What it does
The Doernenburg method classifies dissolved-gas readings into three coarse fault categories — partial discharge, thermal decomposition, or arcing — using four ratios drawn from five gases. It’s reproduced in IEEE C57.104 Annex F, alongside Rogers Ratio, as one of the earliest ratio-based interpretation methods in regular service.
Compared with Rogers, Doernenburg gives a coarser answer — three categories rather than six — but its rules are tight enough that when it does commit to a category, the call tends to be confident. It’s a useful sanity check on Rogers and the Duval Triangle, especially when those two disagree.
The four ratios
R1 · CH₄ / H₂
Thermal vs. electrical signature. Same as Rogers’ R1. Above 1 leans thermal; below 0.1 leans corona.
R2 · C₂H₂ / C₂H₄
Arcing indicator. Same as Rogers’ R2. Climbs as acetylene becomes a meaningful share of the ethylene-acetylene pair.
R3 · C₂H₂ / CH₄
Secondary arcing check. Specific to Doernenburg — elevates the call to arcing when acetylene rises against methane.
R4 · C₂H₆ / C₂H₂
Inverse arcing check. Ethane presence argues against arcing — the larger this ratio, the more the profile looks thermal rather than electrical.
The three fault categories
A category matches only when all four of its ratio constraints hold. Samples that satisfy none are reported as inconclusive — Doernenburg’s narrower rules mean it abstains more often than Rogers, by design.
| Code | Fault | R1 | R2 | R3 | R4 |
|---|---|---|---|---|---|
| PD | Partial discharge | < 0.1 | — | < 0.3 | > 0.4 |
| T | Thermal decomposition | > 1 | < 0.75 | < 0.3 | > 0.4 |
| D | Arcing | 0.1–1 | > 0.75 | > 0.3 | < 0.4 |
When to reach for it
Doernenburg has an applicability prescreen: at least one of the key gases — H₂, CH₄, C₂H₂, or C₂H₄ — must exceed twice its IEEE C57.104 L1 typical-concentration threshold (200, 100, 35, or 100 ppm respectively). Below that level the standard says the method shouldn’t be applied at all, and this app honours that — it abstains rather than force a call on a sample with only trace concentrations.
Above the prescreen, Doernenburg is most useful as a confirmation step. If Rogers Ratio and Doernenburg agree on a category, there’s strong evidence for the call; if Doernenburg goes inconclusive while Rogers commits, the sample is probably near a zone boundary and trend data should weigh more heavily than a single reading.
Strengths and limits
Strengths: four ratios is more constrained than Rogers’ three, so when Doernenburg matches, it tends to match cleanly. The applicability prescreen also stops the method from talking authoritatively about samples it shouldn’t be classifying.
Limits: three categories is genuinely coarse — Doernenburg can’t distinguish the thermal severities (T1/T2/T3) the way the Duval Triangle or Rogers Ratio can, and it can’t separate high-energy from low-energy arcing. The narrow rule set means inconclusive results are common in field practice; the method is designed to abstain rather than guess. Treat it as a confirmation step rather than the primary call, and combine it with the Duval Triangle and Rogers Ratio for the finer-grained call.
References
- IEEE Std C57.104-2019, Guide for the Interpretation of Gases Generated in Mineral Oil-Immersed Transformers (Annex F covers Doernenburg).
- E. Doernenburg and W. Strittmatter, “Monitoring of transformers by gas analysis,” Brown Boveri Review, 1974.