New Model Forecasts Renewable Growth for Distributed Energy Resources
In the last decade, the rapid proliferation of distributed energy resources (DERs), including rooftop solar panels, energy storage systems, and electric vehicles, has transformed the modern energy landscape. In a new study, researchers propose a model to forecast renewable energy growth at both the circuit and substation levels.
The study, by researchers at Carnegie Mellon University, appears as a working paper. The paper was selected as a finalist for the Best Applied Paper Award at the 2025 Institute for Operations Research and the Management Sciences (INFORMS) Data Mining and Decision Analytics Workshop. A shorter version of the paper received the Best Paper Award at the Institute of Electrical and Electronics Engineers (IEEE) Power and Energy Society General Meeting.
“The growth of DERs presents both opportunities and operational challenges for the management of electric grids,” explains Shixiang (Woody) Zhu, assistant professor of data analytics at Carnegie Mellon’s Heinz College, who coauthored the study. “Accurately predicting who will adopt DERs is critical for planning infrastructure, but the inherent uncertainty and spatial disparity of DER growth complicate traditional approaches to forecasting.”
Among the challenges of DERs are that excessive penetration in local areas can strain the distribution infrastructure, leading to fluctuations in voltage, reverse power flows, and damage to substations and transformers. In addition, it is challenging to reliably predict the growth of DERs because adoption patterns are uncertain due to evolving policy incentives, customers’ preferences, the costs of technology, and local socioeconomic factors.
In their study, researchers worked with a utility in Indianapolis, Indiana, to help analyze its future distributed energy resource portfolio, including adoption of solar energy and electric vehicles. The model, which quantifies uncertainty for DER adoption predictions to ensure validity across a hierarchical grid structure, consistently outperformed existing baselines in both predictive accuracy and uncertainty calibration. It will be incorporated into the Indiana utility’s biannual integrated resource plan.
The researchers’ model is primarily suited for short-term forecasts due to its data-driven nature, which limits its ability to anticipate events not reflected in historical data, such as the effects of DER-related policies scheduled for future implementation. This constraint makes long-term predictions challenging, particularly in the nonstationary and policy-sensitive context of DER adoption.
“DERs are no longer peripheral technologies, but are becoming integral components of the electricity grid,” says Wenbin Zhou, a PhD student in machine learning and public policy at Carnegie Mellon’s Heinz College, who led the study. “Our results underscore the importance and feasibility of quantifying structured uncertainty in supporting reliable and resilient grid planning under the evolving landscape of distributed energy adoption.”
###
Summarized from a working paper, Hierarchical Probabilistic Conformal Prediction for Energy Resources Adoption, by Zhou, W (Carnegie Mellon University), and Zhu, S (Carnegie Mellon University. Copyright 2025. All rights reserved.
About Heinz College of Information Systems and Public Policy
The Heinz College of Information Systems and Public Policy is home to two internationally recognized graduate-level institutions at Carnegie Mellon University: the School of Information Systems and Management and the School of Public Policy and Management. This unique colocation combined with its expertise in analytics set Heinz College apart in the areas of cybersecurity, health care, the future of work, smart cities, and arts & entertainment. INFORMS named Heinz College the #1 academic program for Analytics Education. For more information, please visit www.heinz.cmu.edu.