Don’t Ask me what I want, ask me what I do: the key to valid requirements documentation

1. Don’t Ask me what I want, ask me what I do: the key to valid requirements documentation Mary Altalo Ocean.US

2. Understanding How Climate, Wx, Water Information is used- Getting to Requirements • Information in Federal Resource Management Decisions • Setting codes and standards- building safety • Management of natural resources-water, forests • Regulatory oversight of social systems: energy, transport, water, health • Information in International Trade/Aid Decisions • Treaty and Trade Agreements –Canada Northern Trade Route • Aid priorities and planning- Zimbabwe Power, resettlement issues • in Industry Operations and Planning Decisions • Tactical Operations • Strategic Planning • From Mitigation to Adaptation • From Way of Life (practice) to Business Process Reengineering (new tools and practices)

3. Diagnostic Approach to Assessing Vulnerability and Risk

4. Driving Principles for Managing with Environmental Information • Regulatory • Decision Accountability/Shareholder Value • Safety of Life and property • Market Economics & Competitive Advantage • Risk reduction • Reliability, Efficiency, Sustainability • Corporate Social Responsibility- Indices

5. Critical forecast periods Sub day, 2-4 day, 90 day

7. Power Industry Decision Aid Architecture Weather Decision Support Analysis Tool Set for Power Management Strategic Tactical Planning Planning Infrastructure Revenue Load Balancing Power Pricing Maintenance Planning , Design Forecasting / & siting Stock Pricing Market Demand Equipment Readiness Market Demand Market Demand Regulatory Fuel Type Grid Availability Fuel Type Constraints Emergency Response Fuel Availability Fuel Availability Generation Dispatch Fuel Price Weather Impacts Commitment Management Fuel Price Regulatory Regulatory Constraints Constraints Market Demand Market Demand Transmission Price Options Constraints Available Generation Infrastructure Environmental Maintenance Constraints Availability Congestion Tariff Calling

8. Environmental Information “Flow” on the Operational Decision Process: Risk Reduction Areas INFORMATION KNOWLEDGE ACTION OUTCOMES IMPACTS DATA Met-Ocean Pre- Processing Met-ocean Information Integration Met-ocean Decision Support Met-ocean Information Evaluation Met-ocean Data Collection • scheduling and demand balancing • asset management and replacement • enterprise wide contingency and financial planning • demand reduction and price responsive loads • environmental dispatch • congestion management Met-Ocean Data Formatting Outcomes Demand Forecasting Models Operational Decisions NMHS Other Providers: GODAE team Present Value-Added Service Providers • Improved profit • Increased Efficiency • Improved reliability • Increased safety • Decreased Liability • Decreased Risk • Decreased Exposure Short Mid Long term Met-ocean Data Analysis and IT Services: Quantify, source, cost and reduce weather data error Load Model Error Analysis: Improve Software and Support Decision Analysis, Dependencies and Support Tools Economic/ Performance Valuation of Weather/ocean Error Impacts Situational Awareness----Decision Support----Optimal response

9. U.S. Coastal Utilities “Trial” Areas

10. Key Utility functions/decisions requiring coastal weather/climate/ocean data • Load Balancing-single utility and grid • Generation commitment- fuel mix choice (fossil fuel, hydro, wind) • Dispatch scheduling • Power Marketing • Cash trading • Power pricing • Fuel pricing and procurement • Tariff Scheduling • Natural Gas Storage Management • Revenue Projections • Infrastructure siting • Management strategic planning

11. Electricity Demand Model Error -Neural Net Diagnostics Weather Forecast Model- AVN, MRF, etc or ensemble Power Demand Forecast Model- AANSLFF, RER Metrix, etc. or ensemble Skill of the Environmental Forecast Impacts the Skill of the Power demand forecast

12. Most utilities calculate weather error in MW as well as percentage of variance of the load. Analysis indicates that on some days, variance in the load forecast in MW may be solely due to weather error. This appears to be from events or unmodeled mesoscale features such as back door fronts, sea breeze and afternoon thunderstorms. The cost of such events can be up to $10M/day in wasted generation Urban Utility Case Study Findings 1: Significant load error due to weather

13. Bring in new data and reconfigure and train load model to use it optimally if needed—IT support services Wind Resource Map Produced by SiteWind

14. Key Cost Findings2002-2003 Northeast Energy “The project estimated that the benefits of improving day ahead weather forecast accuracy by one degree F or by reducing forecasting error by 50% for days 2-7 is: • --$20-25 million per year for a regional transmission authority • --$1-2 million/year for a major distribution utility. Optimal use of weather information could yield savings of $8–18 million/year for a major university system (electric and natural gas). If these savings were generalized to other regional transmission organizations, large statewide colleges and universities and regional transmission authorities the total savings would be for the Northeast Region: -- $100-140 million/year for ISO’s --$30-60 million for regional electric distribution companies. -- $38-67 million for Statewide university campuses Furthermore, capturing the “events” on top of this will yield significantly higher savings (millions/day).- seabreeze, backdoor fronts, afternoon showers

15. West Coast Power • Load balancing Problems • Coastal winds • Coastal fog • Microclimates • El Nino

16. Consequences of Electrical Load Error • Underforecast- May • Generation shortfall of 5000MW (load required by 500,000 homes )- near blackout • Buy on spot $200+ with caps up to 1000 w/o caps • One day 5-7% error costs of $4-7M/day • Cause- ocean-related weather forecast error of 4% • Overforecast- September • Generation overcommittment • Similar magnitude • Less costly (previously contracted generation) and less visible impact

