Licensed Professional Engineers |
Search/

Request a free quote

Electrical

Transformer Dissolved Gas Analysis: How to Read and Trend a DGA Report

How to read and trend a transformer DGA report under IEEE C57.104-2019: status from levels, changes and rates, Duval Triangle 1, and a free calculator.

By Dreiym Engineering

Subject Matter Expert: Brian J. Brackensick, M.S., P.E.

Technical review: Brian J. Brackensick, M.S., P.E.

Published 10 min read

Power TransformerPower Transformer
Contents
  1. What a DGA Report Tells You
  2. The Three Checks Behind the Status
  3. Why Age and the Oxygen to Nitrogen Ratio Matter
  4. Why the Sample History Matters
  5. Fault Type: Duval Triangle 1 and Annex D
  6. How the Free DGA Calculator Works
  7. A Screening Estimate, Not a Diagnosis
  8. Related Standards
  9. Frequently Asked Questions
  10. How Dreiym Engineering Can Help

Dissolved gas analysis (DGA) is one of the most informative routine tests for a transformer in service. The lab report lists the gases dissolved in the oil in parts per million: hydrogen, methane, ethane, ethylene, acetylene, carbon monoxide and carbon dioxide, usually with oxygen and nitrogen. The numbers on one report mean little by themselves. What matters is how they compare with typical values, how they changed since the last sample and how fast they are rising.

Key takeaways

  1. A DGA report is read as a history. IEEE® C57.104-2019 sets a status from three checks: the gas levels, the change since the previous sample and the rate of increase across several samples.
  2. The typical values in the guide depend on the transformer’s age and on its oxygen to nitrogen ratio, which reflects whether the oil is sealed from the air or breathes.
  3. Duval Triangle 1 is the guide’s main method for naming a likely fault type. Triangles 4 and 5 and the Duval Pentagons in Annex D refine it under set conditions.
  4. When the status comes only from a change or a rate, the guide calls for a confirmation sample within a month before acting.
  5. The free calculator gives a screening estimate, not a diagnosis. Do not base any decision about a transformer on it until Dreiym Engineering has reviewed the full test history under a written engagement.

This article explains how to read a DGA report and trend it under IEEE® C57.104-2019, IEEE® Guide for the Interpretation of Gases Generated in Mineral Oil-Immersed Transformers, and how the free transformer DGA calculator on this site applies the same steps to your own sample history.

What a DGA Report Tells You

Heat and electrical stress break down the oil and paper insulation, and each kind of fault favors certain gases. Partial discharges make mostly hydrogen. Overheated oil makes methane and ethane at lower temperatures and more ethylene as the hot spot gets hotter. Arcing is the main source of acetylene. Overheated paper adds carbon monoxide and carbon dioxide. Dissolved gas analysis measures those gases in a sample drawn from the transformer, so how the sample was taken and tested matters as much as the numbers.

The 2019 edition of the guide treats two questions as separate steps: is anything unusual (the status, 1, 2 or 3), and if so, what kind of fault might it be (fault identification).

The Three Checks Behind the Status

Clause 6.1 of IEEE® C57.104-2019 sets out the status method. Each sample is checked three ways:

  • Levels. The concentration of each gas against the guide’s tables of typical values, which come from large populations of transformers in service. There is a lower table of typical levels and a higher one for levels that only a small share of transformers in service reach.
  • Change. The increase in each gas since the previous sample, against typical changes between samples.
  • Rate. The rate of increase fitted through several samples over a span of months, against typical rates. The guide sets how many samples and what span the rate needs, so one or two samples are not enough for this check.

Figure 1 shows what each status means. Status 2 calls for a closer look and more frequent sampling, and Status 3 for closer surveillance and a review. The status is a statistical flag. It is not a measure of how damaged the transformer is.

Diagram. On the left, three numbered checks: 1, gas levels, each gas against the guide's typical values for the transformer's age and oxygen to nitrogen ratio; 2, change since the last sample; 3, rate of increase, fitted through several samples over months. Arrows from all three lead to a status column with three boxes: Status 1, probably normal, every level, change and rate below its typical value; Status 2, possibly suspicious, a level, change or rate reached its typical value; Status 3, probably suspicious, a level above the higher level table or a rate above its typical value. Below, an illustrated strip of six sample dates shows Status 1 for samples 1 to 3, Status 2 for samples 4 and 5 and Status 3 for sample 6, with a note at sample 4: Status 2 from a change alone, take a confirmation sample within a month, clause 6.1.1. Footer: procedure from IEEE C57.104-2019, clause 6.1; the values are in the guide.
Figure 1. The three checks behind an IEEE® C57.104-2019 status, and how the status can change from one sample date to the next. Simplified; the values are in the guide.

