The powerand the glory

A special report on energy June 21st 2008

ENERGY.indd 1ENERGY.indd 1 6/6/08 16:10:566/6/08 16:10:56

The Economist June 21st 2008 A special report on the future of energy 1

The next technology boom may well be based on alternative energy,says Geo�rey Carr. But which sort to back?

America’s Central Intelligence Agency al­lies himself with tree­hugging greens thathis out�t would once have suspected ofsubversion, you know something is up. Yetthat is one tack James Woolsey is trying inorder to reduce his country’s dependenceon imported oil.

The price of natural gas, too, has risen insympathy with oil. That is putting up thecost of electricity. Wind­ and solar­power­ed alternatives no longer look so costly bycomparison. It is true that coal remainscheap, and is the favoured fuel for powerstations in industrialising Asia. But the richworld sees things di�erently.

In theory, there is a long queue of coal­�red power stations waiting to be built inAmerica. But few have been completed inthe past 15 years and many in that queuehave been put on hold or withdrawn, fortwo reasons. First, Americans have be­come intolerant of large, polluting indus­trial plants on their doorsteps. Second,American power companies are fearfulthat they will soon have to pay for one par­ticular pollutant, carbon dioxide, as is start­ing to happen in other parts of the richworld. Having invested heavily in gas­�redstations, only to �nd themselves lockedinto an increasingly expensive fuel, theydo not want to make another mistake.

That has opened up a capacity gap andan opportunity for wind and sunlight. Thefuture price of these resources�zero�isknown. That certainty has economic valueas a hedge, even if the capital cost of wind

The power and the glory

EVERYONE loves a booming market, andmost booms happen on the back of

technological change. The world’s venturecapitalists, having fed on the computingboom of the 1980s, the internet boom ofthe 1990s and the biotech and nanotechboomlets of the early 2000s, are now look­ing around for the next one. They thinkthey have found it: energy.

Many past booms have been energy­fed: coal­�red steam power, oil­�red inter­nal­combustion engines, the rise of elec­tricity, even the mass tourism of the jet era.But the past few decades have been quieton that front. Coal has been cheap. Naturalgas has been cheap. The 1970s aside, oil hasbeen cheap. The one real novelty, nuclearpower, went spectacularly o� the rails. Thepressure to innovate has been minimal.

In the space of a couple of years, all thathas changed. Oil is no longer cheap; in­deed, it has never been more expensive.Moreover, there is growing concern thatthe supply of oil may soon peak as con­sumption continues to grow, known sup­plies run out and new reserves becomeharder to �nd.

The idea of growing what you put inthe tank of your car, rather than sucking itout of a hole in the ground, no longer lookslike economic madness. Nor does the ideaof throwing away the tank and pluggingyour car into an electric socket instead.Much of the world’s oil is in the hands ofgovernments who have little sympathywith the rich West. When a former head of

An audio interview with the author is at

www.economist.com/audiovideo

A list of sources is at

www.economist.com/specialreports

Trade windsWind power has come of age. But to make themost of it, electrical grids will have to beoverhauled. Page 3

Dig deepCarbon storage will be expensive at best. Atworst, it may not work. Page 5

Another Silicon Valley?The rise of solar energy, in one form oranother. Page 6

Beneath your feetGeothermal could be hot. Page 7

Grow your ownThe biofuels of the future will be tailor­made.Page 8

The end of the petrolheadTomorrow’s cars may just plug in. Page 11

Life after deathNuclear power is clean, but can it overcomeits image problem? Page 12

Flights of fancyThe world of energy must change if things areto continue as before. Page 13

Also in this section

1

More coverage of alternative energy andother green issues is at

www.economist.com/greenview

AcknowledgmentsIn addition to those mentioned in the report, the authorwould like to thank the following for their help: DanReicher (Google.org); Greg Stephanopolous, AngelaBelcher and Mujid Kazimi (MIT); Vlatko Vlatkovic, DanielleMer�eld, Juan de Bedout and Kelly Fletcher (GeneralElectric Research Laboratory); Alan Shaw (Codexis); JayKeasling (UC Berkeley); and Joel Velasco (Brazilian SugarCane Industry Association).

2 A special report on the future of energy The Economist June 21st 2008

2

1

and solar power stations is, at the moment,higher than that of coal­�red ones.

The reasons for the boom, then, are tan­gled, and the way they are perceived maychange. Global warming, a long­rangephenomenon, may not be uppermost inpeople’s minds during an economicdownturn. High fuel prices may fall as newsources of supply are exploited to �ll risingdemand from Asia. Security of supply mayimprove if hostile governments are re­placed by friendly ones and sources be­come more diversi�ed. But none of the rea­sons is likely to go away entirely.

Global warming certainly will not.�Peak oil�, if oil means the traditional sortthat comes cheaply out of holes in theground, probably will arrive soon. There isoil aplenty of other sorts (tar sands, lique­�ed coal and so on), so the stu� is unlikelyto run out for a long time yet. But it will getmore expensive to produce, putting a �ooron the price that is way above today’s. Andpolitical risk will always be there�particu­larly for oil, which is so often associatedwith bad government for the simple rea­son that its very presence causes bad go­vernment in states that do not have stronginstitutions to curb their politicians.

A prize beyond the dreams of avariceThe market for energy is huge. At present,the world’s population consumes about 15terawatts of power. (A terawatt is 1,000 gi­gawatts, and a gigawatt is the capacity ofthe largest sort of coal­�red power station.)That translates into a business worth $6trillion a year�about a tenth of the world’seconomic output�according to JohnDoerr, a venture capitalist who is heavilyinvolved in the industry. And by 2050,power consumption is likely to have risento 30 terawatts.

some batty, that is indeed reminiscent ofthe dotcom boom. As happened in thatboom, most of these ideas will come tonaught. But there could just be a PayPal or aGoogle or a Sun among them.

More traditional companies are alsotaking an interest. General Electric (GE), alarge American engineering �rm, alreadyhas a thriving wind­turbine business andis gearing up its solar­energy business. Theenergy researchers at its laboratories inSchenectady, New York, enjoy much of theintellectual freedom associated withstart­up �rms, combined with a securesupply of money.

Meanwhile, BP and Shell, two of theworld’s biggest oil companies, are spon­soring both academic researchers andnew, small �rms with bright ideas, as is Du­Pont, one of the biggest chemical compa­nies. Not everyone has joined in. ExxonMobil, the world’s largest oil company notin government hands, is conspicuously ab­sent. But in many boardrooms renewablesare no longer seen as just a way of keepingenvironmentalists o� companies’ backs.

Some people complain that many exist­ing forms of renewable energy rely on sub­sidies or other forms of special treatmentfor their viability. On the surface, that istrue. Look beneath, though, and the wholeenergy sector is riddled with subsidies,both explicit and hidden, and costs that arenot properly accounted for. Drawing onthe work of people like Boyden Gray, a

Scale is one of the important di�er­ences between the coming energy boom,if it materialises, and its recent predeces­sors�particularly those that relied on in­formation technology, a market measuredin mere hundreds of billions. Another dif­ference is that new information technol­ogies tend to be disruptive, forcing the re­placement of existing equipment, where­as, say, building wind farms does not forcethe closure of coal­�red power stations.

For both of these reasons, any transi­tion from an economy based on fossil fuelsto one based on renewable, alternative,green energy�call it what you will�islikely to be slow, as similar changes havebeen in the past (see chart 1). On the otherhand, the scale of the market providesopportunities for alternatives to provethemselves at the margin and then moveinto the mainstream, as is happening withwind power at the moment. And someenergy technologies do have the potentialto be disruptive. Plug­in cars, for example,could be fuelled with electricity at a priceequivalent to 25 cents a litre of petrol. Thatcould shake up the oil, carmaking and elec­tricity industries all in one go.

The innovation lull of the past few de­cades also provides opportunities for tech­nological leapfrogging. Indeed, it may bethat the �eld of energy gives the not­quite­booms in biotechnology and nanotech­nology the industrial applications theyneed to grow really big, and that the threeaspiring booms will thus merge into one.

The possibility of thus recapturing thegood times of their youth has broughtmany well­known members of the �tech­norati� out of their homes in places likeWoodside, California. Energy has becomesupercool. Elon Musk, who co­foundedPayPal, has developed a battery­poweredsports car. Larry Page and Sergey Brin, thefounders of Google, have started an out�tcalled Google.org that is searching for away to make renewable energy trulycheaper than coal (or RE<C, as they de­scribe it to their fellow geeks).

