RSS feed for entries

Next post: »

Less heat, more light: solving the energy crisis

The planet’s big problems relate to energy. Using it is warming the planet. Getting it is causing wars. Running out of it will end in poverty and famine. Now that we’re getting closer to the apocalypse, it looks like the four horsemen are all riding one cloned horse.

The first task is to figure out the scope of the problem. How much energy do we use? How much will we need in, say, 2050? (That’s a favorite year for projections: it’s nice and round, and within many current lifetimes, but not so close that there’s no hope.) The next task is to consider which types of energy could supply the needs. And the final task would be to go out and do it. The system breaks down at that crucial point. There are solutions to the energy crisis. We’ll soon find out if there’s a solution to the “people’s heads in a place where the sun don’t shine” crisis.

(Fair warning: this is another one of my interminable posts…)

Current estimates of global energy use vary, but cluster around 14 terawatts (e.g. MIT Energy Research Council, 2006 (pdf), and related “easier” refs: 1 and 2). Projected needs to 2050 vary, but under a “business as usual” scenario, growth in energy requirements is expected to be around 2% per year, which means a doubling of energy required in 35 years. Within that overall doubling, electricity use is expected to triple.  A 2% growth rate is supposed to provide some improvement in developing countries, allow for expected population growth, and maintain standards in the developed world (IEA, World Energy Outlook 2004 (pdf), and OECD, “Energy: The Next Fifty Years” (pdf), Belcher & Nocera, MIT lecture, and one of the linchpin articles: Hoffert et al., 1998, in Nature as well as a popular take on it.). So we want some 28 terawatts, although some estimates go as high as 60 terawatts if all poorer countries are to have the same standard of living as the richest.

Nuclear energy

Nuclear energy advocates tend to argue for higher estimates of future need, because they seem to feel that a big energy deficit makes their case for a big solution, like nukes. The idea seems to be that the limp-wristed brigade doesn’t understand the scope of the problem and pretends wimpy stuff like solar power could make a dent in it, when the only real solution is man-sized power plants putting out fry-your-oysters BTUs. (Sneering? Me? Never!)

The irony of big estimates to prove the need for big solutions is that nukes haven’t got a prayer of solving more than a tiny fraction of the lowest estimate. Forget the highest. Here, for instance, are the projections from Deutch and Moniz, two MIT professors who are proponents of using nukes to mitigate global warming:

“Reaching a terawatt of nuclear power by 2050 is certainly challenging, requiring deployment of about 2,000 megawatts a month,” they write. “A capital investment of $2 trillion over several decades is called for, and power plant cost reduction, nuclear waste management and a proliferation-resistant international fuel cycle regime must all be addressed aggressively over the next decade or so.”

You noticed that, right? A two gigawatt plant needs to be built every month from here to 2050. That will get us all of one (1!) terawatt out of the fourteen needed. After 2050, we’ll still need to build 2GW a month because nuclear plants have a 30-40 year life span (at best). Old plants will also have to be decommissioned (mega$$$). Also notice the little phrases about waste management and nuclear proliferation. Neither of these problems has even begun to be solved for current levels of waste and fuel, let alone all the hot stuff generated by the new 2GW/month plants. Both waste and proliferation have vast implications for the livability of the planet, now and thousands of years into the future.

Nor have the professors even mentioned that uranium is a finite resource, which would start running out in a few decades under that sort of depletion. The numbers that follow are from a lecture by Buonassisi at MIT. (The link leads to all the original references. That link is unfortunately no longer available.) There are 2,300,000 tons proven uranium reserves. 1TW-hr requires 22 tons of uranium. A year’s worth of power requires 192,720 tons. 2,3 million tons would last 11.93 years, at 1 TW per year. Even worse, the arithmetic shows that uranium would be depleted during the 30-year ramp-up period.  We’d never get our 11.93 years of going full bore. (Edited to add, 2014-03-23: Those are simple assessments based on geology, physics, and estimated reserves, so other references saying much the same thing are plentiful. Just a couple: Perez and Perez, 2009 (pdf). Their citations include a link to a useful slideshow, esp. slides 25-28. Gabriel et al. 2013 noting the same problem, but concluding that breeder reactors are the solution.)

