Modeling Energy Efficiency · Mintzer, Dick Munson, Greg Nemet, Lynn Price, Wendy Reed, Art...
Transcript of Modeling Energy Efficiency · Mintzer, Dick Munson, Greg Nemet, Lynn Price, Wendy Reed, Art...
John A. “Skip” LaitnerSenior Economist for Technology PolicyAmerican Council for an Energy-Efficient Economy (ACEEE)
USAEE/IAEE North American ConferenceHouston, TexasSeptember 19, 2007
Characterizing Energy Efficiency Characterizing Energy Efficiency Opportunities in Energy ModelsOpportunities in Energy Models
Some Acknowledgments
• This presentation draws on the many ideas that have evolved from wide-ranging discussions with a variety of friends, colleagues, and collaborators over the years. I would like to acknowledge the many invaluable insights and thoughts from a very broad community, including: Bob Ayres, Steve Bernow, Fatih Birol, Bruce Biewald, Marilyn Brown, George Burmeister, Penelope Canan, Tom Casten, Ken Colburn, Ruth Schwartz Cowan, Laura Cozzi, Stephen DeCanio, Catherine Dibble, Jerry Dion, Karen Ehrhardt-Martinez, Neal Elliott, Andrew Fanara, Lorna Greening, Bill Halal, Don Hanson, Alan Heeger, John Hoffman, Tina Kaarsberg, Jon Koomey, Amber Leonard, Irving Mintzer, Dick Munson, Greg Nemet, Lynn Price, Wendy Reed, Art Rosenfeld, Matthias Ruth, Alan Sanstad, Elizabeth Wilson, and Ernst Worrell.
• I would also like to extend my deep appreciation to ACEEE’s own Steve Nadel and Bill Prindle who encouraged me to rejoin the research community after an absence of more than a decade; and to those of you here today who have come together to explore critical ideas that will make this forum a very real and important contribution to the dialogue.
An Observation With Four Areas of Suggested Improvements
• My own observations since the 1992 Rio Summit suggest that among the causes for US reluctance to adopt any meaningful climate policies have been what I believe to be inappropriate modeling exercises which preempt the review of a more robust set of energy and climate policy initiatives.
• In other papers and journal articles, I suggest at least four areas of needed improvement in our modeling practices:
1) Technology characterization that is often limited or even inappropriate – for both the demand and the supply-side of the equation;
2) Capital flows that are not sufficiently disaggregated to provide meaningful policy assessments;
3) Modeling assumptions about consumers and firms which may be unrealistic and which may also give misleading insights about policy options; and
4) An economic accounting of investments and technology choices that are limited or poorly represented
• In the limited time here today, I will focus on the first of these items.
One Egregious Example of (at least) Five Models Which Use Some Form of the Following
Characterization of Potential GDP Impacts
Elasticity
old
newt P
PGrowthRateGDPGDP−
⎟⎟⎠
⎞⎜⎜⎝
⎛= **0
So that no matter how cost-effective the policies or the technologies, if there is any kind of net price increase from a given policy initiative, the macroeconomic impacts (by definition) must be negative.
Given today’s understanding of returns on technology and market dynamics, this is not an acceptable characterization.
$40/
tCO
2
Domestic CO2 Emissions to 80% of Reference Case Values
Marginal Cost
Therefore, current US reduction targets based largely on voluntary actions
Re-examining the Conventional Marginal Abatement Cost Curve
Estimate of Resource Cost in 2030:~ $40 * 8115 MtCO2 * 0.8 / 2
~ $130 billion per year
But What If. . . .
• The price signal, in this case, $40/tCO2 is not a highly accurate estimate of resource costs, but only a signal that changes behavior and patterns and investments?
• What if the 20% reductions were energy bill savings:– generated through productive efficiency investments that had
(on-average) a 5-year energy payback– Lowered the non-carbon portion of energy prices by 10%, and – Stimulated other productivity innovations?
• Then a negative $130 billion resource cost might become a $227 billion net savings – not at all a free lunch, but a significant return on more productive pattern of investments.
