Calculations on energy, environment & economics

Basic energy conversion calculations
1. a. b. c. d. e. f. In Sweden, roughly 140 TWh/year of electricity is produced. How many joules is this equivalent to? Express your answer in PJ/year. How many barrels of oil are required to produce this much electricity? How many kilograms of enriched uranium? (the energy released from fissioning a U235 nucleus is roughly 200 MeV (million electron volts). How many kilograms of wood? How many kilograms of coal? How many cubic feet of natural gas (at 25°C and 1 atm)? Assume the following conversion efficiencies: Biomass and coal to electricity, 35% Oil to electricity, 40% Natural gas to electricity, 45% Nuclear to electricity, 33% (Note: do a rough estimate for nuclear power, i.e., you do not have to consider the fact that plutonium is also fissioned). Energy content in fuels: Coal 32 GJ/ton; wood 20 GJ/ton; Oil 6.1 GJ/barrel; Gas 1.1 MJ/m3 2. The capital recovery factor is given by CRF =Y


1/(1 + r Y

Estimate its value for T=30 and r=0.05. What does the CRF value stand for? Compare with say a loan on a house. 3. Rewrite the expression for the capital recovery factor in continuous form, i.e., instead of assuming that the present value of a stream of future cost is equal to

L C, /(1 + r Y you assume that it is given by



f C,«


dt . Expressing the capital

recovery factor is easier for quick calculations.


The figure and table below displays possible emissions of carbon dioxide (C02) during the coming two centuries if we are to meet a climate target of a maximum increase ofthe global mean temperature of 2°C above pre-industrial levels. The reason why the carbon emission pathways differ so much is that the high emission scenario is obtain for a low climate sensitivity, and that the low emission pathway stems from a case with a high climate sensitivity.

a. Calculate how large cumulative emissions of C02 we can allow over the coming century in the three cases (high, middle and low climate sensitivity). Compare this with the carbon in the current reserves of coal, oil and gas - roughly 500, 120 and 80 GtC, respectively.


e 0 'iii

0 0

18 16 14 12 10 8 6 4 2 0 2000


-1.5 C climate sensitivity --~ 3.0 C climate sensitivity .. -4.5 C climate sensitivity

ca .c













C02 emissions to reach max 2°C climate target Climate sensitivity (OCfor CO2-doubling) Ar 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 1,5 (GtG) 7,2 8,6 10,1 11,4 12,4 13,2 13,9 14,4 15,0 15,6 15,6 3 (GtC) 7,2 7,7 8,6 9,1 8,8 7,7 5,8 4,7 4,0 3,3 2,8 4,5 (GtC) 7,2 7,4 7,2 6,6 5,3 3,7 1,7 0,0 0,0 0,0 0,0

Energy scenario (primary energy) (EJ) 409 499 610 718 826 934 1038 1142 1245 1349 1453


b. Assume that we have to follow the lower emission path (high climate sensitivity) to meet the 2°C target. Using the energy demand figures in the table below (taken from the IPCC IS92a scenario, being a Business as Usual (BAU) scenario), calculate how large share of the energy supply that has to come from carbon neutral energy sources in the years 2020 and 2050, under the assumption that the average carbon content of the fossil fuels still used is 20 gC/MJ? c. If the low carbon emission scenario is to be reached, policies increasing the cost of emitting carbon, and hence energy, must be implemented. This will also strengthen the incentive for improvements in energy efficiency. Assuming that energy efficiency can be improved by 0.7 percent/year compared to the scenario

presented in the table, how large share of the energy supply then has to come from carbon neutral energy sources in the years 2020 and 2050? d. Estimate how much we have to increase the installed capacity of carbon neutral energy, in MW per day, on average in the period 2000-2050 (and assuming a load factor ofSO%), in the case where energy efficiency improves by 0.7%/year. How many nuclear power plants does that correspond to per week? The share of carbon neutral energy in the year 2000 was 12 percent. Discussion: Given the allowable emissions of C02 over the coming century, are there reasons to restrict the use of oil and gas today? 1.2 Estimate the emissions of CO2 per kWh heat produced from a natural gas (NGCC) for combined heat and power (CHP) production, and a biomass CHP plant in Goteborg, when assuming that the electricity produced replaces coal electricity "on the margin" (i.e., calculate both the direct emissions from combustion and the reduction in emissions from the reduction in production of coal electricity). The emission factor is 15 gC/MJ and 25 gC/MJ for natural gas and coal, respectively. Assume an alfa factor (the ratio between electricity and heat production) in the natural gas plant of 1.6 and in the biomass plant of 0.6, both having a total efficiency of 90 percent. Further assume that the coal electricity displaced has an efficiency of 33 percent. Discussion: Is it reasonable to assume that coal is always on the margin? The figure below shows the Swedish export and import of electricity in the years 2003-2004. As can be seen Sweden is only at times a net importer of electricity, and the major imports is from Norway where hydropower dominates.
5 1() 15 20 25


35 40

46 60




20 26


35 40

4!5 50




800.-----------------------------------------700~--------------~------------------------600 600 400 300


III Tyskland



1 200



100 0

~-100 -200 ·300

-4oor-----~--~---------------Finland ·5oor-------------------------~ Norge ·600 r-----------------------......:_--\J.I--------II Tyskland ·700 r-------------------------~c---------1300 '-------------------------------------5 10 10 20 25 30 35 40 45 50 ,__-'---'---'--,_ 20()'3 5

II Danrnerk




20 25 30 35 40 45 50 2004.l--_'__"___"___'_'


Discussion: Consider a car company that makes a strategic decision to invest in electric cars as a way to reduce CO2 emissions. What would the implications for CO2 emissions be? Let us assume that certain groups in society protest against these ideas, and base their arguments on the fact that coal is on the margin in most electricity systems in the world. Is the argument reasonable?


If not.8.. while the Avensis costs 180000 SEK and has specific fuel consumption of 7 litrell OOkm. the Nordic electricity system is 400 TWh/year and if we assume that there will be no increased supply from other electricity sources in the system)? Assume that the electricity price prior to the close down is 20 ore/k'Wh. How does the choice of discount rate affect your results? Discussion: Why do people in some cases buy more energy efficient products that are not cost-effective (i. at what price is the Prius competitive? c. ~v Read Jonasson & Sanden in the course literature and discuss with your friends.2 Assume that the long term price elasticity for gasoline is -0. a Toyota Prius and a Toyota Avensis. Discussion: Is this estimated price increase reasonable? What do you think would happen in reality? I) 2.3 Consumers may choose between two cars of similar size. What is the increase in the electricity price (assuming that the generation if the annual output from the plant was 5 TWh. a. fuel price of 12 kr/litre and a 10% discount rate.e. while in other cases they do not? . Assume further that governments would like to reduce the demand for gasoline with 50%.1/ I 'V " '/\1.1 Assume that a Swedish nuclear reactor is closed down and that the short run price elasticity is -0. How large CO2 emissions are produced if the heat is generated through the use of an electric heat pump (with coal on the margin. Assume a carbon emission factor for coal and oil equal to 25 gC/MJ and 20 gC/MJ. Which energy source would be on the margin in the electricity system in the future under the assumption that there is a cap-and-trade system in place with increasingly stringent caps? Assume that the heat demand in a house is 20 000 kWh/year. The Prius costs 230 000 SEK and has specific fuel consumption of 5 litre/l00km. produced at 33% efficiency) or in a boiler that bums oil? Assume that the heat factor in the heat pump = 3. I.Further assume a yearly mileage of 20 000 km. How a high price increase is needed in order to achieve such reduction? Discussion: Do you think that the estimated price increase is reasonable? What else could happen if the price increased by that amount? t) ( ~ 2. does not pay back in economic terms).4 .1. Is the Prius a cost-effective option for the consumer? /o \'\' \) 1/ b. Discussion: Is the answer different if there is a cap-and-trade system in place on all large point sources on CO2 (like power stations) and a tax on emissions from local sources? V2. a car lifetime of 10 years. respectively.

b.htm for fuel efficiency levels and the aviation industry's view on climate change.4 MJ/km/passenger (4. How large share of the current global electricity supply (about 18000 TWh) comes from wind and solar power.wikipedia. .5 MW. How many wind mills would it take to reduce global carbon emissions by 1 GtC? Assume that the wind power would replace current production of electricity from coal (with a global average power plant efficiency of 33 percent) and a wind power plant size of 1. b. How far do you have to transport the pellets before the energy used for the transport equals that in the pellets? Discussion: What implications does this have for the possibility of reducing transports in society by raising taxes (or through rising oil prices)? Consider both the impact on demand and the impact on fuel choices and transportation mode switches? What does it say about the viability of transport and trade in biomass for energy purposes in a future C02-neutral energy system? 2. How much electricity would these windmills produce? I Check efficiency# note-4 or http://www. from Ulricehamn to Goteborg (roughly 100 km).org/whatwedo/environment/fuel efficiency. How large part of the cost of the tomatoes (say 30 kr/kg) does the transportation costs take? Assume that the truck drive on diesel at a price of 10 kr/litre and that the fuel economy of the truck is 35 litrell OOkm. as some have suggested.1 The global installed capacity of wind power and solar PVs today is 59000 MW and 5000 MW. How large C02 emissions do someone traveling from Sweden to Thailand emit for a return trip? The distance is 8641 km (check http://www.2a. How large is the transportation cost in this case. Assume that the fuel consumption when traveling by air is 1. b.geobytes. How much would the ticket increase if you would add a tax on carbon equal to 15 Euro/ton C02? Discussion: Would it be a good idea. The carbon content of coal is 25 gC/ litres/100 km!). Assume instead that the truck transports the same amount of wood pellets for energy purposes within Sweden. respectively.iata.5a.2.htm for trips between your favourite cities). Assume that 20 tons of tomatoes are transported by truck from The Netherlands to Goteborg (roughly 1200 km). to include aviation in the EU emissions trading scheme'? 3. assuming a load factor of 25 percent? 3. compared to the cost for pellets (roughly 2000 kr/ton)? How large is the energy used for the transport compared to the energy content of the pellets? Assume an energy content in the pellets of 17 MJ/kg and an energy content in diesel of 41 MJ/litre.

Again take into account the energy needed for the production of the wind mills. given that a 1. roughly 6 GJ/m2 of solar cells. Estimate the costs for electricity generation from solar PV s. Vm3 • 1uf/4. decreases or does not change) to the electricity price (if we assume that the electricity price equals the marginal variable production cost) when a not negligible amount of wind energy is added to a system with coal condensing generation as the marginal technology? 3. 30 percent/year? d. E (in kWhlyr). How large an area would the windmills occupy.5 40 85 1200 1. where k is a constant depending on the wind speed distribution (here you can assume k=2. Assume that the average wind speed at good locations both in Sweden and Brazil can be 8 rnls and that the wind turbines in a wind farm is spaced 6d apart sideways and 8d apart lengthwise. using the data given in the table below. from a conventional wind turbine can be approximated by the formula E = k . Assume that the energy inputs needed for the cultivation and transportation of the biomass is supplied by ethanol produced from the biomass (to make the . equivalent to roughly half a year of electricity production from the wind mill.3 What happens (increases. Assume a lifetime of 30 years and a discount rate of 10 percent. Assume a solar PV efficiency of 15 percent and that 10 percent of the surface area in a solar PV park is used for roads.5 MW windmill has a rotor diameter (d) of 80 meters and they are spaced 6*d apart in width and 8*d in length? Compare to the total area of Sweden. Solar PV Investment cost ($/kWe) Fuel cost ($/GJ) Efficiency (%) Load factor (%) Wind Biomass Coal 5000 1000 20 25 1300 2.c. 3. at two different locations. of different renewable energy sources. and the carbon price in the EU trading system. i. etc.5). Sweden and Brazil. Vm is the annual mean wind speed in (rnls) and d is the rotor diameter. How long would it take to reach this level of wind energy production if installed capacity would grow at the same rate it has done in the last decade.5a. wind. about 500 US$/tC. What carbon price would be needed to make solar.. wind or biomass electricity competitive with coal? The carbon content in coal is 25 gC/MJ. biomass and coal. Annual electricity production. Assume that the solar cells have a lifetime of 25 years. c. 450 000 knr'. calculate the power density.e. Using' the data given below.5 45 85 b. Compare with the CO2 tax in the Swedish transport sector. which has been close to 100 US$/tC.4a. Also take into account that the energy needed for the production of the solar cells. How does the choice of discount rate affect the results? Does a higher discount rate make the renewable energy sources more or less competitive with coal? 3. with a life time of 30 years. how much power an energy sources produce per area on average in (in kWhlm2/yr).

27 percent natural gas. .whole system carbon neutral) with a net conversion efficiency of 35 percent. There are currently around 400 nuclear reactors in the world. How much would the carbon emissions in the world increase? b. and the fuels respective emission factors are 20 gC/MJ.1. To do this assume that the biomass is produced over a cultivation season of 5 months and that average solar insolation during this time is 70 percent higher than the yearly average. Assume that all the nuclear electricity in the wciRt-w uld be replaced by coal fired power plants. 15 gCIMJ and 25 gC/MJ. excluding nuclear. electricity generation / nuclear is 2 645 TWh/year. from a. How many more would need to be built if we were to replace all fossil fuel based electricity with nuclear power. gas and coal plants is about 37 percent. How many nuclear accidents with large external release will statically occur during the 21th century given an expansion of nuclear power? The risk of accidents can be found in the article "A nuclear solution to climate change". 33 percent and 32 percent. hYdrObaSedelectric~'5uamounts to 2 630 TWh/year and total global electricity supply amounts to(l!6 663 Wh/year.5 GtC/year.2. the load factor is assumed to be 80% and 200 MeV is released for each fission U-235 or Pu-239. Brazil Solar PV Biomass Solar insolation . How much carbon emissions would one then save? The current global electricity mix. c. estimate the efficiency of the photosynthesis. 4. Using the data on the average solar insolation and the gross biomass yield in Sweden or Brazil. Currently. both in Brazil and Sweden. is roughly 14 percent oil based. However. depending the latitude of placement.yearly average (W/m2) Gross biomass yield (dry tons/ha/yr) Energy content (GJ/dry ton) Conversion efficiency to electricity (%) Energy inputs for cultivation and transportation GJ/ha Sweden 250 21 20 40 5 100 12 20 40 10 b. Discussion: Why is the power density of renewable energy sources of interest? Is there a qualitative difference in the area requirements for the different renewables? The value for solar insolation is the light that strikes a horizontal plane. Does this change your estimate of land demand for PV s? Assumptions on nuclear power to use in the exercises below: The conversion efficiency from heat to electricity is 33%. the world emits 6. and 59 percent coal. solar PV s are often placed at an optimal angle. hydro and other renewables. respectively. The global average efficiency of current oil. 4.

A nuclear light water reactor is expected to produce 13.000 MWd [Mega. Assume that 95% of the U-238 can be bred into Pu-239 in a closed loop system.Today 9 EJ electricity is produced annually from nuclear power. given that it can be a cheap way to mitigate climate change? Are large scale accidents. worse that the more low-intensity health effects of. One way to prolong the uranium resource base is to breed U-238 to Pu-239 and thereafter fission Pu-239. what implications do you think this would have for nuclear weapons proliferation? How can one deal with the issue of nuclear weapon proliferation internationally? . Half of plutonium is eventually fissioned and half remains. i. like a nuclear power melt-down.. How long will the current uranium reserve last? b. 0.5a.6 atoms of plutonium are generated for every uranium atom that is fissioned. and there after remain on that level. a.Watt-Day] of electricity per metric ton of nuclear fuel (assuming a conversion efficiency of 33% from thermal energy to electricity. 4. How long will reserves last in this case? Discussion: What are the possibilities to increase the uranium reserve? How sensitive do you think the price of nuclear electricity is to the price of uranium? If the breeder track is chosen as a way of reducing carbon dioxide emissions on a large scale.~ Assume further that the uranium 235 concentration in the spent fuel is(1% 'aad that 70% of the energy released in the reactor comes from the fissioniJig_Qf'D235.3. Estimate how much plutonium is produced annually. What is the energy release per fissioning. Discussion: What are the possibilities for utilizing the plutonium in spent fuel as weapons material? How could you reduce the likelihood of plutonium in spent fuel being used for weapons production? 4. b. Assume that this output increase with 1 EJ a year the next hundred years with reactors with a capacity of 1 GWe.. e. Estimate the amount of uranium that is consumed every year in a 1 GWe reactor.9Mton.g. Finally the global uranium reserves that can be mined for a cost of 130$/kg is assumed to 3. Compare with the amount of plutonium required to build a nuclear bomb.e. air pollution from fossil fuel use? 4.4 In a light water reactor. that 4% of the atoms in the fuel are fissioned. Only OnCethrough cycles are usea. Assume a burnout level of 4%. A common energy unit in nuclear energy analysis is MWd. Assume that electricity from nuclear power increases exponentially from 9 EJ today (2005) to 225 EJ in 2100.wnere the uranium is enriched to_1_~g-2~. consider only the weight of uranium). Discussion: Is the risk of nuclear accidents worth taking.

5 10 50% 12. transports. . comes from oil. c. having a carbon content of20 g C MJ(HHVrl. 5.5 GJ (nr' etanol)". Assume that we want to replace 50% of all gasoline use in Sweden (roughly 6000 million litres).5 80% 2. and the share of process energy not coming from residues.Swedish ethanol produced from wheat and Brazilian ethanol produced from sugar cane . Then the inputs for cultivation. a yield of 10 ton dry matter (DM)/ha/year.0 40% 12. Assume that the energy used for cultivation.3 22.3 b. What is the net ethanol yield per hectare in Brazil and Sweden.5. and a conversion efficiency of the old coal plant of35% and the new biomass plant of 40% from heat to electricity) .2a. respectively? d. taking into account that a gasoline car has a fuel economy that is roughly 30 percent lower than the ethanol car (because of lower energy density of ethanol).5 5. The energy content of gasoline is 34.5 MJ/litre and the carbon content 20 g C/MJ. (Note that in reality. Compare these emissions with the emissions from a gasoline car.0 18.0 17. transports.using the data given in the table below.5 million hectares). 20 GJ/ton DM.5 50% 0. you also get a by-product from the ethanol production process that contains energy. The energy content of ethanol (HHV) is 23. Assume that we want the system to be carbon neutral. This aspect also needs to be considered when making a full net energy analysis of ethanol production from wheat). Is currently sold as animal feed and contribute to the economics of the business. Parameter Wheat ethanol . Calculate the emissions of CO2 per km driven for two alternatives for production of ethanol from biomass . How large areas of plantations would be required? (Assume a load factor of7000 hours. has to come from ethanol.Brazil Biomass yield (tDM ha' (3(1) Energy content of biomass (GJ HHV (tDM)"1) Energy use for cultivation (GJ ha" y(1) Efficiency of ethanol prodoction (in HHV) Energy use in ethanol production (GJ (rn" ethanol)") Share of energy use in process from biomass residues Energy use for transport of biomass & ethanol (GJ (rn" ethanol)") 5. Further assume that the ethanol car has a fuel economy of 10 litre/100km. How large areas would be needed compared to the agricultural area in Sweden (2. if we used Swedish wheat ethanol or if we used Brazilian sugar cane ethanol? Discussion: What conclusions do you draw for wheat based ethanol? Should governments support it or not? Which role might it play to pave the way for new alternative fuels? Sweden has recently introduced a tariff on Brazilian ethanol .Sweden Sugar cane ethanol . and the share of process energy not coming from residues. Assume further that the biomass would come from short rotation forestry in Sweden.1 Assume that the electricity from a 1 GWe coal fired power plant would be replaced with a biomass fired power plant.

respectively? Explain why. will increase the cost of energy. How large area would be needed to supply this if the chosen energy source was wheat or sugar cane ethanol. heat and transportation fuels. the cost of biomass is 3 USD/GJ. All cost data is summarized in the table below. b.5+2) 4 (=2+2) 10 Biomass Oil Oil with tax =6 USD/GJ The total demand for heat is 10 EJ/year. This will have 6. The introduction of a price on carbon emissions. respectively? Use data in exercise of the calculated production yields of ethanol per ha.2. in exercise 3. Assume further that it also costs 2 USD/GJ to convert oil into gasoline or diesel and that the conversion efficiency is 100%.3 . and assume that we also can make use of 70 EJ of residues from the global agricultural and forestry sectors. from wheat or sugar cane. and that there are two energy supply sources. What is the scarcity rent for biomass (per GJ biomass) if it is used for heat production (when the oil tax is employed)? What is the scarcity rent for biomass if used for transportation fuel? Discussion: Have you carried out a descriptive.2 Assume that the global energy demand in the transportation sector in the year 2100 is 200 EJ. Assume that biomass is converted into biofuels with a 50% conversion efficiency and that the conversion cost is 2 USD/GJ.used for blending in gasoline (the major share of ethanol use in the Swedish transport sector). for transportation fuel 4 EJ/year. prescriptive or predictive analysis? Motivate briefly your answer.5 Gha. In the absence of a carbon tax or a carbon policy. Compare the result with the global agricultural land area presently: 1. Heat (USD/GJ) 3 2 8 Transportation fuel (USD/GJ) 8 (=3/0. oil is cost-effectively used in both sectors (cost-effective here means that it is the solution that lowers the total costs for the entire energy system). Is this good? 6. But what does the cost-effective solution look like if the carbon tax is set at 6 USD/GJ oil so that the cost of using oil in the heat and transport sector is 8 and 10 USD/GJ. oil and biomass. The cost of extracting oil is 2 USD/GJ. be it through a carbon tax or a cap-and-trade system. a. c. 6. making European ethanol competitive. but the conversion efficiency to ethanol for this resource base is only 20 percent. All other energy sources are converted into their corresponding final energy carrier at no cost and with 100% efficiency (in order to simplify).1 Assume that there are two end-use sectors in the economy. What is the opportunity cost of using biomass for heat (per GJ of biomass)? Explain why. Assume that the total maximum supply of biomass is 9 EJ/year.

forests for biodiversity production. Assume that a farmer can choose between using his/her land for biomass plantations and cereal production. at a cost ofUS$100/ton. respectively. Assume that in the future biomass could be sold at a price ofUS$9/GJ (which would roughly make it competitive with solar power for hydrogen production). Is this a reasonable approach? Is there a need for policies in the competition for land between.0$/GJ and 2. Ifhe/she can produce wheat with a yield of 5 ton/ha/yr.implications on land prices. assuming a biomass production cost 4. b. different uses? . induced by the competition for land between food and bioenergy production. how much would the wheat price have to be to make wheat production competitive with biomass plantations? How does this compare with today's wheat price ofUS$150/ton? Discussion: What implications could this possible effect of land competition on land and food prices have? In scientific analysis of the global area available for bioenergy production it is often assumed that we first set aside the land needed for food production to feed the world. etc. a. Using the yield data (minus energy needed for cultivation and production) for biomass in exercise 5. then the rest is available for bioenergy planatations. calculate how much the profit per hectare would be for biomass plantations in Sweden and Brazil.5.8 $/GJ in the two countries.

4 times higher 2. 3.2050: 80~ c. c.8 tC02 b. 1. 0.)'1 b.5 kr/ton . 2.211er/kg --'I 0·'. which has to be made upfront . respectively b. 210300 kr c. compared to the investment cost.4 tC/yr.4~.1a. 17. Allowable C02 emissions.2 2. 1. (Alternatively.3a.7 ore/kWh 2.8 tC/yr 2. Energy efficiency 2. net -280 gC/kWhheat Biomass CRP: direct emissions 0 gC/kWhheat (biomass is assumed CO2-neutral). 504 PJ/year 207 Million barrels of oil 18. Cumulative CO2 emissions in the high.Background exercises 1a. ~ty(fKin J. Average increase in installed capacity of carbon neutral energy: 920 MW/day (or roughly 7 nuclear reactors per week globally!) 1.065 CRF =r/(l-e-rr) 1.5a. investments and systems thinking 1. No (the annual cost for the Prius is 48 400 kr. d.1 2. middle and low climate sensitivity cases: 355 GtC. 1. which occurs over the life-time of the car. net -164 gC/k Whheat 1. reduced emissions 164 gC/k Whheat. 0. and 1260 GtC. reduced emissions 436 gC/kWhheat. f. Since a higher discount rate will raise the annualized cost. compared to 45300 kr for the Avensis) b. Heat pump. 2050: 72~ d. one can see it as a higher discount rate will put less weight on the fuel cost reductions. e. 645 GtC. 2.3 1.therefore a higher discount rate disfavors the Prius).6 ton U-235 72 Mton 45 Mton 1018 Gm3 0.4a.4 Oil burner.2 Natural gas CRP: direct emissions 156 gC/k Whheat. b. 27 euro . 2020: 19~. 2020:41~.~ ov= ~ 2. it will make the Prius less attractive for the consumer.

Renewable energy 0. Brazilian sugar cane ethanol: 118 GJ/ha 4.8 c/kWh.2 EJ 13 years 340 000 km2 Since wind electricity has no (or extremely low) running costs. 'L( Pi b. d. b. lowering the marginal cost for electricity (if the supply curve is upward sloping). c. coal: 2. around 1 700 nuclear plants c. 2.1 c/kWh. 0. 41 years b.1a.3 Around 200 MeV /fissioning 4.78% 1.2. biomass: 4. Solar PV: Brazil.4a. 41 gC02/km 177 gC02/km Swedish wheat ethanol: 23.41%years 5. Bioenergy 5.4 Mha for the Swedish wheat ethanol case. Sweden . the electricity produced from wind will end up in the lower part of the electricity supply curve (i.3. whenever the wind is blowing. 3. Energy system scenarios and economics 6. wind: 4.4 GJ/ha.3 accidents 4..1 5. biomass: 60 $/tC c.2a.7 GtC 4.e. it will make the solar and wind less competitive. the electricity produced will be sold. 716 kg U-235.52 kWhlm2 Wind: 21 kWhlm2 Biomass: Brazil.9 c/kWh b. ~9-k-g_:Pu~(ittakes 5-1-0kg of plutonium to build a bomb) 4.230 kWhlm2. c.1 million 13.4.4a.0.5a.2a.7-1% 4.1 3.5 Mha for the Brazilian sugar cane ethanol case 3. b. wind: 96 $/tC. This will shift the supply curve outwards. no matter the electricity price on the market at that time).5 kWhl~. 3. For biomass. Since a higher discount rate will raise the annualized cost and put less weight on the running costs (which occurs over the life-time of the plant). d. 315 000 ha Swedish wheat ethanol: 85 gC02/km. Solar PV: 1 370 $/tC. 3.6 GtC (assuming an efficiency of the replacement coal power plants of 40%) b. Sweden . Nuclear energy 4.2 2.3 kWhlm2 b. this effect is less pronounced since both the investment cost and the running cost is higher than for coal.5a.3 6. 1. Solar PV: 30 c/kWh. since their investment cost is higher.1a Cost-effective solution: Heat (10) Oil 1 Biomass 9 Transport (4) 4 . Brazilian sugar cane ethanol.

1 USD/GJ c. Sugar cane ethanol case: 1.6 Gha 6. Sweden: 1 100 US$/ha.3a.2 Wheat ethanol case: 7.9 Gha. Heat: 5 USD/GJ biomass.b. Sweden: 320 US$/ton. transportation: 1 USD/GJ. 6. Brazil 2 500 US$/ha b. Brazil 600 US$/ton .

