Dissolved Gas Analysis (DGA): The Ultimate Guide for Transformer Health

1. A Blood Test for Your Transformer: Dissolved Gas Analysis (DGA) is the single most powerful diagnostic tool for assessing the health of an oil-filled power transformer, analogous to a blood test for a human.
2. Detects Incipient Faults: DGA can detect developing internal faults—like partial discharges, arcing, or overheating—long before any other monitoring method can, providing a crucial early warning.
3. Works by Analyzing Fault Gases: Electrical and thermal stresses break down the insulating oil and paper, producing specific gases that dissolve into the oil. DGA identifies and quantifies these “fault gases.”
4. Provides a Fault Signature: The types and ratios of the gases found (e.g., hydrogen, methane, acetylene) create a unique signature that helps experts diagnose the specific type and severity of the internal problem.
5. Enables Condition-Based Maintenance: By trending DGA results over time, asset managers can move from costly time-based maintenance to a more efficient and effective condition-based strategy, preventing failures and extending asset life.

1. What Exactly Is Dissolved Gas Analysis (DGA)?

• Dissolved Gas Analysis, or DGA, is a diagnostic procedure performed on the insulating oil of a power transformer. The process involves taking a sample of the oil, extracting the gases that are dissolved within it, and using a gas chromatograph to separate and quantify each gas.
• It is fundamentally a forensic chemical analysis. It identifies the byproducts of thermal and electrical stress on the transformer’s internal insulation system (the oil and solid paper insulation).
• The presence of certain gases, and more importantly their relative concentrations and ratios, acts as a “fingerprint” of a specific type of fault. This allows engineers to understand what is happening inside a sealed transformer tank without ever having to open it.

2. Why Is DGA the Most Important Diagnostic Tool for Transformers?

• Unparalleled Early Warning Capability: DGA can detect the very early stages of a developing fault, known as an incipient fault. Problems like minor partial discharging or localized overheating generate small amounts of gas long before they would cause any change in temperature, pressure, or electrical parameters.
• Provides Diagnostic Information: Unlike a simple alarm, DGA doesn’t just tell you *that* there is a problem; it tells you *what kind* of problem it is. It can distinguish between general overheating, partial discharges (corona), and high-energy arcing, allowing for a targeted response.
• Enables Proactive Asset Management: By performing DGA regularly and tracking the rate of gas generation over time, utility engineers can monitor the health of their transformers, prioritize maintenance, plan for replacements, and ultimately prevent catastrophic in-service failures, which are extremely costly and dangerous.

3. How Are Gases Formed Inside a Transformer?

• The insulating oil in a transformer is made of hydrocarbon molecules. The solid paper insulation is made of cellulose, which also contains hydrogen and carbon. When subjected to sufficient energy, the chemical bonds in these molecules break apart.
• The energy source can be thermal (overheating) or electrical (sparks, arcs, corona). The amount of energy input determines which bonds break and how the resulting fragments recombine.
• For example, low-energy events like overheating produce low-energy gases like hydrogen and methane. Very high-energy events like arcing provide enough energy to form acetylene. These newly formed gas molecules are then dissolved into the surrounding oil, where they can be detected by DGA.

4. What Are the Key Fault Gases and What Do They Indicate?

• Different fault types produce different gases. The main gases monitored are:
• Hydrogen (H₂): The very first gas to appear. It’s a key indicator of partial discharges (corona) and can also be produced by low-temperature overheating.
• Methane (CH₄) & Ethane (C₂H₆): Indicate low to moderate temperature overheating of the oil. Methane is formed at lower temperatures than ethane.
• Ethylene (C₂H₄): Indicates higher temperature overheating of the oil, typically above 300°C. A sign of a more serious thermal fault.
• Acetylene (C₂H₂): The most critical fault gas. It requires a large amount of energy to form and is a definitive indicator of high-temperature arcing (>700°C). Even small amounts of acetylene are a major concern.
• Carbon Monoxide (CO) & Carbon Dioxide (CO₂): These are formed from the decomposition of the solid paper insulation. A high CO/CO₂ ratio indicates overheating of the paper, which is very serious as paper aging is irreversible.

5. How Is DGA Traditionally Performed? (Lab Analysis)

• The traditional method involves sending a trained technician to the substation to draw a physical oil sample from the transformer. This must be done carefully using a clean, airtight glass syringe to avoid contaminating the sample with atmospheric air.
• The sealed syringe is then carefully packaged and shipped to a specialized laboratory for analysis.
• At the lab, the dissolved gases are extracted from the oil sample (using methods like vacuum extraction). The gas mixture is then injected into a gas chromatograph (GC), which separates the individual gases and measures the concentration of each one in parts per million (ppm).

6. What Is Online DGA Monitoring and What Are Its Advantages?

• Online DGA monitoring involves installing a permanent device directly onto the transformer. This device continuously samples the oil, extracts the gases, and analyzes them on-site in near real-time.
• Continuous Monitoring: Its biggest advantage is providing constant vigilance. A fault can develop rapidly between periodic manual samples. An online monitor will detect the change almost as it happens, providing a much earlier warning.
• Trend Analysis: Online monitors provide high-resolution data, allowing for very accurate trend analysis of the gas generation rate. This rate of change is often more important than the absolute value in diagnosing the severity of a fault.
• Reduced Human Error: It eliminates the risks associated with manual sampling, such as sample contamination, handling errors, and delays in shipping and analysis.