You can quibble about the numbers, but even if you say there’s 6 million tons of uranium instead of 2.3, you’ve still only bought an extra couple of decades. An extra couple of decades at all of one (1) terawatt a year.  There’s lots more U in sea water, but if you think we should try the environmental disaster of mining seawater — to get 1TW of radioactive energy — you probably got that idea via the fillings in your teeth.

So, could we put the nuclear fantasy to rest, once and for all? Nukes look big, but in reality they’re much too small to deliver the energy needed. They don’t emit greenhouse gases, but that’s where the good news ends. They run on a non-renewable resource. They can’t be built fast enough. The produce radioactive pollution and waste. They facilitate nuclear weapons production. They are hideously expensive at every step and get more expensive with time even after they produce NO energy. Using something with so many side-effects to reduce CO2 is like that old joke about how the operation was a complete success, but the patient died.

Alternative energy

Now for some more bad news. There are almost no sources of energy that can even theoretically meet a requirement for fourteen new terawatts of energy. Nocera, at MIT:

If you gave over every square inch of cropland on the face of the earth to biomass production, you’d only get 7 additional terawatts. Plus, “you couldn’t eat anymore.” You’d still need to add 8,000 nuclear power plants, by building a new plant every 1.6 days for the next 45 years; put wind turbines everywhere; and dam every available river, to approach the 28 terawatt goal.

None of these can be the whole solution, even working all together, even in theory. I think that’s a very important point that’s too often overlooked. What is the point of pouring resources into something that cannot be the solution, no matter what you do?  You can’t fight the laws of physics.  There’s no point even trying.

One energy source Nocera omits is tidal power.  Ocean power is huge, but at about 3 practically recoverable terawatts it is still only a partial solution. (The link is to a 17M powerpoint of a lecture by Lewis at CalTech.)

However, even if most types of clean energy can’t be the whole solution, they can be useful as partial solutions or short term stopgaps.  But then we should include them in our plans as such. (That’s if we had any plans, instead of just hoping for the best.) Biofuels, for instance, have a role to play as a transitional energy source, but they’re not worth huge environmental and human costs, especially when they’re not more than a small part of the cure. The same goes for hydropower, geothermal, and wind.

I am not — not — saying there’s no place for clean, sustainable energy. I am saying that the resources we pour into them need to be proportional to the payoff we can expect from them.

Solar energy

Which brings me to the one known energy source that, theoretically, can provide orders of magnitude more power than even the wildest estimates of need. The sun. That way lies a real solution.  So that’s the direction both for long term planning and for maximum allocation of resources.

Let’s put some numbers on solar energy. On average, the sun hits the earth’s surface with some 6 kilowatt-hours per square meter per day. That includes diurnal factors (e.g. that there’s no sun at night) and atmospheric factors other than clouds. (The energy hitting the top of the atmosphere is more than five times that, about 32 kWh/m2/day, which gives you some indication of just how effective the atmosphere is at screening us.) [Corrected based on Tom’s comment below.]

Six kWh doesn’t sound like much. Furthermore, that energy can’t currently be collected with good efficiency. Let’s say we only get 10%. (All the numbers are conservative. In the desert Southwest, for instance, 9kWh/m2/day is common, and commercially available photovoltaic modules now typically have 11%-13% efficiency.)

In the late 1990s, cities and towns occupied some 32 million hectares of land in the US. (USGS pdf, p. 47) (A hectare is 100 x 100 meters, or 10,000 square meters.) And let’s say we could put photovoltaics on only 5% of that area, or 1,600,000 hectares.

0.6 kWh/m2/day x 10000 = 6000 kWh per hectare or …
6 megawatt-hours per day.
6MWh x 1,600,000 hectares = 9,600,000 megawatts or 9.6 terawatts.

Six kW doesn’t sound so small any more. Roof-top solar panels everywhere (not just on suburban roofs (17M ppt)) could provide multiple terawatts of energy given current technology. This does not assume huge solar farms covering the desert Southwest. Just rooftop solar everywhere. 9.6 TERAWATTS.