Domestic MtC Reductions
So, a Different Result Emerges Using Both Costs and Benefits in the Analysis
Marginal Cost
Marginal SocialBenefit
MarginalCost w/Policies$/tonne
Carbon
**
*Where social benefit minimally includes avoided energy expenditures or financial returns on technology investments
An Emerging Story of Negative Cost Curves
Comparing the Marginal Abatement Cost Curve and the Production Isoquant
GHG Reductions or Energy Savings
$/tC
or
$/Q
uad
GHG Emissions or Energy Use
Cap
ital C
ost/t
C o
r /Q
uad
Marginal Abatement Cost Curve
Production Isoquant
Or the Big MACC. . . .
Which can be based on a discrete set of technology cost and performance characterizations
An Isoquant of Energy Services Showing Relationship Between Capital, Energy, Price Ratio
Cap
ital I
nves
tmen
t
Annual Energy Flows
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
OriginalPrice-Preference
Ratio
Doubling ofPrice-Preference
Tangent
Initial Values of 8.12 for Capital and 1.00 for Energy
Shift in Values to 9.88 for Capital and 0.75 for Energy
Elasticity of substitution whereσ
= 0.70 for this illustration
Changes in Capital and Energy as a Result of Doubling the Price-Preference Ratio
• Under the assumption that– energy prices increase by 50%– while hurdle rates decrease from 20% to 15%
• the price-preference ratio will double – from 1.00 / 0.20 which equals 5.0– to 1.50 / 0.15 which equals 10.0
• In this case– Capital investment will increase from 8.12 to 9.88 (22%)– Annual energy flow will decrease from 1.00 to 0.75 (-25%)
• Project payback will be– (9.88 - 8.12) / 0.25 = 7.04 years under the old energy prices– (9.88 - 8.12) / (0.25 * 1.50) = 4.69 years with the new energy prices
• Note: If this isoquant reasonably describes the entire world economy over time, and in the absence of a price signal, a combination of programs, policies, and changes in consumer preferences would have to reduce hurdle rates on average from 20% to 14% to achieve a 10 percent reduction in energy. This would also require a 16% increase in capital.
≠
Typical CGE Representation Technology-Based Representation
A Thought Experiment: Comparing CES Technology Representation
Production Capital
Energy Capital Energy Use
Energy Services
Utilized Capital
Value Added
LaborCapitalElectricity Non-Electricity
Energy Use
Energy-Value Added Bundle
The conventional CGE representation may generate an inappropriate characterization for two reasons: (1) the base of value-added (which includes both capital and labor costs) is much larger than the actual capital costs anticipated in a meaningful technology characterization, and this forces a larger investment than may be actually needed to achieve a given reduction in energy use; and (2) industries show significantly different elasticities across fuel types than the single elasticity which is generally assumed in standard CES production functions.
Drawing from the LIEF Model (Cleetus 2003) and 2001 Census data for the pulp and paper industry, let us assume that a doubling of the energy price ratio leads to a 12 percent savings of fossil fuels. Let us further assume the following CES functions: (a) The conventional CGE models which impose a substitution elasticity of 0.50 regardless of sector or fuel type; and (b) an actual technology-based representation which suggests an elasticity of 0.38:
Capital required is $6.56 billionSimple payback is 6.91 years
Capital required is $1.35 billionSimple payback is 1.42 years
≠
Note: For the documentation that underpins this review and the full set of preliminary results, see Laitner (2007) soon to be circulated for comment.
Technology-Based Representation
Energy Capital Energy Use
σ = 0.38
Energy Services
Conventional CGE Representation
Value Added
σ =0.50
Energy Use
Energy-Value Added Bundle
A Thought Experiment: Comparing CES Technology Representation
A Useful Hierarchy for Evaluating Efficiency Investments within a Production Function
Energy-Related Capital
Non-Energy Productive Capital
X
1U jU Z
UK 1 1E UjK jE
K~ L
V. . .
M tot
An important distinction if efficiency investments generate, say, a 20% return while other non-energy
capital generate only perhaps 10-12% returns
But What of Emerging Technologies?
• These techniques can do reasonably well for technologies for which there are reasonably known cost and performance characterizations
• But what of emerging technologies, or even technologies for which we have little clear sense of the opportunity?
