assume output from U.S. solar could increase by at least ten-fold (1,000 per cent) over the next decade or so, with further massive price reductions.   The U.S., the Department of Energy, wants to cut the cost of electricity from solar by a further 60 per cent by 2030.

Today, on average solar and onshore wind are the two cheapest forms of electricity.
Or to put it another way, the cost of energy generated by solar is becoming ridiculously cheap.

The problem with renewables and the nuclear-supposed fix
There is a snag, however. As just about every renewables critic will be quick to point out, solar is no good when it’s dark, and wind energy is no good when there’s no wind.

Enter onto the stage: Nuclear.  Nuclear energy, suggest its advocates, provides the baseload. Against that, supporters of renewables say there are multiple fixes to the intermittent problem. We will come to the fixes in a moment.

There is one big possibly fatal problem with nuclear; it is expensive. More to the point, it has been getting steadily more expensive; arguably, it has a negative learning rate.
According to the Our World in Data website, the Levelized cost from nuclear increased from $123 a megawatt-hour in 2009 to $155 in 2019.

That is where small modular nuclear enters the story.  In a nutshell, small modular nuclear provide around 300MW per module (enough electricity for between 30,000 and 70,000 homes) versus approximately five times that amount for conventional nuclear reactors.

The historical record suggests that savings (from learning rates) from small modular nuclear reactors will be inadequate to compensate for the economic challenges resulting from the lower generation capacity

The significant advantage of small modular nuclear reactors is economies of scale. Because they are relatively small, more reactors can be made, creating scale. Linked to that, we may well see a learning rate applied to small modular nuclear reactors because of their modular design, and the more we make, the cheaper they will be.

You may have spotted a flaw with that argument already. Small modular nuclear reactors are small compared to traditional nuclear reactors. However, each unit is still massive compared to wind turbines or solar panels, suggesting that the potential learning rate might not be so great for even small modular nuclear reactors.

Now, a paper from M. V. Ramana of the Liu Institute for Global Issues, School of Public Policy and Global Affairs, The University of British Columbia, Vancouver, Canada, has looked into learning rates applied to Small Modular and Advanced Nuclear Reactors.

The conclusion: “The higher construction and operational costs per unit of electricity generation capacity will make electricity from small modular reactors more expensive than electricity from large nuclear power plants, which are themselves not competitive in today’s electricity markets.”

As for learning rates, it states: “The historical record suggests that savings (from learning rates) will be inadequate to compensate for the economic challenges resulting from the lower generation capacity."

Spending on nuclear will lead to less spending on renewables, thus reducing the potential benefit of a learning rate
Finally, Ramana also says  that the job creation opportunities are “are inadequate to justify constructing the necessary manufacturing facilities.”

Baseload

Okay, so nuclear is expensive, and, even after allowing for learning rates, it always will be costly.
But what about those occasions when it is neither windy or sunny; what then? Don’t we need nuclear as a backup to the grid to support renewables?

Another paper, this one from Mengyao Yuan, from Carnegie Institution for Science, Stanford, even casts doubts on whether nuclear supports renewables.

The paper concluded: “Policies and funding that support particular technologies for low- or zero-carbon electricity generation can inhibit the development of other low-or zero-carbon alternatives.” Or, to put it another way, spending on nuclear will lead to less spending on renewables, thus reducing the potential benefit of a learning rate.

Or, to put it more simply still, the dream of ridiculously cheap electricity created from falling costs of renewables is not compatible with nuclear power. We may have to choose between cheap renewables and expensive nuclear; they are not complementary technologies.

The reason is quite simple. The ideal complement to renewable technologies is an energy source that can be readily turned up and down. The ideal complement is one in which most of the costs of operating are variable. So, if this complementary source of energy is not used, its overall cost is less.

Nuclear is the opposite — it sees massive setup costs and provides energy at a fairly steady rate.  When it is neither windy nor sunny, and renewables do not generate that much electricity, nuclear won’t solve the problem. It is not a form of energy that operators can easily ramp up.

Quite simply, if we go down the nuclear route, then the cost of fighting climate change will be high — that’s guaranteed.
In that case, what’s the answer? This article attempts to answer this question:

The future of energy and why I am not worried about the intermittent nature of renewables

But in summary, renewables need complementary technologies — and the solution to the intermittent nature of renewables comes not in one technology but several. Some are cheaper but less reliable; others are more reliable but potentially pricey such that you only use them in emergencies:
Potential solutions are:

• Solar and wind in combination —  they don’t provide a guaranteed energy supply, but there will be occasions when it is windy but dark or sunny but not windy.
• Smart grids. By using AI, IoT and maybe automation technologies such as RPA, smart grids will be able to channel energy when generated in optimal conditions into less time-sensitive applications, such as  charging up batteries in electric vehicles, storage heaters or ice for use in air conditioning. In time, vertical farms and cultured meat production could benefit.
• Lithium-ion batteries and other related technologies such as solid-state lithium batteries. Like wind and solar, lithium-ion batteries enjoy a steep learning rate; but they may never provide seasonal variety on occasions when it is neither sunny nor windy for an extended period.
• Vehicle to grid energy storage in which the battery in electric vehicles can provide backup to the grid.  This will be especially relevant as the longevity of batteries becomes      greater. But as in the example above, vehicle to grid will not provide backup over extended periods of low energy production.
• Grid level storage can hold energy for longer periods, such as hydropower and vanadium redux batteries.
• Ultra-high-voltage electricity transmission (UHV electricity transmission) is used in China to transmit  electricity over very long distances.  In China,  the Ultra-high-voltage electricity transmission reaches 2,000 miles. By using this technology, the range over which renewables are generated is greater, so it is less likely there will be no wind or sun over such a large area, and can also generate electricity across multiple time zones, so solar can generate electricity in one area for another region even when it is dark in that region.
• Synthetic fuels, which use energy generated at optimal times and when energy is plentiful, creating human-made fuels.

Does that mean there is no need for nuclear? There is probably some need. In remote areas where it is rarely sunny and solar is not effective, nuclear may be essential. Likewise, it may be required for some industrial purposes — although not many.
Image credits, nuclear and wind