Another favorite graphic shows the land area that would be needed in the world’s deserts to supply all our energy:

six pinhead-sized dots in the US SW, Chilean desert, Sahara, Saudi peninsula, Gobi, and Australia could produce some 18TW of energy

The land area is just the bits under the six little dots.  It makes for a nice graphic, but I’m not a big fan of this option.  It requires destroying sensitive habitat and construction and maintenance in remote locations under difficult circumstances. City-based solar looks like a much more elegant, local solution. However, this does show what’s possible.

Here again, you can quibble about the numbers, but you can’t get away from their general significance. Dozens, hundreds, of terawatts are hitting land every day.

Transitioning: technical issues

There is nothing impossible about making the transition from fossil fuel to solar energy. Commercial, industrial-scale technology is available now that only requires government action to be ramped up to the required levels. The action in question is simply an end to the subsidies for fossil and nuclear energy, as well as accounting rules that make sure all the costs are part of the price. (What economists call “including the externalities.”) Actual subsidies for solar would be icing on the cake.

The biggest difficulty seems to be mental. We’re used to thinking that the denser the energy, the better. We want the maximum energy per cubic inch, we want jewelry boxes of antimatter to power whole cities. Solar energy is the opposite. It has to be diffuse, and it requires widespread energy collection and distribution. It’s democratic, open source energy, which seems to be a very difficult concept for the top-down crowd.

There are three technical issues: the photovoltaic cells themselves, the need for excellent energy storage due to the fluctuating power source, and improved distribution of the electricity with minimal loss.

Solar cells have had cost and efficiency problems which they’re overcoming, and still have pollution problems during manufacture.   However, the toxins are produced at controllable point sources, and can be completely recycled out of old solar cells. There are several different kinds of photovoltaic cells based on different combinations of elements, so shortages of the elements involved do not have to be a problem. And some of the newest cells don’t use rare earths at all.

Improvements in solar cell production are everywhere these days: improving efficiency over 30% by making more complex cells, or by using nanowires, or fresnel lens coatings. Thin film arrays can be rolled out in the size and shape needed, and are being researched as well as produced by Nanosolar, and Ovonics (named for Ovshinsky, one of the earliest pioneers in the field).

Storage technology is an essential component of the current grid, but will become even more important in a solar system.  Load levelling and power wheeling will be huge factors. This doesn’t just mean bigger and better batteries, although there have been interesting developments in that technology recently. Ways of load levelling that scale down are needed to reduce the transmission of huge amounts of power at night. Flywheels are a mechanical way to store energy that look better than batteries for some applications, possibly including cars.  And speaking of cars, one that runs on compressed air has just been in the news.  The stored air in that case directly runs pistons, but compressed air can also turn a generator.

Energy storage in the broadest sense includes any fuel. Oil, in that sense, is a way of storing energy till you need it. Hydrogen is a way of storing energy. Fuel cells and photoelectrolysis generally, which can produce hydrogen cleanly from water, are a particular form of energy storage. Advances in that field are happening rapidly, too. For instance, photoelectrolysing thin films of nanowires have been made without expensive trace elements.  Even more elegant: synthetic photosynthesis that allows hydrogen forming straight from water. That’s at the proof of concept stage and has very low efficiencies, but so did regular photovoltaic cells early on.  And last, a process that doesn’t use solar energy directly: microbes can give off hydrogen in the process of digesting all sorts of things, but in the past people have been bored by the low amounts produced.  It turns out that if some electricity is added to the mix, a lot more hydrogen can be recovered. In the case of cellulose (which is most plant waste) the friendly microbes produce hydrogen with 63% efficiency.  That is, to put it mildly, phenomenal.

Hydrogen can also be produced very uncleanly from any hydrocarbon, like coal. That, naturally, is the kind of hydrogen being pushed by Bush. Hydrogen, in an of itself, is not necessarily a good thing, but as part of a clean, sustainable energy industry, it could have it’s place.

