When utility engineers review a dissolved gas analysis report, the hydrocarbon gases typically command most of the attention. Hydrogen, methane, ethylene, and acetylene are the gases associated with active thermal and electrical fault processes, and elevated concentrations or unfavourable ratios trigger immediate concern. Carbon monoxide and carbon dioxide are frequently noted but treated as secondary indicators, checked against ratio guidelines and then set aside.
This prioritisation is understandable but can lead to systematically underweighting some of the most important information in a DGA result. Carbon gases provide the primary oil-based window into the condition of the cellulose paper insulation, the component whose degradation is irreversible, whose failure mode can be sudden and catastrophic, and which more than anything else determines the remaining service life of the transformer.
The Insulation System and Why Cellulose Matters
A power transformer's insulation system consists of two components with fundamentally different longevity characteristics: the insulating fluid (mineral oil or alternative) and the cellulose paper and pressboard that provide mechanical support to the windings and supplement the fluid's dielectric capability.
The fluid can be reconditioned, filtered, dried, or fully replaced at any point during the transformer's service life. Cellulose cannot. The cellulose paper wrapped around transformer windings is accessible only by opening and rewinding the transformer, an intervention that is economically equivalent to replacement for most units. The degree of polymerisation (DP) of the cellulose chains determines its mechanical strength; once DP falls below approximately 150–200 (from a new-condition value of approximately 1000), the paper becomes brittle and may fail mechanically under through-fault current forces. This type of failure is sudden, catastrophic, and not directly predictable from the DGA gas profile without specific cellulose condition indicators.
The rate of cellulose degradation follows the Arrhenius relationship [1]: each 6–8°C increase in hotspot temperature approximately doubles the degradation rate. Sustained operation at elevated temperatures, even modestly above design conditions, consumes cellulose life at a rate that can substantially shorten transformer service life.
CIGRE TB 227 [2] identifies cellulose condition as a primary determinant of remaining transformer life and recommends CO/CO₂ monitoring alongside furan analysis as the primary oil-based indicators of cellulose insulation state.
Carbon Gas Generation: Physical Basis
CO and CO₂ are produced by the thermal and oxidative decomposition of cellulose at elevated temperatures [3]. The mechanism is analogous to the hydrocarbon gas production from oil decomposition, but the substrate is the paper insulation rather than the oil.
Normal ageing background. All in-service transformers generate CO and CO₂ as a background process. Slow thermal ageing of the cellulose at normal operating temperatures produces small amounts of both gases continuously over the transformer's service life. Long-lived transformers with decades of service accumulate substantial absolute concentrations of both gases even without any active fault process. This background accumulation is expected and in itself is not a fault indicator; it is a record of the transformer's service history.
Accelerated thermal degradation. When the cellulose is exposed to temperatures significantly above its normal operating range, either from a localised thermal fault, prolonged overloading, or repeated thermal cycling, the decomposition rate increases substantially. This produces accelerating CO and CO₂ generation that appears as a rising trend in consecutive DGA samples. The rate of increase carries more diagnostic weight than the absolute level, particularly for older transformers with substantial background accumulation.
Active fault involvement. When a thermal fault detected through hydrocarbon gases (ethylene, methane, ethane) is accompanied by rising CO and CO₂, the fault process is affecting the cellulose insulation as well as the oil. This combination is significantly more serious than a thermal fault in the oil alone, because the irreplaceable component, the paper insulation, is being consumed. IEEE C57.104 [4] explicitly identifies this combination as warranting escalated attention.
The CO₂/CO Ratio and What It Indicates
The ratio of CO₂ to CO provides additional diagnostic information about the nature and temperature of the cellulose degradation process.
Under normal low-temperature ageing conditions, the dominant reaction pathway produces more CO₂ than CO. IEEE C57.104 [4] and IEC 60599 [5] both reference a CO₂/CO ratio of approximately 11 as a reference value for mineral oil: ratios above this are generally consistent with normal background ageing; ratios below this (particularly below 3) suggest more acute thermal stress on the cellulose, or conditions approaching partial combustion, which occurs at temperatures significantly above normal thermal ageing conditions.
