Lessons from Historical Energy Transi4ons and Implica4ons for Energy Technology Innova4on Systems arnulf.grubler@yale.edu Economic Challenges for Energy Madrid 22-24 January 2014
Energy Transi4ons - Main Messages Extremely long 4me constant of change, but: measurement biases (inputs instead of outputs) Change pervasive and systemic, but: end- use dominance ignored by policy Scaling across all scales (unit plant industry), but: granular technologies less risky and more successful Impact of policies mixed, but: how ma=ers more than how much
World Primary Energy Transi3ons changeover 3me Δts: 80-130 years 100 World Primary Energy Substitution Begin of energy policy focus: Δt s >2000 yrs Percent in Primary Energy 75 50 25 traditional biomass coal Δt +130 yrs Δt -130 yrs Δt - 80 yrs modern fuels: oil, gas, electricity Δt +90 yrs 0 1850 1875 1900 1925 1950 1975 2000 2025 Source: GEA, Chapter 24, 2012
Input vs. Output Measures of Growth (example ligh4ng services UK, index 1700=100) 100.000.000 10.000.000 1.000.000 100.000 10.000 Light service/useful energy normalized (lumen- hours) Light primary energy normalized 1.000 Light (real) expenditures normalized 100 10 1 1700 1750 1800 1850 1900 1950 2000 Source: Grubler (in press) based on Fouquet, 2008
UK Transi4ons in Energy Services for Light 1.0 Δt= +76 yrs Δt= - 76 yrs Δt= - 25 yrs Δt= +25 yrs Δt= - 126 yrs Source: Grubler (in press) based on Fouquet, 2008
Importance of Energy End- use Least efficient part of energy system, with vast improvement poten4als Dominant in terms of installed capacity Dominant form of energy investment (and GDP & employment mul4pliers!) Short- ribed by systemic innova4on porcolio biases
Capacity of US Energy Conversion Technologies GW (rounded) 1850 1900 1950 2000 sta tiona ry therm al (furnace s/boilers) 300 900 1900 2700 end-use mechanical (prime movers) 1 10 70 300 electrical (drives, appliances) 0 20 200 2200 mobile animals/ships/trains/aircraft 5 30 120 260 end-use automobiles 0 0 3300 25000 sta tionary thermal (pow er plant boilers) 0 10 260 2600 suppply mechanical (prime movers) 0 3 70 800 chemical (refineries) 0 8 520 1280 TOTAL 306 981 6440 35140 Energy end-use = 30 TW or 87% of all energy conversion technologies = 5 TW or 50% when excluding automobiles Source: GEA Chapter 24, 2012
World Energy Technology Innova4on Investments (Billion $) innovation market diffusion (RD&D) formation End-use & efficiency >>8 5 300-3500 Fossil fuel supply >12 >>2 200-550 Nuclear >10 0 3-8 Renewables >12 ~20 >20 Electricity (Gen+T&D) >>1 ~100 450-520 Other* >>4 <15 n.a. Total >50 <150 1000 - <5000 non-oecd ~20 ~30 ~400 - ~1500 non-oecd share >40% <20% 40% - 30% * hydrogen, fuel cells, other power & storage technologies, basic energy research Source: GEA Chapter 24, 2012
Public Policy- induced ETIS Investments billion US$2005 Source: Wilson et al. Nature CC 2012
Scaling of Energy Technologies Across scales: units plant industry markets Prolonged experimenta4on needed (pre- mature scaling- up is risky) Scale dominant source of cost improvements (and not learning by doing ) Improvement poten4als and markets largest for granular technologies
Unit- level Scaling of Wind Turbines Source: Neij et al., 2014
Post Fossil Technologies Cost Trends Source: Grubler & Wilson, 2014