Vinod Khosla, one of the founders ofSun Microsystems, is turning his consider­able skills as a venture capitalist towardsrenewable energy, as are Robert Metcalfe,who invented the ethernet system used toconnect computers together in local net­works, and Mr Doerr, who works atKleiner Perkins Cau�eld & Byers, one of Sil­icon Valley’s best­known venture­capital�rms. Sir Richard Branson, too, is getting inon the act with his Virgin Green Fund.

This renewed interest in energy isbringing forth a raft of ideas, some bright,

1A dance to the music of time

Source: BP

Sources of US energy supply, %

1850 1900 1950 20000

20

40

60

80

100

Wood

Coal

Hydro

Oil

Gas

Nuclear

Renewables

The Economist June 21st 2008 A special report on the future of energy 3

2

1

former White House counsel, Mr Woolseyestimates that American oil companies re­ceive preferential treatment from their go­vernment worth more than $250 billion ayear. And the Intergovernmental Panel onClimate Change (IPCC), a United Nations­appointed group of scienti�c experts, reck­ons that fossil fuels should carry a tax of$20­50 for every tonne of carbon dioxidethey generate in order to pay for the envi­ronmental e�ects of burning them (hencethe fears of the power­generators).

So the subsidies and mandates o�eredto renewable sources of power such aswind turbines often just level the playing�eld. It is true that some subsidies amountto unwarranted market­rigging: examplesinclude those handed by cloudy Germanyto its solar­power industry and by Amer­ica to its maize­based ethanol farmerswhen Brazilian sugar­based ethanol is farcheaper. Others, though, such as a require­ment that a certain proportion of electric­ity be derived from non­fossil­fuel sources,make no attempt to pick particular techno­logical winners. They merely act to stimu­late innovation by guaranteeing a marketto things that actually work.

If the world were rational, all of thesemeasures would be swept away and re­placed by a proper tax on carbon�as isstarting to happen in Europe, where the

price arrived at by the cap­and­trade sys­tem being introduced is close to the IPCC’srecommendation. If that occurred, wind­based electricity would already be com­petitive with fossil fuels and others wouldbe coming close. Failing that, special treat­ment for alternatives is probably the leastbad option�though such measures needto be crafted in ways that favour neither in­cumbents nor particular ways of doingthings, and need to be withdrawn whenthey are no longer necessary.

The poor world turns greener tooThat, at least, is the view from the richworld. But poorer, rapidly developingcountries are also taking more of an inter­est in renewable energy sources, despiteassertions to the contrary by some West­ern politicians and businessmen. It is truethat China is building coal­�red power sta­tions at a blazing rate. But it also has a largewind­generation capacity, which is ex­pected to grow by two­thirds this year, andis the world’s second­largest manufacturerof solar panels�not to mention having thelargest number of solar­heated rooftophot­water systems in its buildings.

Brazil, meanwhile, has the world’s sec­ond­largest (just behind America) andmost economically honest biofuel indus­try, which already provides 40% of the fuel

consumed by its cars and should soon sup­ply 15% of its electricity, too (through theburning of sugarcane waste). South Africais leading the e�ort to develop a new classof safe and simple nuclear reactor�not re­newable energy in the strict sense, but car­bon­free and thus increasingly welcome.These countries, and others like them, areprepared to look beyond fossil fuels. Theywill get their energy where they can. So ifrenewables and other alternatives cancompete on cost, the poor and the richworld alike will adopt them.

That, however, requires innovation.Such innovation is most likely to come outof the laboratories of rich countries. At a re­cent debate at Columbia University, whichThe Economist helped to organise, MrKhosla defended the proposition, �The Un­ited States will solve the climate­changeproblem�. The Californian venture capital­ist argued that if cheaper alternatives tofossil fuels are developed, simple econom­ics will ensure their adoption throughoutthe world. He also insisted that the innova­tion which will create those alternativeswill come almost entirely out of America.

As it happens, he lost. But that does notmean he is wrong. There are lots of tera­watts to play for and lots of money to bemade. And if the planet happens to besaved on the way, that is all to the good. 7

ON A ridge near Toledo in Castile­LaMancha stands a row of white wind­

mills. Literary bu�s, even if they havenever been to Spain, will recognise themas the ferocious giants attacked by DonQuixote, Miguel de Cervantes’s �ctional17th­century hero. These days, however,they are dwarfed by legions of modernwind turbines that grind out not �our butpower, helping to make Spain one of theleading producers of wind­based electric­ity in Europe.

Does this amount to tilting at wind­mills? There is no doubt that Spain’s windturbines would not have been built with­out assistance from the highly visible handof a government that wanted to prove itsgreen credentials. But wind power is no il­lusion. World capacity is growing at 30% ayear and will exceed 100 gigawatts thisyear. Victor Abate, General Electric’s vice­

president of renewables, is so convincedthat by 2012 half of the new generating ca­pacity built in America will be wind­powered that he is basing his businessplan on that assumption.

Wind currently provides only about 1%of America’s electricity, but by 2020 that �­gure may have risen to 15%. The one part ofthe United States that has something ap­proximating a proper free market in elec­tricity, Texas, is also keener than any otherstate on deploying the turbines. In May, T.Boone Pickens, one of the state’s most fam­ous oil tycoons, announced a deal with GE

to build a one­gigawatt wind farm�theworld’s largest�at a cost of $2 billion.

What was once a greener­than­thou toyhas thus become a real business (GE aloneexpects to sell $6 billion­worth of turbinesthis year)�and one with many advantages.For example, as Lester Brown, the presi­

dent of the Earth Policy Institute, a think­tank in Washington, DC, points out, afarmer in Iowa who gives up a tenth of ahectare (a quarter of an acre) of land to aturbine might earn $10,000 a year from it(about 3% of the value of the electricity itproduces). Planted with maize, the sameland would yield a mere $300­worth ofbioethanol.

Moreover, wind farms can be builtpiecemeal, unlike most power stations. Ahalf­�nished coal­�red or nuclear powerplant is a useless waste of money, but ahalf­�nished wind farm is simply a windfarm half the size originally intended�andone that has been providing revenue sincethe �rst turbine was completed.

One consequence of this rapidly grow­ing market is a virtuous circle of techno­logical improvement that is pushing wind­generated electricity closer and closer to

Trade winds

Wind power has come of age. But to make the most of it, electrical grids will have to be overhauled

4 A special report on the future of energy The Economist June 21st 2008

2

1

solving Google’s cheaper­than­coal equa­tion. The �rst turbines were cobbledtogether from components intended forships. Now the engineers are borrowingfrom aircraft design, using sophisticatedcomposite materials and equally sophisti­cated variable­geometry blades to makethose blades as long as possible (bigger isbetter with turbine technology) and assmart as possible (a blade that can �exwhen the wind blows too strongly, andthus �spill� part of that wind, is able to turnwhen other, lesser turbines would have tobe shut down for their own safety). Thetheoretical maximum e�ciency of a tur­bine, worked out in the early 20th centuryby Albert Betz, is 59.3%. Modern turbinesget surprisingly close to that, being about50% e�cient.

They are also more reliable than theirpredecessors. According to Mr Abate,when GE entered the turbine business in2002 the average turbine was out of com­mission 15% of the time. Now its downtimeis less than 3%. As a result, the cost of theenergy cranked out by these turbines hascome down to about 8 cents a kilowatt­hour (kWh) and is still falling.

That makes wind power competitivewith electricity generated by burning nat­ural gas. Coal power is still cheaper, atabout 5 cents a kWh. But according to astudy by the Massachusetts Institute ofTechnology (MIT), that would rise to 8cents if the CO2 from coal­�red power sta­tions had to be captured and stored under­ground (see box on the next page)�or, forthat matter, if a carbon tax of $30 a tonnewere imposed.

The power companies that buy the tur­bines are also getting smarter. They em­ploy teams of meteorologists to scour theworld for the best places to put turbines. Itis not just a question of when the windblows, but also of how powerfully. A dif­ference of as little as one or two kilometres(one mile) an hour in average wind speedcan have a signi�cant e�ect on electricaloutput. And another lot of meteorologistssit in the control centres, making detailedforecasts a day or two ahead to help a com­pany manage its power load. For one pro­blem with wind is that if it stops blowing,the turbines stop turning. After cuttingcosts, that is the second great challenge ofthe spread of wind power.

The third is that people do not necessa­rily live where the wind blows. Indeed,they often avoid living in such places. Sol­ving these problems, though, is a task notfor the mechanical engineers who buildthe turbines but for the electrical engineers

who link them to places where power iswanted. That means electricity grids areabout to become bigger and smarter.

