Touch here for mobile friendly version

Friday, February 10, 2017

Globally, new nuclear power stations are becoming one of the lowest cost sources of energy

Figure 1: Global LCOE from IEA Projected Costs of Generating Electricity, 2015 Edition


I was skeptical when I first saw the nuclear data (encompassing 11 new nuclear power stations). Being a joint venture between the IEA and the NEI, I wanted to check for pronuclear bias. And of course, any projection into the future is suspect but this one only went out to 2020, which is just three years away. So, I went looking for verification. I crosschecked the above values for the energy sources listed with those found by other sources, such as the EIA (not to be confused with the IEA) and found that they were reasonably consistent.

I then crosschecked the LCOE values for other countries from different sources and found them to also be similar in value.

Turns out that the cost to build nuclear power varies greatly from country to country. But when you look at the global range, average, and median LCOE (levelized cost of energy) for the new nuclear power stations built in the last five or so years, they're amazingly competitive. Hydro and coal are still shown to be the cheapest source at the 7% discount rate shown in Figure 1, but because hydro can't, and in my opinion, shouldn't scale up appreciably in the last remaining river ecosystems in the last biodiverse regions of the planet, I'm hoping its low cost does not lead to more of it. The study assumed a $30/tonne carbon penalty which makes coal look more expensive than it actually is ...because there is no global $30/tonne carbon penalty. The study also provided results for 3%, 5%, and 10% discount rates. 

Case in point; a South Korean company will bring on line a 1,400 MW reactor, Barakah 1, (the first of the four being built in series for the United Arab Emirates) this year after starting construction in July of 2012. All four are ahead of schedule for completion by 2020, which is an average of one nuclear reactor every two years. Two years is the same time frame used by Lazards to calculate the LCOE (levelized cost of energy) for wind and solar. The LCOE for these Korean reactors being built in the UAE is in the lower portion of the nuclear range in Figure 1.

One of the main costs of nuclear is the interest being paid on loans while it is being built (number of years without any income to start paying off debt). All else being equal, the faster you can build one, the cheaper it is. South Korea is proof that nuclear power stations can be built very rapidly and cost effectively once a company has acquired the necessary level of  engineering and manufacturing expertise (along with its suppliers).

From an article in The Economist regarding the Barakah nuclear power station:

It is using three times as much concrete as the world’s tallest building, and six times the amount of steel. Remarkably, its first reactor may start producing energy in the first half of this year—on schedule and (its South Korean developers insist) on budget. That would be a towering achievement.

Great pun ...now let's put that into perspective.  Solar PV would take roughly 80 times more tons of mined materials per unit energy. Wind, about ten times more. See Figure 2.

Figure 2: Material/TWh needed to build and operate nuclear, wind, and solar


It would take roughly 6,000 2.3 MW wind turbines to produce the same amount of energy in a year, but because wind doesn't always blow, it would come in fits and spurts instead of as a steady, dispatchable supply. Wind power certainly has its place as a means of reducing fuel used by natural gas plants, but it can't be used  to perform the baseload role in a power grid system (just as nuclear can't perform the peak power role--and neither can wind). See Figure 3.
Figure 3: Number of wind turbines to match output of Barakah 1-4.

Wind? This is the United Arab Emirates. Maybe they should have used solar which has a very low value for LCOE in such a sunny place, although it would consume roughly 40 times more land, and of course, you get nothing at night.

However, you can bet that natural gas, wind, and solar will all play major roles in their power grid along with nuclear because nuclear is best for baseload and limited load following, natural gas is best for load following and peak demand, and when available and when it is economical to do so, wind and solar can reduce how much natural gas is burned. 

Wind and solar enthusiasts, almost without fail, will point to the lowest LCOE on the latest Lazards summary chart and proclaim:
All solar and wind costs going forward will soon be that low everywhere!
Of course, if that is a valid argument for solar and wind which are so dependent on local values of wind speed and solar insolation, then it is especially valid for nuclear as well (see Figure 1).
All power should come from the two sources with the lowest LCOE (onshore wind and utility thin film PV)!
Why not the three lowest, or four, or five lowest, and if utility thin film is the cheapest solar and utility onshore the cheapest for wind, why are we building residential solar PV, solar thermal PV, and off shore wind which all make U.S. nuclear look cheap?

Two more problems with that reasoning:

The Lazards LCOE does not include the costs of integrating intermittent sources into a grid system, which tend to be much higher than for conventional sources.

The Lazards summary chart on page 2 (shown below in Figure 4) for solar is only for solar in the very sunny Southwest (not to mention, the LCOE values on the 2015 report were higher than they are for the 2016 report) and the low wind value is for a capacity factor of 55%. So, unless you live in a really windy place, it won't apply to you.
 
Figure 4: Lazards 10.0 sheet 2, summary graphic with markups

I'm trying to keep my articles short and therefore readable.

