Random Thoughts

U.S. DOE EIA versus MAB...2019 and 2024

Evaluation of IEA and U.S. EIA fossil fuel use projections to 2050

Roger Pielke Jr. posted an analysis of the International Energy Agency (IEA) World Energy Outlook (WEO) predictions of global use of fossil fuels from 2023-2050, versus the U.S. Energy Information Administration (EIA) International Energy Outlook (IEO) for the same time period. The fossil fuels under consideration were coal, oil, and natural gas, and the units of use are exajoules.

The IEA WEO 2023 predicted global use of each of the three fossil fuels would peak before 2030, and the IEA WEO 2023 predicted that global coal use in particular would decline significantly by 2050, to a value of approximately 60 percent (i.e., 0.60 times) the value in 2023.

In contrast, the U.S. EIA IEO 2023 predicted that global use of all three fossil fuels would increase to 2050. For example, if the values for EIA IEO 2023 values for coal, oil, and natural gas are taken as a value of 1.00 in 2050, the values for coal, oil, and natural gas in 2023 would be 0.96, 0.85, and 0.79, respectively.

I decided to develop my own predictions for global use of coal, oil, and natural gas from 2023 to 2050, and compare them to the IEA WEO 2023 and the EIA IEO 2023.

With respect to coal, I also think global use will "peak" prior to 2030, but I think global use during the entire period will be essentially flat, so it will be difficult until very near the end of the period to determine whether a peak has truly occurred. I predict the "peak" coal will occur in 2029, but if that peak has a value of 1.00, I predict that the value in 2023

 

 

Let's talk Xinjiang

Over at the Ron Unz Review, another commenter asked me to comment on the Uighurs in Xinjiang. I wrote the following. I'm always open to reading other opinions.

Hi,

OK, I know little about the Uighurs in Xinjiang beyond reading this:

https://www.bbc.com/news/world-asia-china-22278037

Here’s my take. Read Thomas Jefferson words (but don’t forgive his hypocritical actions):

We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness.–That to secure these rights, Governments are instituted among Men, deriving their just powers from the consent of the governed,

I don’t know what he meant by “all men are created equal” but I think/hope he meant, using current language, is “all people should be equal in the eyes of the law.” Then he says the purpose of government is to protect rights, including “Life, Liberty, and pursuit of Happiness.” And the governments get “their just power from the consent of the governed.”

It’s seems obvious that the government of the People’s Republic of China isn’t coming anywhere near that in Xinjiang: Uighurs and Han Chinese do not appear to be equal in the eyes of the law; the government isn’t coming anywhere near protecting the rights of every person in Xinjiang to “Life, Liberty, and pursuit of Happiness.” And they’re not operating with the consent of the governed, so their power is not “just” (legitimate).

I guess the only other take I have is that it looks like Uighurs are pretty concentrated in one area of the country. Letting Xinjiang break away would be a possibility. But if the Han Chinese minority in Xinjiang then became below the Uighurs in the eyes of the law in Xinjiang, that would just be replacing one wrong with another wrong.

P.S. I see that apparently they’re still *hand-picking* cotton in Xinjiang. That’s really terrible.

Is the RCP 8.5 scenario plausible? Was it ever plausible?

Executive Summary: "No." and "No." (All you executives can leave. Non-executive nerds can hang around for details.)

In 2011, a paper titled, RCP 8.5--A Scenario of Comparatively High Greenhouse Gas Emissions was published. The text in the paper stated, "...RCP8.5 depicts thus a relatively conservative business as usual case..."

In this post, I explore whether RCP 8.5 represents a "business as usual" scenario in 2023; whether it represented a "business as usual" case in 2011; whether it is even plausible in 2023; and finally, whether it was even plausible in 2011?

The answers to those questions are: "No." "No." "No." and "No."

The RCP 8.5 scenario was, even when the paper introducing it was published, a complete fantasy. What made it a fantasy was the assumptions of coal usage in the 21st century. Here is a table of the coal usage assumed in the paper. One column has the global coal use in exajoules (EJ), the units in which coal consumption was presented in the paper. The next column has the coal use converted from EJ to millions of metric tons (MMT), using a conversion factor of 1 EJ = 34.2 MMT of coal.

