看看这个：监管解决方案以实现大型负载的更好预测

Executive Summary - In the U.S., electricity demand has begun to grow after decades of stagnation, with utilities projecting a 20% increase in load from 2023 to 2035, up from previous estimates of 7% in January 2021 [10][25][41] - Accurate load forecasting is crucial for managing affordability and reliability risks, as well as for considering all available investment options [12][13] - The report emphasizes the unique characteristics of new large loads, such as data centers and advanced manufacturing, which should be integrated into modern forecasting processes [13][20] Load Growth and Forecasting - Utilities have revised their five-year peak load forecasts from an expected increase of 23 GW to 128 GW between 2022 and 2024 [25] - The Integrated Resource Plans (IRPs) covering 48% of U.S. electricity sales predict a 20% load growth by 2035, indicating a significant upward revision in forecasts [10][25][27] - The report highlights the need for improved forecasting methods to accommodate the unique characteristics of new loads, which differ significantly from traditional load types [20][25] Regulatory Measures for Improved Forecasting - Regulatory agencies play a critical role in establishing forecasting guidelines, reviewing utility forecasts, and approving investment cost recovery based on these forecasts [21][22] - Recommendations for regulators include enhancing understanding of new load drivers, revising planning guidelines, and coordinating with state and local governments [22][23] - The report suggests that frequent updates to long-term load forecasts and transparent data sharing can improve forecasting accuracy and stakeholder engagement [22][23] Characteristics of New Loads - New large loads, particularly data centers and industrial manufacturing, present unique challenges for load forecasting due to their rapid growth and specific operational characteristics [60][68] - Data centers are expected to account for a significant portion of electricity load in certain states, with projections indicating they could represent 46% of Virginia's load by 2030 [61] - The flexibility potential of different load types varies, with some, like cryptocurrency mining operations, being more price-sensitive and flexible compared to traditional industrial loads [62][63] Best Practices for Load Forecasting - The report outlines best practices for improving load forecasting, including scenario-based forecasting methods and integrating end-use demand forecasts with econometric models [17][18] - Ensuring consistent application of load forecasts across planning processes is essential for informed decision-making [19] - The integration of new large loads into forecasting processes is still in its early stages, necessitating further development of modern forecasting systems [20][21] Conclusion - The report concludes that improving load forecasting can expand the options available to utilities and regulators when addressing load growth [36] - It emphasizes the importance of adapting forecasting practices to account for the rapid changes in load characteristics and the associated risks [36][37]