Bigger means transcontinental, at leastfor people like Vinod Khosla. His analogyis America’s interstate highway system,built after the second world war. The newgrids would use direct, rather than alter­nating, current. AC was adopted as stan­dard over a century ago, when the electri­cal world was rather di�erent. But DC isbetter suited to transporting power overlong distances. Less power is lost, even onland. And DC cables can also be laid on theseabed (the presence of all that waterwould dissipate an AC current veryquickly). In the right geographical circum­stances that eliminates both the di�cultyof obtaining wayleaves to cross privateland and the not­in­my­backyard objec­tions that power lines are ugly. Indeed,there is already a plan to use underwatercables to ship windpower from Maine toBoston in this way.

Rewiring the planetAs it happens, Europe already has the em­bryo of a DC grid. It links Scandinavia,northern Germany and the Netherlands,and there is talk of extending it across theNorth Sea to the British Isles, another no­

toriously windy part of Europe. By con­necting distant points, this grid not onlydelivers power to market, it also allows thesystem some slack. It matters less that thewind does not blow all the time because itblows at di�erent times in di�erent places.The grid also permits surplus power to beused to pump water uphill in Norwegianhydroelectric plants (a system known aspumped storage), ready for use when de­mand spikes.

Smarter grids, however, would help tosmooth out such spikes in the �rst place.The ability to accommodate inherently in­termittent sources such as wind is onlyone of several reasons for wanting to dothis, but it is an important one.

A smart grid will constantly monitor itsload and (this is the smart bit) take particu­lar consumers o�ine, with their prioragreement and in exchange for a lowerprice, if that load surges beyond a preset le­vel. For this purpose, a consumer may notnecessarily be the same as a customer. Thegrid’s software would be able to identifyparticular circuits, or even particular appli­ances, in a home, o�ce or factory. Theirowners would decide in what circum­stances they should shut down or boostup, and the smart grid’s software wouldthen do the job. Water heaters and air­con­

The Economist June 21st 2008 A special report on the future of energy 5

2

EVEN in the most alternative­friendly fu­ture imaginable, coal is unlikely to go

away. It is cheap, abundant and often lo­cal. So what can be done to make coal’suse more acceptable?

One much­discussed possibility is car­bon capture and storage, or CCS, which in­volves burying the carbon dioxide deepunderground. The generating companieshave high hopes of it (see chart 3). Thereare just two problems. No one knows if itwill work (in other words, if the CO2 willstay buried). And everyone knows that,whether it works or not, it will be expen­sive�so much so that the alternatives startto look rather attractive. The one seriousattempt to investigate its use in an actualpower station, the FutureGen project,based in Illinois, was cancelled in Januarybecause the expected cost had risen from$830m to $1.8 billion.

The �capture� part is not that hard. Car­bon dioxide reacts with a group of chemi­cals called amines. At low temperaturesCO2 and amines combine. At higher tem­peratures they separate. Power­station ex­haust can thus be purged of its CO2 by run­ning it through an amine bath before it isvented, and the amine can be warmed torelease the gas where it will do no harm.Better still, the coal can be reacted withwater to produce a mixture of CO2 and hy­drogen in which the carbon dioxide ismuch more concentrated than in normal�ue gas, so it is easier to scrub out. What isthen burned is pure hydrogen.

All this processing is expensive, butthere is no reason why it should not work.An experimental plant in Denmark thatuses monoethanolamine as the captorhas been running for two years. Alstom, aFrench �rm, has almost �nished building

one in Wisconsin that uses ammonia.It is what comes next that is the pro­

blem. The disposal of carbon dioxideneeds to be permanent, so a lot of condi­tions have to be met. To be a successful bu­rial site, a body of rock needs to be morethan 1km underground. That depth pro­vides enough pressure to turn CO2 intowhat is known as a supercritical �uid, aform in which the stu� is more likely tostay put. The rock in question also has tohave enough pores and cracks in it to ac­commodate the CO2. Lastly, it needs to becovered with a layer of non­porous, non­cracked rock to provide a leakproof cap.

So far, only three successful CCS pro­jects are under way. The Weyburn­MidaleCO2 project is burying carbon dioxidefrom a coal gasi�cation plant in North Da­kota in a depleted oil �eld in Saskatche­wan. The Salah gas�eld project in Algeria,run by BP, strips CO2 from local naturalgas and injects it back into the ground.And Statoil, a large Norwegian oil and gascompany, performs a similar trick at twoplaces in the North Sea. None of these pro­jects is actually linked to generating elec­tricity. Still, a few years ago they weretouted proudly. But the touting has be­come more nervous, and no new projectshave come on stream.

The scale of the problem is awesome.The three showcase projects each dumpabout a million tonnes of CO2 a year. ButAmerica’s electricity industry alone pro­duces 1.5 billion tonnes, which wouldmean �nding 1,500 appropriate sites, andnobody knows whether the country’s ge­ology can oblige. Even transporting thatamount of gas would be a huge task.

As to the cost, a report published lastyear by MIT reckons on $25 a tonne to cap­

ture CO2 and pressurise it into a super­�uid, and $5 a tonne to transport it to itsburial site. It therefore suggests that powerstations which dump CO2 into the atmo­sphere should be charged $30 a tonne, a �­gure conveniently near both the middleof the IPCC’s suggested carbon price andthe actual price in Europe. Another report,by a consultancy called Synapse EnergyEconomics, notes that American powercompanies are already starting to employcarbon prices in their internal accounting,using a range of $3­61 a tonne. Again, themiddle of that range is about $30.

Such a charge, whether a tax or a sys­tem of tradable permits to pollute, wouldchange energy economics radically. Buteven the most optimistic proponents ofcarbon capture and storage doubt it willbe a serious alternative much before 2020.And by then both the physical and thepolitical climate may look rather di�erent.

Carbon storage will be expensiveat best. At worst, it may not workDig deep

ditioners might stock up on heat or cold inanticipation of such shutdowns. Fridgeswould know how long they could managewithout power before they had to switchon again.

Reducing spikes in demand that waywill cut the need for what are known in theindustry as �peakers��small power plantssuch as pumped­storage systems that existsolely to deal with such spikes. Parts of

America’s existing dumb and fragmentaryelectricity grid are so vulnerable to load va­riations that their owners think they maybe able to cope with no more than about2% of intermittent wind power. Clearlypeaks will never be eliminated entirely.However, Mr Abate reckons that a combi­nation of smart grids and gas­�red peakersshould push the potential for wind powerup a long way.

To prove the point, GE is collaboratingwith the government of Hawaii, a statewhich is served by a series of small, iso­lated grids highly vulnerable to disruption.The �rm’s engineers reckon that clever gridmanagement will allow up to 30% of localpower to come from wind without anyblackouts. If that improvement can betranslated to the grids on the mainland,wind’s future looks assured. 7

6 A special report on the future of energy The Economist June 21st 2008

1

WIND power works, and will workbetter in the future. But wind is only

an interim stop on the way to a worldwhere electricity no longer relies on fossilfuels. The ultimate goal is to harvest thesun’s energy directly by intercepting sun­light, rather than by waiting for that sun­light to stir up the atmosphere and stickingturbines in the resulting airstreams.

Fortunately, inventors love that sort ofproblem. Ideas they have come up withrange from using the sun to run simpleheating systems for buildings, deploying�reverse radiators� painted black, to thesharpest cutting edge of that trendiest of�elds, nanotechnology, to ensure that ev­ery last photon is captured and convertedinto electricity. The most iconic form of so­lar power, the photovoltaic cell, is cur­rently the fastest­growing type of alterna­tive energy, increasing by 50% a year. Theprice of the electricity it produces is falling,too. According to Cambridge Energy Re­search Associates (CERA), an Americanconsultancy run by Daniel Yergin, a kWhof photovoltaic electricity cost 50 cents in1995. That had fallen to 20 cents in 2005and is still dropping. Not RE<C, but headingin the right direction.

Photovoltaic cells (or solar cells, as theyare known colloquially) convert sunlightdirectly into electricity. But that is not theonly way to use the sun to make electricalpower. It is also possible to concentrate thesun’s rays, use them to boil water and em­ploy the resulting steam to drive a turbine.These two very di�erent approaches illus­trate an unresolved question about the fu­ture of energy: whether it will be gener­ated centrally and transported over longdistances to the consumer, as it has been inrecent decades, or generated and con­sumed in more or less the same place, as itwas a century ago.

