The Challenges of Using an Online DGA Platform

The Challenges of Using an Online DGA Platform

Dissolved Gas Analysis (DGA) is one of the most important monitoring tools power utility companies can embrace. As the new realities of power usage are more demanding than ever, ensuring the health of power transformers becomes more challenging. However, DGA on its own doesn't provide an answer to the monitoring needs of the modern power facility. Relying on DGA without addressing its limitations can be as disastrous as not having DGA at all.

The first problem lies with the data that the platform provides. A single power transformer with DGA will provide engineers with a status on demand, but that status is just a snapshot of the moment of testing. There is no context for the test without a case history, which engineers would have to create on their own. The snapshot alone can hardly point to trends that allow predictions and early warnings of transformer catastrophes.

Even more severe, abnormalities that may occur between testing can go undetected. The importance of identifying abnormalities cannot be understated. Abnormalities are an indicator of the transformer's health and without being able to see abnormalities, engineers will not gain a complete understanding of the grid's status.

New York Power Authoroty Case Study

Compounding these issues in the number of transformers in the station; As It's extremely hard to place the data for just one transformer to get a sense of trends and history, this gets more complicated with multiple transformers. In short, DGA sensors are far from supplying an adequate management system for multiple transformers.

The challenges of using an online DGA platform can be resolved by taking the data extraction away from engineers and allow them to focus on reaction and treatment. The data from the platform can be drawn automatically and allow a constant stream of data to become a trend as well as detects anomalies as they happen.  In addition, an advanced system can establish correlations with other variables such as load and temperature, resulting in a complete analysis of the transformer's health.

A dedicated software solution can go even further. Using machine learning, a program that could aggregate the data and form prediction and recommendations to ensure proper maintenance without over-costing in man-hours.  The strengths of the online DGA management platform are put to use with this solution, while the challenges are all answered.

Leave a comment

You are commenting as guest.