Google.org (not the Google.com everyone is familiar with) just released a plan "to wean the U.S. off of coal and oil for electricity generation by 2030." I have to say I'm a bit disappointed with their plan because even though nuclear power is somewhat a part of it, Google doesn't give nuclear power much consideration.What's their plan?
In order to achieve much of their goal, Google.org calculates that the U.S. should build 380 gigawatts (GW) of wind, 250 GW of solar and 80 GW of geothermal by 2030. This is a total of 710 GW of renewable capacity which would generate 56% of the country's electricity. As well, efficiency is projected to reduce electricity demand 33% by 2030 while plug-in vehicles increase demand by 8%.
Their plan will supposedly cost a total of $4.4 trillion while saving $5.4 trillion through efficiency gains and avoided fossil-fuel use. The efficiency and renewable capacities alone are estimated to cost nearly $2.5 trillion.
What do the numbers look like for nuclear plants?
Based on some simple arithmetic, nuclear power plants could achieve the 56% fuel share at a cheaper cost and half the capacity than Google's proposed plan.
According to EIA's latest Annual Energy Outlook (Google bases their plan on this report), US electricity generation is projected to reach about 5,000 billion kWh by 2030 (2007 was about 4,000 bkWh). If nuclear plants were to provide 56% of the country's electricity by then, they would have to produce 2,800 bkWh.
The average nuclear plant operating 90% of the time produces about 7.9 bkWh in a year (equation: 1 gigawatt of capacity times 90% capacity factor times 8,760 hours in a year). Therefore, about 350 GW of nuclear capacity would be needed by 2030 to provide 56% of the electricity. That's less than half the capacity needed compared to Google's plan!
Not only that, we really only need to build 250 GW of nuclear capacity because the nuclear plants operating today already provide 100 GW of capacity. As well, today's operating nuclear plants aren't projected to begin retiring until 2029.
What's the cost for all of those nuclear plants?
Let's assume a one GW nuclear plant costs $8 billion (just to be on the conservative side). Building 250 GW would therefore cost $2 trillion. That's half a trillion dollars less than the estimated costs for efficiency and renewables in Google's plan.
Let's be realistic.
Okay. 250 GW of new nuclear capacity most likely couldn't happen by 2030 because our infrastructure isn't quite ready for such a large build-out. But as the Wall Street Journal points out, the same is said for wind, solar and geothermal.
I could appreciate the level of effort the Google authors must have spent putting this together. But I'm disappointed that they overlooked one of the biggest technologies that could help achieve their goal. My numbers above are simple and compelling, and Google should take another look at what nuclear energy has to offer. They could start by googling "nuclear energy." :-)
Comments
I noticed that Eric made the same mistake that many renewable advocates make in thinking that if enough renewable sources were amassed together paired with a smart grid that it would solve the problem.
He continues to hammer in his point about wind energy by saying "[the wind never stops]"..."right, right, right [...] you see my point", as if that is the end of that argument. I didn't hear him mention anything about wind capacity factors.
I also did the back-of-the-napkin math to figure out how many new reactors would be needed to fit the bill with a few different scenarios. Given the expertise, planning and licensing required of nuclear, I think it would be wise for congress to pass a package to put a new fleet of generation 3+ reactors to an online fast track.
This need not put a lot of taxpayer money at risk if done correctly. If indeed new nuclear can produce electricity cost effectively (and it can), there should be no good reason why energy companies would not want to pursue it with private investor funds.
What would you include in the infrastructure needed?
Has given us the dot-com boom and bust
Produces and uses equipment ("hardware") with relatively short useful life spans, most of which is being "recycled" in China under horrific working conditions, exposing its workers (mostly women) and the local population to toxic and carcinogenic materials and heavy metals.
Has an environmental record that is both irresponsible and disturbing, resulting in 29 Superfund sites in Silicon Valley alone, the highest concentration of Superfund sites in the country. Most (19) of these sites were the result of manufacturing computer chips for high-tech companies.