A hot tin roofThe idea of solar cells is to keep things lo­cal. They are like wind turbines, only moreso, in that even a single solar panel can pro­duce power immediately. Put a few onyour roof and, if you live in a reasonablysunny place, you can cut your electricitybill. Indeed, you may be able to sell elec­tricity back to your own power company.

The problem is that at the moment youmay need to take out an overdraft to payfor the solar panels, and you will not getyour money back for a long time.

Many engineers, however, are workingto change that. One of them is EmanuelSachs of MIT. Some engineers look for big,exciting technological improvements inthe way solar cells work, but Dr Sachs pref­ers incremental change. As he sees it, it issuch change that drives Moore’s law, thatwell­established description of the rapidimprovement in the power of computerprocessors.

Moreover, the analogy is appropriate.Traditional solar cells are made of silicon,like computer chips, and for the same rea­son. They rely on that element’s propertiesas a semiconductor, in which negativelycharged electrons and positively charged�holes� move around and carry a currentas they do so. In the case of a solar cell, thecurrent is created by sunlight knockingelectrons out of place and thus creatingholes. Dr Sachs’s �rst contribution to theincremental improvement was a tech­nique called the string ribbon, whichhalved the amount of silicon needed tomake a solar cell by drawing the element(in liquid form) out of a vat between twostrings. That invention was marketed by a�rm called Evergreen Solar.

His latest venture, a �rm called 1366Technologies (after the number of watts ofsolar power that strike an average squaremetre of the Earth’s surface), aims to fol­low this up with three new ideas that

should, in combination, bring about a 27%improvement in e�ciency. He and his col­leagues have redesigned the surfaces ofthe silicon crystals on a nanoscale in orderto keep re�ected light bouncing around in­side a cell until it is eventually absorbed.They have also managed to do somethingsimilar to the silver wires that collect thecurrent. And they have made the wiresthemselves thinner as well so that they donot block so much light in the �rst place.

Dr Sachs says that these innovationswill bring the capital cost of solar cells be­low $2 a watt. That is closing in on the costof a coal­�red power station: a gigawatt(one billion watt) plant costs about $1 bil­lion to build. The price, of course, is a di�er­ent matter. As Paula Mints of NavigantConsulting, a �rm based in Palo Alto, Cali­fornia, points out, price is set by marketconditions. These�particularly the gener­ous subsidies given to solar power in someEuropean countries�have kept prices wellabove costs in recent years. Nevertheless,as chart 4 shows, the price of solar cells hasfallen signi�cantly, too.

Other researchers back a newer tech­nology known as thin­�lm photovoltaics.Thin­�lm cells can be made with silicon,but most progress is being made with onesthat use mixtures of metals, sometimes ex­otic ones, as the semiconductor. Thesemixtures are not as e�cient as traditionalbulk­silicon cells (meaning that they donot convert as much sunlight to electricityper square metre of cell). But they use farless material, which makes them cheaper,and they can be laid down on �exible sur­faces such as sheets of steel the thicknessof a human hair, which gives them widerapplications.

At the moment, the commercial leaderin this area is a �rm called First Solar,which uses cadmium telluride as the �lm.But First Solar is about to be given a run forits money by companies such as Miasolé, asmall Californian �rm, that have gone for amixture of copper, indium, gallium and se­lenium, known as CIGS. This mixture isreckoned to be more e�cient than cad­mium telluride, though still not as good astraditional silicon. And it has the public­re­lations advantage of not containing cad­mium, a notorious poison�though First

Another silicon valley?

The rise of solar energy, in one form or another

4Selling cells

Sources: Navigant Consulting; Paula Mints

Photovoltaic modules, price per watt, $

1980 85 90 95 2000 05 070

5

10

15

20

The Economist June 21st 2008 A special report on the future of energy 7

2

1

THE Philippines are not generally asso­ciated with the cutting edge of techno­

logical change. In one respect, though, thecountry is ahead of its time: around aquarter of its electricity is generated fromunderground heat. Such heat is free, inex­haustible and available day and night.

It is also part of a geology that seesparts of the country devastated by vol­canic eruptions from time to time. Thegeysers that turn the generators aremerely the gentlest manifestations of thisvolcanism. The question that exercisesJe�erson Tester, a researcher at MIT, iswhether it is possible to have the onewithout the other. The Earth’s depths are,after all, hot everywhere. So if there is nonatural volcanism around to bring thisheat to the surface, his answer is to createcontrolled, arti�cial volcanism�what isknown as an engineered geothermal sys­tem (EGS). Instead of relying on naturalhot springs, you make your own.

In principle, this is easy. Drill two paral­lel holes in the ground, a few hundredmetres apart, and carry on drilling untilthe rock is hot enough (say 200°C). Thenpump cold water down one hole and waitfor it to come back up the other at a suit­ably elevated temperature. The super­heated water turns to steam which you

use to power a generator. In Dr Tester’sview, the reason this source of power isneglected is that it is invisible. Everybodyfeels the wind and the sun, but only min­ers notice that the Earth’s interior is hot, sono one thinks of drilling for that heat.

Dr Tester reckons that spending about$1 billion on demonstration projects overthe next 15 years would change that. Itwould provide enough information to al­low 100 gigawatts­worth of EGSs to becreated in America by 2050, at a commer­cially acceptable price.

In principle, much more could bedone. The recoverable heat in rock underthe United States is the equivalent of2,000 years­worth of the country’s cur­rent energy consumption, according to areport he and his colleagues publishedtwo years ago. A similar assessment of Eu­rope’s heat resources from the Earth sug­gests that they could be used to generateas much electricity as all of the continent’snuclear power stations produce now.

Rock­hardExtracting this subterranean energy is notas easy as it sounds. Until the term EGS

was coined, the �eld was known as hot­dry­rock geothermal energy, a name thatencapsulates the problem precisely. A

century of data collected by oil compa­nies suggest it is impermeable rocks suchas granite that are the most e�ective reser­voirs of heat. Their very dryness increasestheir heat capacity. But to get the heat outyou have to make them permeable. Hencethe �engineered� in the new name.

Some of Dr Tester’s $1billion would bespent working out how to drill cheaplyand e�ectively through this sort of rock�something that oil companies tend toavoid because impermeable rocks do notcontain petroleum. A lot of the moneywould go on �nding ways to force open�ssures in the granite to let the water �owfrom the injection hole to the exit.

The Cooper Basin in South Australiahas the hottest non­volcanic rocks of anyknown place in the world, and Australialeads the �eld in exploiting subterraneanheat, with seven �rms snooping aroundthe area. One of them, Geodynamics, re­cently completed what it claims is a com­mercial­scale well. And the turbines willalso turn soon at an experimental non­commercial project at Soultz, in France.

If it can be made to work, EGS has gotthe lot. No unsightly turbines. No need tocover square kilometres of land with vastmirrors. And it is always on. Anybody gota billion dollars handy?

Geothermal could be hotBeneath your feet

Solar’s �lms carefully lock the cadmiumup in a way that renders it harmless.

At the moment thin­�lm solar cells arebeing packaged and sold as standard solarpanels, but that could easily change. FirstSolar applies its �lms to glass, but Mia­solé’s boss, Joseph Laia, points out that hissteel­based products are �exible and light­weight enough to be used as building ma­terials in their own right. Greener­than­thou Californians who wish to fall in withtheir governor’s plan for a million solarroofs, announced in 2006, currently haveto bolt panels onto their houses�an ugly, ifvisible, show of their credentials. If MrLaia has his way, they will soon be able touse sheets of his company’s CIGS­coveredsteel as the roo�ng material itself.

Supporters of solar­thermal energytend to look askance at solar panels. Cad­mium telluride and CIGS may be cheaper

than silicon, but glass and steel, on whichsolar­thermal relies, are cheaper still. Thetechnology’s proponents think big:square­kilometres big. They want to �ll thedeserts with steel and glass mirrors anduse the re�ected sunlight to boil water andgenerate electricity, then plug into the long­distance DC networks developed for windpower to carry the juice to the cities.

Desert songThose who worry about the political sideof the world’s dependence on oil will beless than delighted to �nd that one countrythinking seriously about such systems isAlgeria. With the power­hungry marketsof Europe to its north, across the Mediterra­nean, and a lot of sunshine going to wastein the Sahara desert to its south, Algeria’sgovernment is looking for ways to connectthe two. It is now building an experimental

solar­thermal power station at HassiR’mel, about 400km south of Algiers,which if all goes well will open next year.In April work started on a similar project atAïn Béni Mathar, in Morocco, and othersare in the pipeline elsewhere in north Af­rica. Fortunately for people like Mr Wool­sey, the ex­CIA man, America has desertsof it own which are about to bloom withmirror­farms too.

