Intelligent substation monitoring

Intelligent substation monitoring

Special Edition - Substations

Intelligent management of substation assets

Abstract

Transformers are critical and expensive assets in substations that require meticulous monitoring and control to ensure grid integrity. Various tests that are available can determine the condition of transformer by diagnosing the health of active components such as insulating paper and mineral oil, windings, bushings, tap changers, etc. Some of these tests can be time-consuming, expensive, intrusive, and require some level of expertise. They can also lead to accumulation of large volumes of data that may be challenging for the service engineer to process without suitable scientific backing. In this context, artificial intelligence (AI) finds immense popularity among asset holders, researchers and service engineers who are looking to improve transformer performance by efficient and economical means. AI-based methods imitate the human brain in order to process and convert raw data to draw meaningful inferences. They are typically data-driven i.e. they can identify, locate, and eradicate redundancy in input data, thereby reducing the computational burden. When a large fleet containing both new and old transformers is concerned, such tools can be remarkable in prioritizing the grid management strategies and evaluating the cost-effectiveness of various assets. It can also reduce the time needed for knowledge transfer by end-to-end digitization of the utility data and convert it into simple linguistic inference that can assist the service engineer in deciding what are the necessary preventive or corrective actions quickly and efficiently. AI-based tools show tremendous potential in becoming an intelligent alternative for substation asset management despite their challenges. This article is a discussion on such aspects of transformer management in context of AI applications along with its strengths and limitations.

Keywords: transformer management, condition monitoring, artificial intelligence, data-driven modelling, asset maintenance activities

1.  Introduction

Substations are the nodal points in electricity transmission and distribution network containing key assets such as transformers, circuit breakers, switch gears etc. Transformers are particularly critical and expensive assets in substations with high performance and life expectancy. Transformer performance diminishes dramatically with deterioration of its insulation health under various operational stresses and it ultimately affects the integrity and reliability of substation grids.