7. What’s the Difference Between Single-Gas and Multi-Gas Online Monitors?

• Single-Gas Monitors: These are simpler, more cost-effective devices that monitor for only one or two key gases. Most commonly, they monitor just hydrogen (H₂), as it is the first sign of almost any fault. They act as an excellent “first alert” system.
• Multi-Gas Monitors: These are more sophisticated and expensive instruments that measure a full suite of key fault gases (typically 7 to 9 gases). They essentially have a miniaturized gas chromatograph or a photo-acoustic spectrometer inside.
• Multi-gas monitors provide not just an alert but a full diagnosis. By analyzing the ratios of all the gases, they can use diagnostic tools like the Duval Triangle to tell you the likely type of fault, providing much richer information for asset management decisions.

8. How Do You Interpret DGA Results? (Duval Triangle)

• Interpreting DGA results is a specialized skill. It involves looking at the absolute values of the gases, their rate of change over time, and, most importantly, the ratios between key gases.
• Several diagnostic methods use these ratios to pinpoint the fault type. These include the Key Gas Method, Rogers Ratios, and the Dornenburg Ratios.
• However, the most widely used and graphically intuitive method is the Duval Triangle. It uses the percentage concentration of three key gases—methane (CH₄), ethylene (C₂H₄), and acetylene (C₂H₂)—to plot a point within a triangle. The location of this point falls into a specific zone that corresponds to a particular fault type.

9. What Is the Duval Triangle and How Does It Work?

• The Duval Triangle is a powerful graphical diagnostic tool developed by Michel Duval. It helps classify a fault type based on the relative percentages of three hydrocarbon gases: Methane (CH₄), Ethylene (C₂H₄), and Acetylene (C₂H₂).
• First, you calculate the total concentration of these three gases. Then, you find the percentage of each gas relative to that total. For example, %CH₄ = [CH₄ / (CH₄ + C₂H₄ + C₂H₂)] * 100.
• You then plot these three percentage values onto a special triangular graph. The triangle is divided into seven distinct zones, each corresponding to a specific fault type:
  • PD: Partial Discharges
  • T1: Thermal Fault, < 300°C
  • T2: Thermal Fault, 300°C to 700°C
  • T3: Thermal Fault, > 700°C
  • D1: Low Energy Discharge (sparking)
  • D2: High Energy Discharge (arcing)
  • DT: A mix of Thermal and Electrical faults

10. What Does a High Level of Hydrogen (H2) Mean?

• Hydrogen is the simplest gas molecule and requires the least amount of energy to form. Its presence is a highly sensitive indicator that *some* fault activity is occurring.
• The most common source of hydrogen is Partial Discharge (PD), also known as corona. This is low-energy electrical discharging that occurs in voids or defects in the insulation.
• Slow, low-temperature overheating can also produce hydrogen. Because it is the first gas to appear for most fault types, online monitors that specifically track hydrogen are excellent early warning systems. A sudden increase in the generation rate of hydrogen is a clear sign that a new fault has initiated or an existing one is worsening.

11. Why Is Acetylene (C2H2) Considered the Most Critical Gas?

• Acetylene is the most significant fault gas because its formation requires a very large amount of energy, corresponding to temperatures above 700°C.
• The only event inside a transformer that can produce this level of energy is high-energy electrical arcing. An arc is a sustained electrical breakdown, like a continuous lightning bolt inside the tank.
• Arcing is extremely destructive. It rapidly degrades oil and paper, and can lead to a pressure buildup and catastrophic failure of the transformer. Therefore, the presence of any amount of acetylene, even just a few parts per million (ppm), is a critical alarm that requires immediate attention and investigation.

12. How Often Should DGA Sampling Be Done?

• The frequency of manual DGA sampling depends on the criticality, age, and condition of the transformer.
• For a new, healthy transformer, an annual sample is typically sufficient. For older transformers or those with a history of issues, the frequency might be increased to semi-annually or quarterly.
• If a fault is suspected or if gas levels are trending upwards, the sampling frequency should be increased dramatically—perhaps to monthly, weekly, or even daily—to closely monitor the rate of gas generation. This is a primary reason why online DGA monitors are so valuable, as they provide this high-frequency data automatically.

13. Who Are the Top 10 Best Manufacturers of DGA Analyzers?

• The market for DGA equipment includes both laboratory instruments (Gas Chromatographs) and a growing number of online monitors. Choosing a manufacturer with proven technology and analytical reliability is crucial for effective asset management.