There are four competing designs: para­bolic­trough mirrors, parabolic­dish mir­rors, �power towers� which use an array ofmirrors to focus the sun’s rays on to an ele­vated platform, and Fresnel systems,which mimic a parabolic trough using(cheaper) �at mirrors. All either heat upwater to make steam, which drives a gener­ator, or heat and liquefy a salt with a lowmelting point such as sodium nitrate that isused to make steam.

8 A special report on the future of energy The Economist June 21st 2008

2

1

All four of these designs are now eitheroperating commercially in the deserts ofsouth­west America or are undergoingpre­commercial trials. Although the totalcapacity at the moment, according toCERA, is a mere 400 megawatts, this willgrow tenfold over the next four years if allprojects now scheduled come to fruition,and probably a lot more after that. More­over, those plants that melt a salt are able todivert part of the heat they collect into athermal reservoir that can keep the gener­ators turning at night. The main objectionto solar power�that it goes o� after sun­set�is thus overcome.

From little acornsThe engineers clearly think they can de­liver the technology. But can the technol­ogy deliver the power? A back­of­the­en­velope calculation suggests that it can. Twoyears ago a task force put together by thegovernors of America’s western states

identi�ed 200 gigawatts­worth of primesites for solar­thermal power within theirterritory (meaning places that had enoughreliable sunshine, were close to transmis­sion lines and were not environmentallyor politically sensitive). That is equivalentto 20% of America’s existing electricity­generation capacity: not a bad start.

Robert Fishman, the boss of Ausra, anAustralian­American company based inPalo Alto, California, reckons that his �rm’sFresnel arrays combined with its propri­etary heat­storage system can produceelectricity for 8 cents a kWh. That matchesGE’s wind turbines, and mass productionshould bring it down further. It is notcheaper than �naked� coal (Ausra will be­ne�t from various state governments’ re­quirements that their power utilities buyrenewable power)�but if there were a car­bon tax of $30 a tonne, or a requirement tocapture and bury CO2, Ausra would beable to match the coal­�red stations’ prices.

The most intriguing technology of all,though, belongs to SUNGRI, a �rm basedin Los Angeles. This uses mirrors to con­centrate sunlight, but focuses it on a solarcell rather than a boiler. The system is saidto turn 37% of the light into electricity. InApril the �rm claimed it would be able toproduce electricity for the magic �gure of 5cents a kWh.

That claim has yet to be put to the test,and should be viewed with some scepti­cism until it has been. But it is a good indi­cation of the way the �eld is going. Solarpower now seems to be roughly wherewind was a decade ago. At the moment itcontributes a mere 0.01% to the world’soutput of electricity, but just over a decadeof 50% annual growth would bring that to1%, which is where wind is at the moment.If SUNGRI is to be believed, and the pointwhere RE is indeed <C is close, the rise to 1%might happen even faster. After that, thesky is the limit. 7

BURIED in the news a few weeks agowas an announcement by a small Cali­

fornian �rm called Amyris. It was, per­haps, a parable for the future of biotech­nology. Amyris is famous in the world oftropical medicine for applying the latestbiotechnological tools to the manufactureof artemisinin, an antimalarial drug that isnormally extracted from a Chinese vine.The vines cannot produce enough of thestu�, though, so Amyris’s researchers havetaken a few genes here and there, tweakedthem and stitched them together into a

biochemical pathway enabling bacteria tomake a chemical precursor that can easilybe converted into the drug.

But that is not what the announcementwas about. Instead, it was that Amyris wasgoing into partnership with Crystalsev, aBrazilian �rm, to make car fuel out of canesugar. Not ethanol (though Brazil alreadyhas a thriving market for ethanol­poweredcars), but a hydrocarbon that has thecharacteristics of diesel fuel. Technically, itis not ordinary diesel, either: in chemist­speak, it is an isoprenoid rather than a mix­

ture of alkanes and aromatics. But thedriver will not notice the di�erence.

The point of the parable is this: biotech­nology may have cut its teeth on medi­cines, but the big bucks are likely to be inbulk chemicals. And few chemicals arebulkier than fuels. Where Amyris is lead­ing, many are following. Some small �rmswith new and interesting technologies aretrying to go it alone. Others are teaming upwith big energy �rms, in much the sameway that biotech companies with a pro­mising drug are often taken under the

Grow your own

The biofuels of the future will be tailor­made

The Economist June 21st 2008 A special report on the future of energy 9

2

1

wing of a large pharmaceutical company.The big �rms themselves are involved, too,both through in­house laboratories and bygiving money to universities. Biofuels,once seen as a cross between eccentricgreenwash and a politically acceptableway of subsidising farmers, are nowpoised to become big business.

Grassed upThe list of things that need to be done tocreate a proper biofuel industry is a longone. New crops, tailored to fuel rather thanfood production, have to be created. Waysof converting those crops into feedstockhave to be developed. That feedstock hasthen to be turned into something that peo­ple want to buy, at a price they can a�ord.

All parts of this chain are currently thesubjects of avid research and develop­ment. Some biofuels were already compet­itive with oil products even at 2006 oilprices (see table 5). The R&D e�ort willbring more of them into line, as will anylong­term rise in the price of crude oil.

As far as the crops themselves are con­cerned, there are three runners at the start­ing gate: grasses, trees and algae. Grassesand trees are grown on dry land, but needa lot of processing. The idea is to take thewhole biomass of the plant (particularlythe cellulose of which a plant­cell’s wallsare made) and turn it into fuel. At the mo­ment, that fuel is often ethanol. Hence theterm �cellulosic ethanol� that has gainedrecent currency. Algae, being aquatic, aremore �ddly to grow, but promise a high­quality product, oil, that will not needmuch treatment to become biodiesel.

One of the leading proponents of bettergrasses is Ceres, a �rm based in ThousandOaks, California. The species it has chosento examine�switchgrass, miscanthus,sugarcane and sorghum�are so­called C4

grasses. These are favourites with the bio­fuel industry because they share a particu­larly e�cient form of photosynthesis thatenables them to grow fast. Ceres proposesto make them grow faster still, using a mix­ture of �smart� breeding techniques (inwhich desirable genes are identi�ed scien­ti�cally but assembled into plants by tradi­tional hybridisation) and straightforwardgenetic engineering.

The chosen grasses also thrive in arange of climates. Switchgrass and miscan­thus are temperate. Sugarcane and sor­ghum are tropical. Ceres proposes to ex­tend their ranges still further by creatingstrains that will tolerate heat or cold ordrought or salt, allowing them to be grownon land that cannot be used for food crops.

That will make them cheaper, as well as re­ducing the competition between foodsand biofuels.

Trees, meanwhile, are the province of�rms such as ArborGen, of Summerville,South Carolina. Like Ceres, ArborGen isworking on four species: eucalyptus, pop­lar, and the loblolly and radiata pines. It isapplying similar techniques to those usedby Ceres to speed up the growth of thesetrees and to increase their tolerance ofcold. Although creating raw materials forbiofuels is not this company’s only objec­tive (paper pulp and timber are others), itsees such fuels as a big market.

Algae, too, are up for modi�cation. Oneproblem with them is harvesting the oilthey produce. That means extracting themfrom their ponds, drying them out and

breaking open their cells. This process is sotedious that some companies are consider­ing the idea of burning the dried algae inpower stations instead.

One �rm that is not is Synthetic Ge­nomics, the latest venture of Craig Venter(the man who led the privately funded ver­sion of the Human Genome Project). DrVenter hopes to overcome the oil­collec­tion problem by genetic engineering. Syn­thetic Genomics’s algae have been �ttedwith genes that create new secretion path­ways through their outer membranes.These cause the algal cells to expel the oilalmost as soon as they have manufacturedit. It then �oats to the surface of the pond,allowing it to be skimmed o� like creamand turned into biodiesel. The algae arealso engineered to make more oil thantheir wild counterparts.

Harvesting useful fuels from vascularplants, as grasses, trees and their kind areknown collectively, is a trickier business.These plants are composed mainly ofthree types of large molecule. Besides cel­lulose, there are hemicellulose and lignin.Each is made of chains of smaller mole­cules, and all three are often bound to­gether in a complex called lignocellulose,particularly in wood. There are many waysthese long­chain molecules might beturned into fuel, but all of these processesare more complex than for algae.

