DGA in Transformers With Alternative Insulating Fluids: What Changes?

The vast majority of power transformers in global service use mineral oil as the insulating and cooling fluid, and the DGA interpretation frameworks in IEEE C57.104-2019 [1] and IEC 60599:2022 [2] were developed and validated for mineral oil. The threshold values, fault classification methods, and ratio criteria in these standards reflect the gas generation chemistry of mineral oil under thermal and electrical stress.

However, the use of alternative insulating fluids, particularly natural esters (vegetable oil-based) and synthetic esters (dibasic acid or polyol esters), has grown substantially over the past decade, driven by three primary advantages: significantly higher fire point temperature than mineral oil (reducing fire risk in indoor installations and building-integrated substations), biodegradability (reducing environmental liability from oil containment or spill), and in some formulations, enhanced moisture tolerance that can extend paper insulation life in humid environments.

As utilities deploy more ester-filled transformers, the question of how to conduct DGA monitoring on them correctly becomes operationally significant. The answer is clear: the mineral oil frameworks do not apply, and using them produces systematically unreliable results.

Why Ester Fluids Behave Differently From Mineral Oil in DGA

The chemical composition of ester fluids, either triglyceride fatty acid chains from vegetable oils (natural esters) or synthetic fatty acid ester chains (synthetic esters), differs fundamentally from the predominantly paraffinic and naphthenic hydrocarbon chains in mineral oil. This difference in molecular structure produces different gas generation behaviour under thermal and electrical stress.

Higher baseline CO and CO₂. This is the most practically significant difference. Ester fluids contain ester functional groups (–COO–) that decompose more readily than the C–C and C–H bonds in mineral oil hydrocarbons, producing CO and CO₂ at significantly lower temperatures. Normal thermal operation in an ester-filled transformer produces CO and CO₂ concentrations that would be flagged as abnormal against the mineral oil thresholds in C57.104 [1] and IEC 60599 [2]. The elevated carbon gases are a property of the fluid, not an indicator of cellulose insulation degradation.

A utility applying mineral oil thresholds to ester transformer DGA will find that the majority of its ester-filled units appear to have elevated carbon gases, a systematic false alarm that creates unnecessary maintenance activity and erodes confidence in the DGA programme.

Different hydrocarbon gas generation patterns. At the same fault temperatures, ester fluids produce different relative proportions of methane, ethylene, ethane, and acetylene than mineral oil. The fault classification zone boundaries in the Duval Triangle [3], which are calibrated for mineral oil gas ratios, do not produce correct fault type classifications when applied to ester fluid DGA results.

Different gas solubility. The Henry's Law solubility coefficients for dissolved gases in ester fluids differ from those in mineral oil. This affects how gases partition between the oil and gas space in transformers with conservator or sealed-tank preservation, influencing the absolute concentrations measured and their interpretation.

Acetylene generation characteristics. Some ester fluids generate acetylene more readily than mineral oil at moderate thermal fault temperatures, a characteristic that means the presence of acetylene in ester transformer DGA requires fluid-specific threshold values to avoid both over- and under-alerting.

The Appropriate Interpretation Framework

CIGRE Technical Brochure 771 [4], produced by Working Group A2.43, is the primary international reference for DGA interpretation in transformers with non-mineral insulating fluids. TB 771 provides:

• Fluid-specific normal background concentrations for CO and CO₂, corrected for the higher baseline generation in ester fluids
• Adapted fault classification guidance for natural ester, synthetic ester, and other non-mineral fluids
• Modified Duval Triangle zones validated for ester fluid gas generation characteristics
• Guidance on the relationship between standard mineral oil ratio methods and ester fluid results

CIGRE TB 443 [5], produced by Working Group D1.32, is the earlier precursor document and remains relevant for utilities with silicone oil-filled transformers or legacy ester units.

Together, these CIGRE publications provide the technical basis for a DGA programme that is valid for ester-filled transformers, whereas the regional standards, by their explicit scope limitation to mineral oil, do not.

Practical Implications for DGA Programmes With Ester-Filled Units

Record oil type correctly in your DGA management system. Transformer Oil Analyst™ (TOA) accommodates different fluid types as a database field, applying appropriate interpretation parameters for each transformer. A transformer recorded as mineral oil in the system but actually filled with natural ester will receive incorrect threshold comparisons. Correct fluid type recording is the prerequisite for correct analysis.

Do not compare ester and mineral oil results directly. A CO concentration of 300 ppm in a natural ester transformer and 300 ppm in a mineral oil transformer represent different conditions. The ester result may be entirely normal; the mineral oil result may be concerning. Cross-fleet comparisons require fluid-type awareness.

Establish a new baseline after fluid conversion. Some utilities are converting existing mineral oil transformers to ester fluids for retrofit fire risk reduction or at transformer refurbishment. The historical DGA record built under mineral oil does not apply to post-conversion ester oil analysis. A new baseline needs to be established after conversion before trend analysis is meaningful; the CSEV calculation for the ester period should be treated as starting fresh from the conversion date.

If deploying ester-filled transformers at scale, validate your interpretation framework. The guidance in CIGRE TB 771 [4] is based on the available population of ester transformer operating data, which is substantially smaller than the mineral oil population that underpins the regional standards. For utilities deploying significant numbers of ester-filled units, maintaining good sample records and contributing data to the growing ester transformer monitoring community improves the collective interpretation framework over time.

The R-DGA methodology underlying TOA [6] can be applied to ester transformer DGA when the appropriate fluid-specific parameters are used, providing the same population-normalised severity assessment for ester-filled units as for mineral oil transformers, once the correct reference values are applied.

For technical guidance on DGA interpretation for alternative fluids, visit the Science page or contact us to discuss your specific fleet situation.