北米地域の電力会社による負荷予測手法：ベンチマーキングおよびベストプラクティス

Abstract

The Load Forecasting Benchmarking and Best Practices project identifies today's best practices in short- and longer-term forecasting to provide actionable methods for improving load forecasting accuracy and processes, and reducing costs. We contacted 29 North American electric or combination utilities to gain a thorough understanding of their load forecasting activities. Interviews covered applications of load forecasting, types of data collected, types of software used, methods of analyzing data, accuracy of forecasts, and resources required for load forecasting.

Table of Contents

• Table of Contents
• Energy Insights Opinion
• Executive Summary
• The Main Points
  • Energy and peak forecasting
  • Day-ahead forecasting
  • Revenue forecast
  • Forecasts of economic factors
  • Weather data
  • Resources
  • Business processes
• Introduction
• Overview
  • Scope of electric load forecasting
  • Electric load forecasting clients
  • Table: Study Respondents
  • Figure: Scope of Electric Load Forecasting Activities
  • Figure: Internal Clients for Electric Load Forecasts
  • Figure: Regulatory Requirements
• Electric Energy and Peak Forecasting
  • Forecast structure
    • Energy forecast structure
    • Figure: Unit of Analysis for Energy Forecast
    • Figure: Forecast Level for Energy Forecast
    • Figure: Energy Forecast Cycle
    • Peak forecast structure
    • Figure: Unit of Analysis for Peak Forecasts
    • Figure: Time Horizon for Peak Forecasts
    • Figure: Peak Forecast Cycle
  • Forecasting approach
    • Energy forecasting approaches and software tools
    • Figure: Modeling Approach for Energy Forecasts
    • Peak forecasting approaches and software tools
    • Figure: Modeling Approach for Peak Forecast
    • Figure: Peak Forecasting Software Tools
    • Forecasting load factor
  • Forecast adjustments
    • Conservation and DSM programs
    • Figure: Utilities That Incorporate Energy Conservation or DSM Programs into Forecast
    • Self-generation
    • Figure: Utilities That Incorporate Self Generation into Forecast
  • Evaluating the forecasts
  • Figure: Accuracy Target for Energy Forecasts
    • Peak
  • Forecast scenarios and risk assessment
  • Figure: Types of Scenarios for Energy Forecasts
  • Forecast review and approval
    • Analyzing variance
    • Weather-normalizing sales
    • Figure: Frequency of Variance Analysis
    • Figure: Approach for Weather-Normalizing Actual Sales
    • Figure: Time Horizon for Day-Ahead Forecasts
  • Accuracy
  • Forecast scenarios
  • Figure: Accuracy Target for Day-Ahead Forecasts
• Forecasting Revenue
• Figure: Time Horizon for Revenue Forecast
• Forecasting Economic Factors
• Figure: Sources of Economic and Demographic Variables
• Forecasting Switching Rates for Customer Choice
• Weather Data
  • Energy and peak forecast
  • Figure: Sources of Weather Data for Energy and Peak Forecast
  • Figure: Location of Forecasting Department Within Organization
  • Staffing
  • Figure: Number of Load Forecasting Staff by Utility Size
  • Budget
  • Figure: Load Forecasting Budget
  • Consultants
  • Figure: Use of Consultants for Energy and Peak Forecasts
  • Conferences and groups
• Business Practices
  • Communication with other utility departments
  • Hot issues now and then
• Appendix - Load Forecasting Benchmarking and Best Practices Study Topic Guide 2005
  • Introduction
  • Screening
  • A. Utility Overview
  • B. Load Forecasting Scope
    • B1. Energy forecasting
    • B2. Peak demand forecasting
    • B3. Day-ahead or next-day forecasting
    • B4. Forecasting Revenue
    • B5. Forecasting Load Factor
    • B6. Forecasting Economic Factors (including Customer Growth)
    • B7. Forecasting Switching Rates for Customer Choice
    • B8. Summary and discussion of other forecasts
    • C. Weather data and other forecast data
    • D. Other forecasting topics
    • E. Staffing and budget
    • F. Wrap up
    • G. Confidentiality
  • Synopsis