Table 1. Assumed Coal Usage in RCP 8.5 Scenario

The problem with the coal usage assumptions shown in Table 1 is that they're insanely high, particularly in the later half of the 21st century. The coal consumption is assumed to just keep rising throughout the 21st century, which is completely ridiculous, because coal use is actually declining in several key coal-consuming regions, such as North America and Europe.The real insanity kick in after 2020. For example, in 2040 global coal consumption is assumed to be 2.74 times what it was in 2000; in 2060 it is assumed to be 4.70 times what it was in 2000; in 2080 it is assumed to 6.52 times what it was in 2000; and in 2100 is it assumed to be 8.43 times what it was in 2000. In contrast, simple observation of current and previous trends show that it is likely that global coal consumption will never hit the assumed 2040 value of 2.74 times the 2000 consumption. That is, global coal consumption will never hit the 8,800 MMT projected for 2040...let alone continue rising to the mind-boggling value of 27,140 MMT assumed in RCP 8.5 for the year 2100.

Table 2 shows actual coal consumption, in millions of metric tons of coal for the year 2000, 2010, and 2021. (Note: The year 2021 was chosen, rather than 2020, because the COVID-19 pandemic lowered coal consumption around the world.)  Based on Table 1, all the regions of the would have to be at 1.83 times their year-2000 consumption in 2021, in order to be doing their "fair share" of coal consumption. Instead, none of the regions except Asia is doing its "fair share" of coal consumption. In fact, many of the regions--including North America, Europe, the Commonwealth of Independent States (CIS), and the Pacific--actually used less coal in 2021 than in 2000!

Table 2. Actual Coal Usage for 2000, 2010, and 2021

If the assumptions for coal use in RCP 8.5 were so unrealistic, why did the paper published in 2011 call it "business as usual"? Did they not know how utterly implausible the coal usage in RCP 8.5 was? I think the answer is pretty clear. None of the authors was a novice regarding analyzing trends in energy usage and carbon dioxide emissions. In particular, Keywan Riahi (the lead author) and Nebojsa Nakicenovic are senior employees with the International Institute for Applied Systems Analysis (IIASA), and have spent their careers doing energy and environmental analyses. So the only plausible explanation is that they deliberately mischaracterized the RCP 8.5 scenario as "business as usual". But why in the world would they do that? The simple reason is: money. Money would have dried up very quickly if they'd presented a realistic scenario for "business as usual."

The characterization of RCP 8.5 as "business as usual" is a lie. There was no basis in science to ever call it "business as usual." And anyone who writes any paper anywhere that labels it as "business as usual," is either incredibly ignorant or lying.

Let's talk storm surge protection

Responding to Ray Lopez on Marginal Revolution

Responding to comments from Ray Lopez on Marginal Revolution:

"Green energy is a First World fad, you realize that?"

Green energy is a "First World fad" about like LED TVs are "First World fad". Just what do you think Africa is going to power up with in the next 30+ years? (Hint: Think solar, solar, and solar. And if you think it will be nuclear, you need to go talk to people who actually know what they're talking about!)

"Jevons and the Coal Question, bone up on that and get back to us."

I have a better idea...you give me your predictions for global nuclear power production (in terawatt-hours, or TWh, for these years):

2018: 2586 TWh (from this wonderful Wikipedia page):

https://en.wikipedia.org/wiki/Nuclear_power_by_country

2030: ???
2040: ???
2050: ???

Just FYI, here are my predictions:
2030: 2700 TWh
2040: 2100 TWh
2050: 1800 TWh

And then tell me which countries are going to be the biggest generators of nuclear power in 2050.

Here is my prediction: Of the 1800 TWh of nuclear power generated in 2050--down from 2586 TWh in 2018--more than 60 percent will be generated in China, India, and Russia, combined. And less than 5 percent will be from the entire continent of Africa.

"The only solution long term is nuclear (for the world,..."

I think that's absolute nonsense, of course (see my previous predictions). But then again, I've only been studying energy issues throughout my entire advanced academic and professional life.

Present value cost of decommissioning Vogtle 3 and 4

"BillyJoe" on the Neurologica blog asked about the decommissioning cost of Vogtle 3 and 4...if they ever even produce electricity, which is not a sure thing. I responded to him twice on the Neurologica blog, but the terrible Disqus software marked both responses as spam. BillyJoe wrote:

I still don't get it though. If the construction cost of 2 nuclear reactors is 27 billion and, if that translates to $80/MWH. Why doesn't the 2 billion in waste disposal for the two reactors translate to $6/MWH?

Where to start?

1) It doesn't make much difference to this particular situation, but the construction cost of $27 billion for the two Vogtle reactors does not "translate to $80/MWh". The $80/MWh is from a completely separate situation...it comes from a generic "desktop" study...not a study specific to the two new Vogtle reactors. The $27 billion for the Vogtle reactors will probably translate to over $100/MWh, even if the two reactors are completed and operated for 40 years. And the cost per MWh could even be infinite if the two Vogtle reactors never generate a single megawatt (which is a possibility).