2006 In this brief paper. solar PV or ethanol from com? What are the energy and environmental consequences of closing down a power plant.. nuclear power or ethanol from com? What are the CO2 emissions associated with nuclear. I first present a series of "typical questions" that are common in energy systems analysis just so as to give you a general feeling for the problems one might confront.g. e. likely to be able to gain enough political support to be implemented? • • • • • • • • • • • • • . 1. Typical questions • • How is the district heating system affected by energy efficiency improvements in residential heating? How are oil security concerns affected by efficiency improvement in transport vs a policy that stipulates that a certain fraction of the transportation energy use should be based on biofuels? How much energy is used in the production process of.Thinking about energy systems Concepts and tools Christian Azar Chalmers October 12. Then. solar PV. e..e. Finally.. fuels and propulsion technologies? What are the implications for the electricity grid of using intermittent energy sources such as wind power? How do environmental policy instruments such as green certificates or a cap and trade system on carbon impact the answer to some of the questions raised above? How is the energy system affected by the introduction of a carbon tax? How is the electricity price affected by a cap-and-trade system? Which policies are effective in reducing CO2 emissions? Which policies are politically feasible. These concepts are also useful when attempting to answer to the "typical questions" presented below. i. Barseback? Is meat from South America consumed in Sweden more or less energy intensive and more environmentally friendly than meat produced in Sweden? What are the energy and C02 consequences of using biomass for heat. I mention a couple of things that are useful to think through when analysing energy systems.g. I present concepts that are useful when thinking about or analysing energy systems in general. electricity or transport fuels? Would a large scale introduction of natural gas in Sweden lead to reduced emissions of CO2 in Europe? What are the CO2 emissions per unit of km driven for different primary energy sources.

at hot summer day.rtypically measured in watt.\l.Ynt (kinetic ./A wherr'it' reaches eitherintefnal-equilibrium-or equilibrium with its surrounding (under the assumption that the surrounding is so large that the system remains constant). Rather it is a general concept for energy per unit of carriers . wind.6 MJ).2. Observe that power does not need to refer to electrical power.a systetl1. for instance that work (J)=power(W)*time(s )=force(N)* distance(m). Think of a cup of water. Heatial1dwork Energy may be transferred from one body to another either in the form of heat or work. The ability to extract work from a system hinges upon whether it is in equilibrium with its surrounding or notl Ifthe surroundings hold a temperature of 100C. since thermal energy is a less ordered form of energy than work. Exergy is measured in joules but note that the exergy and internal energy of a system typically is different. It should also be noted that matter can be converted into energy according to the world's most famous formula. Heat is·only. Make sure that you feel familiar with units. Exergy Thetl1. Work is power (not necessarily electric!) multiplied by time (s). Primary energy .icaLenergy). kinetic. Ifheat is transferred. sunlight. chemical. coal etc). or force multiplied by distance. Try Einsteins formula and see how much energy you could get from annihilating one kilogram of gasoline compared to its energy content. potential. m for mass and c for the light of speed).l11oY(. The Sl unit for energy is joule.the potential to induce movement (p()tyntial energy.chetl1. fossil fuels (oil. Einstein's E=mc2 (where E stands for energy. Note that it is physically incorrect to state that a services Primary energy is the form of energy found in nature and includes uranium and thorium for nuclear fission. We also have geothermal energy and tides. It comes in many forms: gravitational. Concepts Energy Energy is. 40°C. nuclear or Qr .l1l amount . Both heat and work are thus measured in joules. . say a cup of boiling water. I W=J/s. not energy content. the resulting energy form is thermal energy (an increase in the temperature of the receiving body).11)(itl1. hydropower and biomass. then you can extract work. natural gas. Note that one joule of thermal energy can not (!) be fully converted into one unit of work.)tl1. Make sure that you feel familiar with units. only conversion from one form to another. You can extract no work from that cup if the temperature of the surroundings is also 40°C. for instance that power(W)=voltage (U)*current (A). contains a lot of heat. There can be no energy consumption or production. a measure oftQ_e energy flow. Why is the difference so large? Power Energy per unit of time. Electric energy is typically measured in units of kWh (=3.0fworkthatcM be derivedorextracted-from.It can neither be destroyed nor created.

gaseous chemical fuels (natural gas..9. Sysfems··bOtii1dary When analysing a system. "A system is a set of objects together with relationships between the objects and between there attributes". The key thing to observe is that the system consists of! • objects • tl1. When defining a system it is important to be clear about the boundaries of the system. Both approaches are needed in order to get a better understanding of the system.systen« Ei1efgysysfems Energy systems consist of the following set of objects: the energy supply. which uses "hard tools" like matheml. fuelwood) and liqyig cl1.41'lergycarrie12$ include thefl11alcarriers (heat and cold which both can be delivered through district heating or cooling systems).systems analysis. el~Ptriqity£ solid fuels (coal. Then there is "soft" energy systems analysis which relies more on interviews with key actors (like decision makers in government or in industry) or-other-less mathematical. methods to analyse the system.ltipn. if your aim is to analyse the environmental impacts of a gas fired power plant in Sweden. or the consumers.. the transport of goods and people. conversion. EffeFgyservices are of course the ultimate aim of the energy system. a region. hydrogen and biogas). Enel'gY·SYstelllsAnalysis Energy Systems Analysis is the science that deals with the analysis of energy systems.betweentheobjects • andthesystemboundaries which defines what-is included in and excluded fromshe. the people who work at plants.\1:l. e.~:r:ld. ethanol etc). but this one comes from Young 1964).pptimisation.l§~§y§tem§.e·rell.l. (There are many definitions of systems. We do not demand energy as such but the services that we can derive from the energy flows. One can say that there is "l1.system concept mayor may not also include the The institutions that uphold the system (e. Energy carriers are obtained by conversion of primary energy sources. regulatory framework. it is important to define the boundaries. depending on how you draw your boundaries). The scale could be an industry.g. Expanding the system boundaries may yield different results. stprl. .ltips.l. or the manufacture of goods. charcoal.arg"energy. When analysing an energy system one analyses the objects and the relations between the objects. For instance.emipl. a combined heat and power plant with its users. a certain indoor temperature. a country or the whole world. but could lead to reduced emissions if you draw your boundaries around Europe (since the gas-fired power plant may replace electricity from coal in Denmark or Germany).\lfuel§(gasoline. you will find that it will lead to increased emissions if you draw your boundaries around Sweden. diesel.modelsxetc as a means to understand how the system may change if there is a change in a tax or if the price of energy increases.g..lg~.

and there is for that reason a debate as to whether one should use average or marginal electricity when evaluating environmental impacts. However. e.. Hydro is the cheapest source and will be used to supply at its maximum capacity over the course of the year anyway. • Instantaneous marginal electricity: if your lamp is switched on. e. It can be radically different from the marginal electricity. if you drive your car one km less per year or if you switch off your lamp? Average electricity Average electricity refers to the average electricity supply. Annual marginal electricityxif your lamp is switched on. i. Thus.e. if you switch on your lamp what electricity source would then be taken into operation? There is much debate about marginal electricity in society since the environmental impact of an energy decision often depends on the marginal impact.of. on the instantaneous or annual margin. Hydro may be the instantaneous source but since the potential for hydro is limited by the amount of water and the storage capacity we will not get more hydro over the year just because someone switched on the lamp.changesis carried out. we do not have nuclear power. Structural marginal electricity: currently one may say that a decision to phase out nuclear power in Sweden does not lead to import of nuclear electricity from Russia. Even if the average electricity in Sweden emits very little CO2 an increased use of electricity might lead to more coal. there will not be more Russian nuclear just because of our decision. if Sweden decides to make an effort to transport more • • .g. For instance. Marginal electricity is the electricity that is produced on the margin. or marginal electricity that refers to how the system responds to an instantaneous change.what happens if a marginal.g. In that sense one could argue that there is a need for another "marginal" concept. if we increase transportation by train there will be an increase in the electricity demand in society. since their plants already operate at full capacity. coal is (in most cases) the marginal supply electricity. Please note that on the structural margin. Yet. which I here refer to as structural margin. What do you think? It should be noted that there is a difference between marginal electricity on an annual basis. in other countries other sources are used). the increase in supply comes from hydro power (this is the case in Sweden... Marginal analysis Marginal analysis deals with the analysis . it is over the course of a year coal fired power plant that responds to the increase. which is the case in the Nordic electricity system which is dominated by hydro and nuclear power with very little coal. then they might make a decision to build more nuclear plants.Thinking about what is the relevant systems boundary may often be as important as carrying out the actual analysis of the system! Reflect upon that sentence once again. if Russia or Finland sees that Sweden will close down a lot of nuclear power plants. one could also have wind or biomass electricity.

and people found that "unfair" or undesireable . i.e.l1. the gas and ethanol price rose to. ~4 0 2 -supply -demand 4 quantity (unit) 6 8 Consider the electricity price: assume that you live in a country with hydro power that can be produced at say 1 USc/kWh but the maximum output is 100 TWh/year. i. and demand is equal to 5).e.e. fact it actually meant that those involved in biofuels business made some substantial profits)..'ltP. in most cases people only use marginal electricity.thecostofproducingit). In fact.g 0 c.Y§lYYmg¥99§tgf producingit(the last product tends to be more costly to produce than the initial product. Assume further that the demand for electricity rises beyond 100 TWh/year.. Price versus cost Why does a CD player cost. and this might create confusion.goods and people by train. say.. But. Markets clear.. consumers would be willing to pay 7 USD for the first unit in the graph.-2 . and that .e. prices (on competitive markets) are determined by an interplay between both thedemand. There have been calls from both government ministers and others in Sweden to regulate the price ofbiogas and ethanol (when the oil price rose.Yiit) and thesupply(i. This means that the price<ofagoodgenerallyis. one could then say that there is wind and biomass on the structural margin. there is no excess supply or lack of supply. but actually only end up paying 3 at the market equilibrium. when the price is such that the demand equals the supply (in the graph below. annual and structural marginal electricity are not commonly used.ighYrJh§ll. whereas the first unit cost only 1 USD). Would it then be reasonable to say that trains do not use electricity on the margin? What do you think? Please note also that the terms instantaneous. and for that reason I found it useful to introduce more nuanced concepts. then only 1 unit is produced and there will be a scarcity of the good. 100 USD? Is it because that is the cost of producing it (including retailing etc)? Or is it because that is what we as consumers are willing to pay for the product? In fact. and at the same time makes a decision to build more renewable electricity (so as to avoid debates about trains using coal on the margin). it happens when the price is equal to 3. But it also means that consumers often pay a lower price than what they would be willing to accept. Discuss with your friends: what is a reasonable approach to the price of biofuels? Supply and demand E' C ::::I c8 6 . consider the graph below: if governments set a maximum price level at 1 USD. producing quantity five cost 3 USD.forthe products(thewillingnessto pay fOf.

100 lWhla~ 150 ElfOrbruknlng totliit ar 2000 Here: Vind means wind. kostnad SO iirelkWh eo 70 60 51} Gasturblner 40 so 20' Vinci- kratt 10 0 0 $0 ElproduktiOnl. and energy conservation. kraftvarme fj arrvarme means combined heat and power in district heating systems. The dotted black line means the electricity consumption in the year 2000. At a price of 4 USc/kWh the demand is equal to 130 TWh/year. vattenkraft means hydropower. Supply.formaga. 100 prod. kondenskraft kol. curve A supply curve depicts what producers of a good are willing to supply at a certain price level (see hypothetical example in figure above. It is relevant both for energy or electricity supply. kondenskraft oil means oil fired power plants and gasturbiner means. then the electricity price becomes equal to 4 USc/k Wh. means coal fired power plants.4 USc/kWh. . Below you will find a supply curve for electricity in the Nordic electricity system. not surprisingly. and the Nordic electricity system below). Assume further that a carbon tax is implemented that raises the cost of producing electricity from natural gas fired power plants to 4. karnkraft means nuclear power. but some industrialists in Sweden have protested and argued that it is unreasonable that Swedish electricity prices should increase given that most of Sweden's electricity is carbon free. For instance. When drawing a supply curve for electricity or energy conservation it is sometimes common that the explicit technologies are shown. gas turbines. if the price is 6 USc/kWh the electricity consumption drops even further.natural gas fired power plants are being built. It can be seen that there is "coal on the margin" (the annual margin in the Nordic system at present). at say 5 USc/kWh. suppliers of energy efficient appliances are willing to supply products so that the demand for electricity drops by say 10%. kraftvarme industri means combined heat and power in industry. Is it then reasonable that the electricity price should increase as well? At a competitive market it will (and it will be set by the cost of producing electricity from natural gas including the cost of the carbon tax). with a cost level equal to 4 USc/kWh.

e.forecastingisto· make..and (3) asks what. then the option of using it for heat or electricity is foregone i..tnctionofeconolllic·..lre:afuturethatissufficiently far so as to become relatively independefrt'of-the-current state.~ithatarisesfromthe.that future: Thus the method is normative. It takes (1) asits. a shop owner may put bread on her shelf and make money out of it. On a more general level the opportunity stems from the fact that every time one makes a choice in favour of something you always make a choice against something and that there is a cost associated with that choice. The opportunity cost either arises from a scarcity of e. one looks at energy use in the production chain as well as when using and discarding the product. ~lli. Net····ellel'gy·. but also the energy and environmental implications of producing and getting rid of the product. Another example of opportunity costs is related to biomass use. F?r~~~~ti~g: Forecastingi~a. Back .castillg: Back-casting is an alternative method for studying the future. Illegal such markets include the black market tickets for concerts to pop groups that attract more people than there is space in the stadium.a.tB~ll§ce()fpll§t •• treng~. The price of the product is the extraction cost plus the scarcity rent.. forecasting is extremely difficult.the right.8"ll~h. This is the most dominating method in studies of the future.towards. time.g.point ofdepartl. the opportunity cost is the value of the next best purpose the asset could have been used for.Opp()r11lnitx.~t Opporamity-eostds. Galbraith once said: "The . h~. a certain .factthat. but had she kept cookies on the shelves she could have made even more money. the cost of passingJJ.thatonlyfocusesonenergyandemissio11S' i. astrology look respectable(t. space or material.m~th9gtl1a. but also past relationships that are put into computer models.8t~ . t~e us. Life cycle analysis/cradle to grave analysis These are methods that look not only into the implications of the use of the product. oil) where the price may be higher than the extraction cost.desirable ..path. The advocates for this method tend to emphasise that the future . he11lll~inga.J.g. (2) thea-asks-what.g.. there is an opportunity cost associated with that decision.l1.onto.··analysis:l Aformoflife·cycieanalysis.9(). money.. This approach is common and necessary when analysing whether say ethanol from com has a positive or negative C02 balance compared to gasoline.tm9j~.·.p Jl1~... if it is used in the transport sector as ethanol.e.K. It has advanced in the sense that it does not necessarily only refer to extrapolation of past trends.. w decisions For example..>:fM1:lll'~ t .good-may-be scarce(demandishigherthallthepbssiblesupply~[ Legal such markets include the market for exhaustible natural resources (e.refers to •tlWya. if an asset such as capital is used for one needed. The former president ofthe American Economic Association J.n~:x:tbestchQice.e. Scarcityrellt Scarcity rent .future .

costs due to emissions and noise that you perhaps inflict on others. (e. it means the interest rate at which a person. US banks). and that the future is shaped by decisions we make and not a slave under past and current trends.e.ofmoney. a cost of 105 USD the next year is equivalent to a cost of 100 USD today.. Privateversus . if a lender (such as a bank) charges a customer $90 in a year on a loan of $1000. For a person with a discount rate of 5%/year. It could be an industry that emits dust that turn your laundry black.. This meaning of the term is the focus here (and under the heading "net present value"). RealInterest · rate The real interest rate is the nominal interest rate (the value you actually pay) minus the inflation. could also be "negative". then the private cost is what you pay for (gasoline. or a power station that emits CO2 that causes damage world wide over centuries to come. animals. The second meaning is the interest rate at which national financial institutions. Discount . (e. i..g. Interest rates often change as a result of inflation and Federal Reserve policies. It is generally very difficult to make such valuations because of uncertainty and methodological issues. i. company or a government discounts future costs and benefits when attempting to compare costs and benefits in the future with costs and benefits today.g.paid fQ1: theuse. Tend to be ethically very controversial because the ways the value of life. provide Illtetestrate Aratewhich . can borrow from the central bank.oftimepreferenceahd . if you drive a car.rate There are two meanings for the discount rate. ExternaLcost Costs that are caused by an activity that are not paid for: A typical example in energy systems is the emissions of pollutants that cause damage (costs) to others but that are not paid for. social. but the social costs includes both the private cost and the possible unpaid costs that your activity caused on society as a An interest rate is often expressed as an annual percentage of the amount of money loaned (principal). Normally those two discount rates are different. It is calculated by dividing the amount of interest by the amount of money loaned.chargedor. ecosystems and future generations are treated become critical to the result. It could also be noted that external costs. • it is based on the fact that most humans would prefer to get say 100 USD today than next inherently uncertain. the car etc) but the social cost includes private costs and the external costs.. Which value is appropriate for the discount rate when comparing costs and benefits over time? There are two different approaches. For example.e.costs A private cost is what you have to pay for an activity or a good. the US federal reserve). For instance. There are several reasons for that: Firstthesocial. First. then the interest rate would be 90/1000 * 100% = 9%.

or the time value of money. and that we expect costs until the year t=T. Alternatively. Now. it is better to have the money now when we are not so rich o impatience: people are impatient (you simply want things now because "now" means more than some distant point in time) o uncertainty: people may feel uncertain about whether they actually are going to get the money in the future (or they be may think that they are going to be dead by that time) • Second. in the year 2 costs are equal to C2. then the life cycle costs are given by . . Assume that we in the year 1 have costs equal to C 1.o Expected growth in combination with declining marginal value of income: people expect to be richer in the future and since a dollar when we are richer is less worth than we are poor is valued less. Netpresent'Value If the discount rate is r. This way the discount rate may adjudicate between different projects so that society as a whole invests in the most profitable projects.e.Assume that you are going to make an investment in a plant. You will invest 100 USD today and you think that that investment is going to pay offwith 105 USD the next year. there is not enough money for every investment project. the rate of return which you at least expect to get when making alternative investments. From this perspective. The discount rate should in this perspective be equal to the opportunity cost of money. Should you invest? It depends on what other investments are available.. the discount rate arises because money is scarce (if there was no scarcity of money then the value would be zero). theratebfreturnoninvestment.Life cycle cost The life cycle cost is the discounted cost over the entire course of the project. so the discount rate should be set equal to the marginal rate of return on investment in society. then the net present value of a cost in the year t is where C is the cost. i. one may (if one chooses continuous discounting) use the following approach: which gives approximately the same value (check!).ii. If you could invest 100 USD in project B and get 110 USD then you should probably carry out that investment.

the levelised cost approach.::{il1itial investmel1tin$)/(expected yearly savings in $/year) It gives the amount of years it will take until the investment cost is back in your pockets. in this case the net present value of all the electricity that will be produced). What does that imply for the cost of electricity (USD per kWh)? Here.. is zero. Solving for l-« gives: The levelised cost can be defined as the net present value of all future cost streams divided by the net present value of the good that is produced (i. = the levelised cost of electricity per kWh (i. = C * Lf (where C= installed capacity and Lf is the load factor).e. Assume a load factor of 0. one needs to know both the life time of the plant.e.. It means that the net present value of the investment should be equal to the capital cost. You will see that this gives the equation above. the discount rate chosen and the loadfactor (see definition below). The trick is to convert US/kW into USD/kWh.85 and a discount rate of 5%.e. the value we are seeking). If the discount rate. but one could also state the cost in terms ofkW of thermal input). = the electricity supply (kWh per year). r. In order to answer to that question. T is the life time of the plant and E. k W·ofinstalledcapacit¥ Assume that the cost of a natural gas fired power plant is 600 USD/kWe• (Here the index e refers to electricity.Levelisedcapitalcostofelectricityper.. L. and I will present one approach. It is commonly used. is defined as SP:. but not so useful since it does not consider the time . How? Let us start by the assumption that the investor wants to get at least as much money back from the investment as he or she invested in the plant. Note that E. there are different approaches that one may employ. What is the levelised capital cost for electricity per kWh? How does a high discount rate affect the levelised cost of capital? Is that reasonable? Simple payback time. i. where Ie = the investment cost. then one can see more easily that the formula only states how much the owner needs to charge the consumer in order to get his/her money back. SP.

environmental fees. Cost·. then one can be pretty certain that the investment would be worthwhile even with proper discounting (but a more correct analysis is of course preferable).value of money.f~ptor is a factor that gives the share of initial loan one has to pay every year so as to be able to pay off both the loan (the principal) and the interest rate over a given period T assuming the same payment each year.L)/(l + rY where Cf is the cost of the fuel and 11is the conversion efficiency of the fuel into electricity.4 TWh/year).tQ. For a power plant it is defined as the number of kWh produced compared to the maximum output per year. From the formula. the IRR is the discount rate at which the present value of future benefits just equals the present value of all costs.}. we can see that a low conversion efficiency or a high fuel cost gives a high cost. For example. L ce= . say a few years.+ --r--=---1] 1=1 s. The second term is explained above (see title levelised capital cost of electricity). the pay back time is very short. It is given by . CapitaLrecoY~rx. S = salvage value of the investment at the end of the analysis period (parts of the investment may still be valuable and could generate a revenue if sold at the market) and N = period of the analysis. Assume that the load factor of a 1 GWe plant is 85%. refinery (or also an airline or any factor) is ameasllteofho\lV'idosetlieplantis.Aimumcapacity. The·infernaI··tateiofreturl1. insurance premiums etc but less disregards from these latter here). = operating cost and/or revenue in year n. If.dfaC1()r The load factor of a power plant.perkWh In order to calculate the levelised cost of electricity. C. heat plant.ofelectricity. Lifec-ycleeosW LCe.. one also needs to consider the fuel cost (and maintenance cost. where CC = capital investment. How many kWh of electricity is produced per year? (Answer: 7. on the other hand. Is a project with a high internal rate of return more profitable than a project with a low internal rate ofretum? Discuss! EOa.HsmI.e.iTRRg This concept refers to the discount rate at which the life cycle cost of the project is zero i. The levelised cost of electricity is then given by CJ t. is the sum of the present values of all costs associated with an investment over its lifetime.

or ift= infinity (then CRF = r) Why? It may also be useful to look at the equation andto see what the value is ifr=O (then CRF = liT).l. Taking the derivative with respect to E gives the minimum for which __ d dE terms). cost-benefit analysis (CBA) .I.variableand total Fixedcosfs:A cost that does not vary depending on production or sales levels. Costbenefitanaiysis In contra~t to cost-effectiveness analysis. property tax or interest expense (for a power plant the fixed costs refers to the costof the investment and the cost of the staff that you have employed and have to pay salaries to regardless of whether you produce any electricity or not). an emission reduction target). (For a power plant this cost includes for instance fuel costs and additional maintenance costs since depreciation may increase if you run the plant more intensely).mpt~t(lfilld thetarget>( e. or intuitive? Explain! Cost-effectivenessianalysis Method which seeks tQfillclthe least cost to meet a pre-defined objective (e. The method basically seeks to minimize the cost of the sum of the cost of the environmental damage and the cost of reducing the emissions. While the total variable cost changes with increased production. material or overhead that changes according to the change in the volume of production units. CR(E) is the cost of reducing the emissions and E is the emissions. C()stsLfixed. how much the emissions should be reduced). dC dC = __ R dE (here we have taken the absolute value of the We thus have that emissions should be reduced so that the marginal cost of reducing the emission should be equal to the marginal benefit of doing so. TOtal costs: sum of fixed and variable costs makes up the total cost of production..g.!. Is this a reasonable approach for environmental policy? Discuss with your friends! Is it possible to calculate the environmental damage in economic terms? If not. i. Check the calculations. . we have the following task Where Cd (E) is the cost of environmental damage. Is this reasonable. the total fixed costs stays the same. They are very similar. Variable costs: A cost of labor.e.It is worthwhile to compare this equation with that which is used to estimate to levelised cost of capital. what does that imply for this approach.g. such as rent..

Thus Pareto efficiency has very little to do with what is good or bad for society. but most often it refers to the physical efficiency with which the primary energy is converted into either the useful energy carrier (say oil to heat in boiler) or the energy service requested (say fuel to kilometres). say 50%. Sometimes.g.Efficiency: economic efficiency versus energy efficiency In economics.. a word about Pareto-efficiency. if that was the only concern with nuclear power. The. However.l.! (or sometimes future) technologies.e. then there would be little cause for concern..iyy!:!clwithC]J. they may use very high discount rates). In this case one needs to ask whether the person in question refers to the theoretical.l. ()fel1ergy l1yy9yclioObtain the energy service (e. For instance. the term efficiency typically refers to a state in which nobody can get it better without someone else ending up in a worse condition.rPJJ)J. It should be made clear that an economy may be "in such a state without being efficient or optimal in the common usage ofthe terms.g91'lfliQqlpq(gntiqlis the min. For instance.that.when.lhdeffect __direct and indirect. The reason for that is that driving your car has become less . Cost numbers are useful because they reflect both physical conversion efficiencies. The eoanomic. the technical.potentiai refers to.ll:ent technologies with no extra costs. Finally. Potentialfor energy efficiency When someone states that the potential for energy efficiency is. Reb()l.i.efficiencythatc. effects Assume that you buy a car that is twice as efficient as your previous one (but that the price of it is the same). but with a cost of say 2 USc/kWh.). engineers assess technologies solely based on their energy efficiency.PYacb. an economy with one person owing everything and the rest living at the edge of starvation may be Pareto optimal. fuel costs and capital cost in one single indicator of the viability of the I have many times heard people arguing against nuclear power based on the fact that the conversion efficiency is rather low (only 33%) compared to a gas fired power plant which has an energy efficiency of around al§9 take into. then nuclear power would nevertheless be an interesting option. Energy efficiency can be defined in many ways (depending on the system boundaries of the system studied).111()ll!lt. If the conversion efficiency was as low as 1%. Will you then use only half as much energy over the course of the year? Probably not. close to 100% if the service is maintaining a certain in door temperature). The market-potenuabteiece to efficiency that can be achieved.there.technicalpotentiar2refers!0 efficiency that can be achievedwithcurren. account.Pareto optimality. he or she normally means that the energy efficiency of a product or a process might increase by 50%. might be-a lack of informationaboutproducts (that lead to a less efficient use of products) or when you take into account that consumers may sometimes not behave as optimising agents (i. or optimality. since making the poor richer would require that the rich becomes poorer (at least in a static sense). This concept of efficiency is often referred to as Paretoefficiehcy·or. • • • • The)tlJ. the economic and the market potential for energy efficiency.

However. if you buy a lamp that only consumes 20% as much as your previous lamp. or perhaps I should say.n'l i A historical dependency. This is called the direct rebound effect. travelling. If one does not take into account the rebound effect (both the direct and the indirect effect) one is likely to overestimate the full benefit of the energy saving activity. it means that the demand for electricity increases by 1. Path dependency and "technology lock . you find that demand for a good is given by Qd = Qo (P I Pot (carry out the calculation your The mathematical formula is given as follows: B = self!) Th~il1q()11J&ielasticitymeasure how the demand for a certain good increases if the income increases..1% of the income increases by one percent. ). and partly for that reason remain the dominant technology today.g. The formula for the income elasticity looks the same as that for the price elasticity.e. On top of that. bigger apartment ..ey a change in the priceof it (assuming that all other prices remain constant). A key example is the internal combustion engine. never as large as the energy saving activity. If the price elasticity for gasoline is -0. it means that the demand for gasoline drops by 0. as some people sometimes seem to suggest that other options are inherently "better" than the internal .. learning by doing.. Elastieity---price and income The priqg<elasticity measures howthedemandfor·a certain good is affected. cloths. the fact that the use of a technology tends to reduce its cost and therefore strengthen its position in the market even further). If the income elasticity for electricity is 1.3% for a percentage increase in the price of gasoline..costly. which means that you can afford to drive more. i. you are not likely to use the lamp five times as much.. e. dQ dP .. If you solve this equation.where Q is the quantity of (QI P) the good.1.3. (This does not necessarily mean. The term rebound comes from the fact that the a specific energy saving activity might "rebound" and lead to a somewhat less saving than previously expected. it should be recognised that the rebound effect is very rarely.g. and P is the price of it. it should be recognised that the general income increase that takes in most societies should not be seen as a rebound effect. you will be left with more money in your pocket (under the reasonable assumption that you do not spend all of your saved money on buying fuel) which is likely to be spent on (to varying degrees) energy intensive activities (e. Finally.g. which since it was selected in the beginning of the last century has received much more research and investment efforts than all other options. computers. 8 is the elasticity. the observation that a decision to go for say a specific technology in the past tends to lock us into that technology for a long time (there are several mechanisms that leads to lock in. e.

g. but with the time axis reorganised so that the demand is shown in falling order. A load duration curve depicts the same thing as a load curve. Niche/markets Solar cells are too costly to be competitive on "normal" electricity markets. If costs are reduced to say 75% of its initial value for each doubling of output..e. e. . coal fired power plants. L.combustion engine . grid connected markets where they have to compete with e. solar cells might be most competitive option on certain small markets. However. i..earnil1gbydoillg It is common experience in industry that the cost of various products tends to decline over time. and they may be instrumental in helping to bring down the costs of a technology (since they may enable a technology to grow and thereby come down in cost without having to compete on the main markets). It is common to express this progress in terms of a progress ratio (PR). Load. Such markets are called niche markets.only that historical choices live on and tend to be reinforced through a variety oflock in mechanisms). This is partly due to the fact that the more you produce the more you learn from your experience (both mistakes and best practice) and you may for that reason be able to improve the process so that you come down in cost. and b is a parameter describing how fast costs decline as a function of investment.. duration curve and load curve A load curve for the electricity system depicts. the output in GW over a certain period of time (say the year. than PR is 75%. with time on the x-axis). A formula describing how costs have come done with accumulated experience has been developed and proven to work pretty well for a large number of technologies: (2) where C is the cost per unit as a function of output and Co is the initial cost per unit. A is cumulative investment (a measure of experience). pocket calculators or off-grid markets (say in rural areas in developing countries or on sailing boats and the like). The parameter b is then given by b=ln(PR)/ln(2).g. on the y-axis.

LCA. Allocation of energy use in energy systems: the case of multiple productslbyproducts. How should that be factored into the analysis when estimating the net energy cost of producing ethanol from grain? Further.3. However. In. but rarely the latter. then one also has to state what year this refers to since one can expect improvements in technology over time. Where The question of where is important since the production processes may differ substantially (ethanol from Brazil is much more energy efficient than ethanol in Europe and the US) and the impact of certain emissions may depend a lot on where they occur (e. the question of scale also enters this issue: at current production levels.g. In a net energy analysis of beef from Sweden vs Brazil it was found that Brazilian beefwas less energy intensive even if transported to Sweden. but less so if it is being expanded to make a substantial contribution. The intellectual process of scaling up a proposed technology is often enlightening. for instance. Using such information as a guide to the future of transport would be misleading). the CO2 impact of ethanol production. animal feed is produced as a by-product. e. requires that one is explicit about the question of when. the meat market will be saturated. but of ethanol is to playa major role in the energy system. Why is that so? Comparing apples and pears. how should that be factored into the analysis? Deforestation means carbon emissions which are one of the .g. Does the Brazilian cattle system lead to deforestation in Brazil? (Is beef on the margin leading to deforestation?) If so... Issues to think about when analysing energy systems When (time and technical change) Carrying out a net energy analysis. when producing ethanol from grain. acidifying substances). For instance. Scale A certain technology may function very well if the scale is small. well to wheel studies one often finds information in diagrams referring to the present. other factors are also of importance. System boundaries How are these being drawn? Are all relevant aspects included? If the question under consideration is carbon emissions. then a carbon tax or a capand-trade system tend to make the question of system boundaries less important (the carbon price will determine where it is most cost-efficient to build the plant). whereas in reality future values are what matters. If one argues that the net gain is low. often considers the former aspect. (Hydrogen fuel cells cars may come out pretty bad in such assessments but that is partially because it is assumed that hydrogen is produced from electricity where coal is the marginal electricity source. There are allocation problems associated with production processes that have multiple outputs. say. the by-products may be sold to the meat system.

Key examples are: should we implement policies to reduce CO2 emissions. the conversion efficiency of an existing power plant . but deforestation also mean loss of biodiversity. or whether we should have a cap-and-trade system that cover all sources of CO2 or only large point sources? • Scenario <ihalysis aims at investigating "what . If I think that reducing the emissions of acidifying emissions is the most important thing in the world... one has to be explicit about one's values. there is a risk that they enter when one makes assumptions about the future state of a technology (a person in favour of solar energy may have a tendency to make assumptions about solar that are more positive than what a person who advocates nuclear energy would do). The counterargument is that we in the end have to make such trade-offs even if we do not like them. i. or ifnatural gas is better for the environment than nuclear power or vice versa. Values may also enter when it comes to the selection process. Values do normally not enter when it comes to objective physical properties (e.. whether we should or should not do something. i.. On the other hand. some will perceive statements about what is cost-effective as a statement about what we ought to do? Prescriptive. Facts versus values In none of the questions raised above did I ask whether one technology was better or worse than the other. a carbon tax at 50 USD/tC is introduced.. However. questions which seek to find out what is good or what is bad. the receiver may perceive this as an argument that we should substitute natural gas for coal (even if the normative element was never there in the first place).e. • Predictive analysis attempts at predicting what things will be like.e. etc. i. to force the introduction of biofuels..then it is easy to show that somebody made an erroneous assumption).g. scenario and predictive analyses • Descriptive analysis aims at investigating "whatis'~X!questions. which question is addressed? Which method is used? Which technologies are compared? How is the system boundary drawn? Time perspective? Which objective or target is set? Values do also exist in the minds of the receiver of the analysis. How should that be compared with energy? It might be argued that we should not make such comparisons because the world is not so one-dimensional. i. then clearly energy systems analysis can offer you advice on which technological options will bring down these emissions etc. it is not always easy to be free from value judgements.if'equestione. questions of the type what is net energy balance of an ethanol system? • Prescriptive analysissaims at investigating normative.questions. what will the oil price be in the year 2010 etc.reasons why we are concerned with energy.e.. In order to do so.g. descriptive. If somebody points out that the substitution of natural gas for coal leads to reduced C02 emissions. .e. what happens if the oil price increases to 100 USD/barrel. e. Similarly. to decide whether to go nuclear or not. The reason for that is that energy systems analysis per se cannot provide answers to such questions.

when you say that biofuels ought or ought not to be used based on a descriptive analysis. Reason and logic is simply not enough. Consider for instance statements about the cost of reducing global C02 emissions. His point is that without introducing values you cannot draw conclusions about what one ought to do. hardly measurable. Some state that it may costs trillions ofUSD (that is the net present value of all costs over the next century)." Critical thinking Easier to forget than you think . Similarly. economist often state that biofuels are not costefficient (we make the similar case in the paper Azar. e. Be aware! The complex versus the obvious When attempting to analyse very complex systems. Lindgren. Both studies say essentially the same thing but the impression they create in the minds of people is very different. we should be always be careful not to overlook the obvious. Very often environmentally conscious engineers point to Brazil (including myself) and then seem to suggest that the fact that it is possible to do something (a descriptive analysis) is sufficient argument to suggest that it ought to be done (prescriptive).The natural fallacy The concept stems from the Scottish philosopher David Hume who already 200 years ago noted that you cannot note derive an ought from an is. Andersson in the compendium) but this does not necessarily mean that biofuels should not be used. Put things in perspective The way results are presented often tend to have a big impact on how the result is perceived. It can easily be proven that one can power millions of cars with ethanol from sugar cane and that that would lead to significant reductions in C02 emissions. others say that it only amounts to a loss in the economic growth rate by 0. Do you agree? To take one example from energy analysis. One of his more famous quotes are "Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger". This is often useful. advanced computer models and modelling approaches are sometimes employed. The natural fallacy takes place when you do derive an ought from an is..03% per year. There are several examples of this. but as Galbraith once said: "In unraveling the complex.g.

