RESEARCH
Embedded Files
Flexibility Resources
Flexibility Resources
The traditional power system was designed for unidirectional power flow, where centralized generators supplied electricity to consumers through transmission and distribution networks. In this structure, flexibility was largely confined to large-scale generators participating in centralized markets.
Today, the growing integration of small-scale Distributed Energy Resources (DERs)—such as photovoltaic systems, electric vehicles, and energy storage—within distribution networks has created new opportunities for localized flexibility. These distributed flexibility resources can provide valuable services such as voltage regulation, peak shaving, and frequency support. However, their variable and uncertain nature introduces operational challenges, including congestion, overvoltage, and transformer overloading.
To effectively harness these flexible assets, the establishment of a Distribution System Operator (DSO) framework has become essential. DSOs can utilize dynamic operating envelopes (DOEs) to coordinate distributed flexibility in real time, ensuring secure, reliable, and efficient operation of the distribution grid while enabling active market participation from end users.
Machine Learning for Power Distribution Systems
Machine Learning for Power Distribution Systems
Traditionally, meters were used only for billing purposes. However, with the advent of Advanced Metering Infrastructure (AMI), vast amounts of data are now being collected. This data, when combined with advanced machine learning techniques, can significantly reduce modeling errors in the power distribution system.
AMI data serves a variety of purposes, including:
Phase Identification: Determining the phase to which each meter is connected.
Transformer - Meter Pairing Identification: Identifying the connections between distribution transformers and customers.
Parameter Estimation: Estimating the type and length of customer service lines.
Topology Identification: Mapping the configuration of the distribution system from feeders to end-users.
These applications demonstrate the transformative potential of AMI in enhancing the accuracy and efficiency of distribution system management.
DR Baseline Estimation
DR Baseline Estimation
Missing data is a common issue in load profile processing for distribution systems, often caused by temporary communication losses. Analysis of real-world smart meter data (e.g., from Pecan Street and utilities like Duke Energy) reveals that ~70% of missing data segments are under 4 hours, making their restoration critical for enhancing data quality. Improved data quality supports downstream tasks like load forecasting, non-intrusive load monitoring, and load pattern analysis.
Additionally, demand response (DR) baseline estimation can be treated as a special case of missing data restoration. For instance, during Conservation Voltage Reduction (CVR) events, utilities reduce substation voltage to lower peak loads. Estimating the "would have been" load baseline during such events involves restoring missing data for the CVR period, based on pre- and post-event load profiles. Accurate baseline estimation is vital for evaluating DR performance and benefits load service providers by enhancing system analysis and planning.
Page updated
Google Sites
Report abuse