Why You Shouldn't Wait to Install an Online DGA Platform
Five years ago, dissolved gas analysis (DGA) was like getting a blood test at your annual physical. The transformer expert drew a sample of oil once every three to six months, sent it to the lab for analysis, and then anxiously waited for the results.
Online DGA sensors are changing this paradigm. These sensors, which have become more common over the past few years, take periodic (e.g. hourly) samples and provide real-time results, giving operations teams clear situational awareness at all times.
The need for installing online DGA sensors is becoming more acute due to an aging infrastructure that is accelerating physical processes within the transformers. Furthermore, the entire grid infrastructure is currently working at a higher capacity range than a decade ago due to the constant increases in population and electricity consumption. This results in higher stress on your transformers for longer durations and on a more frequent basis.
The growing use of renewable energy resources in many markets also creates additional stress on transformers. Due to the dynamic nature of demand driven by these resources, transformers are experiencing more volatile demand changes than in the past.
Online DGA Sensors Are Not Enough
Online DGA sensors can help you address these challenges. However, to fully leverage the data these sensors provide and to get a complete picture of transformer health, you need to integrate additional sensors and manage the process using an intelligent and comprehensive online DGA platform.
Standards-based DGA platforms measure each sample in a standalone manner against a pre-defined threshold. The downside of these "snapshots" is that you don't get the historical context needed to identify trends and patterns, as well as deviations from these patterns over time.
In contrast, advanced online DGA platforms analyze real-time and historical data to detect abnormalities even when they occur between thresholds, and enable real-time predictive maintenance.
If you're thinking about upgrading and optimizing your transformer health management, you'll need an online DGA platform with the following analytics capabilities:
- Real-time data aggregation and analysis – Collect data from multiple sensors (e.g., online DGA, temperature, load, acoustic, and RFI sensors), incorporate historical data and perform advanced analytics (described below).
- Abnormality detection - Identify abnormalities even if DGA level doesn't exceed the standard-defined threshold. By collecting samples on an ongoing basis, you can monitor behavior over time and identify suspicious trend changes in direction between thresholds that standards-based analysis cannot detect.
- Prediction – Connect the dots, identify trends and predict situations where DGA levels are likely to cross a threshold in the near future. This enables early warning of potentially dangerous situations that warrant immediate maintenance actions.
- Correlation – Correlate DGA sensor data with temperature, load and other sensors and present them on a single graph to provide a unified picture of what's happening within the transformer. This should include correlations of both real-time and historical data.
- Integration - Integrate information from other types of sensors, such as IR cameras for external heat detection, acoustic sensors for detecting partial discharge, and RFI sensors for detecting insulation deterioration. Connecting to and correlating the data from different types of sensors enables a more complete transformer health assessment.
Using an online DGA platform that combines threshold-based standards with advanced analytics and predictive capabilities can help you avoid unplanned downtime, lower maintenance costs and extend the lifetime of mission-critical transformers.