As chart 6 shows, turning sunlight intobiofuel involves three steps, though di�er­ent methods may miss out some of thesesteps. Algae can make the leap from start to�nish directly, whereas vascular plantscannot. One way of dealing with them isto dry them and then heat them with littleor no oxygen present. This is called pyroly­sis and, if done correctly, results in a mix­ture of carbon monoxide and hydrogencalled �syngas� (short for synthesis gas).With suitable catalysts, syngas can beturned into fuel.

This is the approach taken by Choren

6The chain of being

Source: InterAcademy Council

Potential pathways for biofuel production

“Syngas”Thermal

treatmentMicrobes orcatalysts

Consolidated process(microbes or enzymes)

Biomassdepolymerisation

(microbes or enzymes)

SUNLIGHT BIOMASS MONOMERS FUELS

Algae

Feedstockdevelopment

(smart breeding,genetic engineering)

Biofuel production(microbes, enzymes

or catalysts)

Out of thin air

Source: The Royal Society *Taxes included

Biofuel costs compared with prices for oiland oil products, cents per litre

5

Fuel 2006

Price of oil 50-80$/barrel

Petroleum products 35-60pre-tax price

Petroleum products 150-200retail price* in Europe, 80 in US

Long-termBiofuel 2006 about 2030

Ethanol from 25-50 25-35sugarcane

Ethanol from maize 60-80 35-55

Ethanol from beet 60-80 40-60

Ethanol from wheat 70-95 45-65

Ethanol from 80-110 25-65lignocellulose

Biodiesel from 70-100 40-75vegetable oils

Fuels made 90-110 70-85from “syngas”

10 A special report on the future of energy The Economist June 21st 2008

Industries in Freiburg, Germany, andRange Fuels in Treutlen County, Georgia. Inboth cases the feedstock is chippings andother leftovers from forestry and timber­mills. Choren is making hydrocarbon die­sel and Range ethanol. Both factories,therefore, are steps on the road to makingfuel from trees. Syngas can also be turnedinto ethanol by bacteria of the genus Clos­tridium (a group better known for thechemical used in botox treatment). That isbeing done by Coskata, a �rm based inWarrenville, Illinois. General Motors (GM)likes this idea so much it has bought ashare of the company.

An alternative to the syngas method isto break the cellulose and hemicelluloseup into their component �monomer� mol­ecules. That is easier said than done, par­ticularly if lignin is involved, since lignin isresistant to such conversion. The amountof coal in the world is proof of its resil­ience. Coal is composed mainly of ligninfrom plants that failed to decompose com­pletely and were fossilised as a result.

Many �rms, however, have developedenzymes that break down biomass in thisway. Iogen, of Ottawa, Canada, was one ofthe �rst. Its enzymes decompose celluloseand hemicellulose into sugar monomers.(The lignin is burned to generate heat forthe process.) Abengoa, a Spanish �rm thatis also involved in solar energy, uses thisapproach as well.

Sugar and spiceOnce you have your sugar, you can fer­ment it. These days that need not mean us­ing yeast to make ethanol. A whole rangeof bugs, some natural, some engineered,can now be deployed to make a wholerange of products. Amyris’s souped­up mi­cro­organisms (some are bacteria, someyeasts) turn sugar not into ethanol but intoisoprenoids, at a cost competitive with pe­troleum­based diesel. LS9, based near SanFrancisco, uses a similar method but isturning out alkanes (for petrol) and fattyacids (for biodiesel). It, too, is starting toscale up production. Synthetic Genomicsis doing something similar, though the�rm is cagey about which fuel is being pro­duced. In each case, however, what ismade is a chemical precisely tailored to itspurpose, rather than the ad hoc mixturethat comes out of a re�nery. The rival com­panies thus argue that their products areactually better than oil­based ones.

At least one �rm, Mascoma, of Cam­bridge, Massachusetts, employs a singlespecies of bug, Thermoanaerobacteriumsaccharolyticum, both to break down the

biomass and to digest the resulting sugar.Mascoma will use both grass and wood asfeedstocks. In May it signed deals with GM

and Marathon Oil.It is also possible to use puri�ed en­

zymes to do the conversion from sugar tofuel, as well as from biomass to sugar, andat least two �rms are working on applyingthem to the whole process. Codexis, basedin Redwood City, California, has created arange of enzymes by a method akin to sex­ual reproduction and natural selection.Last year it signed a deal with Shell to usethis technique to produce biofuels of va­rious types. And a Danish �rm, Danisco,

has teamed up with DuPont to do the samething with its own proprietary enzymes.

Shell is also involved in a project to turnsugar into hydrocarbons, this time bystraight chemical processing. It is puttingup the money. The technology (the mostimportant part of which is a set of propri­etary non­biological catalysts) is providedby Virent Energy systems, of Madison,Wisconsin.

Which of these approaches will workbest is anybody’s guess. But their sheernumber is proof that the most radicalthinking in the �eld of renewable energy isgoing on in biofuels. It is in this area thatthe most unexpected breakthroughs arelikely to come, says Steven Koonin, BP’schief scientist. BP is backing one of the big­gest academic projects intended to lookinto biofuels, the Energy Biosciences Insti­tute (EBI), to the tune of $500m, which sug­gests that the company’s board agrees withhim. The EBI is a partnership of the Univer­sity of California, Berkeley, the LawrenceBerkeley National Laboratory and the Uni­versity of Illinois.

One of the people involved, StevenChu, the head of the Lawrence Berkeleylaboratory, is a man with a grand vision.This vision is of a �glucose economy� thatwill replace the existing oil economy. Glu­cose, the most common monomer sugar,would be turned into fuels and maybeeven the bio­equivalents of petrochemi­cals�bioplastics, for example�in local fac­tories and then shipped around the world.That would be a boon to tropical countries,where photosynthesis is at its most ram­pant, though it might not play so well toJames Woolsey’s security fears, since itrisks replacing one set of unreliable suppli­ers with another.

However, there is plenty of biomass togo around. A study by America’s Depart­ments of Energy and Agriculture suggeststhat even with only small changes to exist­ing practice, 1.3 billion tonnes of plant mat­ter could be collected from American soilwithout a�ecting food production. If thiswere converted into ethanol using the besttechnology available today, it would addup to the equivalent of 350 billion litres ofpetrol, or 65% of the country’s current pet­rol consumption. And that is before spe­cially bred energy crops and other techno­logical advances are taken into account. IfAmerica wants it, biofuel autarky looksmore achievable than the oil­based sort.And if it does not, then the world’s hithertoimpoverished tropics may �nd themselvesin the middle of an unexpected and wel­come industrial revolution. 7

2

The Economist June 21st 2008 A special report on the future of energy 11

1

NOTHING ages faster than the future. Afew years ago there was general agree­

ment that if the internal­combustion en­gine ever was replaced by somethingclean, that something would be the fuelcell. A fuel cell is a way of reacting hydro­gen and oxygen together in a controlledway and extracting electricity from theprocess. It was to be the precursor of whatwas known as the hydrogen economy, inwhich that gas would replace fossil fuelsand power almost everything.

Leaving aside the problems of trans­porting and storing a light and leaky gas,what no one was very clear about waswhere the hydrogen itself would comefrom. You would have to make it fromsomething else. That something would ei­ther be a mixture of fossil fuel and water(fuels can be reacted with steam to makehydrogen and carbon dioxide, but you stillhave to get rid of the carbon dioxide), orjust water itself, via electrolysis.

But why bother? Why not cut out themiddleman and plug your car directly intothe electricity mains instead? And that, itseems, is what may happen. You don’thear much about the hydrogen economythese days. Nor fuel cells. The buzz­phrasenow is �plug­in hybrid�.

Plug­ins should not be confused withexisting hybrid vehicles, such as Toyota’sPrius, which contains an internal­combus­tion engine as well as two electric ones. Ei­ther sort may drive the wheels. The electricmotors kick in when they can do a more ef­�cient job than the petrol engine, but eventhen the electricity comes ultimately, viabatteries, from burning petrol.

In a plug­in, the electricity comes fromthe mains, via an ordinary electricalsocket. Some intermediate designs retainthe idea of two sorts of engine, but the goalis that the car should be powered by elec­tric motors alone. If the batteries rundown, a petrol­powered generator willtake over. (Existing batteries are too expen­sive to give such a car the range of a stan­dard petrol­driven machine.) But mostcars, most of the time, are used for shortjourneys. Gerbrand Ceder, a battery scien­tist at MIT, reckons that if the �rst 50km ofan average car’s daily range were providedby batteries rather than petrol, annual pet­

rol consumption would be halved. Giventhat the electrical equivalent of a litre ofpetrol costs about 25 cents, that is an attrac­tive reduction.