2) The $2 billion is a number from "thedonster," not from me. The $2 billion number is based on a value of $1 billion for a "nuclear power plant" (not a "nuclear reactor") from "enoch arden." Neither "thedonster" nor "enoch arden" have ever presented any evidence that they have any knowledge of nuclear power, including the economics of nuclear power. In fact, "enoch arden" has demonstrated repeatedly that he is clueless.

3) So what would someone who actually knows something about nuclear power estimate the decommissioning cost to be for two nuclear reactors of the size of the two Vogtle reactors (roughly 1120 MW each) in the year 2020?

Well, here is a website that has the following:

https://www.world-nuclear.o...

An OECD Nuclear Energy Agency survey published in 2016 reported US dollar (2013) costs in response to a wide survey. For US reactors the expected total decommissioning costs range from $544 to $821 million; for units over 1100 MWe the costs ranged from $0.46 to $0.73 million per MWe, for units half that size, costs ranged from $1.07 to $1.22 million per MWe.

If the cost for units over 1100 MWe is $0.46 to $0.73 million per MWe (in 2013 dollars), the decommissioning cost for each Vogtle reactor would range from $515 million to $818 million. That's in 2013 dollars. Using the consumer price index (CPI) to adjust for inflation (not valid, but convenient ;-))...the cost in December 2019 would increase to $575 million to $913 million per reactor. So for two reactors, the December 2019 cost would be approximately $1.2 billion to $1.8 billion.

4) So now we just compare the $1.2 billion to $1.8 billion to the current estimated cost of $27 billion, to come up with 4.4% to 6.7% of the LCOE will be from decommissioning, right? No, that's wrong. The decommissioning probably won't come until after many years of operation. For example, the average reactor in the U.S. is currently about 38 years old. The present value of a future cost is much less, because money can be placed in escrow, earning interest, to pay for the future costs.

5) For example, let's escalate the cost of decommissioning the two Vogtle reactors 50 years into the future, based on the CPI of the last 50 years. (Again, that's not valid, but it's convenient.):

https://www.bls.gov/data/in...

Therefore, the cost of decommissioning the two reactors combined increases from $1.2 billion to $1.8 billion in December 2019 dollars to $8.2 billion to $12.3 billion in 2070.

6) How much would have to be set aside in December 2019 to have $8.2 to $12.3 billion in 2070? Let's assume we invest in the S&P 500 (and re-invest dividends) and the returns of the next 50 years are like the last 50 years:

https://dqydj.com/sp-500-re...

The returns, not adjusted for inflation, from December 1969 to December 2019 are 14550%. In other words, $1.2 billion invested in the SP 500, with dividend re-investment, in 1969 would have produced $175 billion in 2019. So we only need $8.2 billion (in year 2070 dollars), but we have $175 billion (in year 2070. So to get a fund of $8.2 billion to $12.3 billion in 2070, if the SP 500 returns continue for the next 50 years like the past 50, we'd only have to invest $56 million to $85 million in 2019.

7) Of course, all these numbers are simply illustrative, based on data from the last 50 years. But it is important to note that nuclear power plants typically obtain money for decommissioning by charging 0.1 to 0.2 cents per kWh.

Progress on a corollary of "Moore's Law"

"Moore's Law" as originally stated is not nearly as important as a corollary of Moore's Law, which is how many computations per second occur per dollar of computing power. If I go to the Wikipedia article "FLOPS" (floating operations per second), I find data that can be graphed as shown below. Note that the y-axis is logarithmic. That means that a straight line shows the cost in dollars per gigaflop of computer power is going down exponentially. A linear regression line through the data indicates that the cost in dollars per gigaflop has declined (and will decline) by approximately 11 orders of magnitude in the 42 years from 1982 ($100 million per gigaflop) to 2024 ($0.001 per gigaflop). If the human brain is approximately 10 petaflops (10 million gigaflops), then the cost for a computer that can perform the same number of calculations as a human brain in 2024 will be approximately $10,000 (i.e., 10 million gigaflops times $0.001 per g

 

Global warming avoided by hypothetical Gen IV reactors

These are comments I made on the Neurologica blog, regarding the global warming avoided by entirely hypothetical Gen IV reactors potentially being built in the U.S.

Part 1 of 2: I wrote:

Even if--by some frankly unbelievable circumstance--enough Gen IV reactors would be built in the U.S. by 2060 to provide 30% of U.S. electricity, it would have essentially zero effect on global temperatures by the year 2100. (By that, I mean less than 0.01 degree Celsius, though I haven't calculated a particular value.)

