Dissolved Gas Analysis (DGA) is a critical diagnostic tool for identifying faults in transformer oils, ensuring the reliability and longevity of power systems. The accuracy of DGA measurements directly influences maintenance decisions, making it imperative to utilize technologies that provide precise and repeatable results. Among online monitoring solutions, gas chromatography (GC) technology stands out due to its ability to self-calibrate and maintain high measurement fidelity over time.

A common industry question is whether manual sampling DGA or online DGA provides more accurate results. The answer largely depends on the technology used. Traditional manual sampling introduces several potential sources of error, including sample contamination, handling inconsistencies, and delays between collection and analysis. While laboratory-based GC testing is highly accurate, it is inherently limited by the manual process itself.

Online DGA monitoring mitigates these issues by continuously analyzing transformer oil in realtime, eliminating human-induced variability. However, not all online DGA technologies are created equally. Online GC-based systems maintain superior accuracy due to their frequent self-calibration—typically twice per week—ensuring consistent performance over extended periods. In contrast, other online technologies, such infrared-based DGA, rely on factory calibration with a specific oil type and lack an effective infield recalibration mechanism. This limitation makes it challenging to verify their long-term accuracy under varying operational conditions.

A common misconception perpetuated by competitors is that GC-based DGA systems have a disadvantage due to the use of consumables. All online DGA systems, including those from competitors, require consumables; the difference lies in transparency. Some manufacturers downplay or omit details regarding their systems’ reliance on consumables while simultaneously promoting the notion that their equipment does not require recalibration. However, evidence suggests that these systems can only be calibrated at the factory, raising concerns about their accuracy over time and across different oil types.

Accurate concentration values from both laboratories and gas monitors (15% accurate or better) are needed for reliable DGA diagnosis. An accuracy of 15% means that if 100 ppm are measured, the actual value may be anywhere between 85 and 115 ppm. Low concentration values (<5 or 10 times the analytical detection limit of the laboratory or online gas monitor) are usually quite inaccurate and unreliable and should not be used for DGA diagnosis.

The actual accuracy of laboratories and online monitors can be obtained by using gas-in-oil standards, which can be prepared in the lab or may be available commercially. This will allow users to verify that the calibration used for both gas extraction from oil and analysis of extracted gases is adequate. If lab accuracy is worse than 15%, a calculation of diagnosis uncertainty should be performed, and commercial software is available. In practice, if differences of more than 15% are observed between DGA results from different laboratories and/or online monitors, and if this results in an uncertain diagnosis (e.g., arcing or hot spot) and action on the transformer, it is advisable to verify the accuracy of results.

Advantages of On-Line Gas Monitors
• They catch abnormal formation of gases monitored and quickdeveloping faults occurring suddenly within hours.

• They are not affected by sampling errors.

• They are more reliable for evaluating gas rates.

Limitations of On-Line Gas Monitors
• They are more expensive than laboratory DGA.

• Some monitors are not accurate for some gases.

• Some monitors are not calibrationfree and maintenance-free as claimed by their manufacturers.

Multi-Gas Monitors of the Chromatographic Type

• Use the same standardized, NIST-traceable technique as laboratories.

• Provide automatic recalibration at fixed intervals, as laboratories do.

• Require some maintenance (change of carrier gas, calibration gas mixture, GC columns every 3 or 5 years, depending on the model).

Multi-Gas Monitors of the Infrared Type

• Do not require a change of carrier gas and gas mixture.

• Cannot measure H2 and O2 by infrared, requiring the use of relatively inaccurate solid-state sensors for that purpose.

• Some may need recalibration every year because of contaminants in ambient air (SF6, oil vapor, solvents), and lamps fade with time; some cannot be recalibrated in the field and must be sent back to the factory.

Transformer reliability hinges on accurate dissolved gas measurements, making the choice of DGA technology a crucial factor. Online GC-based monitoring offers unmatched precision through its self-calibration capability and robust analytical methodology. While some competitors claim advantages by downplaying their own systems' reliance on consumables and calibration limitations, the reality remains that frequent and verifiable calibration is key to long-term measurement accuracy. As power systems grow increasingly complex, investing in proven, self-calibrating GC-based DGA solutions is the best strategy for ensuring transformer health and minimizing the risk of costly failures.

Emilio Morales is a Technical Application Specialist in Transformer applications at Qualitrol Company LLC. His focus is to support solutions in comprehensive monitoring for Transformer applications. Emilio attended Nuevo Leon State University in Mexico, receiving his bachelor’s degree in Electromechanical Engineering in 1980. He has over 30 years of experience in power transformer design which includes transformers up to 500 MVA and 500 kV, furnace and rectifier transformers and reactors. He is member of the IEEE/PES Transformer Committee, IEC and CIGRE. Emilio joined Qualitrol in June 2012 and previously worked with GE-Prolec , Ohio Transformer, Sunbelt Transformer and Efacec.

This article was originally published in the March issue of the Transformers Technology magazine, which you can access here.

To download the PDF version of this article, click here.