Recent forecasts indicate a period of high electricity growth in PJM and across the United States, highlighting the need for electric forecasting entities to evolve their methods to incorporate major changes in electricity use over the coming decades. While many forecasters have made progress in updating their forecasting methods, significant improvements are still urgently required. The high potential cost of under-forecasting, driven by lagging infrastructure and evidence of a significant upturn in the next few years, underscores the importance of proactive electric system planning. Recent industry conversations have also addressed the risks of over-forecasting, particularly due to data centers queuing shopping or being double counted.
Key considerations for improving electric forecasts include:
- Explicitly evaluating expansion optionality in all generation and transmission plans to address forecasting risks.
- Optimizing the usage of existing infrastructure through the adoption of grid-enhancing technologies.
- Incorporating the full effects of demand-side resources, including flexibility, and non-wire alternatives.
Electric forecasts are changing as forecasters from various organizations (utilities, RTO/ISOs) adapt their methods. These efforts vary in improvement techniques and maturity, with some entities already incorporating new demand drivers, resulting in substantial changes in load forecasts. However, only a few entities include forecasts of these drivers at the distribution level.
The costs of over- and under-forecasting are significant. Traditional practice involves achieving forecasts believed to lie in the center of probable outcomes, identifying uncertainty bands, and weighing probabilities of over-/under-forecasting with associated costs. The current environment introduces additional costs for under-forecasting due to long lead times for new generation projects and extended transmission project timelines.
Process standardization can provide greater certainty by reducing uncertainty and accelerating the interconnection process through a transparent intake process that filters out less serious customers. Year-over-year changes in energy prices do not follow a linear trend and can be difficult to predict.
T. Bruce Tsuchida, a Principal at The Brattle Group, specializes in bridging technology, economics, and regulatory policy, with expertise in assessing the impact of new technologies and regulatory changes on system operations, utility business, and valuations of transmission, generation, and distribution assets.