New storage technology can mean clever use of existing resources, not just inventing or building new storage. Vehicle-to-grid is one such elegant concept. Electric cars have to store energy, but most cars spend most of their time parked. The storage or generating capacity of cars, depending on the type of electric vehicle, could play a part in load leveling. (pdf of slide show by Kempton, one of the pioneers). The utilities wouldn’t have to build huge extra storage capacity, and consumers would receive money as well as pay for electricity. There are problems, of course. Current car batteries aren’t up to the task, and the grid would need work. Neither of those is impossible.  They’re tedious and they require many different efforts to all coordinate, something which is possible for those who have a functioning government.

The electrical grid also needs some changes because the greater need to wheel power means that lines would be carrying more load than they do now. This, too, is tedious, requires coordination, but is not impossible.

One futuristic caveat about all modern power systems is that they’re vulnerable to hacking as might be expected given how complex they are.  A solar system, being distributed, might or might not be less ticklish. Given that large parts of it would be in the homes of regular folks, good, user-friendly security needs to be built in up front, not as an afterthought.

Transitioning: regulatory issues

So, given that solar power is the obvious choice and given that it’s feasible with current technology, why hasn’t it already happened?

What’s missing is a functioning government. Rebuilding the multi-trillion dollar energy industry, and rebuilding it in ways that don’t profit existing interests, absolutely requires government coordination, funding, and regulation. I want to repeat what I said earlier: the main government action needed is an end to the subsidies for fossil and nuclear energy, as well as accounting rules that make sure all the costs are part of the price. (What economists call “including the externalities.”) Actual subsidies for solar would be icing on the cake.

Fossil fuel subsidies are non-trivial.  Globally, direct subsidies are around $210 billion, $235 billion (pdf), or more per year.  That doesn’t include subtle subsidies like foreign “aid” to build power plants run by oil companies, research on oil and gas extraction, and the like. (Nor does it include oil wars.)  In 2001, US taxpayers were estimated to pay some $257 billion per year in auto subsidies alone. In 2003, the estimated cost of the natural disasters that year linked to global warming was around $60 billion (pdf).  Those costs are generally borne by the poorest, are a form of subsidy since the people doing the polluting don’t have to pay for it, and aren’t included either.

And then there are the nuclear subsidies.  Twenty billion in subsidies in 2005. Koplow, the grand old man of government energy support estimates, calculates that nuclear receives between 4-8 cents per kWh in subsidies in the US alone (not including the one extorted from our grandkids who’ll be paying for dealing with the waste). Compare that to the average, you-see-it-on-your-bill residential cost for electricity of 10c/kWh. Just the subsidy is more than half that. It adds up to $30-60 billion per year, since nukes are said to produce about 85.6GW per year.

It’s hard to get your head around how much we’re paying to destroy the planet.

There are important ways in which a transition to solar needs active help as much as it needs a level playing field. For instance, most people can’t afford solar panels. Instead of subsidizing industry to use fossil and nuclear fuel, the government would have to subsidize people to install photovoltaics. Instead of a portion of the utility bill going to build new plants, it would go to build new PV panels. (This, by the way, is another area where the US has squandered its leadership. In 1975, it had 75% of the then tiny industry. In 2005, it was 10%, and Japan and Germany have 75%.) Instead of utility companies receiving a return on their investment based on the amount they spend on plant and equipment, they should be paid based on the amount they spend in grants to install solar and efficiency improvements and on the energy they store and distribute.  (Some states have started down that path already.)

Paying utilities for efficiency improvements and distributed generation would, by itself, do away with the vested interests that are continually pushing for huge centralized plants controlled by utility companies. That would cease to be how utilities made their money. Paying them to build plants is a regulation. It’s not some inescapable market force. It could be changed by … well … regulation.

Globally, the available energy from the sun is a whole lot more than the terawatts from US towns. If the whole world used roof-top solar everywhere, there’d be enough excess energy floating around to launch interstellar rockets. Seriously, according to those who know, about 200TW are practical to recover (pdf) on the Earth as a whole. Others say 600TW (17M powerpoint). Maybe, if we stopped wasting energy as heat, i.e. got serious about efficiency, we could even make that do for a while.

Transitioning: using efficiency

And that brings me to the topic of wasting energy. Or, rather, of not wasting it. Energy efficiency is the second largest “source” of energy after solar.