A quick insight from A quick insight from MITMIT’’s own labs on the s own labs on the coming technologies: coming technologies: Neil Neil GershenfeldGershenfeld’’ss 2005 2005 book, book, FAB: The Coming FAB: The Coming Revolution on Your Revolution on Your DesktopDesktop----From From Personal Computers to Personal Computers to Personal FabricationPersonal Fabrication
25 years of really cool stuff Lemelson-MIT Program and CNN present “Top 25” innovations January 14, 2005; updated January 26, 2006
• 1. Internet 2. Cell phone 3. Personal computers 4. Fiber optics 5. E-mail 6. Commercialized Global Positioning Satellites 7. Portable computers 8. Memory storage discs 9. Consumer level digital camera 10. Radio frequency ID tags 11. Micro-electromechanical machines (MEMS) 12. DNA fingerprinting 13. Air bags 14. ATM 15. Advanced batteries 16. Hybrid car 17. Organic light emitting diodes (OLEDs) 18. Display panels 19. High Definition TV 20. Space shuttle 21. Nanotechnology 22. Flash memory 23. Voice mail 24. Modern hearing aids 25. Short-range, high-frequency radio
Scenarios and forecasts that do not explicitly anticipate or incorporate or reflect these and many other technologies are likely giving us some very misleading insights into both our future problems and our emerging opportunities.
See: http://www.techcast.org
Tracking the Technology Revolution – Suggesting a Reference Case that May be
Significantly Different than Usually Forecasted
Further Evidence: Searching LexisNexis for Stories and Articles Citing New Materials/Technologies Over Time
KeyWord/Phrases Total Journals SciMagsNanotechnology 60 6 0 "energy" 18 2 0 "energy-efficient" 0 0 0 "energy efficiency" 2 1 0
Carbon Nanotubes 0 0 0 "energy" 0 0 0 "energy-efficient" 0 0 0 "energy efficiency" 0 0 0
Quantum Dots 3 1 0 "energy" 1 0 0 "energy-efficient" 0 0 0 "energy efficiency" 0 0 0
Polymers 1000+ 244 0 "light-emitting" 16 2 0 "light-emitting diodes" 11 2 0 "LED" 411 75 0 "organic light-emitting diode 0 0 0 "photovoltaics" 11 2 0
1990Total Journals SciMags998 60 6891 55 547 3 084 8 0
330 44 0139 16 0
8 2 04 1 0
184 27 069 9 02 1 01 0 0
989 76 8339 78 0165 24 0980 144 354 12 080 10 2
2000Total Journals SciMags1000+ 16 2591000+ 33 78
327 24 8620 48 8
995 76 202670 63 5829 4 054 9 0
769 55 316234 21 6811 1 021 5 1
1000+ 35 349619 52 144470 43 114999 103 250155 23 9355 35 8
2006
Note: These search parameters were developed by Skip Laitner to compare emergence of these and other technologies within energy and climate policy models over time. The search on the LexisNexis database was carried out by Melissa Laitner in mid-May 2007. For more information on this or related efforts, contact Skip Laitner at (202) 478-6365, or by email at [email protected].
While the number of citations in the literature and market have grown significantly over time, that does not appear to be the case within economic policy models.
Genetic Algorithms: A Technique That May More Appropriately Capture Emerging Technologies
• The innovation literature shows that technologies evolve, that stochasticity is inherent, and that niches and inter-technology flows of knowledge are crucial aspects of technology development; yet, these essential features of the innovation process are not well-suited to characterization in standard models.
• One approach that we are exploring** is whether genetic algorithms can provide a tractable way to include these features. We are now exploring the use of GA to model technological change for energy efficiency technologies. The technique would include expected evolutionary changes, mutation and crossovers, as well as a range of fitness functions.
**The exploratory team includes Skip Laitner (ACEEE), Greg **The exploratory team includes Skip Laitner (ACEEE), Greg NemetNemet (U(U--WI), Stephen WI), Stephen DeCanioDeCanio (UCSB), Catherine Dibble (UMD), and Karen (UCSB), Catherine Dibble (UMD), and Karen EhrhardtEhrhardt--Martinez (ACEEE).Martinez (ACEEE).
• Among the many questions we are considering is whether an explicit inclusion of multiple attributes, niches and crossover mean that such a model will more realistically capture historical experience.