Learning Rates vs Cumula3ve Experience (# of units produced) Learning rate (% cost change per doubling) 50.0 40.0 30.0 20.0 10.0 0.0-10.0-20.0-30.0-40.0 8 12 9 10 11 7 5 6 4 2 3 N M I J F L - 50.0 1.E+00 1.E+03 1.E+06 1.E+09 1.E+12 1.E+15 1.E+18 Cumulative # of units produced K G H D C E 1 B A A Transitors B DRAMs C Automobiles D Washing machines E Refrigerators F Dishwashers G Freezers (upright) H Freezers (chest) I Dryers J Calculators K CF light bulbs L A/C & heat pumps M Air furnaces N Solar hot water heaters 1 PV modules 2 Wind turbines 3 Heat pumps 4 Gas turbines 5 Pulverized coal boilers 6 Hypropower plants 7 Nuclear reactors 8 Ethanol 9 Coal power plants 10 Coal power plants 11 Gas pipelines 12 Gas combined cycles Source: Wilson et al. Nature CC S1 (2012)
Cumula3ve Experience /Learning Favors Granular Technologies Learning rate (% cost change per doubling) 50.0 40.0 30.0 20.0 10.0 0.0-10.0-20.0-30.0-40.0 8 12 9 10 11 7 5 6 4 2 3 N M I J F L - 50.0 1.E+00 1.E+03 1.E+06 1.E+09 1.E+12 1.E+15 1.E+18 Cumulative # of units produced K G H D C E 1 B Mean of granular end use technologies: LR=20% CumProd= 1 Billion Mean of lumpy supply technologies: LR=10% CumProd= 10,000 A A Transitors B DRAMs C Automobiles D Washing machines E Refrigerators F Dishwashers G Freezers (upright) H Freezers (chest) I Dryers J Calculators K CF light bulbs L A/C & heat pumps M Air furnaces N Solar hot water heaters 1 PV modules 2 Wind turbines 3 Heat pumps 4 Gas turbines 5 Pulverized coal boilers 6 Hypropower plants 7 Nuclear reactors 8 Ethanol 9 Coal power plants 10 Coal power plants 11 Gas pipelines 12 Gas combined cycles Source: Wilson et al. Nature CC S1 (2012)
Mixed Impact of Policies Significant transi4on de- accelera4on since 1970s (inconsistent policy push and pull ) Success stories when policies are: aligned, pa4ent, and systemic, e.g. Japan, Brazil, vs. US Systemic underinvestment in end- use and efficiency: ALL actors ETIS increasingly global, but too few interna4onal tech coopera4on & knowledge spillovers Erra4c policy leads to rapid knowledge deprecia4on
Knowledge Deprecia4on Rates Degree of technological obsolescence (rate of innova3on) Degree of knowledge stock turnover (policy & human capital vola3lity)
Knowledge Deprecia4on Rates (% per year) empirical studies reviewed GEA Chapter 24 (2012) and modeled R&D depreca4on in US manufacturing (Hall, 2007) Degree of knowledge stock turnover (policy & human capital vola3lity) Low High Degree of technological obsolescence (rate of innova3on) Engineering designs US: <5% Aircraft, Liberty ships manufct. US: 40% Service industries: 95% Wind US: 10% Chemicals, Drugs: 15-20% Miscell. >20% PV Japan: OECD 30% nuclear R&D: 10 40% Computers: 32% Electrical, Machinery: 32-36% France breeder reactors: 50-60% High Source: Grubler & Nemet, 2014
R&D Expenditures and Knowledge Stock (all IEA countries, billion $2007) remaining knowledge stock (KS) calculated with 0, 10, 20, and 40% deprecia4on rates respec4vely Source: Grubler & Nemet, 2014 12 Nuclear iea nuclear R&D and knowledge stock 1974-2007 (billion US$2005) 1000 12 Efficiency iea efficiency R&D and knowledge stock (billion US$2005) 1000 R&D expenditures billion US$2005 9 6 3 R&D expenditures billion US$2005 100 10 9 6 3 R&D relacement level R&D 20%dr KS0%DR KS10%DR KS20%DR KS40%DR 100 10 Remaining Knowledge Stock B$ R&D rela KS0 KS1 KS2 KS4 0 1975 1980 1985 1990 1995 2000 2005 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 1 0 1975 1980 1985 1990 1995 2000 2005 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 1 Cumula4ve R&D: 236 B$ Remaining KS: 11-47 B$ Cumula4ve R&D: 38 B$ Remaining KS: 4-12 B$
Summary Change innova4on systems and not just more of isolated push (R&D) or pull (renewable electricity FIT/PS) policies Policy alignment and integra4on key: e.g. - prices/taxes + standards + R&D Mgmt - market crea4on + diffusion + phase- out