14. Why is FJINNO a Top Choice for Online DGA Monitoring?

• Advanced Photo-Acoustic Spectroscopy (PAS) Technology: FJINNO utilizes state-of-the-art PAS technology, which is a significant advancement over traditional methods. This technology provides direct measurement of each gas without the need for carrier gases or complex chromatographic columns, resulting in higher stability and lower maintenance.
• No Cross-Interference: A key advantage of FJINNO’s PAS system is its exceptional specificity. It measures each gas individually without interference from other gases in the sample, leading to lab-grade accuracy and preventing the false readings that can affect some other sensing technologies.
• Long-Term Reliability and Low Cost of Ownership: By eliminating the need for consumables like calibration or carrier gases and designing for long-term stability, FJINNO monitors offer an extremely low total cost of ownership. This reliability makes them a preferred choice for utilities looking to implement a large-scale, “fit-and-forget” online monitoring program.

15. How Do You Properly Take an Oil Sample for DGA?

• Proper sampling technique is absolutely critical for accurate lab results. The goal is to capture a representative sample of the transformer oil without contaminating it.
• Use a clean, new, airtight glass syringe. Before taking the sample, flush the valve and syringe by drawing and expelling oil several times to clear out any stagnant oil or debris.
• Draw the sample slowly and carefully to avoid creating bubbles. Once filled, invert the syringe, expel any small air bubbles, and immediately close the stopcock to seal it from the atmosphere.
• The sample must be clearly labeled with the transformer ID, date, time, and oil temperature, and sent to the lab as quickly as possible.

16. Can DGA Results Be Wrong or Misleading?

• Yes, DGA results can be misleading if not interpreted with care. The most common source of error is poor sampling. If air contaminates the sample, it will show high levels of nitrogen and oxygen, and some dissolved gases (like hydrogen) can be lost.
• Certain operational factors can also influence gases. For example, some on-load tap changers can generate fault gases that migrate into the main tank, leading to a misdiagnosis.
• This is why it’s crucial to look at trends over time rather than a single snapshot. A sudden jump in all gases might indicate a sampling error, whereas a steady increase in one specific gas is a much more reliable indicator of a real fault.

17. What Are the Next Steps After a Bad DGA Result?

• A bad DGA result requires a structured response, not an immediate panic. The first step is to confirm the result by taking a second sample immediately.
• If the second sample confirms the fault, increase the sampling frequency to determine the rate of gas generation.
• Correlate the DGA results with other data. Are the temperature gauges reading high? Has there been a recent electrical event? Perform complementary electrical tests, like power factor and winding resistance, to try and locate the problem.
• Based on the severity (especially if acetylene is present) and the rate of change, a decision will be made to either closely monitor the transformer, schedule a maintenance outage for internal inspection, or, in critical cases, de-energize it immediately.

18. What’s the Difference Between DGA and Oil Quality Tests?

• DGA and oil quality tests are both performed on the insulating oil, but they look for completely different things.
• DGA looks for dissolved gases generated by internal *faults* (overheating, arcing). It is a diagnostic test for the transformer’s health.
• Oil Quality Tests assess the condition of the *oil itself* as an insulator and coolant. These tests measure properties like dielectric strength (breakdown voltage), moisture content, acidity, and interfacial tension. They tell you if the oil needs to be filtered, dehydrated, or replaced.

19. How Does DGA Integrate with Other Monitoring Systems?

• Online DGA monitors are a key component of a comprehensive transformer monitoring strategy. They are rarely used in isolation.
• The data from the DGA monitor is typically fed into a centralized monitoring platform or asset performance management (APM) software.
• This software combines the DGA data with information from other sensors—such as winding temperature (from fiber optics), bushing monitoring, and load data—to create a complete health index for the transformer. This holistic view allows for much more accurate diagnostics and prognostics.

20. What Is the Future of DGA Technology?

• The future is about making online monitoring standard practice. As the cost of reliable online monitors continues to decrease, they will become standard equipment on most new and critical transformers, largely replacing routine manual sampling.
• Advanced Analytics and AI: The vast amount of data from continuous online monitors is perfect for AI and machine learning algorithms. These systems will be able to detect subtle patterns in gas generation that are invisible to human analysis, providing even earlier fault warnings and more accurate diagnoses.
• Sensor Fusion: The real power will come from fusing DGA data with other sensor data in real-time. For example, an AI model could correlate a sudden increase in hydrogen with a small change in the bushing’s power factor, correctly identifying a developing fault in the bushing before it becomes critical.

21. What is Photo-Acoustic Spectroscopy (PAS) Technology in DGA?

• Photo-Acoustic Spectroscopy (PAS) is a highly sensitive and stable method for gas detection used in advanced online DGA monitors, like those from FJINNO.
• It works by using a beam of infrared (IR) light, modulated at a specific frequency, to illuminate the gas sample extracted from the oil. Each type of gas (like methane or acetylene) absorbs IR light at a unique, characteristic wavelength.
• When the gas molecules absorb the pulsating light, they heat up and cool down rapidly, creating a tiny pressure wave—a sound wave. A highly sensitive microphone detects this sound. The intensity of the sound is directly proportional to the gas concentration. By using different IR wavelengths, the concentration of each gas can be measured precisely and with no cross-interference.

Fiber optic temperature sensor, Intelligent monitoring system, Distributed fiber optic manufacturer in China