09-1.184 J (Note that the Calories stated in food ratings are actually kilocalories 1 horsepower (hp) = 745. 1 EJ = exajoule = 1018 J.(jy-19h.6 MJ/litre (higher and lower heating values) Liquified hydrogen 2.7 watts Electron volt.6 million Joules 1 calorie of heat is the amount needed to raise 1 gram of water 1 degree Centigrade.5 MJ/litre (higher and lower heating values) Ethanol 24-21. 1 kilowatt-hour (kwh) = 3.3 GJ/barrel Crude Oil 42 GJIton Gasoline 5.3-34. Conversion factors.4. 1 kilowatt-hour is the energy of one kilowatt power flowing for one hour. It is equal to the force of one Newton acting through one meter. (lyy==1. sizes and efficiency levels Units 1 Joule (J) is the SI unit of energy. 1 calorie (cal) = 4.76 MJ/litre (higher and lower heating values) Natural gas liquids 4 GJ !barrel Natural gas 39 MJ/m3 Biomass 20 GJ/ton dry wood Biomass (crop residues) 15 GJ/ton Carbon Emission factors (approximate) Coal 25 gC/MJ Oil 20 gCIMJ Natural gas 15 gC/MJ . a unit of energy.3 GJ/barrel (159litres) Gasoline 32. Some basic energy data: Units. 1 British Thermal Unit (BTU) = 1055 J 1 Quad = 1015 BTU or roughly 1 EJ (the unit is common in the US) 1 toe = a million tons of oil equivalents approximately 42 GJ 1 barrel of oil = 159 litres 1 electric Watt is the power from a current of 1 Ampere flowing through 1 Volt.6 x 106 J = 3. (E = P t). Energy Content of Fuels Coal 24 GJ/ton Crude Oil 5.that amount of energy that is released when an electron falls in a voltage drop of 1 V). costs. A BTU (British Thermal Unit) is the amount of heat necessary to raise one pound of water by 1 degree Farenheit (F). 1 kilowatt is a thousand Watts. World energy use is roughly 400 EJ.

efficiency penalties are incurred. Typical size of energy conversion facilities Nuclear power plant . Carbon capture rate is assumed to be 90% of the flow (meaning that 90% of the carbon input to the plant is recovered as CO2 for disposal)./ . Climatic Change. For technologies involving carbon capture. although some suggest higher rates.1 GW electric Coal fired power plant 500 -1000 MWe Natural gas fired power plant couple of 100 MWe Refinary 3 GW crude oil input Wind .5 .3 MW electric Imill . as shown in the table.0. Source Azar et a12006.Typical cost and conversion efficiency of future well perfonning energy conversion 1000 75% 55% 55% 500 45% 500 900 80% 1200 30% 1700 60% 400 70% 300 75% 800 55% 70% 60% The efficiencies indicated here are on a higher heating value basis.

~I .

Ralph David. Ayres. ISBN 1-933115-10-6 (cloth) ISBN 1-933115-11-4 (paper) . This book was designed and typeset by Kulamer Publishing Services. HC85. 3. without written permission. : alk. Simpson. All rights reserved. and Robert U. MA 01923. Inc. Environmental policy. 2. or its officers. USA (fax +1 9786468600. em.ISBN 1-933115-11-4 (pbk. Includes bibliographical references and index. An RFF Press book Published by Resources for the Future 1616 P Street NW Washington. its directors.7--dc22 2004030206 The paper in this book meets the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources. interpretations.Copyright © 2005 by Resources for the Future. Requests to photocopy items for classroom or other educational use should be sent to the Copyright Clearance Center.. They do not necessarily represent the views of Resources for the Future. II. p. Suite 910. I. Robert U.222 Rosewood Drive. Michael A.rffpress. It was copyedited by Steven Jent. Toman. The findings. Toman. Michael A. DC 20036-1400 USA www. Natural resources-Management. paper) . Cover photo by Anthony Ise. Danvers. and conclusions offered in this publication are those of the contributors. Getty Images. Sustainable development.copyright.S33 2005 333. editors. The cover was designed by Rosenbohm Graphic Design. Ayres. whether electronic or mechanical.All other permissions requests should be sent directly to the publisher at the address Library of Congress Cataloging-in-Publication Data Scarcity and growth revisited: natural resources and the environment in the new millennium / R. www. David Simpson. Printed in the United States of America No part of this publication may be reproduced by any III. Photo disc Green Collection. paper) 1. ISBN 1-933115-10-6 (hardcover: alk.

and hydroelectric power combined (WEA 2000. and this brings risks for new oil crises and military conflicts. liquefied-coal or hydrogen from solar energy-and this is what creates the great interest in oil. Solar energy can be converted into clean energy carriers. Four hypotheses are put forward: NERGYIS NOT SCARCE. 167). Local and regional pollutants. The solar influx to the earth carries some 10.000 times more energy per year than the current annual global anthropogenic use of fossil fuels. such as sulfur dioxide (S02)' nitrogen oxides (NOx)' and particulate matter. Moreover. Rather the energy problem.' If coal is unacceptable. E • It is technically and economically feasible to meet stringent climate targets.f This chapter will focus on climate change and how policies to deal with climate change might increase competition for land between biomass and food. nuclear power. such as electricity and hydrogen. and environmental factors. Oil is cheaper than its alternatives-for example. geopolitical. Finally. the "energy problem" has nothing to do with physical scarcities of energy per se. not physical scarcity of energy. are the key reasons why two billion people lack access to modern energy carriers in developing countries. are causing health and environmental hazards for billions of people around the planec. institutional. the global society may resort to the plentiful supplies of solar energy. is related to economic. or problems to be correct. environmental scarcities arise because the burning of fossil fuels emits carbon dioxide (COz). 98 . This has tremendous health and environmental consequences.CHAPTER 5 Emerging Scarcities Bioenergy-Food Competition in a Carbon Constrained World Ch ristian Azar There is enough coal in the world to meet any plausible energy demand projections for centuries to come. geopolitical scarcities exist primarily in the context of oil-conventional oil reserves will likely become scarcer over the next decades. and make our energy and transportation systems virtually emissions free. Poverty and various institutional factors. Thus.

unless food prices rose to the point where profits matched the energy sector. land and food prices are likely to be pushed upwards. Gielen et al. .With these higher profits. Thus. almost regardless of whether one is optimistic or pessimistic aboutthe potential biomass supply.. and Azar and Rodhe (1997)-have argued for setting an upper limit on the increase in the global annual average surface temperature at or around 2°C abovepre-industrial temperature levels. I then provide some empirical evidence for the land competition hypothesis and discuss how all this might affect global hunger patterns. and more detailed energyeconomy modeling will follow. But the potential supply is low compared to the required levels of carbon-free energy. this is some 30 percent higher than the pre-industrial concentration.primarily deforestation in the tropics. and Johansson and Azar (2003) for other modeling studies. even 550 ppm. metric tons of carbon).g. may be acceptable' From 1990 to 1999. the UN Framework Convention on Climate Change (FCCC) (United Nations 1992) calls for a stabilization of greenhouse gases at a level that prevents "dangerous anthropogenic interference with the climate system. and food prices by a factor of two to five. Azar and Berndes (1999).Azarand Rodhe (1997) show that a 2°e target may require that concentrations be kept below 400 ppm. or 109. and it is expected to more than double by the end of this century unless carbon abatement policies are introduced. but also that there is some probability that higher concentrations. Arguments supporting the first three hypotheses will be laid out using back-ofthe-envelope calculations to give a feel for the results. the European Union) and scientists-including Rijsberman and Swart (1990).3 GtC/yr from the combustion of fossilfuels and cement production (1 GtC is equal to a billion. but governments (e. In this chapter.Alcamo and Kreileman (1996). the global society emitted. See Walsh et al. McCarl and Schneider (2001). Policy conclusions are offered in the final section. Meeting Stringent Climate Targets Atmospheric CO2 concentration is currently around 375 parts per million (ppm). For that reason. Bioenergy can be expected to playa key role in the transition to low CO2 emissions because it is a relatively low-cost renewable fuel. farmers would have greater economic incentives to turn to bioenergy. the Scientific Advisory Council on Global Change to the Federal Government of Germany (WBGU 1995)." The FCCC does not attempt to define the concept of dangerous interference with the climate system. (2001). more costly carbon-free energy sources will have to be used iflow CO2 targets are reached. an average of 6. (1996). Land use changes. • The more costly carbon-free energy could raise energy prices to a level that would mean higher profits for the bioenergy sector.Thus. it is estimated that land values might increase by an order of magnitude.Chapter 5: Emerging Scarcities' 99 • There is a risk that new land scarcitieswill emerge in response to climate abatement policies. • The socioeconomic impacts of higher land and food prices on the poor and malnourished of the world are complex and could be either positive or negative.

and 550 ppm targets.. Back-of-the-Bnuelope Global Energy Scenario Energy systems models are generally more detailed than transparent. total CO2 emissions would have to drop to around 2 GtC/yr.seque$tra~onfromifossil fuels and biomass.2. a global energy scenario towards a 400 ppm target will illustrate the full dynamics of that transition. and slowed population growth. wind. nuclear. Later. Compare this to current emissions: • the United States with per capita emissions above 5 tC/yr. the European Union. approximately 1 quad).3 tC/yr.7 tC/yr. A per capita perspective may make this even more apparent. and • India and Africa around 0. within a wide uncertainty range. 450 ppm. and the countries of the former Soviet Union in the range 2-3 tC/yr.I 00 • Christian Azar released an additional 1. 1996). in the 1980s (IPCC 2001a.4. including biomass.l. for the 400 ppm.000 EJ1yr (1 EJ = 1018 J. It is beyond the scope of this chapter to make a detailed assessment of these technologies. Lower CO2 emissions can be achieved through a combination of • reduced ellergy user)!viatechnical energy efficiency measures.1999).0. The challenge is daunting. By the year 2100. lifestyle changes.per capita emissions would have to drop to 0.lheemissionsof COk per unit of primary energy -supply via increased use of renewables. Assume that there will be 10 billion people and that each will use around 200 GJI yr (primary energy). Assume that income grows so that by 2100 the average global citizen can enjoy roughly the same material standard of living as we do in the developed world today. the reductions would not need to be as large. and natural gas at the expense of coal and oil. In order to meet a 400 ppm target by 2100. as much energy as a typical OECD citizen.7 GtC/yr. and • carbo1.7 tC/yr. • Japan. solar and hydro. • reductioriiil1. global per capita emissions would eventually have to drop to a level below that of India today. The interested reader is referred to the World Energy Assessment for a . This means that the total energy supply would be as high as 2. Thus. respectively. • developing countries such as China and Latin America at around 0. This is five times as much as the current global primary energy supply. respectively (Wigley et al. If a higher stabilization target had been chosen. But it is important to recognize that emissions eventually would have to drop to levels below 2 GtC/yr even for these higher concentration targets (IPCC 2001a). Here a simple scenario is offered to demonstrate the technical feasibility of meeting stringent carbon targets by the year 2100. A 450 ppm and a 550 ppm target would require annual emissions to drop to 4 and 7 GtC. With a population ofl0 billion people by the year 2100. and 0. 190). so-called reference scenarios suggest that the global society may emit some 20 GtC/yr within a wide range (IPCC 1992. but they would nevertheless require a radical transformation of our energy systems.A back-of-theenvelope calculation hopefully provides a clearer picture of the assumptions underlying many low-carbon emission scenarios and enables readers to check the results themselves.

With a carbon emission target of 2 GtC/yr. but we need not dwell on it here because changes in its supply by a factor of two or more will not affect the extent to which competition for land may arise. fuel cells for automobiles that could approximately double their efficiency.'' it corresponds. well-insulated houses in cold climates that do not require a heating system (such houses exist in Sweden. recent detailed and authoritative analysis of nuclear.7''101year 5-10 I Total 2.see Ayres (1994) and WEA (2000). Often it is of interest as a low-cost option. or down to 100 GJ/yr per capita. higher energy prices or standards).7 percent/yr. the United States. It should be noted that 200 . 100 GJ of primary energy supply could offer the same energy services as a typical OECD citizen enjoys today while exp~ncling200 GJI~~ar. Energy efficiency improvements may reduce this demand' for energy by half. and behavioral changes are assumed to reduce the demand to 100 GJ/ cap/year without compromising the energy services currently provided.: J~yr is a very impressive amount of enery. The potential for large-scale expansion is limited. in E rough terms.. and even if the supply doubled over the next hundred years. bClearly. Source: Author calculations based on WEA (2000).g. Thus. energy efficiency is not a source of energy but is included here to clearly display the enormous potential this option has when compared with supply options.This would correspond to an increase in energy efficiency by 0. Hydroelectricpowersupplies somel0EJl'yr. organizational.paper. furniture) production. it would be less than 10 percent of the total demand for carbon-free energy.OOOEJ/year-=by the year 2100 (seeTable 5-1)." This means that the world needs to find 900 EJ/yr from other sources. the world would get some 100 EJlyr from conventional uses of fossil fuels. throughout this century. Biomass/ii organic materials from agriculture and forestry-is a promising lowcost renewable energy source that could supply perhaps 200 ± 100 EJlyr. and elsewhere).Our task is to provide the world with 100 GJlyr per capita-l. Technical.Chapter 5: Emerging Scarcities • 101 Table 5-1. fossil. On average this would require an energy efficiency improvement by 0. Primary Energy and Negawatts by the Thar 2100a Source Area requirements (million km2) efficiency improvementsb -D. Some suggest that even . to the size of the current global food system. and renewable energy technologies (WEA 2000). but only if policies to enhance efficiency are introduced (i. Technical measures include fuel cellsfor electricity generation that could raise the efficiency to 50-80 percent (current efficiency is 30-40 percent for coal and about 50 percent for modern natural gas-fired combined-cycle plants).e..7 percent/year. For more on energy efficiency.000 "Assuming that everybody in the world will use energy services to the same extent as the average OECD citizen today (primary energy supply in round nu~bers is 200 GJ (Gigajoules) / cap/year). or ten times current commercial roundwood (e.

The cost of capturing and storing carbon is expected to stay below a few hundred US$/ton C (see WEA 2002. Nuclear pOwer:: is currently in a stalemate." But after that. then leakage rates must be significantly lower than 0. 2004). but these estimates tend to be optimistic. let us assume that 500 GtC can ultimately be stored safely. and heat. 2000. are achievable (see WEA 2000 and Berndes et al. but it should be recognized that the introduction of stringent carbon abatement policies will most likely increase electricity prices and therefore enhance the economic competitiveness of nuclear power. It is not possible in this chapter to study the details of the role of nuclear energy in a carbon-constrained world. a word on acceptable leakage rates. The point is that even this level would begin to stretch what can be achieved in a socially and environmentally acceptable way. 2001). If ocean sequestration is ruled out. Such an expansion would substantially increase the risk of accidents and nuclear proliferation. If structural traps are not deemed necessary. Finally. reductions in costs. industry representatives. depleted oil and gas fields and saline aquifers with structural traps can store several hundred GtC (WEA 2000. . saline aquifers could sequester as much as several thousand GtC. 8 For the purpose of this exercise. perhaps twice as high. including in depleted oil and natural gas fields.This would mean that around 200 EJ Iyr of energy could be delivered annually over a period of 100 years without any carbon emissions. One critical issue is the storage potential. if thousands of GtC are stored. as are improvements in performance and. This expansion would correspond to an increase of the number of reactors in the world by a factor of 10. Windandsolarenergytould both offer large amounts of electricity. and researchers argue that we cannot reduce CO2 substantially without resorting to nuclear energy (see Sailor et al. Grimston et al. and the oceans"? The potential for CO2 storage clearly determines the role fossil fuels can play under stringent climate policies.000 MWe and with a load factor of 0. Some politicians. renewables would need to replace this contribution from fossil fuels with carbon sequestration. a leakage rate of 0. CO2 emissions from the use of fossil fuels-and bioenergy-can be captured and stored. but how much can be stored is uncertain. A technically feasible but politically contentious option is storage in geological reservoirs.000 reactors (each at 1.l" For critical arguments about a large-scale expansion of nuclear power. Thus.800 EJ/yr (without any land constraints) and 231EJ/yr (ifless than 4 percent of the land area is used).1 percent/yr would lead to releases of around 2 GtC/yr for several centuries. This would not be compatible with atmospheric CO2 stabilization at the levels discussed in this chapter." Technically.102 • Christian Azar higher yields.8).1 percent. see Abrahamson and Swahn (2000). Lake et al. 2002). 2003 for reviews of other studies). but for nuclear energy to make a substantial contribution. capture and storage technologies are commercially available and proven. but increases in scale are needed. If 2000 GtC were sequestered. around an order of magnitude expansion would be required: a supply of 100 EJ of electricity or hydrogen (heat losses not counted) would require 4. hydrogen. The question is not whether it is technically possible to supply 200 EJ/yr of biomass energy. or 'even twice this amount. it is possible. deep coal beds and saline aquifers. in particular if it involved a transition to breeder reactors. The global annual wind resource is estimated at 5. Azar et al.

and fossil fuels with carbon sequestration could be much more important or possibly much less depending on storage options and political feasibility.~~~~~g. 200 EJlyr fossil fuel use with carbon capture and storage.CO2 storage ca. Nuclear energy has the potential to be important. The Potential Supply of Biomass Energy Biomass for energy can be obtained from dedicated plantations and from residues (e.g. and households).. 1~~~. paper. The key point to understand is that the contribution from bioenergy is likely to be minor compared to the overall demand for carbon-free energy sources.000 EJ/yr.~withothersectors . Thus. and acCidents may virtuallyaeliminaeeeit. and metals for solar cells.paCity. The influx is typically 7 GJlm2·yr in temperate countries.g. but concerns about weapons-proliferations radioactive waste disposal.500--50. but these are omitted from this analysis. an~materials:~~Bt~~' This chapter focuses on food. agriculture. more importantly. fuel cells. say. Perhaps biomass could contribute some 100 EJlyr more or less.uranium. 1 million km2 would be needed to supply 350EJ/yrifiand this corresponds roughly to 3. and nuclear energy and ocean energy flows. Such competition has already started in Sweden. each square meter of land allocated to solar energy would offer 350 Mj/yr. geothermal. Significant contributions could also be obtained from hydro. In the process of solving one problem. 192). forestry. therefore. a prescribed primary energy supply of 1.Chapter 5: Emerging Scarcities • 103 The solar influx to the earth is large enough to supply any reasonable energy demand level.000 EJ/yr by the year 2100 could be met by 100 EJ/yr conventional fossil fuel use.that use Iandsin particular the food.6 percent of the area of the world's deserts. which might induce new scarcities in. Carbon abatement policies may increase the demand for bioenergy and. such as the conversion of forests into cropland or pastures. and a land use efficiency of 50 percent. Assuming a conversion efficiency of l O percent into hydrogen or electricity. This is of course a large area. but climate policies would also affect the economics of other land uses. 163) estimates a minimum and maximum technical solar energy potential at 1. ()rba. even if optimistic numbers are assumed.2g. The supply from plantations includes . and some 500 EJ/yr from solar and wind energy. and half of that in northern countries. Emerging Scarcities-Food versus Fuel Policies intended to reduce CO2 emissions will increase the demand for other technologies and materials.. but not nearly as large as that envisioned for biomass energy plantations in most global energy scenarios. In conclusion. we must be careful not to create others.tteries{see Andersson and Ride (2002) for the metals case). 200 EJIyr of biomass energy. polar regions excluded (IPCC 2001a. Obviously these numbers are not exact and should only be thought of as indications of potential orders of magnitudes. and there have been requests that the government tax the use of biomass for energy purposes. WEA (2000. green bioproductive areas would not need to be targeted.'! or the profitability of using biomass for the production of woody materials or paper.

conclude that "none of the reviewed studies presents an autonomous assessment of degraded land that is suitable and available for plantation establishment. the studies referred to did not focus on availability of degraded land for plantation establishment. For instance.. Some argue strongly in favor plantations. or herbaceous grasses and depends on the yield and the amount of land that can be made available. but some studies suggest that several hundred million hectares of cropland expansion will be required (e.·4e!'Velqpingcoul). Establishing short rotation plantations on previously forested lands is objectionable for several reasons. Berndes et al. ecological (precipitation. (2003). and economic factors. pastures... etc. In order to estimate the acceptable potential supply of biomass. Others would argue that 500 million is far too much.). many as prime candidates for plantations. 1993). .This critique has already started even though the current area of short rotation forests only amounts to a small share of what is hoped for in more bioenergy-oriented scenarios (Carrere and Lohmann 1996)." It should also be kept in mind that "degraded lands" is an ambiguous concept that includes all lands that have lost some quality and that these lands are most often inhabited and cultivated by poor subsistence farmers. and temperate grasslands and shrublands.tries. Hall et a1. Latin America. (1993. Analysis of global food demand suggests that most of the additional supply over the next 50 years will come from increased yields (Dyson 1996). Plantations can be established on agricultural lands. However. and forests-tropical. <!-re!~ . 644) argue that large areas of lands are suitable for plantations in the United States.5 billion to 4. and forestlands. temperate.. one can expect that social resistance against plantations will increase once they cover large areas ofland. This is consistent with a renewable energy-intensive scenario (lohansson et a1. 192).ttopical savannahs and grasslands." However. as well as environmental concerns (conservation and biodiversity protection).5 billion hectares. largely used for grazing. and boreal-cover 4.. However. cover some 3. Tilman et a1. marginal croplands a.2 billion hectares (IPCC 2001a. large tracts of lands could technically and economically be converted to plantations. Europe. are seen by .V Thus.2 billion hectare potential assumed in WEA (2000. Currently. But it is significantly less than the 1. the extent will be determined by a combination of social. and Africa. and this is largely guesswork that reflects the analyst's subjective views.2001). It is clear that if the demand for CO2-free or -neutral energy is extremely high and the price of energy is sufficiently high.jl1. poplar. global agricultural cropland covers someh5billionhectares. but point out that competition with food production could become significant in Asia. having reviewed the literature on global bioenergy potentials.158).I 04 • Christian Azar·\particulardegradeq .+:r~s. intensive efforts to develop such plantations around the globe are warranted.dpast\. Hall et al.ri.. For the purpose of this chapter. soil quality. The amount ofland that will actually be converted depends on the social and environmental acceptability of these plantations. one is forced to make assumptions about social and environmental reactions... They write that "in light of the prospective favorable economics and the environmental and social benefits that can be derived from plantations' biomass energy. Instead reference is made to other studies: .g. I will assume that 500 million hectares can be freed for plantations under socially and environmentally acceptable conditions.1993.

etc. Further. but it might be 20-100 EJlyr (see Berndes et al. India. and carefully observe the social and environmental implications. and rice yields have improved over the past hundred years. Climatic changes and changes in atmospheric CO2 concentration will affect food production as well as the productivity of biomass in general (IPCC 2001b. such as methane from cattle and rice paddies).000 Mt of biomass per year globally. Chile. wheat.Chapter 5: Emerging Scarcities • I05 My assumption that 500 million hectares can be freed for plantations does not necessarily mean that I believe that this much should be used. average cereal yield today is slightly below 3 ton graiu/ha-yr on somewhat less than 800 Mha of lands (FAO 2002. The amount of residues is determined by the demand for food and industrial roundwood. Local decisions will then determine how much bioenergy plantations will eventually produce. and several other countries). precipitation. and greenhouse gas taxes will increase the cost of agricultural activities (inputs like fuel. and thus the profitability of biomass energy . This corresponds to 100 EJ Iyr. we would get 5. But this aspect should not be overstated: we already have experience with managed plantations for pulp and timber production (there are millions of hectares of eucalyptus and pine in Brazil. Impact on Food and Land Prices Climatic change and countervailing policies might affect food production in several different ways. This makes estimates of the potential from residues uncertain. give or take 100 EJ/yr to account for all the uncertainties. it might be reasonable to assume an average yield of 10 t/ha 'yr if very large areas of land are used. In combination with up to 100 EJ Iyear from residues. Thus.000 mml yr on well-managed experimental plantations in Brazil). assuming an average yield of 10 t/ha·yr and an area of dedicated plantations covering 500 Mha. If all cropland in these regions were used for bioenergy plantations (a completely unrealistic assumption just for the sake of illustration). The populations of North America and Europe combined total 876 million. But a CO2 tax will also increase the price of energy. The higher estimate represents very good conditions (precipitation of2. It is of interest to note that 10 t/ha·yr corresponds to 200 GJlha'yr. Annual yields of short rotation crops range from 5 to 20 metric tons dry matter per hectare per yearit/ha-yr). It is too early for strong views. and fertilizers and the cost of emissions. and Wirsenius 2000 for a detailed assessment of biomass flows in the global food system). the potential supply totals 200 EJ/yr in round numbers. which is roughly equal to the per capita primary energy supply in the OECD countries. 2002). the yield would only be equivalent to half of their current primary energy supply. Dyson 1996). electricity. in particular on degraded lands. 2003 for a survey of different studies. chapter 5). and lands used for crops occupy 400 Mha. that is. some of which would probably be short of rain. around half a hectare of cropland per person. Over the next hundred years. The world should experiment with plantations with multifunctional benefits (Borjesson et al. improvements might be expected as a result of research and experience. Rather it should be seen as a starting point for discussions. as corn. depending on production system.13 Thus.

226). In order to estimate the willingness to pay for biomass. . these cost levels cannot be met without technical improvements.). The inevitable conclusion is that other more costly energy sources must enter the picture. They will eventually set the marginal price of energy. With experience. (2001). hydrogen. the feedstock cost for hydrogen production would be 15 US$/G] hydrogen. and]ohansson and Azar (2003). A supply of 200 E]/yr is a very impressive amount-larger than current global oil use--but it would nevertheless be a third or less of the global need for carbon-free energy by the end .400 US$/ha·yr. but this could be optimistic.Azar and Berndes (1999). Back-of-the-Envelope Calculation of the Willingness to Pay for Biomass Biomass energy from dedicated plantations. Studies of food and bioenergy interactions along these lines have been carried out by Walsh et aL (1998). Azar and Larson 2000. WEA 2000. This mechanism would operate under free market conditions. etc. chapter 19. In the preceding section we concluded that the demand for biomass is likely to be higher than the potential supply. the latter cost might fall to 20 US$/G] (IPCC 1996. The cost of producing hydrogen from biomass (assuming 2 US$/G] feedstock cost) can be expected to fall to 8 US$/G] assuming a conversion efficiency of 60 percent (IPCC 1996. which is in line with the cost of solar hydrogen. the profit per hectare comes to 1. electricity. Boijesson et al. then biomass can be sold to hydrogen manufacturers at a price of 9 US$/G] biomass. McCarl and Schneider (2001). Prices would eventually level out at the cost of the back-stop technology (here assumed to be solar energy for heat. If a transition from CO2 emissions were to materialize. If the price of biomass ends up at 9 US$IG]. substantial profits in the bioenergy sector could be expected. but could of course be modified by specific policies if that were considered desirable. ethanol. thereby affecting land prices and the economic conditions for agriculture. for instance in biomass gasification. Currently. the profit would be 7 US$/G].l" If we treat solar hydrogen as the back-stop energy technology that sets the marginal energy price. further technical improvements. costs around 2 US$/G] in the tropics 'and around 3-4 US$/G] in Europe (WEA 2000. Gielen et al. initially as a result of carbon taxes (or a cap-and-trade system) driving up the price of fossil energy. Edmonds et al. and mass production. potentially the largest source of biomass. (1996) have also analysed bioenergy and food in an integrated model.I06 • Christian Azar plantations. chapter 19). Biomass can be converted into process heat. 226. electricity.of the century. and the biomass production cost is 2 US$/G]. which brings the total cost to 20 US$/G]. Given yield levels of 200 GJlha'yr. 2002). and hydrogen production). or essentially any liquid hydrocarbon (including methanol. it is illustrative to compare the costs of hydrogen from biomass and from solar energy. but without the introduction of any carbon abatement policies. At this price and a conversion efficiency of 60 percent. Capital and conversion costs are estimated at around 5 US$/G]. The higher profits in the bioenergy sector would translate into higher land prices and consequently higher food prices.