Let’s start with what we know.  Historically, energy efficiency has improved about 1% per year without any concerted action.  When there has been such action, it improved around 4% in the US after the 1970s oil shocks. None of this even takes into account creative uses of existing energy like cogeneration, regenerative braking, or minor but fun things like piezoelectric clothing, window units that generate current from breezes, and, of course, the essential coffee-based Stirling engine.  (I’m not totally joking about that last one.  Heat engines can capture significant amounts of energy.)

Figuring on an overall 2% gain in efficiency per year is not unreasonable, given historical data.  No doubt, much more could be achieved, since transportation could use over 30% less energy and buildings and lighting could use over 60% less energy than they do now.  It does take years to implement efficiency improvements, and not all old stock can be replaced or retrofitted.  So let’s stay with 2% improvement.  According to Muller, citing a 2001 paper by Rosenfeld et al.:

Two percent compounded over 100 years reduces energy use by a factor of 7.2. By 2100, with a world population of 10 billion people, everyone can be living at the current European standard of living and yet expending half the energy we are using today. [current use is around 14TW]

A factor of 7.2 doesn’t mean 7.2%.  It means 720%.  Without taking efficiency improvements into account, predicted energy needs are supposed to be at least 50TW in 2100.  A 2% rate of efficency improvements would bring that down to around 7TW.  That boggles my mind.  If it doesn’t boggle yours, you have a sturdier mind than I do.  2150 2050, the year I’ve been using, happens to be the bulge in energy needs, but even then efficiency would reduce the projected total need from some 28TW to 21. That’s 7TW “produced” by not wasting energy.

That’s a huge difference. Plus, these terawatts don’t pollute, they raise property values instead of destroying them, they increase local business opportunities, and they cost a fraction of the trillions of dollars that big-boy energy gobbles up while producing much less actual energy.  Not expending every effort to get maximum efficiency is about as stupid as ignoring solar.  Interestingly, it’s another one of these distributed, democratic sources some people just don’t get.


The big stumbling block to achieving a solar, energy efficient future is supposed to be how much it would cost.  Numbers ending in -illions make it all seem so hopeless.  The problem with absolute numbers is that we have no sense of what they mean.  All we know is that trillions is more money than any ordinary human being is ever going to see.

Projections of the cost of implementing the Kyoto Agreement in the near term (when costs are highest) are around 1% of GDP.  (Of course, the number of variables is large and estimates vary.)  The cost estimate does not include the benefits to GDP from local business opportunities, improved property values, reduced need for environmental mitigation, and so on. That’s been estimated to increase GDP by as much as several percent in developing countries, which need it the most. So what is the global GDP? Around $48,244.900,000,000 according to the World Bank. One percent is almost five hundred billion dollars per year.  It sounds enormous.

Let’s look at it on an individual level.  US GDP is over $13 trillion. On a per capita basis, that’s nearly $45,000. One percent of that is $450. So, when we say that switching over to clean energy is unaffordable, we’re saying that a single guy making $45,000 or a two-earner couple making $90,000 couldn’t afford an iPhone. 

That is idiotic. That is criminally idiotic. 

Admittedly, meeting Kyoto targets is only the beginning. We’ll have to do a lot more than that to avoid environmental disasters. Just for grins, multiply the Kyoto projection times four. Now the couple making $90,000 has to buy a fancy widescreen digital TV.

It would be equally idiotic to pretend they couldn’t do that. Especially when, unlike a TV, it gets more expensive on an exponential curve with every passing year. And especially when, unlike a TV, the price of not buying it is losing our world.

So, the take home message is:

We should look for energy where it exists, and not where it doesn’t.  We should take the long-term cheapest alternative, not the most expensive. The only source that fits the bill is the sun, with a big assist from energy efficiency, and contributions from other clean sources.

And that sounds a lot like what the dirty, effing, hippies having been saying for decades. If we could get over the fact that they were right (again!), and if we could get over how much time we’ve wasted and how big we’ve let the problem become, we could start the Solar Manhattan Project that we now need.

Crossposted to Shakesville

Technorati tags: alternative energy, solar, photovoltaic, energy efficiency, conservation, nuclear energy,