Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Data Poisoning: An Overlooked Threat to Power Grid Resilience

Nora Agah, Javad Mohammadi, Alex Aved, David Ferris, Erika Ardiles Cruz, Philip Morrone

TL;DR

The paper examines how data-driven optimization in power grids is susceptible to adversarial disruptions, with a focus on contrasting poisoning and evasion attacks. It surveys the landscape across evasion and poisoning attacks, clarifying problem formulations and the implications for load forecasting, grid operation, and security protocols. A key insight is that evasion has attracted more attention than poisoning, yet poisoning offers broader threat surfaces and currently fewer robust defenses. The authors advocate red-teaming state-of-the-art data-driven methods and incorporating multi-agent perspectives to strengthen grid resilience in renewables-rich, weather-uncertain environments.

Abstract

As the complexities of Dynamic Data Driven Applications Systems increase, preserving their resilience becomes more challenging. For instance, maintaining power grid resilience is becoming increasingly complicated due to the growing number of stochastic variables (such as renewable outputs) and extreme weather events that add uncertainty to the grid. Current optimization methods have struggled to accommodate this rise in complexity. This has fueled the growing interest in data-driven methods used to operate the grid, leading to more vulnerability to cyberattacks. One such disruption that is commonly discussed is the adversarial disruption, where the intruder attempts to add a small perturbation to input data in order to "manipulate" the system operation. During the last few years, work on adversarial training and disruptions on the power system has gained popularity. In this paper, we will first review these applications, specifically on the most common types of adversarial disruptions: evasion and poisoning disruptions. Through this review, we highlight the gap between poisoning and evasion research when applied to the power grid. This is due to the underlying assumption that model training is secure, leading to evasion disruptions being the primary type of studied disruption. Finally, we will examine the impacts of data poisoning interventions and showcase how they can endanger power grid resilience.

Data Poisoning: An Overlooked Threat to Power Grid Resilience

TL;DR

The paper examines how data-driven optimization in power grids is susceptible to adversarial disruptions, with a focus on contrasting poisoning and evasion attacks. It surveys the landscape across evasion and poisoning attacks, clarifying problem formulations and the implications for load forecasting, grid operation, and security protocols. A key insight is that evasion has attracted more attention than poisoning, yet poisoning offers broader threat surfaces and currently fewer robust defenses. The authors advocate red-teaming state-of-the-art data-driven methods and incorporating multi-agent perspectives to strengthen grid resilience in renewables-rich, weather-uncertain environments.

Abstract

As the complexities of Dynamic Data Driven Applications Systems increase, preserving their resilience becomes more challenging. For instance, maintaining power grid resilience is becoming increasingly complicated due to the growing number of stochastic variables (such as renewable outputs) and extreme weather events that add uncertainty to the grid. Current optimization methods have struggled to accommodate this rise in complexity. This has fueled the growing interest in data-driven methods used to operate the grid, leading to more vulnerability to cyberattacks. One such disruption that is commonly discussed is the adversarial disruption, where the intruder attempts to add a small perturbation to input data in order to "manipulate" the system operation. During the last few years, work on adversarial training and disruptions on the power system has gained popularity. In this paper, we will first review these applications, specifically on the most common types of adversarial disruptions: evasion and poisoning disruptions. Through this review, we highlight the gap between poisoning and evasion research when applied to the power grid. This is due to the underlying assumption that model training is secure, leading to evasion disruptions being the primary type of studied disruption. Finally, we will examine the impacts of data poisoning interventions and showcase how they can endanger power grid resilience.
Paper Structure (13 sections, 2 figures)

This paper contains 13 sections, 2 figures.

Table of Contents

  1. Introduction
  2. Problem Formulation
  3. Evasion Literature
  4. Targeting Electricity Consumption and Production
  5. Adversarial Interventions in Grid Operation
  6. Misleading the Detection Algorithms
  7. Poisoning Literature and Shortcomings
  8. Poisoning Production and Consumption Predictions
  9. Poisoning Grid Operation Tools
  10. Poisoning Power System Security Protocols
  11. Conclusion and Future works
  12. Acknowledgments.
  13. Disclosure of Interests.

Figures (2)

  • Figure 1: The main difference between poisoning and evasion lies in where in the process the disruption occurs. This image is adapted from Koffas_2022.
  • Figure 2: There are many components involved in running a power system safely, especially with more cybersecurity becoming necessary. This image is adapted from Human2014Rege.