The widespread adoption of plug­insmight also reduce carbon­dioxide emis­sions, depending on what sort of powerstation made the electricity in the �rstplace. Even energy from a coal­�red stationis less polluting than the serial explosionsthat drive an internal­combustion engine.If the energy comes from a source such aswind or nuclear, the gain is enormous.

Beyond that, the rise of plug­ins has im­plications for the electricity industry itself.If they succeed, they will put an unantici­pated load on the system. In fact, they mayremake electricity as well as transport.

Don’t all recharge at onceThat is certainly the view of Peter Corsellof Gridpoint, a company based in Arling­ton, Virginia. His �rm hopes to make its liv­ing selling the load­management technol­ogy required for �smart grids�. Asmentioned earlier in this special report,there are several reasons why such tech­nology is desirable. Mr Corsell goes onefurther: he reckons it will become essentialif plug­ins arrive in force. At the moment,the grid would be unable to cope if a largenumber of commuters arriving homeplugged in their cars more or less simulta­neously to recharge them. Yet if those samecars were recharged at three o’clock in the

morning, when demand is low, it wouldbene�t both consumer (who would getcheap power) and producer (who wouldbe able to sell otherwise wasted electric­ity). Such cars might even act as micro­peakers�reservoirs of electrical energythat a power company could draw on if acar were not on the road. Managing plug­ins, Mr Corsell thinks, will be the smartgrid’s killer application.

In sunny climes, plug­ins might alsoprovide another use for solar cells. Googleis already experimenting with photovol­taic car parks. These have awnings coveredin solar cells which will shade its employ­ees’ cars and simultaneously rechargethem. That is an idea which could spread.Supermarkets, for example, might �nd thatcar parks with plugs would attract custom­ers who wanted to top up their cars. Andthe more opportunities there are for sta­tionary cars to be recharged, the morelikely they are to be bought.

Plug­ins are moving from idea to realitywith amazing speed. General productionof the Tesla, Elon Musk’s new sports car,began in March (the �rm is Californian, butthe cars are built in Britain). The Tesla is noteven a hybrid. It draws all of its powerfrom lithium­ion batteries (the sort thatpower laptop computers), and it has arange of 350km. It can manage that be­cause its price of $109,000 buys a lot of bat­teries; Tesla owners are not the sort whocount their pennies.

The end of the petrolhead

Tomorrow’s cars may just plug in

12 A special report on the future of energy The Economist June 21st 2008

2

1

IF YOU want to make an environmental­ist squirm, mention nuclear power.

Atomic energy was the green movement’sdarkest nightmare: the child of mass de­struction, the spawner of waste that willremain dangerous for millennia, the ulti­mate victory of pitiless technology overfrail humanity. And not even cheap. Well,times change. The followers of Rachel Car­son and the Club of Rome in the 1960s and1970s had not heard of the greenhouse ef­fect, but today’s greens have. And theyknow that nuclear reactors are the one pro­ven way to make carbon­dioxide­free elec­tricity in large and reliable quantities that

does not depend (as hydroelectric and ge­othermal energy do) on the luck of the ge­ographical draw. What a dilemma for athoughtful tree­hugger.

Patrick Moore, one of the founders ofGreenpeace, faces no such dilemma,though. He is such a convert to the nuclearcause that he now chooses to consult for it.Cynics take him to task for that, but hemakes no apology. His view of the world,shared by James Lovelock, the inventor ofGaia (the idea that the Earth itself has someof the characteristics of a living organism),is that nuclear power�which already pro­vides 15% of the world’s electricity�is the

only possible way out of climate change.Mr Lovelock thinks it is probably too lateanyway, and that Gaia will shake herselfand be rid of the plague of humans thatnow infest her skin. Mr Moore thinks shecan be persuaded not to, if nuclear poweris applied in even larger doses.

Given the widespread concern aboutnuclear energy, how can that be done?Partly, the answer comes, by shifting prior­ities (for today’s youth, climate change iswhat global nuclear warfare was for thebaby­boomers). Partly by the fading of me­mories: the accident at Three Mile Island,which ended America’s nuclear dreams,

Life after death

Nuclear power is clean, but can it overcome its image problem?

Nor is the Tesla the only sports car to godown this road. Electric motors may lack athroaty roar, but they actually do a betterjob than petrol engines in high­perfor­mance vehicles. They have higher torqueat low revs which makes them acceleratefaster. In Britain a new �rm called theLightning Car Company plans to revivethe country’s sports­car tradition with theLightning GT. Mr Musk also faces competi­tion in California, from Fisker Automotive,whose eponymous founder Henrik Fiskerhelped design the Tesla. (Tesla Motors isnow suing Fisker for infringing its intellec­tual property.)

Mass­production plug­ins are not faraway either, and the rising price of petrolmakes them look more attractive by theday. General Motors intends to launch aplug­in hybrid called the Volt by 2010, andToyota plans a plug­in version of the Prius.Most of the other big car �rms are makingme­too noises. Only Honda and Mercedesseem to be sticking enthusiastically to fuelcells. It is all very encouraging. But whatwould really make a di�erence would be abreakthrough in battery technology.

At the moment, lithium­ion batteriesare the favoured variety. This kind of bat­tery uses lithium in its ionic form (ie, withthe atoms stripped of an electron to makethem positive). When the battery is fullycharged, these ions hang around one of itselectrodes, the anode, which is usuallymade of graphite. During operation, theions migrate within the battery from thiselectrode to the other one, the cathode,and electrons (which are negatively

charged) pass between the electrodesthrough an external circuit. It is that currentof electrons which drives the motor. Thecathode may be made of a variety of ma­terials. Cobalt oxide is traditional but ex­pensive. Manganese oxide is becomingpopular. But the future probably lies withiron phosphate, which has less of a ten­dency to overheat, a problem that has re­sulted in battery recalls in the past.

Iron phosphate certainly will be the fu­ture if General Motors has anything to dowith it. GM is collaborating with A123Sys­tems, a �rm started by Dr Ceder’s col­league Yet­Ming Chiang, to develop batter­ies with iron­phosphate cathodes for theVolt. A123’s particular trick is that the ironphosphate in its cathodes comes in theform of precisely engineered nanoparti­cles. This increases the surface area avail­able for the lithium ions to react withwhen the current is �owing, so such batter­ies can be charged and discharged rapidly.

The Lightning, too, is making use of na­notechnology. Its batteries, developed byAltairnano of Reno, Nevada, replace thegraphite anode with one made of lithiumtitanate nanoparticles. The �rm claims thatits batteries are not only safer (graphite canburn; lithium titanate cannot), but can alsobe recharged more rapidly. Using a 480­volt outlet, such as might be found in aroadside service station, the job should bedone in ten minutes.

Dr Ceder reckons he may be able to doeven better than this. His version of aniron­phosphate battery can charge or dis­charge in ten seconds. It, too, could be re­

charged rapidly at a roadside �lling station.He reckons the process would have to becontrolled to stop overheating, but a safere�ll would take only �ve minutes. And hethinks batteries might get better still.

The 30,000­compound questionAt the moment the process of �nding bet­ter electrode materials is haphazard, but DrCeder proposes to make it systematic.Over the centuries, chemists have discov­ered about 30,000 inorganic chemicalcompounds (those that are not basedaround carbon skeletons), almost any ofwhich might theoretically be suitable ma­terial for an electrode. Examining the rele­vant properties of all of them in the labora­tory is out of the question, but Dr Cederthinks he has found a short cut. He is in­volved in something called the materialsgenome project, which takes the knownproperties of inorganic compounds andturns them into extremely sophisticatedcomputer models. These models are ableto calculate the quantum­mechanicalproperties of the chemicals they are mim­icking�and they seem to get it right. WhenDr Ceder has checked the predictions forhitherto untested materials by conductingreal experiments, he has found that the re­sults coincide.

The materials genome project ob­viously has much wider applications thanbattery electrodes, but that is where DrCeder has started. His computer is nowchewing its way through the chemical en­cyclopedia, looking for the likeliest candi­dates. Watch this space. 7

The Economist June 21st 2008 A special report on the future of energy 13

2

1

took place nearly three decades ago, andeven the Soviet disaster at Chernobyl ismore than two decades past. And partly byrede�ning �cheap�. The Electric Power Re­search Institute, an American industrybody, puts the cost of nuclear electricity at6.5 cents a kWh. Not cheaper than coal’s 5cents, but cheaper than coal that has had aprice put on its carbon emissions. The time,then, is ripe for a rethink.