The price of wheat is roughly 100-150 US$/ton grain. would reduce the profits in the bioenergy sector. we evaluate the sensitivity of these results with respect to some critical assumptions. To this one should add subsidies. Let us assume that average yields are 50 percent higher than 10 t/ha-yr and that the land area covered with dedicated plantations is 100 percent larger (1. Consider wheat production: average yields undergood conditions in mature production systems are close to 6 ton grain (dry matter)/year/hectare in the European Union and around 3 ton/year/hectare in the United States. an increase by as much as a factor of three. These estimates strongly suggest that bioenergy might come to compete with food production for bioproductive lands.The world would then enjoy a supply of300 EJ/yr ofbioenergy from dedicated plantations. This would still amount to a minor share of the total global demand for carbonfree energy. a higher cost would increase the biomass scarcity rents. The willingness to pay for biomass increases by roughly 0. Assume in all cases that biomass supplies are constrained and that solar energy is the back-stop technology. Assuming higher yields or larger areas of dedicated plantations will thus not change the basic picture that the potential demand for biomass energy is larger thari the acceptable supply. but are not yet well understood. .6 US$/GJ for every 1 US$IGJ increase in the back-stop energy cost. Below.400 US$/ha'yr or around 280 US$/ton grain (still assuming a willingness to pay 9 US$/GJ for bioenergy).S. But subsidies vary substantially from country to country. • The cost oj back-stop carbon-free energy. say solar hydrogen. and that factor is omitted here. and thus the increase in land values would be less dramatic.000 Mba rather than 500 Mha). Consider a farm with an annual wheat yield of 5 t/ha·yr. The price of wheat would have to rise to around 400 US$/ton. farmers and 600-900 US$/ha'yr in the European Union. which account for a substantial contribution to farmers' income. land rents would increase by a factor of 5 to 10 if these scarcity costs were to materialize. which can supply large amounts of energy at a constant marginal price. so total revenues equal 300-450 US$/ha'yr for U. Sensitivity Analysis These numbers should be understood as no more than indications of the increase in land values that could follow from a concerted effort to reduce global CO2 emissions. If farmers under market conditions produced wheat rather than energy crops. and more costly sources of energy would nevertheless be needed. the wheat farming profits would have to increase to 1. the average value is 170 US$/ha'yr (USDA 2002). Conversely.P Land rents in the United States range from 50 to 300 US$/ha'yr.Chapter 5: Emerging Scarcities • 107 It is of interest to compare these rents with current land values. Northern European land rents typically fall in the range 100-300 US$/ha 'yr (Iordbruksverket 2001). A lower cost for the back-stop energy source. assumed here at 20 US$I GJ unless otherwise stated. The social and political implications of this could be drastic. Thus. • The potential supply oj biomass.

that land rents are 1. natural gas. and this would free lands for biomass cultivation. say 10 t/ha-yr would then lead to a price of 140 US$/ton wheat. Higher grain yields would reduce the demand for land for food production. • Lower bioenergy yield on degraded lands. and that the cost of collecting them is 1 US$IG]. • Carbon capture and storage from fossil fuels as the back-stop technology. and hydrogen-and at the same time remove carbon from the atmosphere (Obersteiner et al. The profit margin for biomass increases linearly with the carbon tax in both . Furthermore.We would then get 320 US$/ha·yr from the sale of residues. 2001). • Where the value if biomass is highest. It would only lower the value of degraded lands and not affect the opportunity cost of growing bioenergy on prime agricultural lands. This is behind the higher bioenergy values in the more detailed models presented below. or around 60 US$/ton wheat. For instance. The grain price would then be 220 US$/ton grain instead of280 US$/ton (in order to achieve profit parity with bioenergy production). then this technology could also be used in biomass conversion facilities.800 US$/ha.400 US$/ha'yr as in the original example. instead of the 280 US$I ton estimated earlier. but if this option became available on a large scale (if storage capacity were large. 400 G]/ha'yr yield would increase land rents to 2. More optimistic assumptions about the recoverability of such residues would further reduce the price increase on wheat (Johansson and Azar 2003). then higher yields could also be expected for bioenergy production systems. This would not have any significant impact on food prices. This would imply that the cost of the carbon-free energy carrier would be much lower. A higher wheat yield.I 08 • Christian Azar • Higher yields of crops or bioenergy or both.and biomass power plants with and without carbon capture is depicted. This would add to the competitiveness of biomass. Earlier we compared biomass to solar hydrogen. if these high crop yields were attainable.On the other hand. Higher profits in the food and bioenergy sectors would offer incentives to increase yields. These sales would make food production more economical. Let us assume that two tons of residues are sold per hectare per year.Biomass energy with carbon capture and storage could make it possible to produce carbon-free energy carriers-heat. The difference in cost is even higher when it comes to process heat and steam production. then the willingness to pay for biomass might reach 20 US$IG] (assuming the same cost for the heat plant). and grain price increases would be less pronounced. it would lead to smaller food price increases because the higher land rents would be based on a larger crop yield. Part of the residues should be left on the soil for soil quality reasons. but part could be sold for its energy content if energy prices escalated enough. • Selling crop residues for energy purposes.and this would increase land values even further. electricity. If solar hydrogen costing 20 US$I G] is replaced by biomass (burnt directly) at a cost of 2 US$IG]. with an energy content of20 GJlton. for instance. that the value of the residues is 9 US$IG]. the cost of electricity from coal. In Figure 5-1. and capture and storage technologies became competitive). Assume. There is a substantial amount of energy in agricultural harvest residues (typically 50 percent of the dry matter of the crop can be eaten by humans).

transportation fuels. Energy demand levels for electricity. These scenarios imply the need for substantial efforts to improve energy efficiency. Source: Azar et al. and land prices evolve over time.Chapter 5: Emerging Scarcities 100 • 10'1 f . '" -20 o 100 200 300 Price of carbon (US$/ton) Figure 5-1.1) 80 ~ ::> .a more detailed. 2004. The economic cost of achieving such low emissions is discussed in Azar and Schneider (2002). ". c 60 .16 The GET model seeks to minimize the energy and transport costs for any given CO2 constraint (tax. cases (Azar et al. model is required. Transportation and storage costs are included.. the C1 scenario in lIASA/WEC (1995). Cost of Electricityfrom Fossil Fuels and Biomass Note: The cost of electricity from fossil fuels and biomass with and without carbon capture is shown as functions of the carbon tax.To evaluate such issues. Figures 5-2 and 5-3 present the results of such modeling efforts using the GET model.. andAzar et al. It is this profit margin that translates into a willingness to pay more for biomass than the actual production cost. and therefore less transparent. A More Detailed Model of Food-Fuel Competition A back-of-the-envelope calculation of global energy supply and demand. may give a transparent overview of the issues at stake. developed by Azar and Lindgren (Azar et al._.g 1. Other more detailed energy scenarios meeting stringent climate targets include those ofIPCC (1996. emission constraint.ij ::J 40 "'C 0.. food. . 0 e --------------~~--------------------------'" '" -----------""'" Coal with carbon capture and storage Biomass 20 8 ~ 0 Biomass with'" carbon capture and storage . chapter 19). and stationary fuel use are roughly in line with IIASA/WEC so-called ecologically driven scenarios. and of the biomass contention between food and fuel. 2004). 2004). 2003. or atmospheric stabilization target). (2003). it saysvery little about progress toward low CO2 targets and how energy.

500 1.500 2.. Solar and wind Biomass with carbon capture 800 Biomass Fossilfuels with carbon capture Fossilfuels Hydro and nuclear 200 0--2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Figure 5M2. World Primary Energy Supply Note: World primary energy supply meeting a 400 ppm concentration target by the year 2100...000 ~ 2.t 1.. " __ ~ _ __ ~ _ __ ~_~_~ 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Figure 5-3.-----. per hectare per year Bioenergy prices. 4. Generated with the GET model (for details about this model. see Azar et al.. Biomass Scarcity Prices.--------------------------.. 2005).. per ton I I I I 2- VI //// .1 EJ (one Exajoule) is equal to 1018 J.. which corresponds to 277TWh. 2005) .. and Food Prices Obtained with the.... _L L_ ---~ /- // -_L / / / . 3. GET Model (for details about this model. see Azar et al..000 Land rents.000 500 e o L-_~ -""'-----.. per 100 GJ Wheat prices...I I0 • Christian Azar 1. Land Rents.----------------------------..000 .-------------------------------' .000 ..500 3..

electricity. and heat. oil. The higher biomass price implies that land values will increase as consequently will the price of food. Biomass. Nuclear and hydro power are exogenously set at roughly the current level throughout the next century. very little carbon abatement takes place in the world. and biomass. whereas 500 US$/ton C leads to crop price increases by almost a factor of cost parameters. pulpwood. Carbon emissions drop to less than 1 GtC/yr toward the end of the century. natural gas. but they look at more modest carbon abatement policies (modeled as willingness to pay 3 US$/G] for biomass). and the full potential can be reached by 2060.) Their results are somewhat lower than the estimates above. atmospheric concentrations of CO2 are kept below 425 ppm throughout the century and drop to 400 ppm by the year 2100. and hydro. oil. and food can be observed in the real world. McCarl and Schneider (2001) specifically analyse the competition for land between sinks.5 percent/yr. Land prices are calculated as the price of biomass times the yield per hectare minus production costs (assumed to be 1 US$/GJ for nonland inputs). and these will be exhausted before the end of the century. The global carbon tax is set at 75 US$/tC in 2010 and then increased by 2. and hydrogen grow in importance over time. For assumptions on energy availability. Current wheat prices are set at 125 US$/ton and biomass prices at 2 US$/G]. Solar energy may be converted into electricity. Ultimately recoverable oil and natural gas resources are assumed to be twice current reserves. However. conversion efficiencies. In a modeling study of the greenhouse gas mitigation options in U. see Azar et al. Carbon capture and storage is assumed technically feasible on coal. The higher cost of solar and fossil fuels with carbon capture implies that the price of bioenergy. Figure 5-2 shows the least-cost scenario under the increasingly higher CO2 tax. then fossil fuels and biomass with carbon capture.Walsh et al. Biomass energy with carbon capture and storage opens up the possibility of turning the entire energy system into a continuous carbon sink. and ultimately solar and wind energy used to produce heat. so this analysis might seem purely theoretical. and storage options for captured CO2. The maximum supply of biomass energy is set at 200 EJlyr. and its potential is many times larger than the current global primary energy supply. hydrogen. Shadow prices for biomass are generated by the model (see Figure 5-3). and bioenergy and conclude that a carbon tax of 100 US$/ton C would cause crop prices to increase by 30 percent. biomass. (1996) also find that crop prices increase as a function of higher biomass prices. 2005. solar. nuclear. will increase. there are a few caseswhere the competition between bioenergy. natural gas.S. (This carbon tax level is probably high enough to basically phase out all carbon emissions and therefore equivalent to our solar hydrogen future described above. but comparable to Johansson and Azar (2003). wind. agriculture and forestry.Chapter 5: Emerging Scarcities • II r Energy supply options include coal. . hence profits in this sector. The wheat price is calculated as the price required to make wheat production as profitable as biomass production (a yield of five r/ha-yr and a nonland-related production cost equal to 425 US$/ha). Empirical Evidence of Food-Fuel Competition Currently. both of which give a land rent of 200 US$/ha·yr.

He answered that his company is concerned that this would lead to higher costs for their raw materials because of competition from the energy sector. and severalAnnex-l countries with carbon targets. The eucalyptus is planted on lands previously used for grazing. . The tax is levied at around 400 US$lton C for households. Brazilian ethanol is seen by many as the most reasonable way to meet the EU biofuels directive. it shows that a subsidy for ethanol production from sugar cane can induce a large-scale transformation of agricultural products' use. Farmers commonly use cereals for residential heating (personal interviews. district heating. but this is primarily driven by large tax reductions for ethanol users and less by the carbon tax. On the other hand. In Sweden. and Swedish pulp manufacturers are complaining that the high carbon taxes make it profitable for energy companies to buy the wood needed for paper production and that this leads to higher pulpwood prices. because of competition for raw materials (Dielen et al. Being the chair of the meeting. 2000). If the tax were applied exclusively in Sweden. for example. October 2002). Although this is not driven by climate considerations. and it has already induced some competition between different end users of the raw materials. During a meeting organized by the Swedish Environmental Protection Agency on corporate social responsibility. 2002). European forest-based industries are worried that the EU targets for bioenergy will cause substantial losses in wood industries. sugar is converted into ethanol. that many mills will be closed. and there are plans to base biogas production on cereals (Kohler 2002). are investigating the possibility of importing ethanol or wood from short rotation forests in tropical countries (Faaijet al. Producers of furniture based on particle board face bankruptcy because of increasing competition for their raw materials from the energy sector. Holland and Japan. Grain is currently converted into ethanol. where even thickly forested Finland imports eucalyptus logs from Uruguay. he noted that if all countries levied the same tax. it would put Swedish paper producers at a disadvantage on the international market.V The Swedish government has announced that it will investigate this situation (Dagens Nyheter 2003). the head of environmental affairs of a large pulp and paper manufacturer pointed out that their products were renewable and could eventually be used for energy purposes without carbon emissions. this would be less of a problem because the price of paper would rise (in the same way as food prices can be expected to rise). It is now being argued that ethanol from Brazilian sugar cane is nearly competitive with gasoline (Moreira and Goldemberg 1999).1 12 • Christian Azar Sweden is one of the few countries that has actually implemented a carbon tax. ethanol manufacturers claim that they would have to close down if ethanol from Brazil were taxed in the same way as domestic ethanol. The paper industry has expressed similar fears about competition with carbon sinks (in the context of the Kyoto Protocol). In Brazil. which stipulates an increasing share ofbiofuels in the transportation sector. Large-scale biomass trade already takes place in the pulp sector. thereby increasing wood prices. and the transport sector. Maximizing sinks means harvesting less. and so on. I asked him whether that meant that he was in favor of higher carbon taxes.

According to rural workers I interviewed. which will lead to higher yields and revenues in a positive feedback loop. whether you live in urban or rural areas. The ethanol program in Brazil is worth examining. and the poor might lose their land through a variety of mechanisms.'? Higher food prices could make more people hungry. and how effectively social security and legal systems work.· Chapter 5: Emerging Scarcities • 113 Socioeconomic Consequences for Rural and Urban Poor The socioeconomic consequences of higher land values and food prices are complex. • Farmers heavily in debt would be in a weak bargaining position. although the world produces enough food for everyone. whereas others refer to it as an "exodus" because it has forced thousands of rural subsistence farmers off their lands. Such a phase-out would raise the price of agricultural crops in poor countries. as a large demand for bioenergy could save OEeD farmers from losing income. In the optimistic perspective. higher land values and profits in crop production might mean higher salaries. Today some 800 million people are chronically hungry and higher food and land prices will benefit the rural poor. • Poor farmers might not possess legally binding contracts to their land-in many countries this is the norm rather than the exception (see de Soto 2000 for a detailed account of property rights in developing countries)-and might simply be expelled. .The impact of higher land values would depend on whether you are rich or poor. but the situation is complex. Here I will concentrate on the impact on the rural poor. In fact. At first one might expect that rural poor landowners would benefit. Tree plantations are generally much less labor intensive than agriculture. a result most see as beneficial. since the sugar barons controlled not only sugar production but virtually every economic transaction in the region. The higher income that this may generate will help farmers buy fertilizers and other productivity aids. in particular in urban areas but also among the landless in rural areas. but uninformed farmers might sell at too Iowa price. Higher land values would make their land more attractive to large landholders. This argument echoes current discussion of phasing out agricultural subsidies in the rich world. higher food prices and demand for bioenergy from marginal lands will generate additional income for rural communities around the world. the price of food was higher in the sugar belt than outside it. I have traveled in the sugar cane regions of northeastern Brazil and witnessed extreme hunger among rural poor and their families.but modernization of agriculture or an expansion of tree plantations on lands previously under subsidence farming will mean a reduction in the labor demand per hectare. Most of the world's poor live in rural areas. But this appealing prospect should not blind us to the potentially negative consequences. For the rural landless poor. 18 The primary reason is the vastly skewed distribution of income (and entitlements in general). an increased demand for bioenergy could actually make the phase-out of subsidies in developed countries politically feasible. Higher food prices will also offer incentives to increase food output per hectare. • It might be bought. Proponents argue that it has created jobs for the rural poor.

Some Policy Conclusions This chapter has argued that climate change. which has forced poor farmers to move into the Amazon region (Fearnside 2001). The expansion of eucalyptus plantations for pulp and paper production throughout the tropics offers empirical evidence of the social and environmental consequences that a large-scale expansion of eucalyptus for energy might have. the competition for land is less troubling. On the other hand. and even if I feel optimistic that technological solutions are available. If legal systems work and economic development benefits everybody. the value ofbioenergy and land can be expected to increase. However. In Brazil.) required to meet long-term targets. governments intervene and raise the price of CO2. human demand for domesticating more bioproductive lands might . I have attempted to show that new energy supply technologies. However. Although the total world area planted in eucalyptus is small. or if. and a wide array of substitutions and efficiency improvements at different levels in society can ease the transition to very low CO2 particular since it is closely related to the overall development process. through a carbon tax or a cap-and-trade system. at least at a sufficiently rapid pace. local and regional environmental pollution. The government policies required in order to pave the way for the new technologies include the introduction and continuous increase of the cost of emitting CO2. This means that policies such as public R&D and the creation of niche markets complementary to the carbon fees are needed today in order to decrease uncertainty.1 14 • Christian Azar A similar example is the growing soybean production in Brazil (in the Cerrado. The implications of heightened competition for land are complex and difficult to assess.This competition for land might have both positive and negative social and environmental consequences. for instance. the state ofEspirito Santo has prohibited the Brazilian pulp giant Aracruz from establishing more plantations.and this may lead to grain price increases by a factor of two to three during this century. Higher land vales will offer incentives to improve the quality of degraded lands and might generate income for the rural poor. fuel cells. New policies are required. improve performance. Cases of contemporary slavery (heavily indebted rural workers forced to work for local landowners) continue to flourish in the Amazon region according to the Brazilian government (Rohter 2002). in particular solar energy. the tax or permit fees required to meet near-term targets (such as the Kyoto Protocol) are too low to offer economic incentives to develop and commercialize the more advanced technologies (solar cells. A second key observation is that when. and eventual resource scarcities associated with the use of exhaustible resources can be dealt with. etc. one of the most biodiversity-rich ecosystems in the world).I am less optimistic about the prospects for effecting the necessary policies. and reduce costs of future technologies (Sanden andAzar 2005). extensive local opposition to plantations in Asia and Latin America is documented in Carrere and Lohman (1996). this transition will not take place on its own.

and the global primary energy supply of around 400 E]lyr (WEA 2000. catalytic converters. the EU negotiating positions and their energy policies need to recognize the possibility that a stringent concentration target as low as. and natural gas are 25.000 E]. This can be compared to the annual use of 90 E]. DanielJohansson. I would also like to thank the Swedish Energy Agency. Notes 1. and Vinnova for financial support. Global food intake equals 17 E]lyr (metabolizable energy). Formas. Improved combustion technologies.g. Thus. 4. if not most. • consider countervailing policies to reduce the impact on urban and rural poor in developing countries that might face higher staple food prices. monitor carefully social impacts of higher land and food prices.000 E]. Dolf Gielen. Policies are required to foster the positive impacts and discourage the negative. Eric Larson. 166).Chapter 5: Emerging Scarcities • 115 lead to the conversion of biodiversity-rich ecosystems into monocultures. and poor people might be evicted from their lands. 2. • encourage multifunctional bioenergy systems and agroforestry where appropriate (e.20. Faye Duchin. OECD countries. This has already happened in many. 5. and Stefan Wirsenius and Janne Wallenius for discussions on specific parts of it. respectively. • protect the integrity and biodiversity of valuable ecosystems. to counteract erosion on sensitive soils). 3. Governments should at least • strengthen the legal rights of the rural poor to lands that they cultivate. and gross energy biomass production aimed at food production (including residues on crop lands and grass on pastures not consumed by animals) is roughly equal to 220 E]/yr (Wirsenius 2000).. Coal reserves amount to 20. and Kristian Lindgren. and Bruce McCarl for detailed comments on the manuscript. One hundred fifty E]/yr . oil. Furthermore. Thus a balanced mixture of these fossil sources would release 2 GtC for every 100 E] of energy supplied. The emission factors for coal. sulfur removal) in response to policy regulations have brought about these improvements despite increases in energy demand and transportation volumes. I would like to thank: Dean Abrahamson. and the resource base is estimated at 200. because too rapid extraction could threaten soil quality and long-term productivity. and 14 gC/M]. and • be prepared to tax biomass energy if the demand becomes very high. Improvements can also be seen in many developing countries. removal oflead as an additive to gasoline. Acknowledgments This paper builds on invaluable collaboration with Goran Berndes.400 ppm may be required (pending the contribution from other gases and climate sensitivity). • introduce rules for the use of residues in agriculture and forestry. and improved fuel qualities (e. but a 550 ppm concentration might mean that the temperature increase could be as high as 5-6°C. Local air pollution could be significantly reduced. The European Union also adopted a maximum of 550 ppm CO2 equivalent target..g. • value the carbon content of biota.

7. 11. however. then this system would deliver carbon-free energy carriers (e. It is currently considered too expensive to "mine" uranium from seawater.g.5 billion m3/yr. but the resource is huge--4. They produce as much fuel as they consume. The project was then invited to Norway.004 US$/kWh electricity \WEA 2000. For instance. particularly in light of the fact that 2 GtC/yr corresponds to the carbon flow in 400 Mha of bioenergy plantations with an average yield ofl0 t/ha· research project where only 20 tons of liquefied CO2 would be injected at the bottom of the ocean floor just outside Hawaii was recently canceled before it started because oflocal opposition (GECR 2002).World reserves of natural uranium are currently estimated at 3. A light-water reactor of 1 GW electric output typically produces 200-300 kg of plutonium per year. current reactors only use a fraction of the available energy in natural uranium (natural uranium consists of 0. would be to rely on uranium from seawater. which corresponds to 0. the emission factor per unit of useful energy could be either higher as a result of energy efficiency losses (due to the energy cost of carbon sequestration) or lower because less carbon intense fossil fuels could also be expected to playa role. If CO2 released from burning sustainably grown biomass were captured and stored. This would greatly expand risks of civilian plutonium being spread to nation-states or subnational groups. A comprehensive analysis of plutonium management options can be found in NAS (1994). Here an average carbon "emission" factor corresponding to 25 gC/M] has been used.5 billion tons-and technological improvements might reduce costs to 300 US$/ton. 2001). In reality. this would require that societies continue to sequester carbon from bioenergy and store it (to compensate for the leaking reservoirs) for hundreds or thousands of years.000 tons of plutonium in the global economy every year. 10. This would not suffice for the large-scale expansion envisioned here.000 US$/ha. This is an order of magnitude larger than the value of high- . and. Breeder reactors.3 percent U-238. The construction of a nuclear bomb would not require more than about 6 kg. 6. and one might be more than skeptical about committing future generations to such a large-scale geoengineering project.1 16 • Christian Azar of biomass from dedicated plantations would correspond to 15 billion m3/yr of wood. 9.A large-scale expansion of nuclear energy might lead to uranium scarcities.9 million tons (recoverable resources at a cost less than 130 US$/kg: NEA 2002) and would be consumed over the next hundred years at current mining rates (36.7 percent U-235 and 99. The leakage of 2 GtC/yr could be compensated by sequestering 2 GtC/yr from bioenergy. and the remainder is from secondary sources. This corresponds to the emission factor of coal. and a breeder reactor around one ton. Storage in the oceans is politically sensitive. 316). a 200 US$/ton carbon price would mean that deforestation in the tropics (which releases some 200 ton Cz'ha) would cost 40. Current commercial roundwood production is roughly equal to 1. through the production of plutonium from neutron capture by U-238 (followed by two beta decays). One way to avoid this plutonium intensive scenario. make use of the U-238 in the natural uranium.000 ton U/yr supplies 56 percent of the reactor requirements. electricity) and at the same time remove CO2 from the atmosphere (Obersteiner et al. On the other hand.The problem is that this plutonium would have to be recovered from the reactor-taken out of the reactor and separated into new fuel. Four thousand 1 GW breeder reactors would put 4. while still expanding nuclear energy. However. such as reprocessed spent fuel). This leakage rate might be acceptable if biomass energy with carbon capture and storage were applied as a counteracting measure.The carbon prices required to meet stringent climate targets (most likely several hundred US dollars) would have a significant impact on the profitability of establishing forests as carbon sinks or protecting standing forests to avoid emissions related to deforestation.. but opposition from Greenpeace and the World Wide Fund for Nature led the government to withdraw an approval granted by the Norwegian Pollution Control Authority (Giles 2002). and current reactors rely almost exclusively on U-235). 8.

Out of the remaining 12. say. Each kilogram of grain yields somewhat less than a kilogram of residues (e. capital. it is addressed in a separate paper (Hedenus and Azar 2003a). OM. but there will also be losses in the production. and transportation costs would have to be added. 13.05 US$/kWh. (1997) have argued in favor of fertilizing certain managed forests in Sweden to increase yields from 2-5 m3/hectare per year (ha-yr) to perhaps 30 m3/ha·yr. there would be enough food for everybody were it not for the choice to use cereals to produce animal products. so assume that 200 kg/year per capita will be made available for energy purposes.200 kg/year. On croplands.000 kcal/day per capita of biomass is being produced. For that reason I have included this aspect here. conversion losses in food processing and. Borjesson et al. (Statistics on food availability generally suggest that 2. . All of this cannot be burned. Note that a decision not to use cereals as feed would not necessarily solve this situation. because it might only result in a lower production of cereals. the conversion of cereals into animal foodstuffs reduce this amount to less than 2. but in order to avoid the additional complexities of carbon sinks. Further assume that every person consumes 400 kg of paper and woody products per year.Chapter 5: Emerging Scarcities • 117 quality cropland in industrialized countries. if a global climate regime came to include carbon emissions from deforestation and carbon sinks in forests. On a US$/GJ basis. On top of that. a 350 ppm target implies that the world GDP is 10 times larger in April 2102 rather than inJanuary 2100. To illustrate these numbers. 15. we have 70 EJ/year.g. straw). 156).Thus. It would imply a significant loss of biodiversity because plantations generally are monocultures. harvest. and processing of the raw material. Assume that the hydrogen is produced through electrolysis. with a global population of 10 billion people.000 kcal/day is edible products.ln total we have around 350 kg or 7 GJ/yr per capita. see Hedenus and Azar 2003b). 14. 17. but this alternative way of expressing the costs put them in a perspective worth highlighting for policyrnakers and the general public. the poorest 20 percent consume around 1 kg/yr. even more importantly..The number can be compared with the current noncommercial use of biomass in developing countries. then this would have an enormous effect on the way countries govern their forests. the issue of biospheric carbon stocks is closely linked to the issue ofbioenergy plantations. Comparisons with cropland values are difficult because these values are distorted by subsidies that may amount to several hundred US$/ha'yr (at least in OECD countries). corresponds to 16 US$/GJ in energy costs (assuming an 85 percent conversion efficiency from electricity to hydrogen). For instance. and the current world average is 363kg/yr (Dyson 1996). 16. The estimates do point toward trillion US$ levels net present value cost.Animal products add some 270 kcal/ day. However. 12. 5. Of course. 3. whereas Africa consumes 150 kg. But more intensively managed forests could become of interest. They show that even pessimistic studies of the cost to stabilize the atmosphere acknowledge that the cost "only" amounts to a few years' delay in achieving an impressive increase in global income levels. and therefore competition can be expected in this sector earlier than in the food sector. with the electricity being produced from solar energy. pulpwood is cheaper than food. An electricity production cost of . 18. Assume that a third of this residue stream is recoverable for energy purposes. and for reasons of space. (per capita paper consumption by the 20 percent in the world that consume most paper is around 200 kg/yr.This is not to say that the costs are insignificant. and this is twice as much food as is needed (Wirsenius 2000).000 kcal/day. Currently the production cost for solar electricity is at least three times as high.700 kcal/day per capita is available. which amounts to some 33-55 EJ/year (WEA 2000.000 kcal/day of which is pasture biomass and animal forage crops. Thus. and it would be uneconomical to convert forestland to pasture or cropland. North America and Europe consume 600-800 kg/yr per capita. assume that the average citizen of the world consumes 500 kg of cereals per year (direct and indirect consumption). South Asia 237 kg.000 kcal/day. 15. Recycling will lower the per capita input demand to.

Borjesson. C. UK: Edward Elgar. Multifunctional Bioenergy Plantations--Final Report to the Swedish Energy Agency (Multifunktionella bioenergiodlingar. Tisdell. and E. Borjesson. and K. London: Zed Books.. R. not to actual intake). and H.]une 12..Ayres. F. and T. 1997. Bioenergy and Land-Use Competition in the Northeast of Brazil: A Case Study in the Northeast of Brazil. H. L.M. Pulping the South. 2002. UK: Edward Elgar. Andersson. and capital that does not lead to increases in overall food production.. Rodhe. Sometimes this observation has been perceived as an argument that a redistribution of income (or land. edited by R.. L. Mollersten. E.A. Slutrapport till Energimyndigheten). Maki.I 18 • Christian Azar see FAO 2002 or Dyson 1996. 1994. But this number refers to what is available at the wholesale level. Lohmann. 1996. K. and G. Andersson. 1996. In Handbook of Industrial Ecology. C. and L.]. Rytkonen. C. B. Biomass and Bioenergy 13: 399-412. and B. Lacour.. Carrere. c. Kaberger. . The Implication of CO2-Abatement Policies on Food Prices. 2005. 2000. D. and Stolp. Lindgren.. G. Swahn. de Soto. References Abrahamson. R. Cheltenham. primarily because democratic governments need the support of their people to be reelected (Sen 1981).. Energy Policy 31: 961-976. Global Environmental Change 12: 253-271. But a more effective approach would be to also increase the production of food. c. The number of people suffering from starvation is much lower and cases of starvation are much more rare in democratic countries than in dictatorial regimes. 2000. Rade. Global Energy Scenarios Meeting Stringent CO2 Constraints-Cost-Effective Fuel Choices in the Transportation Sector. G. Emission Scenarios and Global Climate Protection. Science 276: 18181819.D. Ayres. In press. Eskiilstuna: Swedish Energy Agency. P. Linder.Ayres and L. 2002. and E.. The Political Atom. The Mystery of Capital-Why Capitalism Triumphs in the ~st and Fails Everywhere. Azar. Alcamo. Gustavsson. Azar. 1997. Azar. Kreileman. Christersson. Cheltenham. K. Climatic Change. 19. Azar. Azar. 2000. Azar. capital. Future Production and Utilisation of Biomass in Sweden: Potentials and CO2 Mitigation. land. and J.. L. Dragun and C. 2003. On the other hand. 1999. Targets for Stabilization ofAtrnospheric CO2. 2002. even democratic governments have not been able to eradicate hunger. Larson. London: Black Swan. Berndes.M. Energy for Sustainable Development IV(3): 64-72.Are the Economic Costs of Stabilizing the Atmosphere Prohibitive? Ecological Economics 42:73-80. Guegan..H. 2000. Larson. It is difficult to envision a redistribution of income. P.. -Bulletir: of the Atomic Scientists 56: 39-44. 391-404. Fredriksson. Carbon Capture and Storage from Fossil Fuels and Biomass: Costs and Potential Role in Stabilizing the Atmosphere. Schneider. In Sustainable Agriculture and Environment: Globalisation and Trade Lioeralisation Impacts. and I. as is evident in countries such as Brazil and India. and S. 2003.. and S. Bioenergy and Water-the Implications of Large Scale Bioenergy Production for Water Use and Supply. Dagens Nyheter. Lindgren. Dielen. Environmental and Resource Economics 4: 434-454. Material Constraints on Technology Evolution-The Case of Scarce Metals and Emerging EnergyTechnologies. On Economic Disequilibrium and Free Lunch. Global Environmental Change 6: 305-334. edited by A. work opportunities) is the key step toward achieving an end to hunger. EU Energy Policy Impacts on the Forest-Based Industries: Modeling Analysis of the Influence of the EC White Paper on Renewable .. Paper Industry Battles with Politicians (Pappersindustrin i kamp med politiker). c. Berndes. 2002. Berndes.

