1. NOAA ESRL Renewable Energy Program Melinda Marquis NOAA Earth System Research Laboratory International Visitor Leadership Program August 30, 2012

2. Context: Climate change and projected energy demands (global and U.S.) Recent integration and optimization studies Wind Forecasting Improvement Project (WFIP) 2 Outline

3.  Wind and solar energy are key to meeting growing energy demands and reducing greenhouse gas emissions.  Integrating more wind and solar energy requires more accurate weather forecasts.  The Wind Forecast Improvement Project is designed to improve forecasts of turbine-height winds. 3 Three Take-Home Messages

4. Climate Change 4 Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007

5. 5 Observed changes in global average temperature, sea level, and NH snow cover

6. 6 Climate Change — Attribution

7. 7 Multi-Model Averages and Assessed Ranges for Surface Warming

8. 8 Projections of Surface Temperatures

9. Projected Energy Demands 9 U.S. Energy Information Administration. Independent Statistics and Analysis.

10. U.S. electrical energy demand projected to increase ~ 40 % in the next 25 years (EIA Annual Energy Outlook 2012). This totals ~ 225 GW of new capacity. Improved weather forecasts are critical to integrating weather-driven renewable energy to allow a significant contribution to this demand. 10 U.S. Energy Demand

11. Projected energy mix and growth by 2035 in U.S. U.S. primary energy consumption quadrillion Btu per year Source: EIA, Annual Energy Outlook 2012 11 Energy Information Administration AEO2012, June 2012 HistoryProjections2010 37% 25% 21% 9% 7% 1% 32% 26% 20% 11% 9% 4% Shares of total U.S. energy Nuclear Oil and other liquids Liquid biofuels Natural gas Coal Renewables (excluding liquid biofuels)

12. 12 Projected Global Energy Demand (2035) International Energy Outlook 2011: http://www.eia.gov/forecasts/ieo/

13. Global energy demand is projected to double from 13 TW in 2001 to 27 TW by 2050, and to triple to 43 TW by 2100. This translates into obtaining 1000 MW (1 GW, the amount produced by an average nuclear or coal power plant) of new energy every single day for the next 40 years. Is this happening? Is this possible? 13 Projected Global Energy Demand (2050) Lewis and Nocera (2006), PNAS, 103: 15729-15735. Hoffert, M.I., et al. (1998) Nature, 395: 881-884.

14. China oil demand scenarios based on Japan or S Korea at similar points of development (Source - Steven Kopits. Douglas-Westwood, energy business consultants) EIA’s Estimates of Developing Countries’ Future Energy Demands Could be Low

15. Recent integration and optimization studies 15

16. 20% wind electricity would require about 300 GW of wind generation Affordable, accessible wind resources available across the nation Cost to integrate wind modest Raw materials available Transmission a challenge 16 20% Wind by 2030 Report

17. Annual installed new capacity 17 The 20% Wind Scenario is not likely to be realized in a business-as-usual future. Achieving this scenario would involve a major national commitment to clean, domestic energy sources with minimal emissions of GHGs and other environmental pollutants.

18. High penetrations of wind generation—providing 20% to 30% of the electric energy requirements of Eastern Interconnection—are technically feasible. New transmission will be required for all the future wind scenarios in the Eastern Interconnection. There are no fundamental technical barriers to integration of 20% wind energy into the grid … The 20% Wind Scenario is not likely to be realized in a business-as- usual future. Achieving this scenario would involve a major national commitment to clean, domestic energy sources with minimal emissions of GHGs and other environmental pollutants. 18 EWITS

19. 19 It is operationally feasible for WestConnect to accommodate 30% wind and 5% solar if: Substantially increase BA cooperation or consolidation, real or virtual Increase use of intra-hour scheduling of generation and interchanges Enable coordinate commitment and economic dispatch of generation over wider regions Use forecasts in operations Increase flexibility of dispatchable generation Commit additional operating reserves as appropriate Implement/expand demand response programs Require wind to provide down reserves WWSIS

20. 20 ESRL Optimization Study

21. 21 NREL Renewable Electricity Futures Study Key Findings: Renewable electricity generation from technologies that are commercially available today, in combination with a more flexible electric system, is more than adequate to supply 80% of total U.S. electricity generation in 2050 while meeting electricity demand on an hourly basis in every region of the country. Increased electric system flexibility, needed to enable electricity supply-demand balance with high levels of renewable generation, can come from a portfolio of supply- and demand-side options, including flexible conventional generation, grid storage, new transmission, more responsive loads, and changes in power system operations. The abundance and diversity of U.S. renewable energy resources can support multiple combinations of renewable technologies that result in deep reductions in electric sector greenhouse gas emissions and water use. The direct incremental cost associated with high renewable generation is comparable to published cost estimates of other clean energy scenarios. Improvement in the cost and performance of renewable technologies is the most impactful lever for reducing this incremental cost.