Here's a calculation.

1) In 2018, approximately 100 nuclear reactors in the U.S. provided approximately 20% of the electrical energy generated. For simplicity, let's say that approximately 150 Gen IV reactors would produce approximately 30 percent of the electricity generated in 2060. (This essentially assumes that electrical generation in 2060 is the same in as in 2018, and average reactor sizes and capacity factors are also the same.)

2) Before we calculate the global warming effects, let's look at the simple arithmetic of years and number of reactors. No commercial Gen IV reactor is going to go online before 2030. To get 150 Gen IV reactors by 2050, that would be taking online 7.5 Gen IV reactors a year, every year, for 20 years. To get the 150 Gen IV reactors by 2060, it would be 5 Gen IV reactors a year, every year, for 30 years. There is essentially no chance of either scenario happening. And I'm happy to put my money where my mouth is. Steven Novella, I will be happy to bet you $10, and give you 100-to-1 odds, that 150 Gen IV reactors won't be operating in the U.S. by 2050, and happy to bet you another $10, and give you 50-to-1 odds, that 150 Gen IV reactors won't even be operating in the U.S. by 2060. So if 150 Gen IV reactors are operating by 2050, I would pay you $1500 (because I would lose both bets). And if 150 Gen IV reactors aren't operating by 2050, you'd pay me $10, and an additional $10 if 150 Gen IV reactors still aren't operating by 2060.

3) For global warming effects, let's assume an equilibrium climate sensitivity of 3 degrees Celsius per doubling of CO2. That means increasing from 400 ppm to 800 ppm would eventually result in 3 degrees Celsius of temperature change. Each ppm of CO2 is approximately 7.8 gigatonnes of CO2. So approximately 3120 gigatonnes of CO2 results in 3 degrees of temperature change. So one gigatonne of CO2 results in 3/3120 = approximately 0.001 degrees Celsius temperature rise.

3a) An analysis of the warming in the RCP 8.5 scenario yields the following results by comparison: From Wikipedia, median warming in the RCP 8.5 scenario is 3.7 degrees Celsius over the 95 years from 1996 to 2091. Emissions during the 21st century in that scenario are 1932 GtC...or 7084 gigatonnes of CO2. In 95 years, that would be 1835 GtC, or 6728 gigatonnes CO2. So one gigatonne of CO2 results in 3.7/6728 = approximately 0.00055 degrees Celsius temperature rise.

Part 2 of 2

4) Let's further assume that the Gen IV reactors displace natural gas and coal in the percentages that existed in 2018. In 2018, 1468 billion kWh of U.S. electricity was from natural gas, resulting in 581 million metric tons of CO2 emissions, and 1146 billion kWh was from coal, resulting in 1150 million metric tons of CO2 emissions. So a total of 2614 billion kWh resulting in a total of 1731 million metric tons (1.731 gigatonnes of CO2 emissions).

5) Nuclear generated 807 billion kWh in 2018 (20% of U.S. total). We're saying that 30% of electricity in 2050 is from Gen IV reactors, so that's 1210 billion kWh. That means that 1210/2614 x 1.731 = 0.80 gigatonnes of CO2 will be avoided each year by the (entirely hypothetical) Gen IV reactors. That will result in 0.80 x 0.001 = 0.0008 degrees Celsius of temperature increase avoided each year, using the equilibrium climate sensitivity method, and 0.80 x 0.00055 = 0.00044 degrees Celsius of temperature increase avoided each year, based on the RCP 8.5 projections.

7) Using the equilibrium climate sensitivity method: From 2060 to 2100, the avoided temperature increase from approximately 150 Gen IV reactors in the U.S. generating approximately 30 percent of U.S. electricity and entirely displacing coal and natural gas would be 0.0008 deg C per year times 40 years = 0.032 degrees Celsius of global warming avoided. There's also the amount avoided in the buildup from 0 to 30% of electricity from 2030 to 2060...that's equivalent to a total of 0.012 degrees Celsius of global warming avoided.
So I wrote that less than 0.01 deg Celsius temperature avoided, but it's theoretically more like 0.044 degrees Celsius temperature avoided, using the equilibrium climate sensitivity method. Using the RCP 8.5 scenario results method, the avoided warming is more like 0.024 degrees Celsius.

Conclusion: The point still stands...even building 150 Gen IV reactors in the U.S. by 2060 will have a negligible effect on global temperature in 2100...somewhere between 0.024 degrees Celsius and 0.044 degrees Celsius of temperature increase avoided, which is far less than the year-to-year variation in average global temperature.

Let's see who knows what they're talking about, part deux