J. July 26. McCarl..A. Wageningen: Afocel. McCarthy. Fermented Cereal Grain Can Produce Inexpensive Biogas without Government Subsidies (in Swedish). the Netherlands. Hedenus. FAO (Food and Agriculture Organization of the United Nations).Washington. Soybean Cultivation as aThreat to the Environment in Brazil. ---. Special Report on Emission Scenarios. F.. N. and M. 2002. and R.jordbruksverket.A. 2001. ScientificAmerican 286(1): 70-79. Gielen. Swedish Board of Agriculture Report no.A. Large-Scale International Bioenergy Trade.P.G.Y. Cambridge: Cambridge University Press. Climate Policy 1: 155-177. Hall. and Vulnerability. Khesghi.Work in progress. edited by Houghton et al.H. Callander. ---. The European and Global Potential of Carbon Dioxide Sequestration in Tackling Climate Change. June 17-21. de Feber. Nature 419: 6. andJ. van der Vorst. Schneider. 1996. and Climate Protection. JiliVrld Resources Review 15: 165-175. P. ---. M. Biomass for Energy or Materials?A Western European Systems Engineering Perspective. Williams.A. T. EcologicalEconomics. Global Environmental Change Report.J. R. Leary. Greenhouse Gas Mitigation in US Agriculture and Forestry. 2002. Cambridge: Cambridge University Press.Washington. 2001 b. Schlamadinger. Cambridge: Cambridge University Press. Biomass for Energy: Supply Prospects. White.. IIASA/WEC. FAO Database.J. Dokken.M..K. B.T. Lake. Canziani. 2002. V. Pearson. M.]. Kelly.A.fao. B. Giles. Kotek..World Income and Resource InequalityTrends. Bennett.K. 2003. IPCC (Intergovernmental Panel on Climatic Change). Summary Report. and UA. B. R. Houghton. Fearnside. 2002. R. 1996.F. and].London: Routledge. D.Azar.Adaptation and Mitigation Options. .A.. 2001a.and C. and Leach. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. 2001. D.1996. 1992. London:World Energy Council.1999.J. M.D. Science 294: 2481-2482. 2003a. Nakicenovic. Norway Sinks Ocean Carbon Study.A.. T. .A. Global Energy Perspectivesto 2050 and Beyond. O.F. 2001.A. Environmental Conservation 28: 23-38. report in Swedish with English summary). edited by J. Sands. Industry. Agriculture. 1993. In Renewable Energy: Sourcesfor Fuel and Electricity. Brown.Rosi116-Calle. ---. Climate Change 2001: Impacts.S.Azar. H. M. Fouquet. Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate A. Dyson. Bos. Cambridge: Cambridge University Press. Johansson. and H.J. ---. and S. Edmonds. Analysis of Land Competition between Food and Bioenergy.. edited by N. N. Grimston. 2002. 1995. Utveckling av arrende. Gerlagh.]. 2002. R.Williams. land prices and real estate prices. edited byT. Nangis. Impacts. edited by J. http:// www. 2003b. Ny Teknik 35: 5. D. Carbon Sinks versus Bioenergy-Physical and Economic Perspectives.B. Amsterdam. Cambridge: Cambridge University Press. Population and Food-Global Trends and Future Prospects. 2001. R. GECR (Greenpeace and the World Wide Fund for Nature). Climate Change 2001: The Scientific Basis. http://www.A. 2001.Chapter 5: Emerging Scarcities • I I 9 Energy Sources on the JiliVodSupply to the European Forest-Based Industries. DC: Island Press. andT. DC: Pacific Northwest National Laboratory.C. 1993. Solantausta. In press.N. DC: Island Press. mark och fastighetspriser i Sverige (Trends in agricultural rents.Woods. Paper presented for the 12th European Conference and Technology Exhibition on Biomass for Energy. Wagner.. Energy Policy 29: 291-302.M. Climate Change 1992: The Supplementary Report to the IPCC ScientificAssessment. IPCCWorking Group II. Land Use and Commercial Biomass Energy. Renewable Energy: Sourcesfor Fuel and Electricity. 2001: 8.E. D. Johansson. Jordbruksverket. and K. Faaij.Wise. Karakoussis. Next Generation Nuclear Power. Kohler.A. F. Reddy.Washington.Varney.Johansson et al. P. and C.

unfccc. M.Wc.Azar.D. 1992. 2000. United B. Obersteiner. http://www. C. and R. 1994. ]. 2001. 2000.ornl. 2002..A. S. Economics and Environmental Choices in the Stabilization of Atmospheric CO2 Concentrations. Production. and B. Washington.]. L. Targets and Indicators oj Climatic Change. http:// T. Sen. 2000.]. Schlamadinger. and B. Richels. C.. Eskilstuna: Swedish Energy Agency. United Nations Development Programme and World Energy Council. Trapped Like Slaves on Brazilian Ranches-Forced Labor Clears Forest for Cattle. walsh.A.. S.R. R. Walsh. Department of Agriculture). Edmonds. 2002. 2005. S. Sailor. Price.120 • Christian Azar Moreira. Riahi. Sdence 292: 281-284. Pasture Rents Up Slightly in Wyoming. Near-Term Technology Policies for Long. Forecasting Agriculturally Driven Global Environmental Change.P. March 26. and Demand. Management and Disposition oJExcess Ui!apons Plutonium. Ray. Framework Convention on Climate Change. Energilaget 2001 (Energy situation in Sweden 2001). Graham. Nature 379: 240-243. Wigley. Fargione. http://bioenergy.. Wirsenius.Y. Science 294: 786-787.html.1995. M.. Moreira. Economic Analysis of Energy Crop Production in the US-Location. Department of Physical Resource Theory. 2001.Yamagata.]. WEA (World Energy Assessment). B. van Ypersele. Bodansky. D. Tilman.. Ph. van der Zwaan. Fetter. WBGU. K. NEA (Nuclear Energy Agency). Rohter. de la Torre Uguarte. Uranium 2001: Rijsberman. P. D.htm.S.. and]. October 4-8. 2002. 1999.nass. 2001.. Bremerhaven.. and C.Yan. The Alcohol Program. STEM. K. and J. Paper presented at BioEnergy 1998: Expanding Bioenergy Partnership. USDA (U. R.usda. Energy Policy 27: 229-245.J. NAS (National Academy of Sciences). Mollersten. P.1998. Germany. 1990. . D. Nuclear Power-A Nuclear Solution to Climate Change? Sdence 288: 1177-1178. S.Wolff. Braun. Stockholm: Stockholm Environment Institute. 1981. Shapouri. Chalmers University of Technology. 1998. Paris: Nuclear Energy Agency. Human Use of Land and Organic Materials: Modeling the Thrnover of Biomass in the Global Food System. Quantities. and Impacts on Traditional Agricultural Crops. and H. diss. www. International Herald Tribune. Azar. Poverty and Pamines-=An Essay on Entitlement and Deprivation.Term Climate Targets. and ]. Goldemberg.stem. Kauppi. 1996. Oxford: Clarendon Press. Managing Climate Risk.L. Energy Policy 33: 1557-1576. Swart (eds. RR. DC: National Academy Press. Slinsky. Scenarios for Derivation oj Global CO2 Reduction Targets and Implementation Strategies. D. Read.). Nilsson.


41296 Goteborg. Chalmers University of Technology.azar@chalmers. SWEDEN Email: fossil fuels have passed/pGt()1J.G/yn. In order to avoid serious climate change. or near-free. hydrogen. global carbon emissions may have to be reduced by more than half by 2100. and higher energy the end ofthe 1. BECS stands for biomass with carbon capture and storage. technologies: solar power. Supply 900 800 700 600 EJ/500 400 300 200 100 yr o 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Figure 1 A world energy-supply transformation scenario under a CO2 stabilization target of 350 ppm (for details see Azar. Figure 1 shows a global energy scenario that meets this target (atmospheric CO2 concentrations are stabilised at 400 ppm). INTRODUCTION Global CO2 emission~fr()111~u11l. A major restructuring of the energy system is inevitable if we are to avoid the threat of serious human-induced climate change. compared to today.i/and they are on track to exceed 20 GtC/yr.Cost-efficiency versus political feasibility Christian Azar Department of Physical Resource Theory. In the short run. These technologies need to be developed further before they are employed on a large-scale. it builds on intense use of existing technologies for biomass..Climate policy . 1 . this scenario relies more heavily on new and advanced carbon-free. In the long term. wind. and by around 90 percent compared to projected amounts. et aI2006). and fossil fuels and biomass with carbon capture and storage.

section 2 deals with climate policy in a world where there is agreement that we should act. Prioritization is needed. one could list dozens of policy instruments. since cost-efficiency often clashes with political feasibility. essentially sufficient yet critical for meeting low emission targets at minimum costs. and that more advanced technologies are developed. I consider three categories of policies central to meeting climate targets cost-efficiently. In this paper. In reality. in which policy choices are more difficult. One aim of this paper is to better understand the conditions for costefficient policy-making and when venturing beyond cost-efficiency may be a reasonable choice. 2. The three key policy instruments are: • • economy-wide price incentives (to correct for the market failure of there being unpaid costs associated with emissions). long and complicated answers are given. If there is political will. such as the Kyoto Protocol.jointly. help us protect the climate at the lowest possible cost. but that there is not enough political will to put these policies in place. political feasibility seems inversely proportional to cost-efficiency. nor efforts to reduce carbon emissions from an increase in biota. I first present the climate policies that. The paper draws upon a recent experience of mine as a member of the Swedish Prime Minister's Commission Against Oil Dependency (spring 2006) and observations made concerning the EU Emissions Trading Scheme. On the other hand. in theory. Second. the following three sets of policy options are. Loosely speaking. Sometimes.greenhouse gases. specific technology policies that enhance the development of advanced energy technologies through both R&D support and market diffusion programs (to 2 . I turn to the real world. COST-EFFICIENT CLIMATE POLICY In this section. the question about what needs to be done is not that complicated. I have been at several meetings with ministers and policy makers where experts have been posed questions about what to do about climate change. Often. It is important to note that too much focus on cost-efficiency risks leading to fewer abatement policies being put in place. The paper will not deal with international climate negotiations and treaties. In reality.We need climate policies to ensure both that currently available low carbon technologies are deployed more extensively. The key problem is not that we do not know what to do or that it is too complicated. too little focus on costefficiency may mean that the costs of climate policies become too high for many policy makers. reduction of non-Co. Key examples of this include the fact Sweden has a much higher carbon tax on residential heating and transport than on industry and that the EU decided to grandfather rather than auction permits in its emissions trading scheme. but I think that such approaches may sometimes be more confusing than enlightening. section 3 deals with a situation in which there are strong special interests that lobby against climate policies. CDM projects.

correct for myopic behaviour in industries and difficulties in keeping industry inventions in-house, when long time-scales are involved, leading to underinvestment in R&D and learning-by-doing for these new technologies); standards for energy efficiency (to correct for market failures associated with asymmetric information, skewed incentive structures, and other barriers to energy efficiency).

All three categories of incentives are "critically complementary", in the sense that implementing only one or two of them will not suffice to reach the climate target at minimum cost. For instance, policy that only relies on developing more advanced technologies (e.g., a Manhattan- or an Apollo-style program) is not likely to succeed since such programs do not change the economic environment in which companies make investments. Thus, even if solar PV or hydrogen technologies drop in costs, they may remain more costly than coal for base-load power. Further, when introducing climate policies, e.g., a carbon tax, the economic landscape will change: demand for new technologies will emerge, demand for old technologies will drop. There may be environmental, social, and security problems associated with these new technologies and these problems have to be addressed already from the beginning. Examples of such problems and countervailing policies will be given below. 2.1 Economy-wide price incentives Global climate change is a complex problem. Governments cannot know in detail which technical solutions are best for which company, household, or consumer. For that reason, policies that rely on price signals are in general preferable to policies that regulate individual companies and their choice of technologies. Such price signals work in three complementary ways: • by offering incentives to shift fuels toward less carbon-intensive alternatives (renewables, natural gas instead of coal, nuclear, and fossil fuels with carbon capture and storage); by leading to higher energy prices which increases the benefits of increasing energy efficiency, choosing less energy-intensive ways of life (public transportation instead of private cars, rail instead of trucks), saving energy through life-style changes, and choosing less energy-intensive materials (the benefits of recycling materials increase); by offering, at least if they are part of credible long-term plans to reduce CO2 emissions, incentives for the private industry to develop new and more advanced C02-neutral energy technologies (both on the supply and demand side).

Experience with carbon-based price incentives is limited to a few countries: the Nordic countries and, more recently, the ED as a whole. The Swedish carbon tax (roughly 350 DSD/tonC) was implemented in the early 1990s and led to substituting an increased use of biomass, heat pumps and waste heat, and an expansion of the district heating system, for oil for heat. The district heating system in Sweden is now almost fossil-fuel free (see figure 2).


The Norwegian carbon tax (set in 1996 at around half the Swedish level) gave Statoil the incentive to sequester CO2 emissions from its gas fields. Natural gas contains CO2 that has to be removed before the gas is sold. Normally, the CO2 is just vented into the atmosphere, but in this case it became economically profitable to capture and store it in an aquifer, 1000 meters below sea level. TWh per year


40 30 20 10

1970 1974 1978 1982 1986 1990 1994 1998 2002
Figure 2. Swedish district heating fuel use. Biomass use has been driven by the carbon tax since 1990 and before that date by various excise taxes on fossil fuels. (Source, Energilaget, Swedish Energy Agency, 2004, data available at

The EU first opted for a carbon tax (in the beginning of the 90s), but it proved impossible to reach agreement at the time (the then-Elf Environment Commissioner actually resigned in 1992 for this reason). In 2005, however, a cap-and-trade system was introduced. The cap covers more than 12,000 large point sources of CO2 in the European Union, but point sources such as cars, trucks or residential fuel use are exempted. Overall, roughly 50% of all C02 emissions in the EU are included. It is too early to evaluate the EU Emissions Trading Scheme (ETS). Initially, the carbon price rose to 20-30 USD/t CO2. But individual countries allocated more permits to their corporations than these needed. When that became evident, the price collapsed to less than half of its previous value in the course of a few days, in the spring of2006. In a recent analysis, Watts (2006) pointed out that the EU allocated more permits for the year 2005 than were actually used. (Total allocation amounted to 1,829 Mtons CO2, whereas emissions ended up at 1,785 Mtons CO2. Poland is responsible for two thirds of this excess.) Since the trading period is three years long, these permits may eventually be in short supply, at least if there is an increased demand for electricity or energy intensive materials, but there is also a risk that the price collapses to very low levels and that the scheme will not really lead to any reductions. This was known already when trading began in January 2005, but a low ambition when it comes to permit allocation might have been necessary both to get the system in place and also because the allocation was carried out within a year before the trade started, leaving little time for investments in bringing down 4

emissions. However, it is clear that for future periods, ED governments must reduce the amount of allowances allocated, if the ED is to comply with the targets in the Kyoto Protocol without imposing an undue burden on the non-trading sectors or relying very heavily on international carbon markets (CDM projects, joint implementation, or hot air from Russia).

Emissions trading versus a carbon tax Many companies prefer emissions trading to a carbon tax. Emissions trading may seem more market-based ("it offers an opportunity for the market to solve the problem;" "emissions will be reduced where this is most cost-efficient to do so," etc.). But a tax also leads to emissions being reduced where it is most cost-efficient; a tax is a market-based instrument. Equally important, emissions trading, per se, does not reduce emissions. Government involvement is still required. With a tax regime, governments set the carbon price; with a cap-and-trade regime, they set the cap on total emissions. The cap leads to emission reductions; trading helps to achieve reductions at a lower cost. Same carbon price in all sectors In order to ensure economic efficiency, text-book economics teaches that the carbon price had better be the same in all sectors. This can be obtained by implementing the same tax in all sectors, or through a cap-and-trade scheme that encompasses all relevant sectors (for certain sectors, cap-and-trade may be difficult). Suppose instead, for instance, that the carbon tax is lower for power generation than for oil used for residential heating. Someone who heats hislher house with oil could then shift to heat pumps, powered with electricity from a coal-fired power plant, and pay less for roughly the same emissions. The overall costs to meet the target would increase (assuming increasing marginal abatement cost curves). Similarly, there is no reason why someone who uses an electric car should pay less for his/her CO2-emissions (from the power plant), than those who use gasoline or diesel cars. Further, in many countries there are policies to promote biofuels. But the C02emissions from the production ofbiofuels are often substantial, perhaps half of, or even equal to, the emissions from using the equivalent amount of gasoline (this may be the case for biodiesel based on rape oil, wheat, corn, or wine ethanol). In order to avoid sub-optimization, these emissions should be taxed at the same rate. In systems analysis of fuels and their environmental impacts, a life-cycle perspective (or a well-to-wheel analysis) is often emphasized as a way to understand which fuels have the best overall energy and CO2 performance. How to draw systems boundaries becomes very important since the scale of the analysis becomes critical (should wefor instance include the fuel used for producing the fertilizers?). Interestingly, a uniform carbon price offers a very attractive way of providing the necessary information and incentives to the market for avoiding such sub-optimizations. If the indirect emissions associated with heat pumps or corn ethanol really are in parity with those from oil combustion, then this would show up in the price of the products. 1
1 There are special cases where it may be reasonable to have different carbon prices. This could be the case if awareness about future more stringent climate policies is higher in one sector than in others. If, for instance, the transport sector is more myopic than energy-intensive industries, then this behaviour could be corrected for by implementing a higher carbon tax in the near term in the more myopic


energy efficient cars. or a stringent target for emissions set well in advance of the target date will assist the required technology development but will not be sufficient. and biomass). Policies to bring advanced energy technologies to the market: R&D and market diffusion programs Technological change consists of two processes.2. those set in the ED ETS or the Kyoto Protocol. they are also of critical importance in the process of setting these stringent targets. 3) uncertainty regarding whether stringent climate policies will be put in place and high internal discount rates in private companies entail that too little . we can use existing and mature technologies (e. In addition. to be discussed below) are preferable. Interestingly. In the first. "technology development".g. 6 . price incentives (and sometimes performance standards. The reasons for this are complex but ultimately stem from the facts that 1) it takes a very long time (30-50 years) to take advanced technologies from invention through innovation to mass diffusion. it is likely that policy makers will never dare set and implement such stringent targets unless they can be sure that these can be met without too many sacrifices. or choose small. The largest drop is for nuclear fission. In order to meet near-term climate targets. for the lEA countries). So. e. see figure 1). A tax on carbon emissions. the global trend in publicly-funded energy research has been rather constant or even declining (see figure 3. in the beginning of the 80s.. natural gas..g.from a social point of view . Therefore. industry so as to obtain "correct" investments for the subsequent period where higher carbon taxes will be implemented all around. policies that specifically aim at developing new advanced technologies will not only help us meet stringent climate targets at lower costs. from those already available on the shelf. more advanced energy technologies are needed (for an illustration. wind. The second kind of technological change amounts to bringing new technologies to the shelf. we simply choose a different technology. and somewhat discouragingly. cars.2. But in order to meet more stringent long-term targets-say. which is understandable since the nuclear industry is mature. We may for instance start using gas instead of coal to generate electricity. In order to achieve change the technologies we pick from the shelf. instead of big. reduction targets of 50% by the year 2050-as envisioned by several European governments. what needs to be done? A very common answer is to increase R&D expenditures.research is carried out on developing these technologies. but there was also a drop for renewables. 2) diffusion of knowledge at this time-scale implies that private companies are not likely to reap the benefits of in-house research on advanced technologies.

through innovation and full-scale diffusion. generates investments and learning that lead to further cost reductions. Over the same period.iea..aspx ?id=2 82). in turn. Governments need to develop bridges over what is sometimes called the "valley of death". Germany. There are certain positive trends in the implementation of such policies. while needed. The key is to set in motion a process of self-sustained growth. portfolio standards and public procurement programs (where governments. The installed capacity in the world more than tripled between 1998 and 2002. e.I) large Hydro (> 10 l\·tW) Million IJSD (2004 price. decide to buy only hybrid cars. All these policies aim at creating niche markets that enable learning-by-doing.. implying an annual growth rate above 30% (see figure 4). and Spain. and exchange liiltei.Small Hvdro « 10 ~r~r. Advanced energy technologies are often not costcompetitive when taken from the lab. cost-covering prices.1 2S00r-----~--~--------~------------------------~ 2000 r Geothermal • Biomass D Ocean !Ill 1500- VIlind E3J Solar Thermal. 7 . where cost reductions generate market growth which. This has led to a rapid growth. or municipal bus companies decide to run only buses using fuel cells powered by hydrogen).g. Possible policy instruments for this include investment subsidies. and some other countries have large market-based incentives for the development and expansion of wind power.e.!Ill electric 1000 - SOlar Photovoltaic SOiaI' Heating 0 &Coollng Figure 3 Government-funded energy research. the marketsupport programs for solar PV in Japan and Germany increased the global installed PV capacity by more than 25% per year (see figure 5). development and demonstration in IEA countries (MUSD/year). driven by dynamic learning and scale-effects. will not suffice to bring advanced technologies to the market.. i. Source: International Energy Agency (http://www . org/w /bookshop/add. and there are pros and cons with all these approaches. green certificates. policies that nurture technologies from the stage of invention. Denmark. There is also a need to create markets so that they are brought from the lab to commercialisation. Higher public spending on R&D.

2010 2000 1990 1980 Figure 5 The cumulative global production of solar photovoltaics Source: PV News. EWEA 2003 2 1 0. 8 ..--------.. Photon International 1980-2002."-'----. 1980-2002.-------...35 30 25 20 GW 15 10 5 0 1980 1990 2000 Figure 4 The global wind power capacity. Source: Gipe 1994.

a "technology-neutral" policy is an evasive objective when it comes to developing more advanced carbon-free technologies. Similar observations can be made with respect to policies for introducing biofuels. even though it is widely recognized that these are more promising. DME~ FT fuels. However. To 9 . Thus. I definitely agree with this view. i. However.~the result of an intended search for solutions to a specific problem (e.or not. and all details can not be discussed in this brief paper. This will further strengthen the role of these technologies over solar PV~ despite the fact that solar PV very well could be a lower-cost option in the long run. sugar beets. supporting the development of advanced technologies. when it comes to bringing technologies to the shelf.a serendipitous process? Inventions may be purposive. This means that these technologies will gain training grounds on which they may thrive and experience leaming-by-doing and cost reductions. it will lead to a lot of support for the poorer-performing fuels pathways and must reasonably be complemented with support for the more advanced elusive objective? The process of selecting policies for technology diffusion is complex. So far I have only discussed ways of driving "purposive" inventions. bio. the design ofthe policy instrument.Inventions . governments will "pick winners" almost regardless of how they choose policies. it will lead to one outcome. cap-and-trade systems or carbon taxes must form the backbone of climate policies.e.. improvements in battery technology aimed at consumer electronics and power electronics in trains have been of importance for the emergence of electric hybrid vehicles. if it is initiated now. solar PV~ and small-scale hydro count as renewable in this system. only wind.. ends up supporting only the technologies that happen to be most cost-efficient at present. It is often claimed that governments should not pick winners. the Apollo or Manhattan projects) or be completely serendipitous. wheat.e. In that sense. and this ensures that government involvement in selecting technologies from the shelfbecomes minimal. as is currently the case in the EU~ excess wine. if any.come to play an important role in many others. Further. If a policy is implemented when more advanced technologies are developed. For these reasons. therefore. A "technology-neutral" scheme would promote ethanol from com. improvements in one area may .unexpectedly . the timing of policy implementation is very important. will invest in solar PV. in which the government requires electricity consumers to buy a certain amount of renewable electricity.g. although seemingly technology-neutral among the included energies. Few. Governments should not pick winners . and which have been effective in making use of others. i. since bioelectricity and wind are much less costly. In Sweden. Governments should therefore try to better understand which types of national invention and innovation systems have been successful in generating inventions and innovations. For instance. One issue I want to cover is whether governments should try to pick winners . Think of a certificate system for electricity. or hydrogen) or ethanol from cellulose. many important inventions throughout history have been serendipitous. but not the more advanced fuel pathways such as gasification of wood (to produce methanol. or.

in seeking to lower building costs. zero-heating houses. which is five times higher than today. hybrid cars. construction companies may be tempted to cut expenditures on energy efficiency measures (e.3 Efficiency standards In the climate debate. technologyspecific subsidies to. On this basis. Then I will discuss the need for policies that specifically aim at driving energy efficiency.g. Such improvements can be expected from both higher carbon prices (which drive higher energy prices) and technology policies that aim at developing more energy efficient products (e. all new promising technologies cannot. be supported. improving energy efficiency is central to meeting low emission targets. be as technology-specific as possible. we also need to develop standards for energy efficiency. since there are other less costly ways to meet the constraint). fuel cells.g. wind and solar PV are warranted. we would be roughly 60% more efficient and the energy supply 800 EJ/year (which is in line with the levels shown in figure 1. Primary energy supply per capita in the OECD region is around 200 GJ/cap/year.g. there are already cars being sold that are more than twice. By the end of the century. lighting can be made five times as efficient. Clearly. I will start with a back of the envelop calculation illustrate that point. arguably. etc. Here. and this may result in energy efficiency levels that are not optimal from an economic point of view. Improving the efficiency with which energy is used might be equally or even more helpful than the potential contribution from all energy supply options.). However... Priorities must be set. as far as possible. there are instances where the case for standards is strong. Economists often raise concerns about such standards because they lack confidence that governments are better than the market at choosing technologies. This level is in line with many global energy scenarios. e. insulation) since consumers pay for heating. and should not. 1 TW = 31 EJ/year). most discussion tends to focus on energy supply technologies. as efficient as the current average. For instance. We would then end up at an energy supply of2000 EJ/year. there are houses built in cold climates that do not need external heating systems.avoid. complementary policies should. or three times. This is not a far-fetched objective. Clearly.. primarily in the residential sector (household appliances and heating). whereas subsidies to. 2. com-ethanol are questionable (since it is very unlikely that costs will come down significantly and that com-ethanol would be used in a future with a very stringent carbon cap in place. plug-in hybrids. Assume that 10 billion inhabitants of the world on average would enjoy that level by the year 2100.. etc. e.g. and (ii) the supported technology should have a reasonable likelihood of winning market shares in the future when there is only acap-and-trade or a carbon tax in place. Assume that energy efficiency improves by 1%/year over this period of time. When choosing which technologies to support. But. Information asymmetries may contribute to excessively low levels of energy efficiency in residential electricity appliances. (i) Subsidies should be expected to lead to significant reductions in costs. having governments pick winners. 10 .

Assume for instance that a person buys a $20. that very stringent climate policies will be put in place during the lifespan of the product. or may in some cases be counterproductive. or 20 ton over the lifespan of the car.000 car (with CO2 emissions at 120 g/km) instead of a $50. Still. The implication would be that energy demand would drop and.4 Countervailing policies and back-casting strategies for desired energy futures Efforts to reduce CO2 emissions will change the economic landscape so that demands for new technologies emerge. 2. close to the actual gain. This line of reasoning would support standards for very efficient houses. Let me illustrate how efficiency standards without a cap on total emissions or a carbon price that increases the relative price of carbon over other goods becomes less effective. In the process of solving one problem we should be 11 . i. if market actors are unaware of the fact. This implies that performance standards for energy efficiency are likely to enhance cost-efficiency . so the problem with the rebound effect is much smaller. However. that the price of the permits would drop too.000 USD over the lifespan ofthe car on lower gasoline expenditures). I emphasized the fact that price incentives. If there is a cap-and-trade policy in place. or the likelihood. hence. and Swedish cars are currently at around 200 gC02/km. It is also important to notice that in many cases the economic gains from using a more energy efficient product are much smaller.Information asymmetries may also operate over time. If. on the other hand. since these should be expected to be able to make the best use of the price signals they get from the market. If the price policy is implemented through a tax.e. For instance. efficiency standards are implemented when there is a cap-and-trade policy. This could facilitate the introduction of more stringent caps for the subsequent period. I do not think.13 kg C02IUSD). Now. that performance standards should be implemented for energy-intensive industries.000 USD for the car and roughly 10.000 km implies that the total gain amounts to 2 ton CO2 Iyear.000 SUV (220 gC02/km). assume that the consumer spends this money as the average American consumer (here taken as CO2 emissions per GDP. for instance. the consumer also saves money by buying the smaller car (30. The EU target for 2008 for the average car in the EU is 140 gC02/km.but standards should be implemented with some care. equal to 0. Earlier. energy efficiency standards will lead to lower emissions. then a performance standard for longlived products would be one way of avoiding costly investments. This example should not be read as an argument against buying more energy efficient cars (since a consumer who decides to do that probably has a lower carbon to income spending on other goods as well). Assuming an annual driving distance of20. then overall emissions will remain the same. the example points to the need for a carbon tax or a cap-and-trade system that keeps the overall pressure on the economy's carbon emissions..000 dollars spent on other goods add up to 17 ton CO2. Thus 35. policies to speed up technology development. and performance standards for energy efficiency are complementary in the sense that they need one another to work properly.

farmers risk being expelled or bought off the land for very low prices. In countries with properly functioning property-rights systems. We saw this when sugarcane expanded in Brazil. in the 90s. as farmers. all nations have the right to build uranium enrichment and reprocessing plants. Nuclear energy and nuclear weapons Nuclear energy provides a similar example. With an enrichment plant in operation it takes at most some six months to build a nuclear bomb. we are not likely to witness any strong reemergence of nuclear power. or expelled farmers may move into the woods. are moving in. Questions about how to deal with the link between nuclear power and nuclear bombs in an equitable. not only for how to manage forests acceptably when the demand for timber increases. Back-casting 12 . we need perhaps a ten-fold increase in the number of power plants. in the 70s. biomass will become increasingly profitable. higher electricity prices resulting from carbon policies are expected to increase industry interest. both negative and positive impacts on the world's poor. but elsewhere. One or two more nuclear reactors in countries that already have nuclear power will not make much of a difference. Governments need to think through what this means for global security. but that it must buy the reactor fuel from Iran. in the Amazon. This might lead to higher food prices with potentially. Countervailing policies to protect poor farmers and sensitive ecosystems are clearly warranted. Plans have been put forward to limit the rights of nations to build such installations. but also for how to deal with the increased demand for branches and tops (not normally used by the timber industry but now in growing demand from the energy sector following our carbon tax). with oilpalm plantations being established in rainforests. Biomass for energy: emerging competition with food production and sensitive ecosystems? With higher CO2 taxes. but if nuclear is to make a significant dent in the global carbonemissions trajectory. and secure manner must be resolved before we go nuclear again. Further. and I would like to stress this. and when soybean expanded in the same country. Currently. The former is happening now in Malaysia. or material for a bomb. The difference between an enrichment plant used for producing civilian nuclear reactor fuel. this might not be a problem. Here I will illustrate this with a brief discussion of biomass and nuclear. But. rainforests may be cut down to establish plantations. but will they be accepted by other countries? Imagine if the UK was told it may have nuclear power. Sweden is already establishing regulations. expelled as a consequence of the soybean plantations. the latter. is essentially only a matter of intent. This means that farmers might prefer to grow energy crops instead of food. One particular risk is that the higher profits associated with biomass might lead large agricultural farms to establish plantations on land from which subsistence farmers make their living. In the absence of climate policies.careful not to create others. just.

