Dissolved gas analysis is the most powerful early-warning tool for transformer faults, but the interpretation rules engineers rely on were built for mineral oil. Apply them unchanged to an ester-filled transformer and you raise false alarms on healthy units and risk misreading genuine faults. IEEE C57.155 closes that gap. It is the only IEEE document devoted to DGA interpretation in ester fluids, covering both natural and synthetic esters, and it is the reference an advising engineer reaches for the moment an ester sample lands that cannot be read with the mineral-oil playbook.
What it covers
The guide addresses the theory of combustible gas generation in natural and synthetic ester-filled transformers, the interpretation of dissolved gas results, recommended actions, and a body of supporting laboratory and field data. Its central message is that ester fluids generate the same fault gases as mineral oil but in fundamentally different proportions and quantities — most strikingly, esters produce carbon oxides far more abundantly from the fluid itself, and natural esters can show a distinctive stray-gassing signature soon after energising. It provides reference threshold concentrations drawn from a multi-source database and adapts the Duval Triangle method for ester fluids. It explicitly does not cover mineral oil, silicone or gas-to-liquid fluids, and it addresses gas interpretation only — not fluid quality parameters such as dielectric strength, moisture or acidity, which are handled by the relevant specification and maintenance standards.
Why it matters in practice
The guide is candid about its own limitations, and that candour shapes how it should be used. It rests on a limited amount of in-service data and several laboratory data sets; the synthetic ester population in particular is small, so the higher-percentile thresholds carry wide uncertainty. It also notes that a field-based gas-evolution rate limit for ester fluids has not been established, so there is no equivalent of the rate-of-change guidance that exists for mineral oil. The practical consequence is that trending against a transformer's own history is far more reliable than comparing a single result to population statistics, and that confirmation sampling, not a single reading, should drive any intervention decision. The guide gives interpretation, not condemnation limits, and its values are best framed as reference percentiles rather than pass-or-fail thresholds.
How we use it
For TriboTech this is the primary reference behind every ester DGA interpretation we deliver. We use it alongside the relevant CIGRE work and the ester-specific Duval Triangle and Pentagon methods when assessing dissolved gas in ester-filled units — core scope for the synthetic-ester generator transformers in our offshore wind portfolio and the natural-ester fleet onshore. The guide's pyrolysis data underpins our explanation of why mineral-oil ratio methods fail for esters, and its background-gas guidance keeps us from alarming on the high carbon-oxide levels that are normal in a healthy ester transformer. Where our field experience extends beyond the guide — the role of decomposition products it does not tabulate, or early-service stray-gassing that can distort a first sample — we layer that on top, while keeping the guide's framework as the documented basis for the assessment.