22. 22 NREL RE Futures Study

23. 23 Carbon Dioxide Emissions

24. 24 2010

25. 25 2050

26. Improve short-range forecasts (0-6 h) of wind speed, direction, and turbulence at wind turbine hub-height. Deploy a regional network of upper- air remote sensing observations Combine this network with industry provided tall-tower and wind turbine nacelle meteorological observations Assimilate this data into NOAA’s developmental High Resolution Rapid Refresh (HRRR) NWP model Demonstrate that the improved forecasts can reduce the cost of wind energy and make renewable energy profitable 26 Wind Forecast Improvement Project 915 MHz radar profiler 0.1-4km Surface Flux 10m 449 MHz ¼ scale radar profiler 0.2-8km Sodar 40-200m Lidar 40-200m Tower 50-80m Nacelle anemometers 85m

27. Preliminary Results from WFIP Jim Wilczak NOAA

28. Northern Study Area 9 profilers 5 sodars 1 lidar

29. Southern Study Area 3 profilers 7 sodars

30. 13km Rapid Refresh domain Current RUC CONUS domain 3km HRRR domain RUC – older oper model - 13km Rapid Refresh (RR) – new WRF-based oper model in May 2012 - 13 km HRRR - Hi-Res Rapid Refresh - Experimental 3km - 15h fcst updated every hour - Initialized from RUC/RR All models re-initialized and run every hour, run to at least 15 hs, 3D var data assimilation Hourly Updated NOAA NWP Models

31. 31 Hourly observations (stations for raobs/profiles) # obs N.Amer Rawinsonde (T,V,RH)120 Profiler – NOAA Network (V)21 Profiler – 915 MHz (V, Tv)25 Radar – VAD (V)125 Radar reflectivity - CONUS2km Lightning (proxy reflectivity)NLDN Aircraft (V,T)2-15K Aircraft - WVSS (RH)0-800 Aircraft – TAMDAR (V,T,RH)0-50 Surface/METAR (T,Td,V,ps,cloud, vis, wx) 2200- 2500 Buoys/ships (V, ps)200-400 Mesonet (T, Td, V, ps)4500 GOES AMVs (V)2000- 4000 AMSU/HIRS radiancesUsed GOES cloud-top pressure/temp13km WindSat scatterometer2-10K RUC/Rapid Refresh Hourly assimilation cycle Cycle hydrometeors Cycle soil temp., moisture, snow 1-hr fcst 1-hr fcst 1-hr fcst 11 12 13 Time (UTC) Analysis Fields 3DVAR Obs 3DVAR Obs Back- ground Fields

32. Sodars RR with WFIP assimilation RR no WFIP assimilation

33. OPERATIONAL (NWS)RESEARCH (ESRL) HRRR (w/ assimilation of WFIP obs) Rapid Refresh (RR)RR (w/ assimilation of WFIP obs) Rapid Update Cycle (RUC)RUC (w/ assimilation of WFIP obs)  Same grids, same dynamical core, same physical parameterizations  Different computers, minor differences in implementation Model comparisons Exercise of opportunity – models are similar but not identical. Not ideal! Data Denial Experiment for 30-40 days at end of field program (Model improvement)

34. Impact of data on models: Vertically averaged radar wind profiler vector wind RMSE, with and without WFIP special data, RR and RUC models With WFIP data N Study Area, 9 profiler average, 500-2000m No WFIP data The lower the RMSE, the better! The scores are better (lower) when the WFIP observations are used (red). Early study results show success.

35. Model evaluation using tall tower observations RMSE % Improvement Vector wind Combined Model Obs North Domain South Domain Combined Model Obs North Domain South Domain

36. AWST Truepower Analysis MAE Power Improvement, October 2011 Southern Study Area

37. Preliminary Economic Results—Southern Region Analyses performed for “shoulder” month – October 2011 when load is low and wind speeds are higher Operational Cost Savings are dependent on natural gas prices – average actual price of 3.44 $/MMBtu used for October in Texas Preliminary results show both environmental and cost benefits as a result of improved forecasts

38. WFIP Preliminary Findings Next Steps Run data denial simulations using identical models Develop metrics for ramp events 12% - 5% reduction in vector wind RMSE for forecast hours 1- 6 for combined effect of new observations and new model. Preliminary estimates are that approximately 20% - 60% of this improvement is due to new observations, depending on forecast hour and location. Significant economic and environmental benefits would have occurred with the new forecasts

39. 39 Back to where we started … Photos courtesy of New York Times.