Why Age and the Oxygen to Nitrogen Ratio Matter

The guide splits its tables by the age of the transformer and by the ratio of oxygen to nitrogen in the oil. A low ratio is typical of a sealed transformer, where little air reaches the oil. A higher ratio usually means air reaches the oil, as in a free-breathing unit with a conservator open to the atmosphere through a breather. Gases build up differently in the two, so the same reading can be ordinary in one and unusual in the other. When the lab reports no oxygen or nitrogen, the guide says to use the values for the higher ratio, and a transformer near the dividing line can switch tables between samples.

Why the Sample History Matters

A single sample can only be checked for levels. The change needs the previous sample, and the rate needs several samples over months. A transformer can sit at Status 1 on levels alone while its gases climb quickly toward the tables, and that climb is what the change and rate checks catch. Read the history in date order, from the same sampling point, with oil processing, top-ups and repairs noted, because degassing or new oil lowers the levels and changes how later results compare.

When the status comes only from a change or a rate, with the levels still low, clause 6.1.1 calls for a confirmation sample within a month before anything else is done. One unusual increase can come from sampling, handling or the lab, and a second sample shows whether the trend is real.

Fault Type: Duval Triangle 1 and Annex D

Once the levels are high enough to mean something, the next question is what kind of fault could be making the gas. The guide’s main method, in clause 6.2.3, is Duval Triangle 1. It uses the relative amounts of methane, ethylene and acetylene and places the sample in one of seven zones: partial discharges (PD), low energy discharges (D1), high energy discharges (D2), thermal faults below 300 °C (T1), from 300 to 700 °C (T2) and above 700 °C (T3), and a mix of thermal and electrical faults (DT).

Annex D, which is informative, adds Duval Triangles 4 and 5 and Duval Pentagons 1 and 2. Triangle 4 looks more closely at low temperature faults and Triangle 5 at hotter thermal faults, each used only after Triangle 1 points to the kind of fault it was built for. The pentagons use five gases at once and add zones such as stray gassing and, for thermal faults, possible carbonization of paper.

Two cautions apply. A triangle always names a zone, even for a healthy transformer, so the guide recommends fault identification only when the gas levels call for it. And at very low concentrations the lab’s uncertainty can move a point from one zone to another, so a zone near a boundary, or one computed from a few parts per million, deserves less weight than the status and the trend.

Other methods described in the guide, such as the Rogers ratios, the key gas reading and the ratio of carbon dioxide to carbon monoxide, make useful cross-checks. When they agree with Triangle 1, confidence goes up. When they disagree, the sample, the lab and the transformer’s history deserve another look.

Related service

Electrical & Power

Studies modeled in ETAP® and SKM, field testing and electrical design by licensed PEs.

Request a free quote

How the Free DGA Calculator Works

This free tool, the transformer DGA calculator, applies the IEEE® C57.104-2019 status method, the Duval methods and the cross-checks to your own data. Type or paste up to 20 samples from a spreadsheet, or download the Excel® template, fill it in and upload it. Give the transformer’s age if you know it, the insulating liquid and the compartment the oil came from, then pass the quick check. The calculation runs on our server, and the values you enter are not stored. You get:

  • A status for every sample date, worked out with only that sample and the earlier ones, plus the dates the status escalated and the gases that drove each change.
  • Next sample guidance that follows the guide, including the confirmation sample when the status comes from a change or a rate.
  • A projection: if the current rate continued, which gas would reach the next status level within five years. It is a straight line drawn from your data, not a forecast.
  • Duval Triangle 1, plus the Annex D triangles and pentagons where their conditions apply, with an explanation of every zone you point at or tap.
  • A PDF report with the status, the trends and the Duval figures in one file, and synthetic examples to see how the tool behaves before you enter your own data.

A Screening Estimate, Not a Diagnosis

The calculator gives a screening estimate from published methods. It is not a diagnosis, and it does not replace engineering judgment about a specific transformer. Its results depend on how the oil was sampled, handled and tested, and on facts the calculator never sees: the design, the loading and operating record, maintenance and repairs, and other test results. Do not rely on the results for any decision about a transformer, such as keeping it in service, loading it, repairing it or replacing it, until Dreiym Engineering has reviewed the full test history under a written engagement.