The End

However ...if you have a burning desire to know why nuclear (or wind, or solar, or hydro, or gas) should not necessarily be used just because they may have a low LCOE in a given grid system, feel free to read on.

Why do intermittent power sources tend to have much higher integration costs? From the horse's mouth:

The (biased against nuclear, see Footnote 4) Lazards 10.0 study:

Certain Alternative Energy generation technologies are cost-competitive with conventional generation technologies under some scenarios; such observation does not take into account potential social and environmental externalities (e.g., social costs of distributed generation, environmental consequences of certain conventional generation technologies, etc.)

While the levelized cost of energy for Alternative Energy generation technologies is in some cases competitive with conventional generation technologies, direct comparisons must take into account issues such as location (e.g., centralized vs. distributed) and dispatch characteristics (e.g., baseload and/or dispatchable intermediate load vs. peaking or intermittent technologies).
From the (biased for nuclear) IEA Projected Costs of Generating Electricity, 2015 Edition:
Despite the general relevance of these conclusions, the cost drivers of the different generating technologies nonetheless remain both market- and technology-specific. As such, there is no single technology that can be said to be the cheapest under all circumstances. As this edition of the study makes clear, market structure, policy environment and resource endowment all continue to play an important role in determining the final levelised cost of any given investment.
So, don't shoot the messenger. Below I show some graphics to supplement the above verbiage:

Figure 5: Example of system costs not accounted for in Lazards LCOE calculations
Source

What really matters is what combination of energy sources (baseload + load following + peaking +wind + solar) in a given place in a given grid system will result in the lowest charge to consumers served by that grid.

And finally, some analogies to go with the verbiage and graphics:

It is the combination of electronic components (resistors, capacitors, transistors, diodes, and microchips etc., some costing many orders of magnitude more than others) that create the most cost effective way to make a computer motherboard.

A one size fits all power source is as unrealistic as a one size fits all means of transport (ships, airliners, trains, tractor trailer rigs,  delivery vans, SUVs, cars, bicycles, and on and on).

For a given power source used inside of a grid system, you could start by answering some questions: 

Figure 6: Simple example of a logic flow just to pick the lowest LCOE power source inside of a system (grid).

And after running through that logic chart, all you have found is the lowest LCOE source in a given windy, sunny, wet, or none of these, area. The hard part is to determine if that cheap source can fit into the grid system with all of the other energy sources without increasing electric bills for consumers. Will it require more power lines than another source, or cause the operation costs of other existing and needed sources to increase etc?

You may be scratching your head trying to visualize just exactly how the insertion of a source with a lower LCOE than any other source in the grid could increase overall costs. I created a spreadsheet to show how it can happen. See Figure 7 below (click on it to enlarge):
Figure 7: Spreadsheet used to demonstrate how a low LCOE power source can increase rates to consumers in a given system (grid).
 See Footnote 5 for sources used for above graphic. 

Footnote 1: http://www.world-nuclear.org/information-library/nuclear-fuel-cycle/introduction/nuclear-fuel-cycle-overview.aspx

Footnote 2: http://nature.berkeley.edu/er100/readings/Pacca_2002.pdf

Footnote 3: http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr1640/v2/sr1640v2.pdf  (Page 2 and 37 of 259) 

Footnote 4:

Note (j) on page 3, note (c) on page 14, and note (e) on page 20 of Lazrds 10.0, in reference to the nuclear values says:
Does not reflect decommissioning costs or potential economic impact of federal loan guarantees or other subsidies.

Couple things wrong with those notes that indicate an antinuclear bias:

1) Decommissioning costs are already part of the LCOE of nuclear. Nuclear power stations pay into an audited and monitored fund to make sure future decommissioning costs are covered.

2) Although wind and solar LCOE values also don't include subsidies, and the headings to the charts already say "Unsubsidized Levelized Cost of Energy Comparison" nuclear is treated differently with a unique note to point that out again.

3) Although many renewable energy projects have made use of federal loan guarantees, this note is only applied to nuclear.

And if subsidies are a significant player for nuclear, why did they not include nuclear in their "Sensitivity to U.S. Federal Tax Subsidies" chart on page 4?

Footnote 5:

http://www.oecd-nea.org/ndd/reports/2012/system-effects-exec-sum.pdf

https://www.nei.org/CorporateSite/media/filefolder/Policy/Wall%20Street/WallStreetBrief2014slides.pdf?ext=.pdf

https://www.lazard.com/media/438038/levelized-cost-of-energy-v100.pdf
 

  

No comments:

Post a Comment

Comments that are not respectful of other participants will be deleted, so don't waste your time on a post that will be canned. Feel free to post links to pertinent sources and to your own website as part of your comment. Spam disguised as a comment will also be deleted as will comments that consist primarily of copied and pasted words from other authors (plagiarized content).