40.  Wind and solar energy are key to meeting growing energy demands and reducing greenhouse gas emissions.  Integrating more wind and solar energy requires more accurate weather forecasts.  The Wind Forecast Improvement Project is designed to improve forecasts of turbine-height winds. 40 Conclusions

41. 41

42. 42

43. Back-Up Slides 43

44. 44 U.S. Electricity Net Generation National Renewable Energy Laboratory Innovation for Our Energy Future Net generation for 2006 = 3814 TWhr UCb Source: EIA Annual Energy Review 2007, AEO 2008 2.4% Next several slides are courtesy of Dr. Chuck Kutscher of the DOE National Renewable Energy Lab

45. Key results from the AEO2012 Reference case, which assumes current laws remain unchanged Projected growth of energy use slows over the projection period reflecting an extended economic recovery and increasing energy efficiency in end-use applications Domestic crude oil production increases, reaching levels not experienced since 1994 by 2020 With modest economic growth, increased efficiency, growing domestic production, and continued adoption of nonpetroleum liquids, net petroleum imports make up a smaller share of total liquids consumption 45 Energy Information Administration AEO2012, June 2012 Projected U.S. Energy Demand (2035)

46. 46 U.S. Energy Information Administration, Annual Energy Outlook 2012: http://www.eia.gov/forecasts/aeo/chapter_executive_summary.cfm

47. 47 Background: Population http://esa.un.org/unpd/wpp2008/peps_documents.htm

48.  World population projected to reach 7 billion in 2011 and surpass 9 billion by 2050.  Most growth will be in developing countries.  Population of less developed regions is projected to rise from 5.6 billon in 2009 to 7.9 billion in 2050. 48 Background: Population

49. 49 Background: Projected Energy Demands EIA International Energy Outlook 2010

50. N = global population GDP/N = globally averaged gross domestic product (GDP) per capita 50 /GDP = globally averaged energy intensity Even assuming a decrease in energy intensity, the rate of world energy consumption is projected to double from 13.5 TW in 2001 to 27 TW by 2050 and to triple to 43 TW by 2100. Lewis and Nocera (2006), PNAS, 103: 15729-15735. Hoffert, M.I., et al. (1998) Nature, 395: 881-884. Background: Projected Energy Demands

51. 51 From Lewis and Nocera (2006), PNAS, 103 (43): 15729-15735. Background: Projected Energy Demands

52. IN ME MA Geographic Location of Selected Applications 52 AWS Truepower, LLC, NY WindLogics, Inc., MN

53. What is included (and excluded) in developing EIA’s “Reference case” projections? Generally assumes current laws and regulations excludes potential future laws and regulations (e.g., proposed greenhouse gas legislation and proposed fuel economy standards are not included) provisions generally sunset as specified in law (e.g., renewable tax credits expire) Some grey areas adds a premium to the capital cost of CO 2 -intensive technologies to reflect current market behavior regarding possible future policies to mitigate greenhouse gas emissions assumes implementation of existing regulations that enable the building of new energy infrastructure and resource extraction Includes technologies that are commercial or reasonably expected to become commercial over next decade or so includes projected technology cost and efficiency improvements, as well as cost reductions linked to cumulative deployment levels does not assume revolutionary or breakthrough technologies 53 Energy Information Administration AEO2012, June 2012

54. Major changes in the final AEO2012 Reference case from the early release Incorporation of Mercury and Air Toxics Standards (MATS) issued by EPA in December, 2011 Updated historical data and equations in the transportation sector, based on revised data from the National Highway Traffic Safety Administration (NHTSA) and Federal Highway Administration Revised long-term macroeconomic projection based on an updated long term projection from IHS Global Insight, Inc. New model for cement production in the industrial sector Updated handling of biomass supply 54 Energy Information Administration AEO2012, June 2012

55. Overview of U.S. energy supply and demand 55 Energy Information Administration AEO2012, June 2012

56. Energy and CO 2 per dollar of GDP continue to decline; per-capita energy use also declines index, 2005=1 Source: EIA, Annual Energy Outlook 2012 56 Energy Information Administration AEO2012, June 2012 HistoryProjections2010

57. In the AEO2012 Reference case, energy-related CO 2 emissions never get back to pre-recession levels by 2035 billion metric tons carbon dioxide Source: EIA, Annual Energy Outlook 2012 57 Energy Information Administration AEO2012, June 2012 200520202035 Energy-related CO 2 emissions 6.00 5.43 5.76 % change from 2005- -9.4%-4.0% ProjectionsHistory 20102005

58. In the AEO2012 Reference case, energy-related CO 2 emissions never get back to pre-recession levels by 2035 billion metric tons carbon dioxide Source: EIA, Annual Energy Outlook 2012 58 Energy Information Administration AEO2012, June 2012 2010 ProjectionsHistory Natural gas Coal Petroleum Electric power 2005 20202030 2035 Commercial Transportation Residential Industrial

59. Current U.S. energy consumption is 83% fossil fuels; demand is broadly distributed among the major sectors 2010 total U.S. energy use = 98.0 quadrillion Btu Source: EIA, Annual Energy Review 2010 59 Energy Information Administration AEO2012, June 2012 Primary energy demand by fuelPrimary energy demand by sector Natural gas 25.2% Coal 21.3% Renewable 8.2% Nuclear 8.6% Petroleum 36.7% Electricity – Residential 15.6% Residential and Commercial 11.2% Electricity – Commercial 14.3% Electricity – Industrial 10.4% Industrial 20.4% Transportation 28.1%