[Paper Review] Development and Validation of a Scalable Fast Load Shedding Technique for Industrial Power Systems

Andrea Petriccioli, Samuele Grillo+2 more|arXiv (Cornell University)|Sep 10, 2021

Islanding Detection in Power Systems14 references3 citations

TL;DR

This paper presents a scalable, dynamic model-based fast load shedding (FLS) algorithm for industrial power systems, designed to enhance stability and reliability in offshore platforms. The method computes an optimized shedding matrix via power mismatch analysis and priority-based load selection, validated through dynamic simulations and implemented in IEC 61131-3 Structured Text for deployment on ABB AC500 PLCs, ensuring compatibility and scalability across industrial plants.

ABSTRACT

The work aims to improve the existing fast load shedding algorithm for industrial power system to increase performance, reliability, and scalability for future expansions. The paper illustrates the development of a scalable algorithm to compute the shedding matrix, and the test performed on a model of the electric grid of an offshore platform. From this model it is possible to study the impact on the transients of various parameters, such as spinning reserve and delay time. Subsequently, the code is converted into Structured Text and implemented on an ABB PLC. The scalability of the load shedding algorithm is thus verified, confirming its performance with respect to the computation of the shedding matrix and the usefulness of the dynamic simulations during the design phase of the plant.

Motivation & Objective

• To develop a scalable, topology-adaptive fast load shedding (FLS) algorithm for industrial power systems to improve reliability and performance.
• To enable dynamic simulation-based optimization of key parameters such as spinning reserve and delay time to ensure system stability.
• To ensure compatibility with existing industrial control systems by implementing the algorithm in Structured Text (ST) compliant with IEC 61131-3.
• To validate the algorithm’s effectiveness and scalability through a real-world offshore platform model and PLC testing.
• To minimize load shedding while maintaining frequency stability, reducing operational disruption during grid disturbances.

Proposed method

• The FLS algorithm is structured into two main components: Load Selection (LSE) and Event Detection/Slashing Action (ED-SA).
• LSE computes the power mismatch (PM) for each event by analyzing grid topology and generator status, ensuring sub-network stability.
• LSE selects loads to shed based on priority and power consumption, minimizing the number of switches while ensuring PM is exceeded.
• The output is a shedding matrix (SM), where 1 indicates a load to be shed and 0 indicates retention, updated every 0.5–2 seconds.
• Dynamic simulations are used to model transient behavior under various spinning reserve (SR) and delay time conditions.
• The algorithm is converted into IEC 61131-3 compliant Structured Text using automatic code generation and deployed on an ABB AC500 PLC for hardware validation.

Experimental results

Research questions

• RQ1 How can a fast load shedding algorithm be made scalable across diverse industrial power system topologies?
• RQ2 What is the impact of spinning reserve and delay time on frequency nadir during system disturbances?
• RQ3 How can dynamic simulation models be used to optimize FLS parameters without compromising safety or operational continuity?
• RQ4 To what extent can a model-based approach reduce the amount of load shedding while maintaining system stability?
• RQ5 Can the FLS algorithm be effectively implemented on standard industrial PLCs using standardized programming languages like Structured Text?

Key findings

• The proposed FLS algorithm achieves high scalability by dynamically adapting to grid topology and event conditions using only field inputs and design parameters.
• Dynamic simulations revealed that increasing spinning reserve from 3.1 MW to 8.4 MW allows a safety margin of 1 Hz to 0.5 Hz on the underfrequency threshold, respectively.
• Delay time has a critical impact on frequency nadir: beyond 200 ms, the nadir drops sharply, indicating that minimizing delay is essential for stability.
• A spinning reserve of 6 MW ensures a 0.5 Hz safety margin above the 48 Hz underfrequency threshold, balancing operational continuity and safety.
• The algorithm was successfully implemented in Structured Text and tested on an ABB AC500 PLC, confirming compatibility and real-time performance.
• The model-based optimization method enables precise selection of SR and delay time, minimizing unnecessary load shedding while meeting underfrequency protection requirements.