To ask for that review, request a DGA review and tell us about the transformer and the lab reports you have.

IEEE® C57.104-2019 is written for mineral oil in a transformer main tank, and the calculator applies only its status method, Duval methods and cross-checks. Tap changer compartments, main tanks that share oil with one, and ester or silicone liquids need other guides; bushings and cables have their own practices. These documents cover those cases and the steps around the interpretation (editions as listed by each publisher in November 2026).

  • ASTM® D3612, Standard Test Method for Analysis of Gases Dissolved in Electrical Insulating Oil by Gas Chromatography (D3612-02, reapproved 2026): the lab method behind many DGA reports.
  • ASTM® D923, Standard Practices for Sampling Electrical Insulating Liquids (D923-15, reapproved 2023): how the oil sample is taken.
  • IEC 60567:2023, Oil-filled electrical equipment, Sampling of free gases and analysis of free and dissolved gases in mineral oils and other insulating liquids, Guidance: the IEC guidance for the lab work.
  • IEC 60599:2022, Mineral oil-filled electrical equipment in service, Guidance on the interpretation of dissolved and free gases analysis: the international counterpart to C57.104; background only.
  • IEEE® C57.139-2026, the IEEE® guide for the interpretation of gases generated in liquid-type on-load tap-changers: applies to tap changer oil instead of C57.104.
  • IEEE® C57.155-2014, IEEE® Guide for Interpretation of Gases Generated in Natural Ester and Synthetic Ester-Immersed Transformers: applies to ester-filled units; IEEE® has listed it as inactive-reserved since March 2025, with a revision in progress.
  • IEEE® C57.146-2025, IEEE® Guide for the Interpretation of Gases Generated in Silicone-Immersed Transformers: applies to transformers still filled with their original silicone liquid.
  • IEEE® C57.152-2025, IEEE® Guide for Diagnostic Field Testing of Liquid-Filled Power Transformers, Regulators, and Reactors: field tests read together with DGA.
  • NFPA® 70B, Standard for Electrical Equipment Maintenance (2026 edition): the maintenance program routine oil sampling belongs to; see what NFPA® 70B requires.
  • ANSI®/NETA MTS-2023 and ANSI®/NETA ATS-2025, the NETA® maintenance and acceptance testing specifications for electrical power equipment and systems.
  • CIGRE Technical Brochure 771, Advances in DGA Interpretation (2019): interpretation rules that consider the type and location of faults; background only.

Frequently Asked Questions

What does Status 3 mean on a DGA report?

Under IEEE® C57.104-2019 it means a gas level is above the higher of the guide’s two level tables, or a gas is rising faster than its typical rate. It calls for closer surveillance and an engineering review. By itself it does not mean the transformer is about to fail.

How often should transformer oil be sampled for DGA?

The guide sets no fixed routine interval; the owner’s maintenance program does, usually by the transformer’s importance and condition. It does call for a confirmation sample within a month when a change or a rate sets the status, and for more frequent sampling as the status rises.

Is the Duval Triangle enough to diagnose a fault?

No. Triangle 1 suggests a likely fault type from three gases, and it always names a zone, even for a healthy transformer. It is used when the gas levels call for it, together with the status, the trend, the cross-checks and the transformer’s history.

Can I use the calculator for ester, silicone or tap changer oil?

It shows the samples and trends, but it gives no status, because IEEE® C57.104-2019 covers mineral oil in a main tank only. Ester, silicone and tap changer oil have their own IEEE® guides.

Can I act on the calculator’s result?

Not on its own. It is a screening estimate, not a diagnosis. Do not base any decision about a transformer on it until Dreiym Engineering has reviewed the full test history under a written engagement.

How Dreiym Engineering Can Help

We review transformer DGA histories under a written engagement as part of our electrical power engineering work, and requests from the calculator go to Brian J. Brackensick, M.S., P.E. An infrared scan adds an outside look at connections and cooling, and when a transformer has already failed, our electrical failure analysis practice looks at why; see what engineers look for after a transformer failure.

Start with the free transformer DGA calculator. When you want an engineer’s view, request a DGA review or call (866) 621-6920 for a free quote. We typically respond within one business day.

Talk to an engineer about your project

We typically respond within one business day.

Request a free quoteor call (866) 621-6920