IEC 63585:2026 is the international guide to DGA interpretation for esters. It addresses dissolved gas analysis in natural and synthetic esters and is the reference for ester-filled transformers, where mineral-oil habits can misread a result.
What IEC 63585 covers
It covers natural ester to IEC 62770 and synthetic ester to IEC 61099, in ester-filled transformers and similar equipment. Most experience is with distribution and industrial equipment, traction and offshore included. Wind and solar transformers are named applications; power-transformer experience is still thin. Esters are increasingly used with tap-changers, but switching equipment is excluded because the gas record was insufficient. Mono-esters and blends appear only in an informative annex.
What is new in IEC 63585
This is a first edition, Edition 1.0, dated July 2026 and prepared by the IEC fluids committee. Maintenance documents for esters already existed, but guidance on interpreting their dissolved gases was scarce; this edition supplies it. In Europe it is published as EN IEC 63585:2026, and in Denmark as DS/EN IEC 63585:2026.
How IEC 63585 relates to IEC 60599
It does not replace IEC 60599. The introduction cites that guide, with IEEE C57.104-2019, as existing mineral-oil interpretation, and IEC 60599 stays on the normative list; the new edition adds the ester reading they lack. The introduction describes the earlier IEEE C57.155-2014 as mainly about natural ester and absolute levels. Upkeep of the liquid stays with IEC 62975 for natural ester and IEC 61203 for synthetic ester.
Why esters need their own DGA interpretation
The fault gases are those of mineral oil, but ratios and rates can differ, which can come from the different molecular structure. Water can hydrolyse an ester, and carbon dioxide may rise from it or from oxidation of a synthetic ester open to air. Unlike in mineral oil, those carbon oxides cannot be tied to cellulose ageing, and the source of a high carbon dioxide level is hard to judge. A natural ester with several double bonds, soybean and rapeseed more than high-oleic sunflower, forms ethane and hydrogen, especially when air is present. Unused natural ester is recommended only for closed equipment, because its unsaturated bonds oxidise readily. Ethane is expected and usually levels off; a continued rise can indicate air in the tank.
Partial discharge, stronger discharges and thermal faults still resemble mineral-oil patterns, but the amounts can differ, and not every mineral-oil scheme transfers. The reading keeps the same marker gases, combining reference levels, two-gas ratios, rates of increase and trend. The ratios serve both ester families with one exception: in a natural ester, a thermal fault cannot be called from ethylene against ethane alone, because a highly unsaturated fluid forms ethane without a fault once a little air is present.
Comparative tests of stray gassing show the effect grows with the polyunsaturated share of a natural ester, and inhibited high-grade mineral oil did not show it in that study. The liquid, construction materials or a metal passivator can imitate a fault, and CIGRE TB 927 is the laboratory check named for that case.
Reference levels and Duval methods for esters
Service data give reference levels for natural esters, including soybean and high-oleic sunflower, and for synthetic esters. They prompt a closer look but do not decide that a fault exists. Owners should calculate their own, and when every gas sits well below the published levels, the tools can mislead. No population rate of increase is given, so each unit is trended alone. A single result high in ethylene or acetylene can still mean the unit is removed for examination. In most cases, though, trending is the right step. Abnormal gassing brings tighter sampling and, where wider evidence supports it, removal for inspection.
The standard points to the literature for Duval triangle and pentagon forms for soybean, sunflower, rapeseed and synthetic esters, modelled on the mineral-oil diagrams. The only reference it gives for this point, CIGRE TB 771, includes rapeseed as a pentagon but not as a triangle. The annex works the methods through examples for synthetic ester and soybean-based natural ester, for information only. Field evidence does not yet settle the zones, and every method assumes an experienced reader who allows for sampling problems.
How TriboTech uses IEC 63585
IEC 63585 is our primary reference when we interpret dissolved gas in transformers filled with natural ester or synthetic ester. This includes the synthetic-ester transformers in wind turbines, offshore and onshore. We read it with CIGRE TB 443 and the ester variants of the Duval triangle and pentagon, and take particular care over stray gassing in a natural ester, so a healthy fluid is not reported as a fault. Analyses are carried out at partner laboratories; our role is the engineering interpretation and the advice that follows, described under transformer oil analysis.