Ernest Moniz, MIT’s leading energyguru and himself a nuclear physicist,agrees. He thinks that, on the technical sideat least, the key to a nuclear revival is to gofrom a craft­based approach, in whicheach reactor is a bespoke thing of beauty,to a manufacturing approach, in whichmodules of components are made in fac­tories and simply bolted together on site.

The other modern desideratum, he be­lieves, is �passive safety�. This seems to bethe same as what engineers used to call�fail­safe�, but perhaps the marketing de­partment no longer approves of the word�fail� getting anywhere near a reactor.What it means is that safety measures kickin automatically in an emergency ratherthan having to be activated. That can besomething as simple as con�guring thecontrol rods that regulate the speed of a re­action so that they drop by gravity ratherthan having to be inserted.

Both Dr Moniz’s preconditions are be­ginning to be met. The world’s three largestnuclear­reactor �rms are hoping for salesof reactors whose designs have been up­graded to be more �bolt­together� and pas­sively safe than their predecessors. Accord­

ing to CERA there are plans in Americaalone to build 14 AP1000 Westinghouse re­actors, six General Electric economic sim­pli�ed boiling­water reactors, two or moreGE advanced boiling­water reactors andseven of the French �rm Areva’s latest de­sign, the European pressurised reactor.

New generationIndeed, the idea of modularity can betaken even further. Toshiba, a large Japa­nese engineering �rm, is planning some­thing known as nuclear batteries: factory­made sealed units with an output of 10 me­gawatts and a lifetime of 15­30 years. Whenthey stop working, you simply send themback to the factory for disposal.

The acme of modular, factory­built,passively safe reactor design, however, isfound in South Africa. People there havebeen experimenting with so­called peb­ble­bed reactors for decades. They hope tostart building one for real in 2010. A peb­ble­bed reactor is fuelled by small spheresthat are, in essence, tiny reactors in theirown right. They are made of uranium ox­ide (the fuel) and graphite (a substance thatslows down the �ying neutrons that causenuclear �ssion). Pile enough pebbles to­gether and a chain reaction will start. Noris any complicated pipework required toextract the heat. All you need do is run aninert gas such as helium through the peb­bles and it will collect the heat for you.

The design also looks like the ultimatein passive safety because a phenomenoncalled Doppler broadening, whichchanges the speeds of the neutrons andmakes them less likely to cause �ssion,shuts it down automatically if it over­heats�though critics argue that the graph­ite in the pebbles is a �re hazard, and thathelium is so leaky that there is a risk of airgetting into the system and starting a �re.

None of these ideas deals with thequestion of nuclear waste. But that islargely a political problem, not a technicalone. Though it sounds like a cop­out, thebest answer really is to bury the stu� forthe time being. That should be done inplaces where it can easily be recovered forreprocessing one day when technologyhas caught up. But it is also worth notingthat buried, unprocessed waste cannot beused to make bombs. 7

AS SAMUEL GOLDWYN wisely ob­served, you should never make pre­

dictions, especially about the future. As faras predicting the technological future isconcerned, people almost always eitherovershoot or undershoot. Holidays on themoon by 2000, as forecast in the 1960s?Not exactly. A quick hop out of the atmo­sphere, courtesy of Virgin Galactic, is thelimit of that vision for the moment. On theother hand, a seemingly boring way oflinking computer �les full of data on sub­atomic physics can turn into a world wideweb of information in half a decade.

In retrospect, this special report will no

doubt be proved to have been guilty ofboth over­ and undershooting. It has be­gun from the premise that big changes areafoot in the energy �eld, and has tried topick the technologies most likely to be im­portant. Some outcomes are mutually ex­clusive. A truly electric car would elimi­nate the need for biofuels, except, perhaps,in aircraft. Truly cheap biofuels might priceelectric cars out of the market. A break­through in the capture and storage of car­bon dioxide would bring coal back intoplay with a vengeance. Geothermal maybe better than solar. Solar may be betterthan wind.

The report has ignored some technol­ogies because they will not get anywhere.Fusion, that favourite of fantasists, is 30years away, as it always has been and prob­ably always will be. Giant satellites collect­ing sunlight and beaming the energy toEarth as microwaves are an idea of heroicproportions, but enough sunlight getsthrough the atmosphere to make them ir­relevant. Other technologies may make acontribution, but only on a small scale. Theidea of �oating platforms that capturewave energy is technically feasible, but itseems more trouble than building windturbines. Tidal power works but, even

Flights of fancy

The world of energy must change if things are to continue as before

14 A special report on the future of energy The Economist June 21st 2008

2

www.economist.com/rights

O�er to readersReprints of this special report are available at aprice of £3.50 plus postage and packing. A minimum order of �ve copies is required.

Corporate o�erCustomisation options on corporate orders of 100or more are available. Please contact us to discussyour requirements.

Send all orders to:

The Rights and Syndication Department26 Red Lion SquareLondon WC1R 4HQ

Tel +44 (0)20 7576 8148Fax +44 (0)20 7576 8492e­mail: rights@economist.com

For more information and to order special reportsand reprints online, please visit our website

Future special reportsCountries and regionsThe Koreas September 27th

Business, �nance, economics and ideasTerrorism July 19thThe business of sport August 2ndGlobalisation and emerging marketsSeptember 20thThe world economy October 11thCorporate IT October 25th

Previous special reports and a list offorthcoming ones can be found online

www.economist.com/specialreports

more than hydro, it depends on geography.And the idea of liberating hydro from geog­raphy with small, free­standing turbinesmay have local applications, but maintain­ing such turbines is far more trouble thantaking a spanner to a windmill.

All sorts of wacky but intriguing ideasare being looked into, such as �ying tur­bines that would exploit the high winds ofthe jetstream. And so are perfectly sensibleones, such as ultracapacitors for storingelectricity, that are now niche products butmight suddenly blossom, to the embar­rassment of prophets. Maybe, too, the hy­drogen economy will rear its head again�but only if a way can be found of storingthe gas easily and at high density. Thatwould require a material that can absorblarge volumes of it. One for Dr Gerber’smaterials genome project, perhaps.

This report has also ignored the ques­tion of e�ciency, except in the special con­text of smart grids. The idea of �nega­watts�, as improvements in e�ciency aresometimes known, has always been a fa­vourite of greens. But there is too often agleeful hairshirtedness to their pronounce­ments, which helps to explain why high­pro�le changes such as the introduction ofenergy­e�cient light bulbs are viewedcynically by so many people.

In any case, a lot of e�ciency improve­ments just happen in the background, aspart of most businesses’ continuoussearch for cost savings. Car engines, for ex­ample, are much more e�cient than theyused to be, and are likely to become stillmore so. The reason that American cars aresuch gas­guzzlers is not that their engineshave got worse but that the cars them­selves have got heavier.

Besides, as Robert Metcalfe, the net­working guru, said at a recent conference:�You are not going to conserve your wayout of the problem.� The need to keep do­

ing the same thing�consuming energy inever larger quantities�is a force for change.Price, political security and environmentalpressures are all pushing in the same direc­tion. How quickly that change will happenis hard to tell, but it is wise to remember thepower of compound interest.

Sunlit uplandsIn some �elds, such as information tech­nology, change happens suddenly or not atall. In others, such as energy, it can happengradually to start with, but as the curve ac­celerates upward there comes a pointwhere things move very fast. Ten years agowind turbines were marginal. Now theyare taken seriously, and in another decadethey may contribute as much as a �fth ofthe world’s electricity.

The same could happen to solar energy,which is ten years behind wind, and ge­othermal, with a 20­year lag. Whether itwould happen faster if carbon emissionswere charged for at an honest price is a

moot point. Certainly, that is the only wayto bring about the widespread adoption ofcarbon­dioxide capture and storage. Butfor the rest, the best way might, paradoxi­cally, be what exists now: a threat that isreal enough for electricity generators toprice it into their future calculations with­out a�ecting their existing plants.

The lack of new coal­�red capacitycreates a real opportunity for alternatives,among them renewables. But the lack ofan actual carbon price still keeps the cost ofexisting electricity down, and thus the ne­cessary incentives in place to make Goo­gle’s cheaper­than­coal equation a reality.

If and when such cheaper alternativesarrive, the markets of Asia will open andMr Khosla, an Indian­born American, willsee the fruits of his adopted homeland rollout into his native country. It will be a longtime before King Coal and Queen Oil aredethroned completely, but their reigns asabsolute monarchs of all they survey arecoming slowly to an end. 7