• We are also looking to see if inclusion of these items will also have policy implications that differ from the results of models that include technological change as a function of autonomous improvements, learning-by-doing, or simple price signals.
• Preliminary review suggests that – compared to backstop technologies or other deterministic techniques – the genetic algorithm may provide a more satisfying characterization when service demands, technology characteristics, or adopter characteristics are unknown to researchers ex ante.
• But the work is just beginning. . . .
Genetic Algorithms: A Technique That May More Appropriately Capture Emerging Technologies
The Good News About Energy Efficiency Investments and Climate Change Policies
• It is does not have to be about ratcheting down our economy;
• Rather, it can be all about:– using innovation and our technological leadership;– investing in more productive technologies (including both
existing and new technologies); and – developing new ways to make things, and new ways to get
where we want to go, where we want to work, and where we want to play.
• But again, most economic models appear to assume the former – to the detriment of smart energy and climate policy.
A Selected Modeling and Technology Characterization Bibliography
• Elliott, R. Neal, Therese Langer, and Steven Nadel. 2006. “Reducing Oil Use through Energy Efficiency: Opportunities Beyond Cars and Light Trucks,” Washington, DC: American Council for an Energy Efficient Economy, January.
• Elliott, R. Neal and Shipley, Anna Monis. "Impacts of Energy Efficiency and Renewable Energy on Natural Gas Markets: Updated and Expanded Analysis," Washington, DC: American Council for an Energy Efficient Economy, 2005.
• Geller, Howard, Philip Harrington, Arthur H. Rosenfeld, Satoshi Tanishima, and Fridtjof Unander. “Polices for increasing energy efficiency: Thirty years of experience in OECD countries,” Energy Policy, 34 (2006) 556–573.
• Koomey, Jonathan G., Paul Craig, Ashok Gadgil, and David Lorenzetti. 2003. “Improving long-range energy modeling: A plea for historical retrospectives.” The Energy Journal, vol. 24, no. 4. October. pp. 75-92.
• Laitner, John A. "Skip“ and Donald A. Hanson. 2006. “Modeling Detailed Energy-Efficiency Technologies and Technology Policies within a CGE Framework,” Energy Journal, Hybrid Modelling: New Answers to Old Challenges.
• Laitner, John A. "Skip" and Alan H. Sanstad. 2004. "Learning by Doing on Both the Demand and the Supply Sides: Implications for Electric Utility Investments in a Heuristic Model." International Journal of Energy Technology and Policy, 2004, 2(1/2), pp. 142-152.
• Laitner, John A. "Skip.“ 2004. “How Far Energy Efficiency?” Proceedings of the 2004 ACEEE Summer Study on Energy Efficiency in Buildings. Washington, DC: American Council for an Energy Efficient Economy.
• Laitner, John A. "Skip", Donald A. Hanson, Irving Mintzer, and Amber J. Leonard. 2005. “Adapting in Uncertain Times: A Scenario Analysis of U.S. Energy and Technology Futures.” Energy Studies Review, Vol. 14, No.1, 2005 pp120-135.
• Laitner, John A., Stephen J. DeCanio, Jonathan G. Koomey, and Alan H. Sanstad. 2003. “Room for Improvement: Increasing the Value of Energy Modeling for Policy Analysis.” Utilities Policy, 11, pp. 87-94.
• Martin, Nathan, et al. 2000. "Emerging Energy-Efficient Industrial Technologies," Washington, DC: American Council for an Energy Efficient Economy, 2000.
• Sachs, Harvey et al. 2004. “Emerging Energy-Saving Technologies and Practices for the Buildings Sector,” Washington, DC: American Council for an Energy Efficient Economy, 2004.
• Shipley, Anna Monis and R. Neal Elliott. 2006. “Ripe for the Picking: Have We Exhausted the Low-Hanging Fruit in the Industrial Sector?” Washington, DC: American Council for an Energy-Efficient Economy, April.
Contact Information
John A. “Skip” LaitnerSenior Economist for Technology Policy
American Council for an Energy-Efficient Economy (ACEEE)
1001 Connecticut Avenue, NW, Suite 801Washington, DC 20036
For more information and updates visit:http://www.aceee.org