I think that a rapid and unplanned 90% drop in energy use would be a fair indicator for societal collapse.

That said, we are so wasteful, so unbelievably profligate in our energy use that I think the 60+% Smil’s talking about is a very easy reach given 20 years and smart policies. I think we could do that with only modest changes to our way of life in much of the country (though profound impacts for the most wasteful minority).

But I think that really leaves the door still open for changes in the way we live. I’m not at all convinced that 1950s suburban living and credit-based consumerism has much shelf-life left in it for all sorts of reasons, many of which have nothing directly to do with energy or the environment.

I think many energy efficiency proposals are improved means to unimproved ends. I think we can improve the ends, too. I think we can do it straight up, competing on quality of life and economic returns.

Finally, it’s worth remembering the extent to which our expectations in this whole set of debates have been strongly shaped by those trying to sell us energy, products and lifestyles. Comparing the US/Canada debates even to the debates in the UK and Germany (never mind Scandinavia, NL, Switzerland) is extremely informative. We still tend to define realism in the US by what coal, oil, auto and suburban real estate developers and their PR agencies have told us is realistic.

Now, is 90-95% in 20 years easy? No. Is it possible? Yes, clearly, on technical terms. Is it realistic? Well, no, not unless we change the game, but I think the game is changing without us choosing anyways.

Anyways, provocative post. Thanks.

I agree completely that a whole set of ideas, technologies, and institutions from the past make it difficult to know what will be possible in the future. That’s part of my point, that we need visions for the future that aren’t based purely on historical data or some argument submerged in a spreadsheet.

To be a uniter here, I’d say that just having the argument about where we should be aiming is progress. Smil says the number 110 is GJ/person. Is that fixed in perpetuity? Nah. I think 110 is just a nice stand-in for the current amount of energy usage that the present inefficient system needs to deliver performance on certain important metrics. I think that we both believe technology and systems design can improve upon that and push the number down. How far? I’m not sure.

On the other hand, I (obviously) think that history has lessons about energy use. The U.S. and Great Britain haven’t used 110 gigajoules a person for about one hundred years. Granted, back then, we were burning a lot more coal for each unit of electrical energy, etc. But even in the late 1940s — as power plant efficiencies were plateauing — we were already far beyond that threshold. It might be possible to reduce U.S. energy consumption from 300+ to 30 over the next 100 years, but I’d be willing to bet Ray Kurzweil that it isn’t happening between now and 2050 without collapse.

The basic transformations of energy are all at least 100 years old. The idea that useful scientific knowledge is expanding exponentially just isn’t true in energy. We’re learning a lot, but our knowledge of the world’s basic physics hasn’t changed nearly as radically over the last fifty years as it did in the late 19th century.

While I spent some time attacking the energy forecasting methodology of the “miss highers,” all kinds of energy advocates have missed high with their specific technologies. Nuclear, natural gas, solar, wind, and efficiency have all had their speed of diffusion overestimated over the last six or seven decades. These energy transitions just don’t move quickly, even if they don’t move linearly. Even the fastest introductions of energy sources, ones that were convenient and cheap and heavily pushed by the richest companies in the world — oil and nuclear power — took decades upon decades. You can bet against that trend, but I find it hard to.

Also, re: your point that “unless you assume we never get better at technology, efficiency, design, or culture of living.” My assumption is that such advances bring their own problems and that a design or efficiency advance can sometimes open up other energy intensive options or living styles. Technology — or more precisely, the people who make and use technology — aren’t relentlessly pushing in one direction. Even as exurbs are abandoned and dense living increases, other trends militate against locking in systemic reductions in American and world energy usage. The culture of energy, muscle cars and Christmas trees, is still alive and well. The technological sublime is still a major part of our experience of the world. Unless energy gets ridiculously scarce and expensive, I don’t think you can expect all Americans to start optimizing every decision for energy usage. Do you? I know I don’t and I don’t think I’d really want to. The same goes for companies and other institutions.

That said, I’m also excited by the 100 GJ vision. Putting in renewable and clean generation becomes a hell of a lot easier when you only have to come up with 1/3 of the energy. That usage level would easily allow the United States to make the necessary carbon cuts to prevent further derangement of the atmosphere and stop the most destructive kinds of mining and resource exploitation.

We know that a huge variety of energy uses were locked in by history, rather than being inherently “best.” We also know that a great many uses could be traded for other lower-energy-use practices w/o impact on standard of living (e.g., hard to argue that well-designed urbanism is less prosperous than contemporary suburban living). A sweet spot implies that going below that consumption level is bad or impossible: not sure that’s true, or at least not sure it’s true unless you assume we never get better at technology, efficiency, design or the culture of living…

Iow, I’m not sure that we couldn’t set a sweet spot at 10% of current American energy use, or 5%, and be just as right in the near future.

I think that 100 years of a culture that glorified new energy sources (and measured prosperity in part by the amount of stuff and energy used) has left us so unbelievably bad at thinking about energy use that history offers little relevance for determining what’s actually possible. It’s like asking Soviet economists to predict entrepreneurial contributions to Russian GDP in 1990.

So, Kate Galbraith writes that data centers used 1.5% of U.S. electricity in 2006. The U.S. used 3,669,919 gigawatt hours of electricity in that year, according to the EIA. Making the transformations: that’s 55,048 gigawatt hours of electricity used by data centers or 198 million gigajoules. There were about 298 million people in the U.S. that year, so data centers accounted for about 0.66 gigajoules per person. Data centers remain a bit more than a rounding error.

Let’s bump up Vaclav’s numbers and give ourselves an extra 10 gigajoules per person, just to account for growth in datacenters over the next 50 years.

Because we are all so intimately connected with our computers, it’s hard to conceptualize the difference in scale between all of the computers and IT in the world and established, energy-intensive processes like driving to work or building/living in houses. But the gap is huge. In energy terms, computing and data are still just a blip.

As I said in a post at Wired: Googling Is Not the Issue, Dude.