Clearly. too. what pressure should be the standard in hydrogen tanks and other distribution systems. It would raise the price of electricity from coal fired power plants by around 1 USc/kWh but it would also generate global revenues equal to 357 billion USD..e. Here back-casting exercises may be useful. we might get locked into nuclear. the establishment of infrastructures for gaseous fuels. e. one could mention the need for more information (about the climate problem as well as about the energy performance of various products). There is also a need for demand-side management.. e. In addition. Oil in the Middle East costs a few dollars per barrel. investments in that technology will bring down costs and may enable such a future. the cost to develop the new technologies can be expected to be lower than the cost to implement them at a large scale).The key issue illustrated by the biomass and nuclear examples is that energy policy is not only about selecting the lowest-cost option. Currently. etc. city planning and the provision of better-performing public transportation systems and railway networks. 2. but ought to be lower than the total cost of an ambitious abatement programme (i. more broad-ranging questions such as what we want society to look like in the future enter the picture. It may be interesting to dwell a little bit about the cost-relations between these three policies. One sometimes hear that the revenues from carbon taxes should be spent on energy R&D so as to increase the acceptability of carbon taxes. a solar future seems more difficult and expensive to acheive than a nuclear future. and removing them is clearly worthwhile.g. or one that relies on large establishments of plantations. ending subsidies to fossil fuels (see Kammen & Pacca. much more difficult to estimate since they both involve direct government expenditures and indirect costs following requirements for firms to produce certain amount of renewable electricity etc.g. on the other hand. Assume that a carbon tax of 15 USD/tC02 (which is the current EU ETS trading price) would be implemented world wide.e. but it should be kept in mind that R&D spending is rather minor to the potential revenues a carbon tax would generate. it should be kept in mind that their removal would not likely change the picture fundamentally. policies to initiate markets when there are chicken-and-egg-like problems (e.' or the simultaneous establishment of biomass plantations and biomass fired power plants). are higher than R&D expenditures. some 30 times global public energy R&D spending. although this policy has been criticized for being too costly. or for alternative fuels for transport. Rather. participants think of what a desired energy future sufficiently far away from the existing infrastructure could look like. If we currently opt for nuclear. we consider a solar-based energy future more attractive. and coal for electricity generation is still likely to be less costly than many renewables. and the establishment of standards for new fuels. The cost for market diffusion programmes. i.3 2 The Swedish government has required that all gas stations have ethanol fuel available. In such exercises. 13 . 3 Although these subsidies are large.g. technology policies. and cars that run on these fuels.5 Some additional reflections So far I have only discussed three key categories of policy instruments to reduce C02 emissions: price incentives. If.. with little investments in solar.. 2004) would be both environmentally friendly and economically beneficial. and efficiency standards.

we need to think more about how to manage special interests. I tried to outline climate policies in an ideal world where one does not have to confront special interests. Such policies could deliver deep reductions in C02 emissions. or resort to prohibition of certain technologies which we simply do not think fit into a "climate friendly society"? The question is important since one of the aims of the establishment of many policies today is to prepare for the long-term future. Climate policy proposals need to be addressed in this context. lack of sufficient support from the general electorate (although most voters.g. Will we then use price-based incentives.e. so that support for climate policies is created. but I won't address these here. For instance. resistance from industry groups and other special interests who think . Finally. and how to make the (currently) impossible politically feasible. NGOs and policy makers should spend as much time thinking about how to create public acceptance for climate policies. acute than what many argue now. I think these policy instruments are important. In section 2. Further. but they are likely to be more costly than policies based on price incentives. and markets do sometimes fail to exploit cost-efficient options for energy efficiency. even in the US. but the three categories of policies considered in this paper must form the backbone of any effective climate policy regime. one could think of alternative potentially successful policy strategies or frameworks. there are policies to reduce other pollutants or improve energy security. information campaigns and explicit discussions with the general public. researchers. perhaps most importantly. The key reasons why so little happens are a lack of political will in government circles.There may also be a need to carefully consider how "other" policies affect energy use.rightly or not . in Sweden. POLITICAL FEASIBILITY VERSUS COST -EFFICIENCY We have technologies to reduce emissions of CO2• We know which policy instruments to put in place to make sure this happens. It should also be noted that policies to enhance energy security might lead to higher CO2 emissions. if a country is rich in coal or heavy oils and decides to support these energy sources at the expense of conventional oil or natural gas. We should spend the same 14 . Further.. 3. the use oflegal instruments to rule out the use of new coal-fired power plants without carbon capture and storage.. One interesting question is how society will react if it turns out that the climate problem is as. This involves agenda-building.that they have much to lose from climate policies. tax law for company cars has led to Swedes driving heavier and 10% more gas-guzzling cars than our Nordic neighbors. e. and. i. Why? Because they directly address the key reasons why we "overemit" CO2: the emitter does not pay for the damage caused.g.. e. incentives to develop new technologies when the incumbents are expected to remain cheaper are weak. in these early stages when we are just starting to implement climate policies. in combination with strict efficiency standards for cars. However. think that climate change is a serious issue). or even more. the use of coal for the production of synthetic diesel. These could lead to some CO2 reductions at negative costs.

Below. science is no better at answering 15 . In many other EU countries. From a text-book economics perspective. prior to negotiations in Kyoto. 3.amount of time thinking about how to get a stringent cap in place as we do on how to design the cap-and-trade system in an effective way. The Swedish carbon tax for households and transportation is five to ten times the carbon permit price in the EU ETS (as oflate spring 2006). the real-world choice that developed countries faced in early 1997 was not one between cost-efficient reductions or an inefficient Kyoto Protocol. 3. But would including the transport sector be desirable? Clearly. The reason. Thus. This means there is one carbon price for those sectors. with 95 votes to 0. is that developing countries' participation would lower the cost to meet any given global reduction target. and we ended up with an even less cost-efficient solution to the problem. I offer a few such examples where I think that sacrificing cost-efficiency for the sake of political feasibility may be reasonable during an initial phase and a few examples where I instead think such sacrifices· are less attractive. steel mills. and others for other sectors. and power generation). Alleging that Kyoto is flawed because of its costinefficient features strengthened those who did not want to do anything about climate change. but given that the US emits roughly ten times more CO2 per capita than China and 20 times more than India. the EU ETS encompasses large point sources (primarily large industries. both before and after the Kyoto negotiations. and that these and other developing countries struggle to meet basic human needs. This implies that cost-efficiency cannot be the main criterion by which climate policies should be assessed. but one between a scenario with reductions in developed countries (Kyoto or something similar) and a scenario with no abatement at all. In 1997. plants. the prospects for getting China and India to accept reduction targets in Kyoto were rather bleak. will be cost-inefficient during an initial phase for the simple reason that the costefficient policies may be politically infeasible. it is not up to science to state definitively what is desirable or undesirable. The economists' objections were used to justify the positions of energyintensive companies and others who rejected Kyoto for less lofty reasons. the US Senate adopted a bill. refineries. The purpose of this section is to shed light on this challenge.which sectors to encompass Currently. this is economically inefficient. The reason for this is that many policies. there is no carbon price at all. to say the least. argued that developing countries should face targets. urging the US government not to accept any reduction targets unless key developing nations committed themselves to meaningful action and participation. and the like. Balancing costefficiency and political feasibility is a delicate challenge. by necessity. Many economists. of course.1 Global carbon abatement policies .2 The EU Emissions Trading Scheme . Calls for global cap-and-trade regimes are common. I have no doubt that this is true.

for instance. Putting all sectors under the same cap implies that all sectors will face the same stringent policy. carbon taxes approaching 100 USD/t CO2 would have a substantial impact on production costs for energy-intensive industries. whereas steel producers and many other energy-intensive industries do. and it became clear that the overall ambition of the EU governments was not as great as many had hoped for. Second. the impact on these industries is likely to be much greater. When the first cap was set. so as to maintain the incentive to reduce emissions). if I did not think it would affect the overall cap negatively. All sectors under the same cap may mean both that the overall cap is watered down and that we lose the possibility to introduce more stringent policies in the transport sector. 3. If climate taxes apply to households or manufacturing industries. through a structured analysis we may sort out the key arguments for and against inclusion.. Given that transport does not compete on international markets. border-adjusted taxes (e.these questions than are.g. by allocating permits freely to energy-intensive industries (as the EU has done). policy makers. or by subsidies to the negatively affected industries (e. The key argument against including all sectors is that the ability to pay for permits is much higher in the transport sector than in industry.g. but perhaps only after we have seen that the current scale works well. it is less easy to get away with this excuse. this is most often not the case. However. they would face the same marginal price on carbon. we have to consider what this means for the ultimate cap. If.04 USD/kWh (almost a doubling of the production cost) and a cost increase approaching 30 percent for steel producers (from iron ore). some kind of protection of energy-intensive industries may be needed. setting a cap on the total amount of allowances would explicitly demonstrate the real ambition of policy makers. on the other hand. However.g.. including as many sectors as possible increases cost-efficiency. all sectors are under the cap. Under such a scheme there is a risk that the energy-intensive industries will lobby even harder against strict policies than they did during the first round of the EU ETS. e.. or more specifically. First of all.3 Protection of energy-intensive industries Climate policies are frequently claimed to reduce international competitiveness for countries that introduce them unilaterally. A carbon price of say 50 USD/ton CO2 implies a cost of roughly 0. 16 . Let me begin with the positive arguments. based on production levels. by implementing an implicit carbon tax on imported energy-intensive goods). I would be for inclusion. In order for governments to set stringent targets. Should countries with unilateral climate policies try to protect their energy-intensive industries? This can be done through tax differentiation (Sweden.1 USD/liter (a 10 percent increase for European consumers) whereas it would imply a cost increase for electricity by roughly 0. policy makers could continue to argue that they would be able to meet their overall targets by referring to planned policies and measures in other sectors. has a much lower carbon tax for industry than for other sectors). When deciding whether to include transportation in the trading scheme.

and such groups can often block carbon policies). If the aim is to protect the capital owners. Clearly. • making climate policy politically feasible (large energy-intensive industries often make a very powerful lobby. But it is clear that cost-efficiency alone cannot be the sole criterion. too: • reducing carbon leakage (without protection. and for reasons related to simplicity and equal treatment (currently. steel manufacturing. In addition. It is difficult to weigh these arguments. we should ensure that they become neither selfperpetuating nor excessive. • avoiding loss of capital. possibly. but then production may move out. cotton subsidies in the US.the so-called "Double Dividend"). if protective policies are introduced. it became possible to soften objections to the EU trading scheme. actually ended up allocating more permits to the energy-intensive industries than was needed. This would happen if it were more profitable to close down the plant and sell the permits than to continue operations. freely). The main foes were 17 .4 Emissions trading . through border-adjusted taxes.grandfathering versus auctioning The key issue in allocation of permits in the EU ETS is free allocation (currently carried out in the EU system through grandfathering) versus auctioning. But in that case. However. to take a concrete example. An additional drawback with protecting industry is that once protection is put in place. higher CO2 emissions per ton of steel). However there are reasons for protection. we need to explicitly consider whether we aim to protect production (employment) or capital.From an economic perspective. whereas Sweden gives 60%. it might grow entrenched so that it is difficult to get rid of. may move out ofthe country into a region with. such protection is not warranted if the aim for the country is to reduce its emissions at the lowest possible cost. • avoiding job losses in sensitive regions or sectors (although. we would still suffer from the carbon leakage problem. We need only think of agricultural policies in Europe. that the country is serious about its climate ambitions. • sending signals to the rest of the world. 3. even when the policy is no longer justified. by choosing to allocate allowances for free based on emissions history. this can be done by allocating permits freely for the relevant plant for a long period of time. although a strong advocate for stringent climate policies and auctioning of permits in the EU trading scheme. Auctioning is preferable from an economic perspective (the revenues from auctioning can be used to reduce other taxes and therefore reduce the social cost associated with the climate policy . the allocation principles are different in different countries in the EU. Sweden. in the long run trade policies will not determine the number of jobs or the level of unemployment in the country as a whole). even for similar installations: Germany gives 100% or so to their power plants. and I lean toward the following conclusion: carbon leakage and political feasibility arguments speak strongly in favor of some protection. anyway. and subsidies to coal in Germany.

a 1000 MW power plant nets an extra 60 million Euros per year. We presented our report in mid-June. for instance. and support for developing advanced energy technologies (in particular gasification of biomass for the production of advanced biofuels). The ED should seriously consider limiting the import of electricity (e. If so.g.. Another absurd aspect of the ED trading scheme is how it deals with new versus old plants. less costly.thereby bought over to the "proper" side. England. 2006. for the next 14 years! This has resulted in several companies currently investing in coal-fired power plants in Germany. After many rounds of compromises.5 The Swedish oil commission and free trade inbiofuels In December 2005. 18 . In order that incumbents not be unduly favored. Perhaps this kind of thing is required in order to make certain difficult changes acceptable. the Swedish Prime Minister initiated work on how to break oil dependency in Sweden by the year 2020. but the price of electricity goes up (even free permits have opportunity costs). but the idea is not new. I wrote an addendum to the report stating that the Commission should have supported free trade in tropical sugar-cane based ethanol. There are four reasons: it would be hypocritical not to do this given that Sweden is and has been a strong supporter of free trade and tropical sugar cane ethanol is more area-efficient. Allowances for free means that coal-fired power plants. new plants also receive free allowances. Grandfathering permits to coal-fired power plants is to my mind simply too much. which produce and emit the most carbon dioxide per kWh power generated. higher energy efficiency in households and for cars (Swedish cars are 20% more energy-consuming than the ED average). A commission was set up to analyze possible options on how to achieve this aim. enhance international competitiveness for Swedish firms working with energy and environment related products. by not expanding grid capacity) if companies should start importing power from regions without carbon policies. improving energy security. But there has to be a limit. when slavery was outlawed in the colonies. there would be no risk that the closing of coalfired power plants leads to increased emissions elsewhere. If the price of electricity goes up by 1 eurocent/k Wh due to the trading scheme. reap huge rewards. and preparing Swedish households and industries for a future with higher energy prices. we were able to reach agreement in all areas except when it came to biofuels based on traditional Swedish agricultural crops and import duties for ethanol. 3. and more energy efficient than domestic ED ethanol. Grandfathering permits to coal-fired power plants turns the polluter pays principle (PPP) on its head: higher profits for the polluters (HPP). Germany has gone so far as to promise that new plants will receive as many allowances as they need. in particular in light of the fact that there is very little cross-ED border trade with electricity. They obtain virtually all the allowances they need. and in doing so. compensated Caribbean slave owners (!) in 1833. Key proposals included a tightening of the ED Emissions Trading Scheme. Paying those who do wrong in order that they do the right thing is objectionable. Key reasons for doing this were reducing Sweden's climate impact.

that these policies to a large extent rely on support from the farm lobby and the notion that such subsidies will create income and jobs in rural areas. but they may . that this market is driven by tax subsidies. since industries face international competition. they are not easy to do.still be reasonable given the difficulties associated with introducing sufficiently high carbon taxes. but it is important to be explicit about them. This 3+1 approach must form the backbone of any successful climate policy.g. in the real world the implementation of such a policy faces severe obstacles. and households do not. assessments of policies to reduce CO2 emissions must consider both costefficiency and political feasibility. This argument points out that Sweden has a rapidly growing market for ethanol (both as a blend and as a fuel for ethanol (E8S) cars). When assessing approaches that are not fully cost-efficient. the following key criterion better be fulfilled: the approach should pave the way for either more advanced technologies or more cost-effective policies. For these reasons. cost-efficiency objectives and concerns about the long-term political consequences of building up a lobby that can be expected to lobby for its own interests (and not for CO2 policies in general) convinces me that the free trade approach is preferable (even if there is a risk that it would lower the political interest and support in biofuels in the near term). there would be less support for the ethanol policy. It is not easy to tax CO2 emissions from households and industry at the same rate.. most importantly. In addition. 4. and. I have also emphasized the need for countervailing policies to make sure that new problems are not created in the process of avoiding old ones. e. For instance. However. in this case. and we might end up in a situation without ethanol at all.but the target would be lower! These are the trade-offs that have to be made. lower VAT for public transportation. 2) technology-specific policies to drive technological development so that future targets can be met. Conclusion In this paper. and 3) performance standards for energy efficiency.In this paper. I have stressed the need for 1) economy-wide price instruments for achieving low-cost emissions reductions. etc. These kinds of arguments are worth considering. If all of our ethanol were based on tropical sugar cane. I'll focus on one. Thus. subsidies for introducing pellets burners. but I think that the risk of getting locked into a situation where more ethanol factories lead to even stronger lobbying against free trade and in favor of continued subsidies is real and undesirable. we may accept 19 .in some cases . Lowering the tax on households would increase costefficiency in meeting a domestic CO2 target . There is a similar debate in the US where import duties are in place and where the farm lobby has been instrumental in bringing the com ethanol system in place. I will not mention all possible counter arguments. this will likely remain a problem. There is also a tendency to find other policy instruments (instead of economy-wide price incentives). tax exemptions for biofuels. Such approaches have their problems and may be cost-inefficient. As long as the rates with which countries implement climate policies are different.

by auctioning a minor share of the permits right from the start).g. subsidies to individual technologies may be acceptable as a way of bringing down costs but only if these technologies are expected to one day be able to compete without the subsidies. Thomas Sterner. Further. The challenge is to select policies that form a platform upon which it becomes easier to implement new more stringent and more cost-efficient policies. 20 . but it is important that policy makers then clearly state that the long-term objective is auctioning (e. and Per Kageson for discussions about climate policy over the years that form the basis for this paper.. Thanks to Formas and Swedish Energy Agency for financial support. Acknowledgements I would like to express my gratitude to Paulina Essunger and Kerstin Astrand for detailed comments on the manuscript and to Bjorn Sanden. Tomas Kaberger.grandfathering of permits in the interest of getting a policy in place.

" Haroon S. USA. MA 02139.11<. Wigley16 effort..z. New York.lack the.. Montreal. L.)1I coral fed' bleach- thermohaline circulation shutdown.\-i~Yn. CA 92697. University of California.nth (1) In the 20th ccruurv.J? John S. S\1('h 88 55() ppm. USA.dI6-1~dd (::).' below j ~){)Olevels ::()(lg to. . USA. DC 20375.11. ihtl.' il cannot be CO. could . McGill University. schleslnger.3 Gregory Benford. Livermore.. at 550 ppm is (1 major [lower consurnprion is ~ J 2 TW. Houston. 14Plasma Physics Division. CA 94550.ust' the Ie"l:]>.. withdrew 1(\]' jill.bined. NY 10003..S.''j(J ppm byWigley emission .14 John C.T. too weak because much grc:Jkr emission rcductipn:.'. Cambridge. emission reductions hv lla!iOl1~ llmt at. IL 61801. Lackner.: l'y.iJr!lpled and primary power consumption in. . NJ 08801.}ol.111<1 as \'1'(: better understand the links between fossil fuel burning. AI 85721.'1.]lly comparable in magnitude but opposile in sign it) tile otthe 1«51 lee (4 I- al. (3. E- tion on Clin1at\. 15NASA Headquarters. Quebec H3A 2T7.yoto Proi(lcol C<lII~ Ior iJ.1* Ken Caldeira.1r. "Centre for Climate and Global Change Research.12 H. New York University. Quebec H3A 2K6. Washington.'vc'lltll. Washington.\nHn.' Change aims to stabilize gas conccntrotions at levels that ulltlm)tlt)!J. could Herculean SCIENCE VOL 298 'itF1clincd broad rangt~ (>j' 'Iabiliz~lti. energy supply Oil fossil fuels I Han~("11 t'I M more than a century ago. 9ExxonMobil Research and Engineering Company. (i 1'::.REVIEW: ENGINEERING Advanced Technology Paths to Global Climate Stability: Energy for a Greenhouse Planet Martin I. 'Institute of Space Systems Operations.. University of illinois at Urbana-Champaign. jaln. Canada. initial .tater] reason). 'iSO. Douglas Lightfoot. viroumental impacts C'Jr_ Atmospheric CO.)I<.nmospherc euough to warm [.i in Cll\'rgy (energy usc [KT gmss domestic Much cuts than those called Cor in ih..: :. USA. . Irvine..h·· Tile United Nations Framework Couvert1Department of Physics. Kheshgl. if 1. Even challenge. our www. are as an economic burden by SOllW (the United St<lk.H 450 ppm not energy technology (()1"<'S\." Klaus S. from fossil fuel huming e'Jnki raise 11K'infra I'ed opacity of the . 10 4'\(l ppm. . on i'otnl elnis. USA 16Na_ tional Center for Atmospheric Research. "Lunar and Planetary Laboratory.cduciiollS arc d(.6 Howard Herzog. policy C8 un stabilization be needed . University of Arizona..> Pnli(lC(. USA 3Lawrence Livermore National Laboratory. Boulder.Paradox"is roo weak and tun strong: Too strong kC. 7MIT Laboratory for Energy and the Environment. (itf Tile 1'0%:1 fuel theory has become credible as observations accumulate .m (ppm). Y(?. CO 80307.. USA.sciencemag. storage. hv H. USA.C'llic interference with \l·"c' climate svsiem m of energy production. the human population qu. New York. John Perkins. and _.:limai\. TX 77204. Criswell.]y.l \'111'11 a decline uf j ". has increased !)'()J1) ~275 to ~J70 p.hoftertsenyu.:enlhHbe g:l:.Iahiiizaliun could preI't'Ht developing nations from bHsillg their 3W.we! of the West Wigley and co]" emission scenarios I() stahilm. "Department of Applied Physics and Applied Mathematics. NY 10027. The l. 450 ppm or below over rhe next lilt) 'This statement emission-free docs power 0'1'. it will pass 550 ppm this century Cl imatc models and palco. Arrhenius put forth the i(kl rhat CO. ofwhich .450.a~('.1U(. Hoffert.'11 the IPCC'~ repori!.. Unchecked.15 Michael E.13 Wallace Manheimer.) means U. will be (lnd \Vr". h..1 are necdedluier. 'Department of Biology.5% is fossil-fueled.. Till')' minimized early ernis!'lS<' frolU . . year.4 David R." Atul K.~t:. climate change. alld ( should be addressed. University of Houston. Srabilizalion at 550. 10Department of Earth and Environmental Engineering. it'." 5°'.l business-ns-usuai scenario HWl combines econOli'lL' of 2 to 3'. call for 1'('dllC1wJ1S ill methane and black s. Naval Research Laboratory.11 L. Mankins.)n levels. Urbana. disu+bution. 650... Annandale. 6Department of Economics. McGill University.en level tal'gds (0 as 10\'\ .. fud greenhouse warming .~inbk bot de nut address fossil.s Christopher Green. Tucson.lrts per nulli.'\'(. Columbia University. Mauel.t((' tee Sheet (6). Canada. Dna conversion The need to research on such tcchnolonow j. USA. b:. which also cause warming Sudl l. USA. "Department of Atmospheric Sciences. Lewis.jon controls initially following .2 Tom indicate 111m ')50 ppm." Michael E." Tyler Volk. or 750 ppm ( 1 NOVEMBER 2002 . Montreal. at which Sl~ibiIiZt'.\. USA.Inri certainlV~'\(1 pPlll. 'Department of Physics and Astronomy. DC 20546. USA ·To Whom correspondence mail: rnarty. ~ilm')~jJlwtic al 350.

Opportunities still exist to improve efficiency in power generation and end-use sectors: transportation. Fuel cells may replace heat engines but will likely run on hydrogen. (A) Fossil fuel electricity from steam turbine cycles. about 30 to 40%. 19).gnsdo not exist operation" ally or as pilot plants.6 1. theoretical peak independent of cell or ecosystem is =8%). could foster emission-free electricity and hydrogen production. manufacturing.2 1. motor). Most efficient are large electric generators (98 to 99% efficient) and motors (90 to 97%)." =59%). as well as well-to-wheels efficiency [(torque X angular velocity at wheels)/(fossil fuel power in)]. 80 to 85%.~.1. and aerodynamic drag.9% present. How much can energy efficiency improve? In a given technology class.SCIENCE'S COMPASS sources Jhatqan<produce·l 00 to )0. including transmission losses) yield the nominal thermal-to-electric power conversion: 3 kW (thermal) "'" 1 kWe (electrical). The efficiencies of mature technologies are well characterized (14. Steam-cycle efficiencies (39 to 50%. and heat flows). theoretical "Betz limit. These are followed by rotating heat engines that are limited by the second law of thennodynamics: gas and steam turbines (35 to 50%) and diesel (30 to 35%) and internal combustion (15 to 25%) engines. 80 to 85%. The development of fusion could be similar: The best experiments are close to balancing power to ignite the plasma. IA and 4A). fluorescent (10 to 12%).6 1. charge-discharge cycles.0 5-6 7-8 9 -10 Carbon sequestration rates to produce 10 TW CO2-emission-free from fossil fuels nesium carbonate bricks J Fig. Power trains are typically 18 to 23% efficient for internal combustion (IC). But consumer demand for sport utility vehicles (SUVs) has driven the fuel economy o~ the U.S. car and light truck fleet to a 21-yearl low of 8. motor). and 30 to 37% for fuel cell-electric (75 to 80%. fission fuels. Lifestyles also affect emissions. A seamless transition would use H2 extracted A steam water condensate or sequestered C02 out + B Fossil fuel cnergy content Carbon content [GTC] [TW-yr] (cruel/C) [TW-yr/GtC] (E/C) [TW-yr/GtC] Sequestration rate [GtC/yr] Gas Oil Coal 1200 1200 4800 570 750 3690 2. Electrolyte and electrode materials and catalysts limit electrochemical fuel cells (50 to 55% now. . including combined cycles and cogeneration) and overall primary energy-to-electricity efficiency (30 to 36%. power is carried off by fusiongenerated neutrons. reformer. and incandescent (2 to 5%) illumination generate more heat than light. fuel cell. limited by chlorophyll absorption bands (most productive ecosystems are about I to 2% efficient.3 1. 70% eventually).1 1.9-1. driving patterns. More efficient automotive power conversion is possible (18. higher for multiband cells. Improving Efficiency Efficiency is the ratio of usable energy output to energy input. 30 to 35% for IC-electric hybrid (higher efficiency from electric power recovery of otherwise lost mechanical energy). low emissivity windows. =24%. grows for decades to centuries. wind. and/or solid carbonate sequestration.4 . Environmental Protection Agency highway driving cycle (EPA hwy)]. High-pressure sodium vapor (15 to 20%). The earliest gas turbines could barely turn their compressors. Photosynthesis has a very low sunlight-to-chemical energy efficiency. "Renewables" are primary energy in natural fluxes (solar photons. 982 1 NOVEMBER 2002 VOL 298 from gasoline or methanol in reformers (75 to 80%). (8) Collecting CO2 from central plants and air capture. 15). 21 to 27% for battery-electric (35 to 40%. losses that in many cases engineers have already expended considerable ef- fort to reduce. and (indoor) climate conditioning (I3). and levels off at some fraction of its theoretical peak (J 6). Impressive reductions in waste heat have been accomplished with compact fiuorescents. and fusion fuels.grl?ig()~Gl'(. followed by subterranean. Ultra fuel-efficient cars are available today that can travel up to 29 km liter-1 [68 miles per gallon (mpg) U.sciencemag. Emissions depend on vehicle mass. theoretical peak for single bandgap crystalline cells. Energy conversion always involves dissipative losses. efficiency normally starts low. It took 300 years to develop fuel cells from 1%-efficient steam engines. 50 to 55%.2 -1.S. but no net power output has occurred yet. Renewable energy converters include photovoltaic (PV) cells (commercial arrays.5 km liter-1 (20 mpg EPA !nVY)j SCIENCE www. and cogeneration (17). central power plant. Primary energy in metastable chemical and nuclear bonds includes fossil fuels. Can we produce enough emission-free power in time? Here we assess the potential of a broad range of technologies aimed at meeting this goal. Fossil and nuclear fuels are much closer to their limits (Figs. but huge processing and sequestration rates are needed (5 to 10 GtC year " to produce 10 TW emission-free assuming energy penalties of 10 to 25%). lower for more cost-effective amorphous thin films) and wind turbines (commercial units. about 15 to 20%.?t1sutnption •• ithout w greenhouse eillissi. electricity. 80 to 85%. water. ocean.

SCIENCE'S (19). the effects of such efficiency could be overwhelmed if China and India follow the U. Thus. soils. path from bicycles and mass transit to cars. trees. episodic. (The resistivity of HTS wires vanishes below the 77 K boiling point of nitrogen available from air. Unfortunately. ocean injections can sub- Renewables Renewable energy technologies include biomass. Energy can be transferred to H2 with an efficiency of about 72% from gas. Buckminster Fuller with modern computerized load management and high-temperature superconducting (HTS) cables could transmit electricity from day to night locations and foster low-loss distribution from remote. Also being explored is longer term CO2 sequestration in the deep sea (31). Decarbonization and Sequestration Reducing the amount of carbon emitted per unit of primary energy is called decarbonization. Thus. has also been proposed (30). Decarbonization is thus intimately linked to sequestration (25). However. The long-term trend has been from coal to oil to gas. and solid mineral carbonates (Fig. Substantial research investments are needed now to make this technology available in time. the decarbonization of fuels alone will not mitigate global warming. This reaction (calcination) is a key step in making cement from limestone. A liquid nitrogen chiller B equllateral-trlanqle- Icosahedron: An faced solid that reduces map projection errors -Buckminster Fuller's Global Electrical Grid Fig. COMPASS stantially decrease peak atmospheric CO2 levels. Sequestration reservoirs include oceans. IB). more CO2 is produced by making H2 from fossil fuel than by burning the fossil fuel directly. and 55 to. Turnover of iron fertilization-enhanced plankton uptake (27) can be similarly fast if organic detritus oxidizes near the surface. doubling (or more) efficiency is quite feasible. The main advantage of sequestration is its compatibility with existing fossil fuel infrastructures. Continuation of the trend would lead to use ofH2. but H2 does not exist in geological reservoirs. but the capacity to absorb CO2 is limited. Recovery of fossil fuel CO2 emitted from decentralized sources (like cars) may be needed. 34). The simplest air capture is forestation. carbon-neutral fuels or CO2 "air capture" may be the best alternatives to develop. After heat extraction. Emission-free H2 manufactured by water electrolysis that is powered by renewable or nuclear sources is not yet cost effective.) SCIENCE VOL 298 1 NOVEMBER 2002 www. the CO2 is sequestered and the H2 used for transportation or electricity generation (24). Processes requiring energy are needed for its synthesis. carbon sequestration could be a valuable bridge to renewable and/or nuclear energy. The underlying problem is providing 10 to 30 TW emission-free in 50 years. 76% from oil. wind. For a given emission scenario. or dangerous power sources. Biological sequestration approaches to longer term storage include sealing undecayed trees underground (28) and sinking agricultural residues to the deep ocean (29).60% from coal (21). The energy can come from fossil fuel feedstocks. 23). when the net photosynthesis-respiration reaction is to the right: hv + CO2 + H20 ~ CH20 + 02' Historical CO2 data and models imply a temperate forest carbon sink today of I to 3 billion tons of carbon (GtC) year-I (3). The exchange time of CO2 with trees is ~7 years. One vision of "clean" coal incorporates CO2 capture and sequestration: Coal and/or biomass and waste materials are gasified in an oxygen-blown gasifier.sciencemag. Per unit of heat generated. Backdiffusion and pH impacts of ocean CO? disposal could be diminished by accelerating carbonate mineral weathering that would otherwise slowly neutralize the oceanic acidity produced by fossil fuel CO2 (33. deep saline aquifers. H2 is produced today by steamreforming natural gas (2H20 + CH4 ~ 4H2 + CO2). followed by oil and gas (22. then enormous sequestration rates could be needed to stabilize atmospheric CO2 (Fig. which remove atmospheric CO2 over geologic time scales (35). with CO2 recovery by heating CaC03 in a retort to produce CaO and CO2. Air capture by aqueous calcium hydroxide [Ca(OH)2] in shallow pools. Uptake occurs during growth of or- ganic matter (CH20). with each fuel emitting progressively less CO2 per joule of heat (20). but some models show forests reversing from sinks to sources later this century as global warming increases soil respiration (26). including CO2 injections for enhanced recovery from existing oil and gas fields and capture of CO2 from power plant flue gases.S. although all cases eventually diffuse some CO2 back to the atmosphere (32). and the product is cleaned of sulfur and reacted with steam to form H2 and CO2. 2. (8) The global grid proposed by R. Continuing the trend to lower carbon fuels requires disposing of excess carbon because the trend opposes the relative abundance of fossil resourceshigh-carbon coal being most abundant. lB). coal seams. a carbon-neutral . Even with SUVs. solar thermal and photovoltaic.) As a result. (A) Mass-produced Widely distributed PV arrays and wind turbines making electrolytic H2 or electricity may eventually generate 10 to 30 TW emission-free. depleted natural gas and oil fields. (Asia already accounts for >80% of petroleum consumption growth. Tree and soil sequestration does not require combustion product separation or more fuel. if other emission-free primary power sources of 10 to 30 TW are unavailable by mid-century. but breaking the Ca-C02 bond requires substantial energy. A far-reaching removal scheme is reacting CO2 with the mineral serpentine to sequester carbon as a solid in magnesium carbonate "bricks" by vastly accelerating silicate rock weathering reactions.

