Corrosive sulphur is one of the quieter ways a transformer can fail. Certain organo-sulphur compounds in mineral oil react with the copper windings to form copper sulphide, which migrates into the paper insulation and builds conductive deposits between turns. The result is internal arcing, often after years of normal service. IEC 62535 lets the industry ask the right question first: can this oil do this?
What it covers
IEC 62535 specifies a laboratory method for detecting potentially corrosive sulphur in used and unused mineral insulating oil. A short length of paper-wrapped copper winding conductor is sealed in a glass vial with a measured volume of oil and held at elevated temperature for several days. Afterwards the copper is examined for the dark discolouration of copper sulphide, and the wrapped paper is examined for the characteristic metallic deposits that signal sulphide migration into the insulation.
The result is deliberately binary: the oil is reported as either potentially corrosive or non-corrosive, with no graduated scale. What makes the method distinctive is that it includes the paper layer and runs hot enough and long enough to expose species that older copper-strip and silver-strip tests miss entirely. It was developed because those earlier methods could pass an oil that later caused failures in the field.
The standard is scoped to mineral insulating oil — the wrapped-conductor method was developed and validated for it. For DBDS-specific monitoring in natural esters, the gas-chromatographic route applies instead.
Why it matters in practice
The damage pathway is what makes corrosive sulphur dangerous rather than merely undesirable. The conductor itself can tolerate some tarnish; the problem is the conductive sulphide that ends up in the cellulose, bridging the dielectric gap — and by the time conventional indicators react, the deposit may already be present.
A few judgement points matter at the bench. Metal passivators, sometimes already present in service oil, can suppress the reaction on the test conductor and make a problematic oil look clean — so a non-corrosive result on a passivated sample is not proof of a sound oil. Batch-to-batch variability from the same refinery is real, which is why we test the delivered batch rather than trusting a product grade. And visual interpretation depends on good lighting and an experienced eye; where the paper reading is ambiguous, microscopic elemental confirmation settles whether a deposit is copper sulphide or ageing discolouration.
How we use it
We run IEC 62535 as part of new-oil acceptance, before an oil is committed to a transformer, and as a targeted in-service check when there is reason to suspect a problem — unexplained gas trends pointing to paper involvement, an oil of uncertain provenance, or a fleet screening after corrosive sulphur turns up at one site. We pair it with a quantitative DBDS measurement so we can distinguish "never corrosive" from "corrosive species already consumed by ongoing attack." Where an oil tests positive, the conversation moves quickly to risk assessment and mitigation rather than a simple pass or fail.