MTLF Winter Storm Elliott Review & Load Forecasting Analysis
This document provides insights into the MTLF Winter Storm Elliott review conducted by Calvin Opheim, focusing on load forecasting and analysis. It includes information on MTLF models, software upgrades, corrective action plans, and independent forecasting systems used for load forecasting. The document also details the various weather forecast options available for the Itron models and important notes on specific MTLF models like E, A3, A6, and M. Additionally, it highlights the reliability of different forecasting models and their respective backup strategies.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
MTLF Winter Storm Elliott Review Calvin Opheim Load Forecasting & Analysis OWG 2/16/2023
Agenda MTLF Model Review Software Upgrade Winter Storm Elliott Corrective Action Plan 2 PUBLIC
MTLF Models Two Independent Forecasting Systems Itron (creates the E, E1, E2, and E3 forecasts) GE (creates the A3, A6, and M forecasts) Residing on separate computer systems provides redundancy in case one system is unavailable 4 PUBLIC
MTLF Models E, E1, E2, E3 Based on neural network and linear regression models with a separate model for each hour 49 weather stations (expanded set) Can easily switch between any available weather forecast provider 5 PUBLIC
MTLF Models A3 Notes Uses recent history to update the model coefficients Limited to a single weather forecast provider Requires coding changes to switch to alternative weather forecast Provides a reasonable backup to the Itron models during the normal range of weather conditions 6 PUBLIC
MTLF Models A6 Notes Uses recent history to update the model coefficients Only uses one weather station per weather zone Limited to a single weather forecast provider Requires coding changes to switch to alternative weather forecast Provides a reasonable backup to the Itron models during the normal range of weather conditions 7 PUBLIC
MTLF Models M Notes The weighted average of the 6 other models Historically has shown to be the most accurate Day Ahead forecast Provides a good benchmark for other models For more details https://www.ercot.com/calendar/04042022-SAWG-Meeting- by-Webex https://www.ercot.com/calendar/06242022-SAWG-Meeting- by-Webex 8 PUBLIC
MTLF Available Weather Forecasts The following weather forecasts are available for the Itron models: European Model (Euro) Global Forecast System (GFS) Global Forecast System Ensemble (GENS) North American Model (NAM) 3 vendor models ERCOT was the first ISO that incorporated raw weather forecasts into their operational load forecasts ERCOT is actively working on increasing the number of raw weather forecasts to those mentioned above Particularly of interest are the Canadian and HRRR models 9 PUBLIC
Software Upgrade and Impacts A software upgrade for Itron went live on 12/6/2022 A change in how holidays are accounted for was not uncovered during testing Resulted in holidays being significantly under-forecasted Rendered E, E1, E2, and E3 unusable for most of the week All enhancements/improvements previously implemented in the Itron created MTLFs were unavailable because these forecasts were unusable Ability to select from any available raw weather model (like Euro, GFS, NAM) Ability to apply error correction Extreme weather forecasting improvements 11 PUBLIC
Software Upgrade and Impacts ERCOT Forecasts for 12/23 created 12/17 80,000 70,000 60,000 50,000 MW 40,000 30,000 20,000 10,000 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour A3 A6 E E1 E2 E3 Actual 12 PUBLIC
Weather Forecasts DFW Forecasts for 12/23 @0800 30 25 20 Deg F 15 10 5 0 Dec-14 Dec-15 Dec-16 Dec-17 Dec-18 Forecast Date Dec-19 Dec-20 Dec-21 Dec-22 Dec-23 NAM GENS MDA Actual NAM was the preferred forecast but was unavailable to the A3 and A6 forecasts 14 PUBLIC
Weather Forecasts IAM Forecasts for 12/23 @0800 35 30 25 20 Deg F 15 10 5 0 Dec-14 Dec-15 Dec-16 Dec-17 Dec-18 Forecast Date Dec-19 Dec-20 Dec-21 Dec-22 Dec-23 NAM GENS MDA Actual NAM was the preferred forecast but was unavailable to the A3 and A6 forecasts 15 PUBLIC
Weather Forecasts AUS Forecasts for 12/23 @0800 30 25 20 Deg F 15 10 5 0 Dec-14 Dec-15 Dec-16 Dec-17 Dec-18 Forecast Date Dec-19 Dec-20 Dec-21 Dec-22 Dec-23 NAM GENS MDA Actual NAM was the preferred forecast but was unavailable to the A3 and A6 forecasts 16 PUBLIC
Load Forecasts Selected Forecast Peak Demand for 12/23 @0900 76,000 73,910 74,000 72,000 70,898 70,000 69,324 69,313 MW A3 67,943 67,574 68,000 Actual 67,238 66,000 64,000 62,000 12/17/2022 12/18/2022 12/19/2022 12/20/2022 Forecast Date 12/21/2022 12/22/2022 12/23/2022 Forecasts were created at 8:02 am each day 17 PUBLIC
Forecast Error Sources During Elliott Using the A3 model resulted in an increase in forecast errors Ramifications of using A3 Based on MDA s weather forecast instead of the weather forecast (NAM) recommended by our meteorologists Seeing that the NAM forecast was better than MDA s forecast in the days leading up to 12/23, the result was higher forecast errors than if the Itron models had been operational Limited features/flexibility as compared to the Itron models Limited historical data on cold weather events Uri data after 2/14/21 was of no value due to system outages Weather forecast error Load forecasting model error 18 PUBLIC
ERCOT Load-Weighted Temperatures 19 PUBLIC
Corrective Action Plan
Corrective Actions Currently Planned Objective 1 Reduce forecast errors from the alternative load forecasting source (currently the GE models A3, A6) Add an external load forecast that will complement ERCOT s Itron models Must have the ability to seamlessly switch weather forecasts Objective 2 Develop an innovative process for developing extreme cold weather forecasts Develop a process for creating specialized extreme cold weather models Based on a robust set of historical cold weather data 21 PUBLIC
Corrective Actions Currently Planned Objective 3 Reduce weather forecast error Add more raw weather forecasting models If the Canadian forecast was available, the day(s) ahead forecast would have been better If the HRRR forecast was available, the intraday forecast would have been better Objective 4 Improve Itron software upgrade testing process Create a more robust testing environment Update testing procedures to include holidays and other outlier phenomena 22 PUBLIC
Questions? 23 PUBLIC