3C)] (50). More cost-effective PV panels and wind turbines are expected as mass production drives economies of scale. Even if PV and wind turbine manufacturing rates increased as required. comparable to all of human agriculture. 49).S. (8) An SPS in a low Earth orbit can be smaller and cheaper than one in geostationary orbit because it does not spread its beam as much. storage. shallow gravitational potential well of the Moon). Fuller envisioned electricity wheeled between day and night hemispheres and pole-to-pole (43). A massive (but not insurmountable) scale-up is required to get 10 to 30 TWequivalent. Fig. 3. All renewables suffer from low areal power densities. . and the Earth-Sun L2 Lagrange exterior point [one of five libration points corotating with the Earth-Sun system (Fig.j> 5 X 10-3 kg Pt m-2 (41) (a 10-TW hydrogen flow rate could require 30 times as much as today's annual world platinum production). -20 kWe m-2 for proton exchange membrane (PEM) cells (21). ones that are close to users. Biomass plantations can produce carbon-neutral fuels for power plants or transportation. The Lagrange interior point L.6 W m-2) for biofuels to contribute significantly to climate stabilization (14. Advanced electrical grids would also foster renewables. For solar energy. 3. Spanning the world electrically evokes Buckminster Fuller's global grid (Fig.S. but other materials can be limiting. smaller satellites. "Balance-of-system" infrastructures could evolve from natural gas fuel cells if'reformer H2 is replaced by H2 from PV or wind electrolysis (Fig. Space solar power (SSP) ( . and tidal (36). Meeting local demand with PV arrays today requires pumped-storage or battery-electric backup systems of comparable or greater capacity (40). Indonesia. [10 TW equivalent (3. If theoretical microwave transmission efficiencies (50 to 60%) can be realized. With the exception of firewood and hydroelectricity (close to saturation). Even before the discovery of hightemperature superconductivity (42).. Worldwide deregn1ation and the free trade of electricity could have buyers and sellers establishing a supply-demand equilibrium to yield a worldwide market price for grid-provided electricity.000 km") (38). lines of sight. Such networks need to be reengineered. all the PV cells shipped from 1982 to 1998 would only cover -3 km2 (39).1/4 the area of surface PV arrays of comparable power. Department of Energy (DOE) studied an SSP design with a PV array the size of Manhattan in geostationary orbit [(GEO) 35. A major challenge is reducing or externalizing high launch 0 984 1 NOVEMBER 2002 VOL 298 SCIENCE www. Reversible e1ectrolyzer and fuel cells offer higher current (and power) per electrode area than batteries. energy consumption may require a PV array covering a square -160 km on each side (26. lunar materials. ocean thermal. 2B). Papua New Guinea. and power conditioning. U. the Moon (48. but photosynthesis has too Iowa power density (-0.800 km above the equator] that beamed power to a 10-km by 13-km surface rectenna with 5 GW output. these are collectively < 1% of global power. Alternative locations are 200. 37). cloudy Earth. solar power satellite (SPS). and the Maldives have agreed to participate in such experiments (52).3 TWo) requires a surface array -470 km on a side (220. could deflect 2% of incident sunlight. PEM cells need platinum catalysts. (A) The power relay 10.] Other architectures. and Colombia on the Pacific Rim. (10 TW from biomass requires >10% of Earth's land surface. Solar flux is -8 times higher in space than the long-term surface average on spinning. 2A). and lunar power system all exploit unique attributes of space (high solar flux. Wind power is often available only from remote or offshore locations. Tanzania. Capturing and controlling sun power in space.SCIENCE'S COMPASS hydropower. existing grids could not manage the loads.000-km altitude satellite constellations (47). (C) Space-based geoengineering.:. A 2000-km-diameter parasol near L.3 TWe) requires 660 SSP units. but it does not appear fixed in the sky and has a shorter duty cycle (the fraction of time power is received at a given surface site). 75 to 100 We could be available at Earth's surface per m2 of PV array in space. However. as could aerosols with engineered optical properties injected in the stratosphere. geothermal. In the 1970s. Potentially important for CO2 emission reduction is a demonstration proposed by Japan's Institute of Space and Aeronautical Science to beam solar energy to developing nations a few degrees from the equator from a satellite in low equatorial orbit (51). the National Aeronautics and Space Administration (NASA) and the U.sciencemag. and Malaysia. provides an opportunity for radiative forcing to oppose global warming. 45). Ecuador.) PV and wind energy (-15 We m-2) need less land. The electrical equivalent of 10 TW (3.000 krrr'). Present hub-and-spoke networks were designed for central power plants. A and B) exploits the unique attributes of space to power Earth (44. Brazil. But renewables are intermittent dispersed sources unsuited to baseload without transmission. and newer technologies were explored in the NASA "Fresh Look Study" (46).

sciencemag. water-cooled (RBMK) reactors.eV) (55). (8) The most successful path to fusion has been confining a D-T plasma (in purple) with complex magnetic fields in a tokamak.T reaction. Breakeven requires that the "plasma triple product" satisfy the Lawson criteria: n X 'I' X kl'r= 1 X 1021 m-3 s keV for the D. How long before such reactors run out of fuel? Current estimates of U in proven reserves and (ultimately recoverable) resources are 3. 4B). Bombarding natural U with neutrons of a few eV splits the nucleus. lithium bred to T could generate 16. Runoff and outflow to the sea from all the world's rivers is 1. MUltiplying by the oceans' huge volume (1. Boltzmann's constant (66. Recoverable U may be underestimated. Even with 100% 235U extraction. Still. releasing a few hundred million eV 35U + n ~ fission products + 2.000 TW-year in 235U. this would only last 6 to 30 years-hardly a basis for energy policy.3year half-life) would also be bred in the lithium blanket (n + 6Lj ~ 4He + T + 4. '1'.3 MeV) is of interest because it yields charged particles directly convertible to electricity (70). graphite-moderated. Studies of D-3He and D-D burning in inertial confinement fusion targets suggest that central D.7 MeV). 4. Breakeven occurs when the plasma triple product (number density X confinement time X temperature) attains a critical value. Best results from Princeton (Tokamak Fusion Test Reactor) and Europe (Joint European Torus) are within a factor of two (68).000 TW-year (69). (A) The conventional LWR employs water as both coolant and working fluid (left).2 X 10-6 kg dissolved U m-3 e made toward "breakeven" with tokamak (a toroidal near-vacuum chamber) magnetic confinement [Q == (neutron or charged particle energy out)/(energy input to heat plasma) = 1] (Fig. This represents 60 to 300 TW-year of primary energy (60). 0. Experimental work on advanced fusion fuel cycles and simpler magnetic confinement schemes like the levitated dipole experiment (LDX) shown are recommended. SSP could perhaps be demonstrated in 15 to 20 years and deliver electricity to global markets by the latter half of the century (53. Available reactor technology can provide CO2 emission-free electric power. the flow rate needed to make reactor fuel at the 10 TW rate is five times as much as this outflow (64).org . with 30. Higher Qs are needed for power reactors: Neutrons penetrating the "first wall" would be absorbed by Conventional nuclear fission reactor molten lithium.72% of natural U. If D.4 billion metric tons U and 80. and gas-cooled graphite reactors. modular nuclear fission reactor is theoretically immune to loss-of-coolant meltdowns like TMI and Chernobyl (right). and excess heat would be transferred to turbogenerators.8 MeV). 67).T reaction or the harder-to-ignite D-D reactions (~ 3He + n + 3. heavy water (CANDU) reactor. The existing ~500 nuclear power plants are variants of 235U thermal reactors (56. The focus has been on the deuterium-tritium (D-T) reaction (~ 4He + n + 17. confinement time.T ignitors can spark these reactions.a 235U energy density of 1. twice the thermal energy in fossil fuels. Sailor et al. the most promising long-term nuclear power source is still fusion (65). Despite enormous hurdles.0 MeV). but it would be a big stretch. Tritium (12.37 X 1018 m") gives 4. temperature. graphite-moderated. Loss-of-coolant accidents [Three Mile Island (TMI) and Chernobyl] may be avoidable in the future with "passively safe" reactors (Fig. The 235U isotope. is often enriched to 2 to 3% to make reactor fuel 40-year reactor lifetimes. The main problem with fission for climate stabilization is fuel. it would be imprudent (at best) to initiate fission scaleup without knowing whether there is enough fuel. Getting 10 TW primary power from 235U in crustal ores or seawater extraction may not be impossible. With adequate research investments. The D-3He reaction (~4He + p + 18. 57): the light water reactor [(LWR) in both pressurized and boiling versions].2 X 106 m3 S-1 (63). pebble-bed. 54). SCIENCE VOL 298 1 NOVEMBER 2002 www. What about the seas'? Japanese researchers have harvested dissolved U with organic adsorbents from flowing seawater (61). Recent tokamak performance improvements were capped by near-breakeven [data sources in (68)]. 4A). D in the sea is virtually unlimited whether utilized in the D.8 MJ m-3. (58) propose a scenario with 235U reactors producing ~ I 0 TW by 2050. At 10 TW. where n is number density. Steady progress has been 1015 1975 1985 1995 Year achieved Fig.SCIENCE'S costs. COMPASS Fission and Fusion Nuclear electricity today is fueled by 235U. respectively (22) [Ores with 500 to 2000 parts per million by weight (ppmw) U are considered recoverable (59)]. Pebble bed modular nuclear fission reactor A B 10 [MW] [neutrons s-1] 1018 A (62). Oceans have 3. though it poses well-known problems of waste disposal and weapons proliferation. and k.43n + 202 M.T reactors were operational. T.4 and 17 million metric tons. LWRs (85% oftoday's reactors) are based largely on Hyman Rickover's water-cooled submarine reactor (58). The helium-cooled. like Chernobyl.2 MeV and ~ T + P + 4.



Ignition of D-T -fueled inertial targets and associated energy gains of Q 2: 10 may be realized in the National Ignition Facility within the next decade. Experiments are under way to test dipole confinement by a superconducting magnet levitated in a vacuum chamber (71), a possible D-3He reactor prototype. Rare on Earth, 3He may someday be cost-effective to mine from the Moon (72). It is even more abundant in gas-giant planetary atmospheres (73). Seawater D and outer planet 3He could power civilization longer than any source other than the Sun. How close, really, are we to using fusion? Devices with a larger size or a larger magnetic field strength are required for net power generation. Until recently, the fusion community was promoting the International Thermonuclear Experimental Reactor (ITER) to test engineering feasibility. Enthusiasm for ITER waned because of the uncertainty in raising the nearly $10 billion needed for construction. The U.S. halted ITER sponsorship in 1998, but there is renewed interest among U.S. fusion scientists to build a smaller-sized, higher-field, nonsuperconducting experiment or to rejoin participation in a half-sized, redesigned ITER physics experiment. A "burning plasma experiment" could produce net fusion power at an affordable scale and could allow detailed observation of confined pi asrna during self-heating by hot alpha particles. The Fusion Energy Sciences Act of 200 I calls on DOE to "develop a plan for United States construction of a magnetic fusion burning plasma experiment for the purpose of accelerating scientific understanding of fusion plasmas (74)." This experiment is a critical step to the realization of practical fusion energy. Demonstrating net electric power production from a self-sustaining fusion reactor would be a breakthrough of overW he 1m ing

The neutron would be used to breed 233U from Tb in the fusion blanket. Each fusion neutron would breed about one 233U and one T. Like 235U, 233U generates about 200 MeV when it fissions. Fission is energy rich and neutron poor, whereas fusion is energy poor and neutron rich. A single fusion breeder could support perhaps 10 satellite burners, whereas a fission breeder supports perhaps one. A related concept is the particle accelerator-fission hybrid breeder (56): Thirty 3-MeV neutrons result from each 1000-MeV proton accelerated into molten lead; upon injection to a sub critical reactor, these could increase reactivity enough to breed 233U from Th, provide electricity, and power the accelerator efficiently (-10% of the output). The radiotoxicity of hybrid breeder reactors over time is expected to be substantially below LWRs. These ideas appear important enough to pursue experimentally, but both fission and fusion are unlikely to play significant roles in climate stabilization without aggressive research and, in the case of fission, without the resolution of outstanding issues of high-level waste disposal and weapons proliferation. Geoengineering No discussion of global warming mitigation is complete without mentioning "geoengineering" (78, 79), also called climate engineering or planetary engineering on Earth and terrafonning on other planets (80). Geoengineering in the climate change context refers mainly to altering the planetary radiation balance to affect climate and uses technologies to compensate for the inadvertent global warming produced by fossil fuel CO2 and other greenhouse gases. An early idea was to put layers of reflective sulfate aerosol in the upper atmosphere to counteract greenhouse warming (81). Variations on the sunblocking theme include injecting sub-micrometer dust to the stratosphere in shells fired by naval guns, increasing cloud cover by seeding, and shadowing Earth by objects in space (82). Perhaps most ambitious is a proposed 2000km-diameter mirror of 10-f.Lm glass fabricated from lunar materials at the L) (83) Lagrange point of the Sun-Earth system (84) (Fig. 3C). The mirror's surface would look lik t t ld d fl t 201 1 e a permanen sunspo, wou e ec 10 of solar flux, and would roughly compensate for the radiative forcing of a CO2 doubling. Climate model runs indicate that the spatial pattern of climate would resemble that without fossil fuel CO2 (84). Engineering the optical properties of aerosols injected to the tr t h t d . f I' ti S a osp ere 0 pro uce a vanety 0 c una lC effects has also been proposed (85). Our assessment reveals major challenges to stabilizing the fossil fuel greenhouse with energy technology transformations. It is only prudent

surance policy should global warming impacts prove worse than anticipated and other measures fail or prove too costly. Of course, large-scale geophysical interventions are inherently risky and need to be approached with caution. Concluding Remarks Even as evidence for global warming accumulates, the dependence of civilization on the oxidation of coal, oil, and gas for energy makes an appropriate response difficult. The disparity between what is needed and what can be done without great compromise may become more acute as the global economy grows and as larger reductions in CO2-emitting energy relative to growing total energy demand are required. Energy is critical to global prosperity and equity. If Earth continues to warm, people may turn to advanced technologies for solutions. Combating global warming by radical restructuring of the global energy system could be the teclmology challenge of the century. We have identified a portfolio of promising technologies here-some radical departures from our present fossil fuel system. Many concepts will fail, and staying the course will require leadership. Stabilizing climate is not easy. At the very least, it requires political will, targeted research and development, and international cooperation. Most of all, it requires the recognition that, although regulation can playa role, the fossil fuel greenhouse effect is an energy problem that cannot be simply regulated away.

1. S. Arrhenius, Phila. Mag. 41, 237 (1896). 2. J. R. McNeill, Something New Under the Sun: An

References and Notes

Environmental History of the Twentieth Century (Norton, New York, 2000). 3. j. T. Houghton et aI., Eds., Climate Change 2001: ScientifiC Basis (Cambridge Univ. Press, New York, 2001). 4. M. I. Hoffert, C. Covey, Nature 360, 573 (1992). 5. ~~!~:e,N;:~~n~/:;:~;:::n;~~~e~!~~;9n2,c~~a~: janeiro, Brazil (United Nations Environment Program-World 6. r20~2~'Neill, Meteorological M. Oppenheimer, Organization, Climate

importance but cannot be relied on to aid CO2 stabilization by mid-century. The conclusion from our 235U fuel analy-

sis is that breeder reactors are needed for fission to significantly displace CO2 emissions by 2050. Innovative breeder technologies include fusion-fission and acceleratorfission hybrids. Fissionable 239pu and/or 233U can be made from 238U and 232Th (75). Commercial breeding is illegal today in the United States because of concerns over waste and proliferation (France, Germany, and Japan have also abandoned their breeding programs). Breeding could be more acceptable with safer fuel cycles and transmutation of high-level wastes to benign products (76). Th is the more desirable feedstock: It is three times more abundant than U and 233U is harder to separate and divert to weapons than plutonium. One idea to speed up breeding of 233U is to use tokamak-derived fusion-fission hybrids (68, 77). D-T fusion yields a 3.4-

Change Secretariat,

Geneva, 1992).

Science 296, 1971

7. T. M. L. Wigley, R. Rlchels.]. A. Edmonds. Nature 379, 240 (1996). J 8. M. I. Hoffert et al., Nature 395, 881 (1998). 9. 8. 8olin, H. S. Kheshgi, Proc. Natl. Acad. Sci. U.S.A. 98 4850 (2001). I 10. j. Hansen, M. Sato, R. Ruedy, A, Lacis, V. Oinas, proc).

Natl. Acad. Sci. U.S.A. 97,9875


11. R. B. Cheney et el., National Energy Policy: Report 0 I the National Energy Policy Development Group (The White House, Washington, DC, May 2001); availabl~ at WWw.whitehouse.gOV/energY/National-EnergyJI Policy.pdf. 12. B. Metz et al., Eds., Climate Change 2001: Mitigationl (Cambridge Univ. Press, New York, 2001), p. 8. 13. National Laboratory Directors, Technology opportu' l nities to Reduce U.S. Greenhouse Gas Emissions (DOE) Washington, DC, 1997); available climate_change/climate.htm. at

986 M,V alpha particle and a 14M,V n::: EM





,~,,~mg 'I '''P""~


SCIENCE'S Global Climate Change, R. G. Watts, Ed. (Lewis, Boca Raton, FL, 1997), pp. 205-259. 15. V. Smil, Energies (MIT Press, Cambridge, MA, 1999). 16. J. H. Ausubel, C. Marchetti, Daedalus 125, 139 (1996). 17. A. H. Rosenfield, T. M. Kaarsberg. J. Romm, Phys. Today, 53, 29 (November 2000). 18. M. A. Weiss, J. B. Heywood, E. M. Drake, A. Schafer, F. F. AuYeung. On the Road in 2020: A Life-Cye/e Analysis of New Automobile Technologies, EL00-003 (MIT Energy Laboratory, Cambridge, MA, 2000): available at 19. K. H. Hellman, R. M. Heavenrich, Light-Duty Automotive Technology and Fuel Economy Trends: 1975 Through 2001, EPA420-R-Ol-008 (U.S. Environmental Protection Agency, Washington, DC, 2001): available at rOl008.pdf. 20. C. Marchetti, Nue/. Sci. Eng. 90, 521 (1985). 21. C.-], Winter, J. Nitsch, Eds., Hydrogen as an EnergyCarrier (Springer-Verlag, New York, 1988). 22. N. Nakicenovk, A. Griibler, A. McDonald, Global Energy Perspectives (Cambridge Univ. Press, New York, 1998), p. 52. 23. Energy in recoverable fossil fuels ("reserves" plus "resources" from conventional and unconventional sources) Is -~4800 TW-year for coal and -1200 TW-year each for oil and gas [converted from values in (22): 1 billion metric ton oil equivalent = 44.8 EJ = 1.42 TW-year: natural gas Includes liquids, excludes methane hydrates]. Total energy in recoverable fossil fuels ranges from 6400 to 8000 TW-year, two-thirds of which Is coal. Coal resources could roughly double if energy in shales could be extracted cost-effectively. 24. J. P. Holdren et el., Federal Energy Research and Development for the Challenges of the Twenty-First Century (President's Committee of Advisors on Science and Technology [PCAST], Washington, DC, 1997): available at 25. D. Reichle et al., Carbon Sequestration Research and Development (DOE, Washington, DC, 1999); available at www.ornl.govlcarbon_sequestratlon/. 26. P. M. Cox, R. A. Betts, C. D. Jones, S. A. Spall, l. J. Totterdell, Nature 40B, 184 (2000). 27. J. E. Martin, S. E. Fitzwater, R. M. Gordon, Glob. Biogeochem. Cye/es 4, 5 (1990). 28. F. J. Dyson, G. Marland, Workshop on the Global Effects of Carbon Dioxide from Fossil Fuels, Report CONF-770385 (DOE, Washington, DC, 1977), pp. 111-118. 29. R. A. Metzger, G. Benford, Clim. Change 49, 11 (2001). 30. K. S. Lackner, P. Grimes, H.-J. Zlock, Carbon Dioxide Extraction from Air: Is It an Option? Technical Report LA-UR-99-583 (Los Alamos National Laboratory, Los Alamos, NM, 1999); available at energy/ecology /carbon/ docsl AIrExtraction.pdf. 31. H. Herzog, B. Eliasson, O. Kaarstad, Sci. Am. 2B2, 72 (2000). 32. H. S. Kheshgl, B. P. Flannery, M. I. Hoffert, A. G. Lapenis, Energy 19, 967 (1994). 33. K. Caldeira, G. H. Rau, Geophys. Res. Lett. 27, 225 (2000). 34. Carbon dioxide dissolves in water and slowly corrodes carbonate sediment, resulting in the storage of fossil fuel carbon primarily as bicarbonate ions dissolved in the ocean. Left to nature, the net reaction CO2 + H20 + CaC03 --> Ca2+ + HC03- would occur on the time scale of -6000 years. 35. K. S. Lackner, C. H. Wendt, D. P. Butt, E. L. Joyce jr., D. H. Sharp, Energy 20, 1153 (1995). 36. T. B. Johansson, H. Kelly, A. K. N. Reddy, R. H. Williams, Eds., Renewable Energy (Island Press, Washington, DC, 1993).

10 TW

37. H. S. Kheshgi, R. C. Prince, G. Marland, Annu. Rev. Energy Environ. 25, 199 (2000). 38. J. A. Turner, Science 285, 687 (1999). 39. H. C. Hayden, The Solar Fraud: Why Solar Energy Won't Run the World (Vales Lake Publishing. Pueblo West, CO, 2001), p. 161. 40. J. K. Strickland [r., Sol. Energy 56, 23 (1996). 41. B. C. H. Steele, A. Heinzel. Nature 414, 345 (2001). 42. J. G. Bednarz, K. A. Muller, Z. Phys. B 64, 189 (1986). 43. R. B. Fuller, Critical Path (SI. Martins, New York, 1981). 44. P. E. Glaser, Science 162, 857 (1968). 45. , F. P. Davidson, K. I. Csigl, Eds., Solar Power Satellites (Wiley-Praxis, New York, 1997). 46. J. Mankins, Aerosp. Am. 35, 30 (1997). 47. M. 1. Hoffert, S. D. Potter, in Solar Power Satellites, P. E. Glaser, F. P. Davidson, K. 1. Csigi, Eds. (WileyPraxis, New York, 1997), pp. 231-254. 48. D. R. Criswell, In Innovative Solutions to CO2 Stabilization, R. Watts, Ed. (Cambridge Unlv. Press, New York, 2002), pp. 345-410. 49. __ , Ind. Physicist B, 12 (April/May 2002). 50. G. A. Landis, in Proceedings of Solar Power Satellite '97 Conference, 24 to 2B August 1997, Montreal, Canada (Canadian Aeronautics and Space Institute, Ottawa, Canada, 1997), pp. 327-328. 51. M. Nagatomo, Sol. Energy 56, 111 (1996). 52. P. Collins, Y. Purwanto, X. C. Ji, Space Energy Transp. 4, 47 (1999). 53. Committee for the Assessment of NASA's Space Solar Power Investment Strategy, Aeronautics and Space Engineering Board, National Research Council, Laying the Foundation for Space Solar Power: An Assessment of NASA's Space Solar Power Investment Strategy (National Academy Press, Washington, DC, 2001). 54. S. J. Smith, N. Mahasenan, J. F. Clarke, J. Edmonds, Space Solar Electric Power: Future Market Share Under Competition (Battelle Memorial Institute, Washlngton, DC, 2001). 55. H. G. Graetzer, Am. J. Phys. 32, 10 (1964) [translation from O. Hahn, F. Strassmann, Naturwissenschaften 27, 11 (1939)]. 56. A. M. Weinberg, in Technologies for a GreenhouseConstrained Society, M. A. Kullasha, A. Zuker, K. J. Ballew, Eds. (Lewis, Boca Raton, FL, 1992), pp. 227237. 57. R. L. Garwin, G. Charpak, Megawatts and Megatons: A Turning Point in the Nuclear Age (Knopf, New York, 2001). 58. W. C. Sailor, D. Bodansky, C. Braun, S. Fetter, B. van der Zwaan, Science 288, 1177 (2000). 59. M. A. Adelman, J. C. Houghton, G. Kaufman, M. B. Zimmerman, Energy Resources in an Uncertain Future (Ballinger, Cambridge, MA, 1983), p. 334. 60. Fission energy per ton (J t-') 23SU is 235 arnu

1.37 X 10'8 m' 8 X 104 TW-year


1year _ 6 ,_, 3.15 X 107 s - 5.4 X 10 m s . 65. T. K. Fowler, The Fusion Quest (Johns Hopkins, Baltimore, MD, 1997). 66. J. D. Lawson, Proc. Phys. Soc. London B 70, 6 (1957). 67. G. H. Miley, Fusion Energy Conversion (American Nuclear Society, LaGrange Park, IL, 1976). 68. W. Manheimer, Fusion Tecbnol. 36, 1 (1999). 69. L. M. Lldsky, In Energy Technology Handbook, D. M. Considine, Ed. (McGraw-Hill, New York, 1977), pp. 5-134. 70. G. L. Kulcinski, J. F. Santarlus, Nature 396,724 (1998). 71. J. Kesner, L. Bromberg, M. Mauel, D. Garnier, J. M. Dawson, The Dipole Fusion Confinement Concept: A White Paper for the Fusion Community (MIT Plasma Science and Fusion Center, Cambridge, MA, 1998); available at!ldx/pubs/dipole_ wp.pdf. 72. L. J. Wittenberg, J. F. Santarius, G. L. Kulcinski, Fusion Technol. 10, 167 (1987). 73. J. s. Lewis, Space Power 10, 363 (1991). 74. Fusion Energy Sciences Act of 2001, H.R. 1781 (107th Congress). 75. 239pU and 233U don't exist in nature. They must be bred from their feedstocks, 238U (-99% of U) and 232Th (-100% of thorium), with the use of suitable neutron sources (238U + n -> 239pU;232Th + n --> 23'U). Fission energy per unit mass Is comparable to 23SU (239pU + n --> fission products + 2.9n + 210 MeV: 233U + n -> fission products + 2.5n + 198 MeV). We estimate breeding could increase fission energy resources by factors of 60 and 180 for 239PU and 233U, respectively. Breeding rates depend on feedstock mining rates and availability of neutrons. 76. Committee on Separations Technology and Transrnutation Systems, Board on Radioactive Waste Management, Commission on Geosciences, Environment, and Resources, National Research Council, Nuclear Wastes: Technologies for Separations and Transmutations (National Academy Press, Washington, DC, 1996). 77. H. Bethe, Phys. Today 32, 45 (May 1979). 78. S. H. Schneider, Nature 409, 417 (2001). 79. D. W. Keith, Nature 409, 420 (2001). 80. M. J. Fogg, Terraforming: Engineering Planetary Environments (Society of Automotive Engineers International. Warrendale, PA, 1995). 81. M. I. Budyko, Climate Changes, R. Zelina, Transl., L. Levin, Ed. (American Geophysical Union, Washington DC, 1977). 82. Panel on Policy Implications of Greenhouse Warming, Committee on Science, Engineering, and Public Policy, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, Policy Implications of Greenhouse Warming: Mitigation, Adaptation and the Science Base (National Academy Press, Washington, DC, 1992), pp. 433-464. 83. J. T. Early, j. Br. Interplanet. Soc. 42, 567 (1989). 84. B. GOVindasamy, K. Caldeira, Geophys. Res. Lett. 27, 2141 (2000). 85. E. Teller, L. Wood, R. Hyde, Global Warming and Ice Ages: Prospects for Physics-Based Modulation of Global Change, Report UCRL-231636/UCRL JC 128715 (Lawrence Livermore National Laboratory, Livermore, CA, 1997). 86. We thank D. Harvey, J. Katzenberger, M. MacCracken, G. Marland, J. Palfreman, V. Smil, and P. Werbos for valuable discussions. K.C, H.H., and A.K.J. thank the DOE-Office of Biological and Environmental Research for support in this work.

202MeV 1.661 X 10-27 kg/amu

X ts

1.602 X 10- J 10' kg 1 MeV X 1 ton

= 8.3 X 10

J t"'U

61. 62. 63. 64.

