Nuclear Renaissance? Probably not…

“Renaissance” is a term frequently used to describe the growth in nuclear power plant construction worldwide.  I recently completed an examination of the nuclear power market and determined that this word (meaning “reborn” in French) has been misapplied. 

No nuclear power plants began construction in the US between 1979 and 2011 – now there are four: 1 in 2011 and three in 2012.  What’s unclear is whether there will be any more.  Fifteen additional plants have been announced, but none have firm construction starts and a few of them have been “indefinitely deferred” or cancelled.  During the 90’s and 2000’s a few plants in Europe began operation, mostly in France.  There are only two currently under construction in France and Finland and both are considerably delayed and over budget. So I would not call what’s happening in Western Europe and the US a “renaissance” by any stretch.  In Russia and Asia, however, it’s not a period of resurging nuclear interest.    Nuclear power plant construction certainly slowed for a while, but by no means did it stop completely.  In fact, a huge growth spurt is happening in South Korea, China, and Russia.  In fact, from 2008 through 2012 China began construction of 30 GW of new plants.  Japan had a number of plants in construction as well, but all have been suspended since Fukushima. 

At present there are 372 GW of nuclear power plants.  They contributed about 12% of the world’s electricity production.  Official forecasts predict an additional 117 GW to be installed by 2020, or an increase of about 23% at a cost of about $344 billion.  Globally, the most likely new nuclear capacity installed by 2020 will be between two thirds and three fourths of the official forecasts and the invested capital about 75% to 85% of official.  In fact, on a probabilistic basis, the official forecasts are higher than the highest metric on the distribution curves. 

Declarations of a renaissance typically originate from organizations whose primary purpose is to promote the technology or are based upon data provided by those organizations.  The source forecasts from the World Nuclear Association (WNA), (NEI) Nuclear Energy Institute, and the International Atomic Energy Agency (IAEA) are the most frequently cited reports.  The US Energy Information Administration (EIA) and the International Energy Agency also produce outlooks however these seem to rely on data similar to others. 

Additionally, many countries publish their own national outlooks and in many cases, these forecasts reflect official energy policies – policies that can be as much about intent as national pride.  These forecasts are part and parcel to the IAEA and WNA forecasts.

Given the sources of these outlooks and forecasts (with the possible exception of the EIA), they represent what appeared to me as extreme optimism and worthy of considerably deeper analysis.  In addition, some of the drivers that impact such a forecasts are rather uncertain themselves, especially global and national economic conditions.  I therefore chose to use three different forecasting methodologies: a compendium of all the official forecasts to establish the “as announced” case using the announced operation dates and capital costs where available; a scenario forecast that incorporated some assumptions about economic futures where the as announced case reflected the highest growth; and a probabilistic analysis that sought to capture the uncertainties in the scenario analysis and provide a likely range of installed capacity and capital investment.

Other findings

In addition to the conclusion that considerably less construction of plants than announced plans is likely, and at a higher unit cost, the analysis showed that:

  • Utilities installing nuclear power plants in the US and Europe are paying a premium for this technology over other electricity generation sources. Assuming the capital costs experienced by the units currently in construction in the US and Europe, the levelized all in cost for nuclear is just about 16 cents per kilowatt-hour. That’s more than any other alternative base loaded generation.  In fact a recent Congressional Budget Office report cited a range of costs for a new coal plant with carbon capture and storage between 9 and 15 cents/kWh.  GTM cites a number of solar power purchase agreements at between 7 and 9 cents.  Since these metrics are quite public, the question is not whether nuclear power is economic, but rather how much of a premium over alternatives are utilities paying to make use of the technology.
  • Outside the US nuclear appears to be marginally economic however official capital cost estimates from those countries are not transparent – the extent of subsidization not included in those estimates cannot be determined.  China just announced that all nuclear power generation should meet a cost target of about 7 cents/kWh – this is not a number that is likely to be achieved in Western countries.
  • Small nuclear plants may be coming, but none will be commercial before 2020 and they are likely to be as expensive as their big brothers. A number of countries and private firms are working on several different concepts for small modular nuclear plants.  These designs integrate all major components in a single encapsulated system that will shut itself down without human intervention under certain circumstances.  They are small relative to the gigawatt plus sized units in construction, but are all in at least the tens if not hundreds of megawatt size.   The hope is that the smaller units might be less capital intensive and easier to construct and license but it is not clear at al they will be any less expensive (on a per unit basis) than their large cousins.  I generated probability distributions for the installed capital costs for some of the front runners and compared them with the big units- any differences were too small to be meaningful.  Over the next two or three years a few demonstration units will begin operation, but this new class of reactors will not be commercial at any time through the period of the report – 2020.
  • Decommissioning expenditure forecasts for nuclear, however substantial now, probably understate what will actually happen.  Plans are being put in place to decommission all the plants in Europe, some at the end of their license lives and some much earlier than that.  Some reactors in the US may receive license extension, but others will proceed to decommissioning.  In Japan the new licensing process is just about complete and it is not certain which of the currently shut down reactors can comply with the new regulations.  Recently a number of plants have begun decommissioning and the actual costs are considerably higher than previously considered.  The range of costs are between $1,000 and $4,000 per kWe, much higher than the originally installed costs of these units.  Long term expenditures for decommissioning are likely to exceed investments in new nuclear installations.
  • Several major issues associated with nuclear power from its inception remain unsolved and unresolved over the last 20 years.

Probabilistic risk analysis proved to be a far better means to assess a market with the inherent uncertainties of this one.