A low CO₂/CO ratio combined with elevated hydrocarbon gases and rising trends is a more serious finding than elevated carbons alone, because it indicates that the thermal degradation is more intense than normal low-temperature ageing.
Important caveats: The CO₂/CO ratio is influenced by the transformer's oil preservation system (conservators lose CO₂ to the atmosphere more than sealed tanks), by the age of the oil between changes, and by the presence of other organic materials in the transformer. These factors should be considered when applying ratio guidelines to specific cases.
Carbon Gas Interpretation in Practice
Always read carbon gases alongside hydrocarbons. A thermal fault classification (T2 or T3 in the Duval Triangle) is meaningfully more serious when accompanied by rising CO and CO₂ than when the carbon gases are stable. The combination indicates that the fault is consuming both oil and insulation paper.
Track CO and CO₂ trends, not just levels. Because background accumulation is expected in ageing transformers, the absolute concentrations of CO and CO₂ are less informative than their rate of change over consecutive samples. A transformer with CO₂ at 4,000 ppm that has been stable for five years is less concerning than one at 1,500 ppm that has doubled in the past year.
Review carbon gases for transformers with high CSEV. The CSEV metric in R-DGA methodology [6] integrates fault severity across all fault gases, including the contribution of CO and CO₂ generation to cumulative severity. Transformers with high CSEV driven substantially by carbon gas accumulation have a different maintenance implication from those with high CSEV from hydrocarbon gases: the former may indicate extensive insulation degradation that changes the life extension calculus.
Use furan analysis to corroborate. Furfuraldehyde and related furan compounds are produced by cellulose degradation and dissolve in the oil. Furan analysis provides a direct measure of cellulose degradation that is complementary to, but more specific than, the carbon gas profile. CIGRE TB 227 [2] recommends furan analysis alongside DGA carbon gas monitoring for comprehensive insulation life assessment.
For further technical detail on DGA interpretation, including carbon gas analysis, visit the Science page and the Learn page. For product information, visit the TOA page or contact us.
References & Further Reading
- [1]McNutt, W.J., “Insulation Thermal Life Considerations for Transformer Loading Guides” IEEE Transactions on Power Apparatus and Systems, 1992.
- [2]CIGRE Working Group A2.18, “Life Management Techniques for Power Transformers” CIGRE Technical Brochure 227, 2003.
- [3]Duval, M., “A Review of Faults Detectable by Gas-in-Oil Analysis in Transformers” IEEE Electrical Insulation Magazine, 2002.
- [4]IEEE C57.104-2019, “IEEE Guide for the Interpretation of Gases Generated in Mineral Oil-Immersed Transformers” IEEE, 2019.
- [5]IEC 60599:2022, “Mineral oil-filled electrical equipment in service — Guidance on the interpretation of dissolved and free gases analysis” IEC, 2022.
- [6]Dukarm, J.J., Draper, D., Arakelian, V.K., “Improving the Reliability of Dissolved Gas Analysis” IEEE Electrical Insulation Magazine, 2012.

Delta-X Research develops Transformer Oil Analyst™ (TOA), the market-leading tool for managing and interpreting insulating fluid test data for high-voltage apparatus. Founded in 1992 and based in Victoria, BC, Canada, the team applies Reliability-based DGA methodology to help utilities worldwide assess transformer health and prioritise fleet maintenance decisions.
Related Articles
Rogers Ratios, IEC Ratios, and R-DGA: A Practical Comparison
Three major analytical approaches dominate transformer DGA interpretation: Rogers Ratios, IEC 60599 key gas ratios, and Reliability-based DGA. This article explains the design purpose, theoretical basis, and known limitations of each method, and how they work together in a comprehensive diagnostic framework.