17. Retrospective Analysis:Relationship of Coastal Weather Uncertainty and Cost

18. Cal ISO Mean Daily Forecast Error • Delta breeze and weather/load forecast errors contribute to major errors in prediction of Delta Breeze effects. • Delta breeze is defined as the conditions when the wind speed is > 12 knots and the direction is between 190 degrees and 280 degrees. • Delta Breeze can change load by 500MW • Direct Costs: 250k per breeze day; 40 events per year • An overforecast problem Delta Breeze events

19. Principle Causes of Uncertainty on Energy Operations and Planning • Uncaptured WINDEvents • Coastal Delta Breeze- Cal ISO • Lake effects- Salt Lake City- Pacificorp, Great Lakes- SUNY Buffalo • Seabreeze- NE ISO • Coastal Frontal passage- 2-4 day • Uncaptured PRECIPITATION Events • Rain vs. snow/ice • Regional day ahead error in precipitation- Pacificorp • Afternoon thunderstorms • Marine Layer, fog- SDG&E • Drought and flood, flash flood • Uncaptured CLIMATE Events • Climate outlooks –weather events frequency • El Nino and seasonal events • Decadal ocsillations- NAO • RESOLUTION- Spatial, temporal • Sub grid level • Targeted watershed level, Nodal, congestion and population • Topographic Effects- microzones • Hourly changes during events • Load Model Error • 50% load error at certain event periods • Can’t incorporate probabilities/ ensembles • Sub-optimal Use

20. Requirements

21. Activating the Information Link to an Engineering Requirement “Codifying”

22. Turning a “Parameter” into a “Factor” in an Engineering EquationThe key to “activating” observing system Information Case study 1: Activating rain and wind data in reservoir management. Improved precipitation, and winds feed into actions for runoff conservation, reservoir management, water quality • The Revised Universal Soil Loss Equation (RUSLE) due to rainfall A=RKLSCP • soil erosion the equation due to Wind E = f (I K C L V) Better ocean observations and Ocean-atmospheric models lead to better precipitation forecasts R rainfall erosion index, which includes the amount as well as the “force” of the precipitation Better ocean observations and models lead to better wind forecasts Climatic factor C determined by wind velocity and soil surface moisture

23. Case Study Two: Wave, sea level, storm surge, wave height and Beach Erosion and sediment transport 4 Affected Hydrodynamic Processes which are measures by GOOS Storm Surge Tidal Ranges and Currents Waves With rising sea level, there is increased impact of coastal storms. Improved wave prediction as well as the “inundation” parameters are required in the mitigation strategies of beach nourishment and armament Activating Inundation and Sea Level Information The relationship between sea level rise and waveheight (Dean,1986) Relationship between wave height and beach erosion Dean (1986)

24. Link to Decision Support Tools and Management Scenarios

25. Linking Forecast Simulation Tools with Emergency Response Simulation Tools can aid in Severe Weather Emergency Energy Management Expert “Grid” Management Situational Awareness and Power Restoration Management Tool Emergency preparedness with “CATS” (consequence assessment tool set) Locate critical energy assets, estimate damage and position for relief Storm Tracking with simulation tool- predict hurricane landfall Data-Information Knowledge Action and Outcomes

26. Economic Valuations for Risk Reduction

27. The Value of a Seasonal Seabreeze Forecast to Energy Demand Forecasting 4. Using ensemble forecasting techniques can enhance information content of the forecast (probabilities) adding “spread” around forecast 1. Major sea breeze events (May-September) are not captured adequately in the day ahead temperature forecast 2. Major Sea Breeze Events Cause Significant Electricity Demand Error as power demand drops. Savings of up to 2 million per year for increased accuracy of forecast 3. Bringing in more offshore observation data can enhance forecast accuracy 5. Improving power management for national grid operators

28. Ocean Obs/ Models/Thorpex type experiments enter here T &Ppt forecast Prediction of Power and Water Demand from Irrigation Load Power Purchasing on contract vs spot Grid stability and Cost savings (.25M/yr for I utility) Soil moisture Crop type Planting time Value of Improved Seasonal Precipitation Forecasts to Agriculture Irrigation Pump Load Forecasting and Power Grid Stability 1. Timing and total irrigation time varies from season to season depending on soil moisture causing unpredicted draw on electricity grid due to start of pumps

29. Using ensembles to improve day ahead Weather forecasts for A Southern Power Utilitywith L. Smith of the LSE • A 30% reduction ( real dollar savings of $300,000 ) over two very different days

30. Value of Ensemble Forecasting to Electricity Load Forecast Accuracy 2. Ensembles have more information- Precision & Probability 1. Existing weather error is costly 3. Works with asymmetric cost curve 4. Saves Utility network operator $15M during summer peak period over present method

31. The GOOS Challenge

32. GOOS IS ARRANGED AROUND ITS VALUE STREAM Defining the Program Elements Based on Observations Data Information Knowledge “Know-how” -Maritime Safety -Homeland Security -Natural Hazards -Climate Change -Public Health -Ecosystem management -Marine Resources - Scientific Understanding e.g. NDBC Feedback for EVOLUTION Alpha test Beta test Technology performance metrics Societal performance metrics

33. “Wx/Climate , Water Information Supply Chain”Knowledge Management as well as Data Management Regional Alliance Approach

34. “Activated” Environmental Information The Management Decision Process

35. GOOS Information Products in the EU Market Value Chain GOOS/GEO