slide1 l.
Skip this Video
Loading SlideShow in 5 Seconds..
Presentation Outline PowerPoint Presentation
Download Presentation
Presentation Outline

Loading in 2 Seconds...

  share
play fullscreen
1 / 20
Download Presentation

Presentation Outline - PowerPoint PPT Presentation

erika
412 Views
Download Presentation

Presentation Outline

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. Generation and Analysis of Alternative Technology Scenarios Using MARKAL-MGA:Application to the Transportation SectorDan Loughlin, US EPAGary Goldstein, IRG, Ltd.International Energy Workshop, Paris, France, June 24th, 2004

  2. Presentation Outline • Context • US EPA’s Global Change Air Quality Assessment • MARKAL and Simulation • Simulation versus Optimization • Can MARKAL be used for simulation? • Modeling to Generate Alternatives (MGA) • Theory • Initial application to the transportation sector • Critique of MGA & MARKAL • Next steps

  3. Global Change Air Quality Assessment Assessment Goal For 2050, explore the relationships among: • Meteorological change • Technology change • Economic growth • Emissions growth • Air quality

  4. Technological change Economic Modeling Productivity Vehicle Technologies Economic growth By region/sector Emissions growth Mobile Modeling Emissions Projection Emissions growth By region/sector Base year Emissions Inventories Projected Mobile Emissions factors Regional Temperatures Emissions Processing Scenario Assumptions Regional Temperatures Model-ready emissions Regional Meteorology Global Circulation Air Quality Modeling Global Meteorology Regional Meteorology Ambient AQ Benefits Assessment Monetized benefits Global Change Air Quality Assessment Scenario Assumptions Economic Assessment

  5. Global Change Air Quality Assessment Scenario Assumptions Incorporating Technological Change Energy Model Energy Demands Economic Modeling D Energy Prices Vehicle Technology Penetration Economic growth By region/sector Point/Area Source Emissions Factors Mobile Modeling Emissions Projection Emissions growth By region/sector Base year Emissions Inventories Projected Mobile Emissions factors Regional Temperatures Emissions Processing Scenario Assumptions Regional Temperatures Model-ready emissions Regional Meteorology Global Circulation Air Quality Modeling Global Meteorology Regional Meteorology Ambient AQ Benefits Assessment Monetized benefits Economic Assessment

  6. Global Change Air Quality Assessment Scenario Assumptions Selected models MARKAL Energy Demands REMI or EMPAX D Energy Prices Vehicle Technology Penetration Economic growth By region/sector Point/Area Source Emissions Factors Projected VMT MOBILE6 EGAS Emissions growth By region/sector Base year Emissions Inventories Projected Mobile Emissions factors Regional Temperatures SMOKE Scenario Assumptions Regional Temperatures Model-ready emissions MM5 GISS II’ & PCM CMAQ Global Meteorology Regional Meteorology Ambient AQ BenMAP Monetized benefits REMI or EMPAX

  7. Potential Roles of Technological Modeling How will things turn out? Simulation: • What set of technologies are expected to be adopted, given various policies and the respective cost-effectiveness of technology options? Optimization: • What set of technologies most cost-effectively meets energy demands and emissions constraints? • What policy options most cost-effectively meet energy demands and emissions constraints? What should I do? Q: MARKAL is an optimization model. Can it be used for simulation? A: Yes, but carefully.

  8. One Objective Optimization Problem Two Objective Optimization Problem Objective 2 Objective Objective 1 What the LC Solution Does Not Tell Us • The least cost (LC) assumes optimal behavior by individual consumers and companies • This outcome likely cannot occur due to unmodeled issues, uncertainties, unpredictability of behavior, etc. • The real outcome likely will appear suboptimal

  9. Modeling to Generate Alternatives • Explore sub-optimal solutions • Identify maximally different alternative solutions • Analyze, combine, and compare solutions • Utilize in: • Decision support • Exploratory modeling • Complementing sensitivity analysis

  10. Implementation • Solve the LC formulation • Set total system cost <= (1 + d) * LC • Set the objective to maximize some metric of distance from the LC solution • For example, minimize the occurrence of all technologies from the LC solution • Generate MGA alternative 1 • Repeat iterative, each time maximizing the difference from the LC & all previous MGA solutions • Terminate when a target number of MGA solutions have been reached or where difference is small

  11. Set of Feasible Solutions Objective: Minimize Cost Least Cost Solution Constraints N-dimensional objective space Least Cost Optimization

  12. MGA Alternatives Cost Constraint N-dimensional objective space Modeling to Generate Alternatives (MGA) Least Cost Solution Constraints

  13. Illustrative Example Application • Examining US personal automobile technology penetration through 2030 • Considering: • Conventional vehicles • Advanced gasoline-fueled internal combustion engines • Advanced diesel engines • Hybrid vehicles • Advanced hybrid vehicles • Hydrogen fuel cell vehicles

  14. Least Cost Solution Legend FCV Hyb3x Hyb2x AdvD AdvG ConvD ConvG Note: These are illustrative results only.

  15. Application of MGA • Set total system cost <= 1.01 * LC • Maximize difference in: • Personal vehicle transportation technologies capacity • Hydrogen generation infrastructure • Generate LC and 25 alternatives

  16. MGA Alternatives Legend FCV Hyb3x Hyb2x AdvD AdvG ConvD ConvG Note: These are illustrative results only.

  17. MGA Alternatives Fuel Cell Vehicle (FCV) Penetration • Fraction of scenarios with FCV penetration: 23% • When FCV penetration occurred: Maximum: 53% Minimum: 1% Average: 16% Std. Dev: 19% • Generation of Hydrogen Primarily centralized generation Wide range of possible technologies (e.g., SMR, CG) Typically comes online in 2025 or 2030 Note: These are illustrative results only.

  18. Other Potential Applications • Alternative pathways to achieve a target FCV penetration level • Alternative emissions outcomes • Alternative trading outcomes in an emissions trading program

  19. Critique of Experience Advantages • Easily integrated into MARKAL • Readily applied • Interesting results Challenges and Issues • What should cost relaxation be? • How should results be interpreted? • Is ‘expected value’ result meaningful? • What are appropriate distance metrics? • With overall cost relaxation, small players can behave oddly

  20. Next Steps • Add sector-level cost constraints • Application and testing • Explore questions on previous slide For more information, please contact: Loughlin.Dan@epa.gov 919-541-3928