So when a guy with a PhD in computer technology says that energy is "just a math problem" (in a presentation with the latest state-of-the-are computer graphics, of course), I start to get skeptical.
To tell you the truth, I don't think I want these guys getting into the energy game. I'd hate to think about the damage that they could do if they ever got into the business of manufacturing something larger than a computer chip and "playing" with something that lasts longer than the source code of the latest "killer app."
In a talk about powering the grid, Dr. Schmidt of Google says, "it sure sounds to me like personal computers." That's exactly what I'm afraid of.
Before you start feeling all bitter about Google-plan, don't jump the gun and start condescending on computer scientists.
I am a computer scientist, and worse, I work in computer graphics.
That doesn't mean I am uncapable of the basic math that needs to be done for discussing the future of energy sector.
And in fact, I do like nuclear power. Most people versed in science and engineering like nuclear power, because they think.
If some people have concerns, it is not because they are some geeky-flashkids.
Most computer scientists admire science fiction. They dream about starships powered by nuclear fusion. They do not belong to the typical luddite lumberjack opposition that nuclear faces.
If you see a computer scientist talking about something, pay attention.
Oh and the energy efficiency part of the plan would be a total failure as well.
Dr. Jeffery Greenblatt, Ph.D. joined Google.org in March 2008 as Climate and Energy Technology Manager. He reviews renewable energy proposals for grants and/or investment for the REC (Renewable Energy Cheaper Than Coal) initiative. He also advises the team on climate change science, energy use and greenhouse gas emissions forecasts, and a broad range of climate mitigation strategies. He is currently focused on the Clean Energy 2030 proposal and ways for Google to help create a 21st century electricity grid.
Before coming to Google, Dr. Greenblatt was High Meadows scientist at Environmental Defense Fund, where he evaluated the technical, economic and environmental aspects of a wide range of energy technologies. He developed "wedge" climate stabilization scenarios for California, the Midwest, and the US, and he was also the technical lead editor on "Earth: The Sequel," a book about the emerging clean energy field, by Fred Krupp and Miriam Horn.
He received his training in climate and energy at Princeton University, where he was on the research staff for four years, working on many projects including ocean carbon cycle modeling, the economics of wind energy and energy storage, and the development of the "wedge" climate stabilization concept with professors Robert Socolow and Stephen Pacala.
He received his Ph.D. in chemistry from the University of California, Berkeley in 1999.
Enough said!
I'm not criticizing Computer Science or computer scientists. Most people who know me know that I'm quite the computer geek myself, and I've contributed my share of spare time over the years helping to develop the Free Software that Google uses for its core business.
I've been following computer technology, the computer industry, and the trends of both for well over two decades now, which is exactly why I am skeptical.
The electronics and computer industry has done great things and have really made an impact on all of our lives (for example, providing the opportunity to discuss this online), but it also known for exaggerating, sensationalizing, and over-hyping "nifty" ideas. A few examples include "the paperless office," virtual reality, and thin clients (a cheap, energy-efficient alternative to the traditional PC that was heavily promoted by Larry Ellison, Oracle's CEO, about 10 years ago -- how many of you are reading this on an efficient "thin client"?).
No, Charles has nailed it on the head. Schmidt is taking his queue from Amory Lovins et al. It's pretty obvious from his presentation in San Francisco, which Jason links to (thanks, Jason).
In answering the very first question after his presentation, Dr. Schmidt explains that, in Google's "distributed" energy plan, there will be "a relatively large number of gas turbines," but also "a pretty good mixture of smaller sources of generation, of which solar thermal will be one." Lots of natural gas ... hmm ... sound familiar?
He goes on to say that "since we're not going to solve the grid problem in the next year or two, we're going to be forced to having more of these [natural gas turbines] than we'll like."
In his answer to the next question, he mentions a 2004 report by the Rocky Mountain Institute.
Good research, Charles.