On Statistical Modeling of Load in Systems with High Capacity Distributed Energy Resources

TL;DR

This paper explores the use of generalized mixture models to statistically represent the uncertain and variable load patterns in power systems with high-capacity distributed energy resources like electric vehicles, addressing new challenges in system reliability.

Contribution

It introduces a novel application of generalized mixture models for modeling aggregated loads in systems with high-capacity distributed energy resources.

Findings

Generalized mixture models effectively capture load uncertainties.

Modeling improves system reliability assessments.

Electric vehicle load patterns are highly variable and complex.

Abstract

The emergence of distributed energy resources has led to new challenges in the operation and planning of power networks. Of particular significance is the introduction of a new layer of complexity that manifests in the form of new uncertainties that could severely limit the resiliency and reliability of a modern power system. For example, the increasing adoption of unconventional loads such as plug-in electric vehicles can result in uncertain consumer demand patterns, which are often characterized by random undesirable peaks in energy consumption. In the first half of 2021, the electric vehicle sales increased by nearly 160%, thus accounting for roughly 26% of new sales in the global automotive market. This paper investigates the applicability of generalized mixture models for the statistical representation of aggregated load in systems enhanced with high capacity distributed energy resources such as plug-in electric vehicles.