The 23SU energy per ton natural U is therefore 0.0072 X B.3 X 10'6 J t-' = 6 X 10'4 J t-'. and the energy in proven U reserves and ultimately recoverable resources is -·61 and 300 TW-year, respectively (1 TW-year = 31.5 EJ). S. Nakamura, K. Oguma, Y.lto,J. Chem. Eng. Jpn. 2B, 660 (1995). P. Henderson, Inorganic Geochemistry (Pergamon, New York, 1982), p. 282. J. P. Peixoto, A. H. Oort, Physics of Climate (American Institute of Physics, New York, 1992), p. 271. The flow rate through a hypothetical seawater extraction plant to yield 23'U fuel at 10 TW is


VOL 298


edu (S.5% annual carbon emissions growth of the past 30 ..importaut not. 500. We. and industrial know-how to solve the carbon and climate problem for the next half-century. industrial-scale mitigation options also exist for subordinate gases. we divide the stabilization triangle into seven equal "wedges. IA).sciencemag.llS.inn}i.developtherevolutionary mitiglltiqn .. the quantitative details of the stabilization target. Pacala'* and R. USA. for the next 50 years. The BAU and stabilization emissions in Fig. What Do We Mean by "Solving the Carbon and Climate Problem for the Next Half-Century"? Proposals to limit atmospheric CO2 to a concentration that would prevent most damaging climate change have focused on a goal of 500 ± 50 parts per million (ppm). but readers will disagree about which to reject on such grounds. scaling \iP What we already know how to do. The current concentration is ~375 ppm. the Intergovernmental Panel on Climate Change (IPCC) has claimed that "technologies that exist in operation or pilot stage today" are sufficient to follow a less-thandoubling trajectory "over the next hundred years or more" [(1). Every element in this portfolio has passed be:ton~!h~. and revised estimates of the size of the stabilization triangle. Thus. It is important to understand that each of the seven wedges represents an effort beyond what would occur under BAU. We focus exclusively on CO2." located between the flat trajectory and BAU. But not filling the stabilization triangle will put 500-ppm stabilization out of reach.beyond. p. NJ 08544. The CO2 emissions reductions necessary to achieve ally such target depend on the emissions judged likely to occur in the absence of a focus on carbon [called a business-as-usual . and a very slow decline thereafter that matches the declining ocean sink. the oceans and terrestrial biosphere).org ." A wedge represents an activity that reduces emissions to the atmosphere that starts at zero today and increases linearly until it accounts for I GtC/year of reduced carbon emissions in 50 years. Stabilization is represented by a "flat" trajectory of fossil fuel emissions at 7 GtC/year.Rili~ . then ~ 10 wedges would be needed instead of 7. Although no erement Is a credible candidate for doing the entire Job (or even half the job) by itself. *To whom correspondence should be addressed. To develop the revolutionary technologies required for such large emissions reductions in the second half of the century. even though they are currently on course to more than double (Fig. Princeton. basedsolarelectricity. in one simple model (9) that begins with the stabilization triangle but looks beyond 2054. Very roughly.tiglltjol)technplq~y.thatfundamental research is vitaLto. and an interval for which the technology is close enough to envision. IA are explained in Section I of the supporting online material (SOM) text. and climate problem over the next half-century" means to deploy the technologies and/or lifestyle changes necessary to fill all seven wedges of the stabilization triangle. stabilization at 500 ppm requires that emissions be held near the present level of 7 billion tons of carbon per year (GtC/year). to "solve the carbon VOL 305 SCIENCE 968 13 AUGUST 2004 www.). a continuation of the historical rate of decarbonization of the fuel mix prevents the need for three additional wedges.P. the lifetime of a power plant. Policies designed to stabilize at 500 ppm would inevitably be renegotiated periodically to take into account the results of research and development. If carbon emissions were to grow 2% per year. In this paper.) The Stabilization Triangle We idealize the 50-year emissions reductions as a perfect triangle in Fig. It is the length of a career.e. experience with specific wedges.socolow@princeton. On the other . believing that the corresponding deployment is certain to occur in BAU. The debate in the current literature about stabilizing atmospheric CO2 at less than a doubling of the preindustrial concentration has led to needless confusion about current options for mitigation. tobecp1!le!Jeg\ljle. edu (R. and ongoing improvements in energy efficiency prevent the need for eight additional wedges. and BAU is represented by a straight-line "ramp" trajectory rising to 14 GtC/year in 2054.SITat<:gjesneege<i. removes exactly onethird of BAU emissions. IA are near the center of the cloud of variation in the large published literature (8). enhanced research and development would have to begin immediately. The next 50 years is a sensible horizon from several perspectives. For example.oftechnologicalreadiness'vand calls for "revolutionary changes'.8~ revolutionarytechnology. but eventually drop to zero. s of this centuryand.the first half of this centurysimply by. Socolow2* Humanity already possesses the fundamental scientific.lch<. A portfolio of technologies now exists to meet the world's energy needs over the next 50 years and limit atmospheric CO2 to a trajectory that avoids a doubling of the preindustrial concentration. Stabilization at any level requires that net emissions do not simply remain constant. or less than double the preindustrial concentration of 280 ppm (3-7).. such as methane and N20. On one side. and the future behavior of natural sinks for atmospheric CO2 (i. To keep the focus on technologies that have the potential to produce a material difference by 2054. Department of Ecology and Evolutionary Biology. The "stabilization triangle. The calculations behind Fig. followed by a linear decline of about two-thirds in the following 50 years. Hurnanitycan solve the carbon and climate problem . then ~ 18 wedges would be required (Section I of SOM text). In that same simple model (9). This historic trend in emissions has been accompanied by 2% growth in primary energy consumption and 3% growth in gross world product (GWP) (Section 1 of SOM text). It thus represents a cumulative total of25 GtC of reduced emissions over 50 years.. (BAU) trajectory].'. lB.9bytheP8§si. because it is the dominant anthropogenic greenhouse gas. On the ather side." w~furrrnaustria[ scale. 2Department of Mechanical and Aerospace Engineering. many are already implemented some.fusion. Most readers will reject at least one of the wedges listed here.TOWARD REVIEW A HYDROGEN ECONOMY Stabilization Wedges: Solving the Climate Problem for the Next 50 Years with ~urr~"nt T~chD~Qlo~Les S. technical.· and artificial photosynthesis (2).agree. our list of mitigation options is not exhaustive. and if carbon emissions were to grow at 3% per year. Our BAU simply continues the 1. the portfolio as a whole is large enough that not every element has to be laboratory bench~~"~"~de~~ati~np!!lJ!!:~. 8].illfue econdhalf . a recentreview in Science asserts that the IPCC claim.S. Princeton University." ppm stabilization is achieved by 50 years of flat emissions.B ul it is . 50 years of BAU emissions followed by 50 years of a flat trajectory at 14 GtC/year leads to more than a tripling of the preindustrial concentration. Emall: pacala@princeton. demollstrates"misperceptions .

High-Latitude Exchange interior Diffusion Advection (HILDA) ocean model telecommuting.emissions required for stabilization. The power shifted to Wedges can be achieved from energy effiOption 2.) '~"carbonin:ensiw~oaJ: However. the Unitcoal-based electricity ed States announced the goal of decreasing its in 2054 were procarbon intensity (carbon emissions per unit duced at 60% instead GDP) by 18% over the next decade. u. 2 billion cars (roughly ~our that its efficiency has GtC/year in 2054..TOWARD A HYDROGEN ECONOMY l( '\/ What Current Options Could Be per gallon (mpg) on conventional fuel. m 2054. t~ l~ore as large from natural of stabilization curves described m (11). (B) ldeallzatlon of !A). and the total area of the stabilization triangle is 175 Gte. and refiigertions could be scaled up to two or more ation in residential wedges. year out of 6. Carbon capture and storage energy. hensive but complex "integratedassessments" (1) produced about oneof carbon mitigation by letting the full-scale exfourth of all carbon amples that are alreadyin the marketplacemake a emissions: 1. from the decarbonization of the supwedge would also be achieved if the average current gas-based power. An entire wedge "'Xdte/fi5'11: Decorwoul~ be create~ if tl~eUnited States were to bo~iiatiolliJf . The arrow at the bottom right of the stabilization triangle points downper year (as they do today). the less the available savAbout half of potenings from greater efficiencyof electricityuse.carbon emissions as CO from fossil fuel combustion and cement manufacture: 1. a wedge would be Scaled Up to Produce at Least One 2054 averaged 60 mpg. (CCS) technology prevents about 90% of the ological storage in forests and agricultural Opti9n$:Monu1fi9ient buildings. costs.yremenelo its own of the SOM text. ply of electricity and fuels (by means of fuel fuel economy of the 2 billion 2054 cars were Option 6: Storage of carbon captured in shifting. Ric~71s.) plant in 2054 has in. buildings.e details InSe~tlOll 20fthe~0l\1 .org SCIENCE VOL 305 13 AUGU5T 2004 969 .water heating. a de. The area between the two curves re~res. times as many as today) average 10. iniifficiency.and Ed~ond~ lWRE) family catal~sts . 1.2004. gas power plants as the SOM text. instead of averaging 30 miles such plants emit car.>In2000. and from bi5000 miles instead of 10. Here.b. bonized. Improvements coal. and tial savings are in the vice versa.and coal power plants.15% perye~Ejp. from buildings by The more the electricity system becomes decarabout one-fourth. our operating on average analysis is intended to complement the compreat 32% efficiency.· Carbon emis.below 7 GtC/year after 2054 to achieve stabilization at 500 ppm. carbon capture and storage. With linear growth. natural gas plants.>" (See references and every country were to follow suit byaggin!5fhe details in Section 3 same 0. effic~encyand OptionStSubsti.sciencemag. The stabiliOption 1: Improved fuel economy. Below.ef~clency. re. The ..Eleereset Its carbon intensity goal to a decrease of .aches 1 ~ose that in 2054.public acceptance. modified as described m Section 1 of efficient lighting and insulation for buildings. (A) The top curve is a representative BAU emissions path for global conservation options are I~ss tangible than tutingnatural gas for those from the other categones. These apBecause the same BAU carbon emissions proaches reduce midcannot be displaced twice. a wedge is the so a wedge would be provided by the instalpIes of options that are already deployed at an difference between pursuing and not pursuing lation of CCS at 800 GW of baseload coal industrial scale and that could be scaled up ''known and established approaches" to energyplants by 2054 or 1600 GW of baseload further to produce at least one wedge (sumefficientspace heating and cooling.tfiCiryGnd Fuels 2. The bottom curve assumes an ocean uptake calculated with the to the gro~1:h of the service e~onomy and from coal plants.000 miles.(12) and a constant net land uptake of 0. with baseload gas by 2054.000 miles improved to 50%.5 GtC/year (Section 1 of the SOM text). Wigley.!iYSup. for example.\ of 40% e~ crease of 1. As. 1 www. A gas forthis wedge is four times as large as the total ciency.:Imrequire some price trajectory for carbon. . ~arbon many po~sible . or even half of it.7 GtCI simple case for technologicalreadiness.11 % per year and extend it to 50 years. eluding future fuels prices.allowed emissions are fixed at 7 GtC/year beginning In 2004. and renewable energy).ott?m in energy efficiency will come from literally sions per unit of elec. Also. in 2002. the deproved-power-plant tails of which depend on many assumptions. cars in bon at a rate of 1 GtC/year. Interactions among wedges are disbuildings in developcussed in the SOM text. Fig. om focus is not on ing countries (1). /\ccording fossil carbon from reaching the atmosphere. from 7:0 GtClyear in. the total avoided emissions per wedge is 25 GtC. Although several oplighting. In general. One wedge would Because 700 GW of ward to emphasize that fossil fuel emissions must decline substantially be achieved if.d e~lsslons (blue). andif. each o! which. alone. soils. T~.:A stabilization less . (See references and age baseload coal triangle of avoided emissions (green) and all~we. but the annual distance traveled were power plants.the achievementof a wedge will Opti{)n<4: ·. we discuss 15 different examto a 1996 study by the !PCC.text. For example. we provide four of sume that the capaci.comparisons of greater and ty factor of the aver. with fuel type and achievedby displacing 1400 GW of baseload coal Wedge? distance traveled unchanged. cost reductions by means of1earning.and. The most likely approach marized in Table 1). A wedge would Improvements in efficiency and conservation be created if twice toprobably offer the greatest potential to proday's quantity of vide wedges.curve is a CO emissions path consistent With atmospheric CO~ stabllizatl?n hundreds of innovations that range from new tricity are about half at 500 ppm by 2125 akin to the. we doubt that any could fill the and commercial stabilization triangle.2 GtCI Category I: Efficiency and-Conservation year. chemi~al pr~cesses.creased to 90% and zation triangle is divided into seven wedges.96% per year.ents the avoided.Instead. achieving one century emissions wedge often interacts with achieving another. nuclear 30 mpg.' Reduced reliance on cars.5% per year grO-Mh starting.

..__ . followed by (ii) geologic storage.~_t~_~_~u_rr:~~t_~~~~:L~. biodiversity from reduced deforestation reforestation.•. so a wedge would be a flow to storage IS and 20 times as large as the current flow.sciencemag.1rI\9I1e!]1i~sh)J1r~te inl()H~y1. Sleipner project in the North Sea strips CO2 from natural gas offshore and reinjects$i9!iS/$GDP)(e. plant Technology already in use forH.(twke the (#r~!'it@p..9. Reduceduse of vehicles 3.c.. coal PQw~r introduce ccs at plants pri...tworisrn.. The capture part of a wedge '..:~Pture CO.S· .~ __ . Biomass fuel for fossil fuel productlon. ihfra.. . ".01 GtC/year of carbon as CO2 is injected into geologic reservoirs to spur enhanced oil recovery.~~~~!~ttl?_~_ " .. t~~letP~ten~ialWe4g~s: Sfrit~iie~~vajla!)le toreduce the~. by~arbonpP\ity ••• . Increase fueleconomy for 2 bllUoncar~ from 30 to G() mpg .. Hydrogen production from fossil fuels is already a very large business.>dvcing250 MtH. and a carbon flow of 1 GtC/year (whether as methane or CO2) is a flow of 69.Wer Urban deslgri. ·~~n b~tuneJ . Nu'... •. 1. .. . was~e·._. infrastructure Add 4 million 1-MW-peak windmills (tootlmes the 12.~tJ·t.· . Create 3S(jO Sleipners ..~?_~Le-_'P:cI~@J~2-~T.~.. The scale of the storage part of this wedge can be expressed as a multiple of the scale of "... Hz safety. Gas baseload power for coal baseload power Cornpetlng demands for natural gas ~:<.~~ Etof\pm)Hi"icl~ c~tbi.• .t!~.• . Introduce q::S at synfuels plants..: . Eff9t{ ~}I':~05~rf~f_p.e. pei' ye~rt02.-.Wed~~. with the use of 2S0X 1Q6ha (one-sixth of world cropland) Forests and agricultural soils Land demands of agriculture.. Capture (02 at H.Weak lncentlves Advanced high-temperature materials 5.~~~_ .._.••..~::.. . .....1li%wye~r .000 Bscf/year (190 Bscf per day)....{)ri •.5..~_~~./year.. Today.S._:. if half of feedstock carbon is avaHable for capture . if synfuels are produced Without CC..·.... .. Duriltll~?torage..~.. -:. . 2()o4tQ2()~4~y??ptq. Norway's -_' ". PY production cost· . The capture part of a wedge of CCS electricity would thus require only a tenfold expansion of plants resembling today's large hydrogen plants over the next 50 years... . 01' the first geological storage demonstration project.Q60/0f~d. in which the waste CO2 is injected into subsurface geologic reservoirs..<l1 ofl .structLire 13./year today fromall sources] .Capture CO2 atcoal-to-synfuels plant (jeqlqgicill storage .Winc:l power for. .. . plus Decrease tropical deforestation to zero instead of . afforestation. Option 7: Storage of carbon captured ill hydrogenplants.".. • MUltiple·usesoflandbec~use windmills are Add2rnil!ion1·MW·pe~kwindmllls(S()tlinesthe currerit qapacity) "occupying". Energy ~ftis/entY ~i1d ~oriseIVIiH6i1 . and establish 300 Mha of new tree hew plantations. To smooth out seasonal demand in the United States. larger projects) over the next 50 years.~i~~. '.. has two steps: (i) precombustion capture of CO2.tele~ornrriuting from 10. .city) Increased CO.producing ~o million barrels a day from co<\1(zOO tlmes 5a~01).. ... Decrease car travel for 2 bilUo.. .' . .lh-int~fl$ity .-~_. plantations (twice thecurrentrate] J_:. .v. ·'W?eWiibleeli!ctficitY~rlcf Nels 970 13 AUGUST 2004 VOL 305 SCIENCE www. or current seasonal storage of natural gas.'..~+~---..~f/'~< ·t.·..ed lander bffshi..••.110/0per y~~r) .J-... ih¢re. Wind Hz In fuel-cell car for current capacity) gasolineln hybrid car Blodlverslty. .. plant . r~l. Globally. <:Miye!ii'of..... in which hydrogen and CO2 are produced and the hydrogen is then burned to produce electricity.".. . 3() X 1O~h<)._}~~e!.•.~~.. __~P-P'lx.CS at aOoGWtoal or 1600 OW natural gas (compared with 1060 oW coal in 1Q99) . the natural gas industry annually draws roughly 4000 billion standard cubic feet (Bscf) into and out of geologic storage.:_. EfflClentvehicles 2..n~:. . ~~c_. competing land use Add 100 times the current Brazil or U./ye!Jf from natural gas (compared with 40 MtH. PV power-fer coal power Mel 2000 GWcpeak PV (70Q times the current capacity] on 2X 106 ha •. mass translt.._..000 to SQOOmilesper year .1 GtC/year of CO2.. A worldwide effort is under way to assess the capacity available for multicentury storage and to assess risks of leaks large enough to endanger human or environmental health..Ttu: hydrogen resulting from precombustion capture of CO2 can be sent offsite to displace the consumption of conventional fuels rather than being consumed onsite to produce electricity.•.. Car sl~~. 7.-:". so a wedge would be 3500 Sleipner-sized projects (or fewer.~pti?n. ethanol 13.• . Effident baseload coal hict~~seiedy¢tlpnby adoitioTi~IQ.• •wic:lelyspac. at baseload power .•.• . '.. Nqd~~rt#$i¢n Add 7QQG.A""I.)..()./Year from coalor 500 MtH.W . about 0.B..• Cut carbon emlsslons byone-fourth in tluilolngs and appliances projected for 2054 Producetwlce today's coal power output at 6()% instead of 40% efficiency (compared with 32% today] ... . . Reduced deforestation.p. . successful permitting· N~~le~rpr()\lt~ri\tion..y¢\i9(1.a$~\. to reduce~ilrbph etni~siq~s f(QJl1 qi!inh1eill:s. benefits to 14..~.:I~~~X.~:. and ().. .3 million tons of carbon a year (MtC/year) into a non-fossil-fuel-bearing formation... Fue! shift Replace 1400 GWSO%-effident coal plants with gas plants (four times the current production of gas-based power) (:02 Capture oM Swage (CCS) Introduce C:. hydrogen plants consume about 2% of primary energy and emit 0. gO._..S Gtc:.le~r p~WerforFR~1 p(i\fiJet 1().. Efficient b\1ilqings 4.TOWARD A HYDROGEN ECONOMY current enhanced oil recovery. so a wedge of geologic storage requires that CO2 injection be scaled up by a factor of 100 over the next 50 years..:. prcductlon Hi safety.

Looming over carbon manageabout 2 million hectares of land in 2054. I.000 barrels per day spectacular. biofuel to power one's car. 5. or at quire 2000 GWp of installed capacity that natural gas plants producing 500 MtH/year. A wedge of PV electricity would Op~9n1J:Storageof carbon captured in require 700 times today's deployment. 981 (2002). Press. An ethanol wedge would require 250 ity currently deployed.. One can buy hydrogen produced with the chemistry of carbon capture. the world's largest synfuels efficient or the fuel-cell vehicle is less facility. Installed wind capacity has been a wedge or more without a lot of new growing at about 30% per year for more than research. PV hydrogen Index. conservation tillage practices had been adopted on 110 million hectares of the world's 1600 million hectares of cropland. R. A second half-wedge would www. some on land and some offshore. At least Because windmills are widely spaced. the Option 5 estimates. land one half-wedge would be created if the current with windmills can have multiple uses. however. seeds are drilled into the soil without plowing). 1971 (2002). a wedge being halved. and facilities that inject carbon into geologic reservoirs. References and Notes 1. the use of cover crops.. per year. Conclusions In confronting the problem of greenhouse warming.) Currently. A CCS in 2054. The storage part of this electricity has been growing at a rate of 30% option is the same as in Option 6. Contribution to the Third Assessment Report of the tnterqovemmentei Panel on Climate Change (Cambridge Univ. or about twice the nuclear capaccom. Practices such as conservation tillage (e. production of synthetic fuel (synfuel) from coal.Pv array. we assume the same value for the esting factor-of-two carbon-saving advanflow in synfuels and allow for imperfect tage of wind-electricity over wind-hydrocapture. Eds. need not exceed newable electricity can produce carbonthose associated with fuel refined from crude free hydrogen for vehicle fuel by the elecoil if synfuels production is accompanied by trolysis of water. Watson et a/. we presented a part of the case for action by identifying a set of options that have the capacity to provide the seven stabilization wedges and solve the climate problem for the next half-century. M. M. almost all of which can be attribthe same 90% capacity factor assumed for the uted to Brazilian sugarcane and United States coal plants. 2. produce 165. An even larger area would be required to sion in nuclear power requires restoration the extent that the biofuels require fossil-carbon of public confidence in safety and waste inputs. Assuming that half of the carbon entering by 4 million l-MW p windmills in 2054. Option 10: Wind electrici(v. Science 296. Climate Change 2001: Miti9ation. ReCarbon emissions. by 2054. or 2 ment in 2054 is the possibility of large-scale to 3 m2 per person. production of about 34 million barrels per day Option 9: Nuclearfission. T.htm.We account_')::'ategOl:V llI: Natural Sinks for the intermittent output of windmills by / Although the literature on biological sequesequating 3 GW of nominal peak capacity (3 tration includes a diverse array of options and GW p) with I GW of base load capacity. rn(lllqgement. Compared with Option between capturing and venting the CO2 from 10. Option l~:. if a 2054 synfuels plant leaves as fuel but the used in high-efficiency fuel-cell cars. (The flow of carbon in tricity that achieves a wedge by displacing 24 million barrels per day of crude oil is I high-efficiency baseload coal. Geneva. here we restrict our attention to the deployment of 2)109 G}y" that displaces coal pair of options that are already implemented electricity in 2054W2 miJIion I-MWp wind at large scale and that could be scaled up to turbines). Royal Commission on Environmental Pollution.Dti the basis of of ethanol in 2054 that could displace gasoline. a wedge requires 200 Option 13: Biofuels..illle can buy electricity from a ·wInd turbine. Here. or nuclear power plant. 2001). if it continequal to about one-sixth of the world's cropued for 50 years (10). displaces coal electricity in 2054. (See Section 4 of the SOM text 10 years and is currently about 40 GWp' A for references and details. B. (IPCC Secretariat. primarily because annual tilling increases the rate of decomposition by aerating undecomposed organic matter. About 55 Gte. If conservation tillage could be extended to all cropland. IPCC. and international security agreevested biofuels crops is also often suitable for ments governing uranium enrichment and conventional agriculture. B. from coal. Option 12: Renewable hydrogen.g. gas turbine. respectively). En- 13 AUGUST 2004 971 . 2001). This about 50 times larger than today's global prowould require 700 GW of nuclear power with duction rate. available at www. This ethanol production rate would be cient base load coal capacity in 2054. Cambridge. a wedge of nuclear provided the ethanol itself were fossil-carbon electricity would displace 700 GW of effifree. Thus. Oppenheimer. the choice today is between action and delay. some very large estimates of the global poa wedge of wind electricity would require the tential. Every one of these options is already implemented at an industrial scale and could be scaled up further over 50 years to provide at least one wedge. UK.Agricultu"aL§()ils.Fossil-carbon fuels can Sasol-scale coal-to-synfuels facilities with also be replaced by biofuels such as SCIENCE VOL 305 A HYDROGEN ECONOMY be created by reforesting or afforesting approximately 250 million hectares in the tropics or 400 million hectares in the temperate zone (current areas of tropical and temperate forests are 1500 and 700 million hectares. 4. Because land suitable for annualJy hardisposal. The hydrogen produced CCS. This interGtC/year. the capture would achieve a wedge of displaced gasopart ofa wedge in 2054 would be the difference line or diesel fuel.. has been lost historically in this way. biofuels production plutonium recycling. up to one-half of the soil carbon is lost. A third half-wedge would be created by establishing approximately 300 million hectares of plantations on nonforested land. and hundreds of devices that improve energy efficiency. could compromise agricultural productivity. ment of temperate and tropical forests. Switzerland. The wind turtive assumptions lead to the conclusion that at bines would "occupy" about 30 million hectleast one wedge would be available from reares (about 3% of the area of the United duced tropical deforestation and the manageStates). accompanied by a verification program that enforces the adoption of soil conservation practices that actually work as advertised. Although The former is six times the current rate of only 3 GWp ofPV are currently installed. farms practicing conservation tillage. from photovoltaic (PV) electricity would reat coal plants producing 250 MtH/year. C.TOWARD would require the installation ofCCS. Metz et al. None of the options isa pip~~dIea11l0r all unproven Idea:-TOOaY. an area to 1990 would yield a wedge.) wedge of wind electricity would thus require Option 14: Forest management. One can visit tropical forests where clear-cutting has ceased. When forest or natural grassland is converted to cropland. Substantial expanland. Science 298. O'Neill. the Sasol plants in gen is still larger if the coal plant is less South Africa.Exmeavx50 times today's deployment. other half can be captured as CO2. By 1995. Thus. rate of clear-cutting of primary tropical forest Option 11: Photovoltaic electricity: Simwere reduced to zero over 50 years instead of ilar to a wedge of wind electricity.grida. Hoffert et a/. 3. or two wedges' worth.sciencemag. this is twice as many I-MWp windmills coal synfuels plants producing 30 million baras would be required to produce the elecrels of synfuels per day. and erosion control can reverse the losses. The storage part of this opwedge of biofuel would be achieved by the tion is again the same as in Option 6. The global pace of million hectares committed to high-yield (15 nuclear power plant construction from 1975 dry tons/hectare) plantations by 2054. Climate Change 2001: Synthesis Report. a good case could be made for the IPCC's estimate that an additional half to one wedge could be stored in this way. and synfuels plants.

J. G. Sarmiento. Eds. R. Additionally. and just how many people can they support (6-8)7 Technologies for Hydrogen Production Hydrogen can be generated from water. Today. 100. Similarly. in press. hydrogen is mainly produced from natural gas via steam methane reforming. Wigley. SP.pdf.availableat www. In addition.rcep. Adequacy of Commitments-Avoiding "Dangerous" Climate Change: A Narrow Time Window for Reductions and a Steep Price{orDelay(2002). Williams at Princeton. a number of critical material properties must be satisfied to meet the required stability under the operating conditions of Hl 972 13 AUGUST 2004 SCIENCE www.pdf. and R.environmental defense. D. Future energy systems require money and energy to build. Biomass can easily be converted into a number of liquid fuels. StatisticalReview of World Energy (2003). T. and many of these fossil-derived processes can be adapted for use with a large variety of biomassderived feedstocks. Because energy is required to extract hydrogen from either water or biomass so that it can be used as an energy carrier. However. M. and although this process can sustain an initial foray into the hydrogen economy." Since that time.prlnceton. and biomass processing (using a variety of technologies ranging from reforming to fermentation).availableatwww. as a chemical feedstock. It is clearly not sustainable. 10. available at www. Vancouver. What energy-producing technologies can be envisioned that will last for millennia. if the United National Renewable Energy Laboratory. Golden. The authors thank J. recognizing the finite supply of coal and the possibilities of hydrogen derived from water electrolysis. Coal gasification could produce considerable amounts of hydrogen and electricity merely because of the large size of available coal deposits (9). CMI ( Is sponsored by SP and Ford.}. Proceedings of the Seventh International Conference on Greenhouse Gas Control _time. hydrogen is required in the form of ammonia for fertilizer. and U. One must question the efficacy of producing. Geophys. or as an energy source. 7. Francis Bacon suggested using hydrogen as an energy storage system (4). Rudolf Erren in the 19308 suggested using hydrogen produced from water electrolysis as a transportation fuel (3). in The Carbon Cycle. in a world that is both food-limited and carbon-constrained. nuclear. can address issues of President Bush proposed "$1. available at www.grlda. E-mail: jturner@nrel. the 'energy required tor the necessary sequestration of CO2 would increasethe rate at which coal reserves are depleted. 258-276. a number of thermal chemical cycles have been identified that can use lower temperatures and produce hydrogen and oxygen in separate steps. Special Report on Emissions Scenarios (2001). His goal was to reduce automotive emissions and oil imports into England. 13. whether the best use of biomass is for food." NRP Rep. Res. J.TOWARD A HYDROGEN ECONOMY 9. pp. Although biomass is clearly (and necessarily) sustainable. Given that the United States has a finite supply of both. and pyrolysis oil. CO . and energy security. and C. Socolow. M. Couple that to a modest economic growth rate of -I%. L. The vision of using energy from electricity and electrolysis to generate hydrogen from water for transportation and energy storage to reduce environmental emissions and provide energy security is compelling. Mottershead at BP. Biomass processing techniques can benefit greatly from the wealth of research that has been carried out over the years on refining and converting liquid and gaseous fossil DC1 SOM Text Figs. the amount of hydrogen needed just for U. Press. articles both pro and con have buffeted the whole concept. environmental emissions. States chooses a hydrogen-based future it needs to think carefully about how much energy we need and where it is going to come from. K.because of its relatively low cost. and geothermalenergy as major resources for sustainablehydrogen production. Greenblatt.2659 (1995). L.sciencemag. Environmental Defense.2 billion in research funding so that America can lead the world in developing clean. or (after gasification)!energy. available at www. The hydrogen production pathways from these resources include electrolysisof water. (Cambridge Univ. and this path must be followed with a sustained and focused effort. such as com to ethanol. supporting Online Material www. VOL 305 250-year coal reserves drop to 75 years or so (6). but as yet remains unrealized. L. Cambridge.S. biodiesel. The one that has received the greatest attention involves sulfuric acid (H2S04) at 850°C and hydrogen iodide (HI) at 450°C (l J). Jules Verne. Replacing our current energy carrier mix with a sustainable fuel is one of the key pieces in that system. In 1874. including methanol. however.S to 9 September.htm. Shaffer. S. and distributing that much hydrogen. It remains to be seen. Greenblatt.wau. 12. 6. Issues relating to hydrogen production pathways are addressed here. transportation needs would be about 150 million tons per year (5). which is not at all sustainable. For the high-biomass-yield processes. biomass. Hotinskl. The next generation of fission reactors includes designs that can provide the necessary heat. sustainability must be a hallmark of any proposed future Index. hard decisions must be made about the path forward. 11. thermal chemical cycles using heat. it is often cited as the best resource for economically producing large quantities of hydrogen. Hydrogen as an energy carrier. natural gas. If one assumes a full build-out of a hydrogen economy. Canada. Some of these processes require considerable amounts of 2004. convertingthe vehiclefleet to electricvehicles and generatingthat electricityfrom "cleancoal" or making hydrogen as a possible energy carrier would accelerate that depletion. it cannot supply hydrogen in the amounts required. S. made the comment that "water will be the coal of the future" (2).do?categoryld=95&contentld= 2006480. The hydrogen economy (1) is not a new idea. 51 and S2 Tables S1 to S5 References ergy: The Changing Climate (2000). 954281 (2002). storing. 8. U. In his 2003 State of the Union Address.htm. it represents only a modest reduction in vehicle emissions as compared to emissions from current hybrid vehicles. hydrogenpowered automobiles.That leaves solarderived. Pacala. Wigley. primarily derived from water. wind. Turner Identifying and building a sustainable energy system are perhaps two of the most critical issues that today's society must address. Schimel.bp. VIEWPOINT Sustainable Hydrogen Production John A. This paper is a product of the Carbon Mitigation Initiative (CMI) of the Princeton Environmental Institute at Princeton. "Climate optlons for the LongTerm (COOl) synthesis report. ethanol. USA. and ultimately only exchanges oil imports for natural gas imports. 2000). Keller at Penn State. T. R. Because the direct thermal splitting of water requires temperatures of >2000°C and produces a rapidly recombining mixture of hydrogen and oxygen (10). which could be transported and used to generate hydrogen on site.

Sign up to vote on this title
UsefulNot useful