Rogers Ratios, IEC Ratios, and R-DGA: A Practical Comparison

Asset managers evaluating DGA software or reviewing their organisation's analytical methodology will encounter three major interpretive frameworks: Rogers Ratios, the IEC 60599 key gas ratio method, and Reliability-based DGA (R-DGA). Each was developed for a different purpose, operates on different principles, and answers a different question about a transformer's condition. Understanding those differences is essential for using them appropriately, and for understanding why combining them within a single analytical platform produces the most complete picture.

Rogers Ratios: Fault-Type Classification From Gas Ratios

The Rogers Ratios method was developed by R.R. Rogers of the Central Electricity Generating Board and formalised in a 1978 IEEE paper [1]. The method uses ratios of four key dissolved gases to classify the type of fault:

• CH₄/H₂ (methane to hydrogen): discriminates between thermal faults (higher CH₄) and partial discharge (higher H₂)
• C₂H₂/C₂H₄ (acetylene to ethylene): indicates arcing when elevated; essentially zero in most thermal fault conditions
• C₂H₄/C₂H₆ (ethylene to ethane): differentiates thermal fault intensity; ethylene generation dominates at higher temperatures
• C₂H₂/C₂H₆ (acetylene to ethane): used in some formulations for additional arcing discrimination

Each combination of ratio values maps to a fault-type category: low-energy partial discharge, high-energy discharge (arcing), low-temperature thermal fault, and high-temperature thermal fault. The underlying physical basis is the pyrolysis chemistry of transformer oil: different fault temperatures and energy densities produce characteristic distributions of decomposition products. Partial discharge generates predominantly H₂ through high-energy electron bond dissociation at low total energy; thermal faults above 700°C generate C₂H₂ by acetylene formation from higher-energy aromatic decomposition; intermediate temperatures produce C₂H₄ preferentially; and low-temperature thermal faults generate CH₄ as the dominant hydrocarbon [2].

The Rogers Ratios method is straightforward to calculate and has been widely used since its publication. However, CIGRE TB 296 [3] documents a well-established limitation: a significant proportion of real transformer gas profiles, with estimates in the literature ranging from 15–25% of cases, fall outside the defined ratio category boundaries, producing unclassifiable results. This occurs when gas profiles reflect multiple simultaneous fault types, when concentrations are low enough that measurement uncertainty materially affects the ratio values, or when gas generation reflects benign background chemistry rather than fault activity. Ratio methods work most reliably when at least one gas is elevated significantly above background levels.

IEC 60599 Key Gas Ratios: The International Counterpart

IEC 60599:2022 [4], the international standard for DGA interpretation, employs a three-ratio approach using the same gas pairs as Rogers:

• C₂H₂/C₂H₄: primary arcing indicator
• CH₄/H₂: thermal versus discharge discrimination
• C₂H₄/C₂H₆: thermal fault temperature range

IEC 60599 [4] assigns specific numerical ranges to each ratio for five fault categories: partial discharge (PD), low-energy discharge (D1), high-energy discharge (D2), thermal fault below 300°C (T1), thermal fault 300–700°C (T2), and thermal fault above 700°C (T3). The ratio boundaries differ from Rogers Ratios in several zones, particularly the PD/D1 boundary and the T1/T2 thermal boundary, reflecting different calibration choices and, in part, the different database of field cases used in each standard's development.

IEC 60599 [4] also provides guidance on typical concentration values for normally operating transformers, specifically background values for each key gas that allow engineers to assess whether absolute concentrations are elevated relative to population norms before applying the ratio classification. This background guidance is a practical enhancement over the ratio-only Rogers approach, though the concentration ranges provided are broad by design to accommodate the wide population of transformer designs and service histories.

Like Rogers Ratios, IEC 60599 is a fault classification method. IEEE C57.104-2019 [5] takes a parallel approach for the North American market, with its own condition-level thresholds and ratio-based guidance. Both standards classify what type of fault is indicated; neither quantifies how serious the fault is relative to the population of transformers, or how rapidly it is progressing.

R-DGA: Severity Quantification and Fleet Risk Ranking

Reliability-based DGA, developed by Jim Dukarm beginning in the early 1990s [6] and formalised in peer-reviewed form by Dukarm et al. [7], takes a fundamentally different approach. Where Rogers Ratios and IEC 60599 ask "what type of fault is this?", R-DGA asks "how severe is this transformer's condition relative to the population, and is it getting worse?"

The two core metrics are:

CSEV (Cumulative Severity) integrates the gas generation rate over the transformer's complete history of DGA results, normalised against a validated reference population of transformer histories [7]. Rather than evaluating the most recent sample against a fixed threshold, CSEV asks: how unusual is this transformer's entire gas record relative to all transformers in the population database? A transformer that has been generating gas at a low but consistent rate for 20 years accumulates CSEV over time even if no individual sample ever crossed a threshold boundary. Conversely, a transformer with a single elevated sample but no history of sustained generation has limited CSEV accumulation.

This history-integrating property makes CSEV particularly valuable for ageing equipment, where the pattern of slow accumulation over many years is often more informative than any single sample result. Dukarm et al. [7] demonstrated that CSEV-based assessment detects both the false positives produced by threshold methods (transformers flagged for elevated concentrations that are normal for their age and design) and the false negatives (transformers with consistent trajectories that never cross any single threshold).

HF (Hazard Factor) maps the CSEV value onto the empirical relationship between condition severity and observed failure probability across the population database [7]. HF translates a severity measurement into a risk ranking, a fleet-level metric that allows direct comparison across all transformers in a programme. Two transformers with the same gas concentrations but different histories will have different CSEV values and therefore different HF rankings, even though threshold methods would treat them identically.

R-DGA does not classify fault type. This is a feature, not a limitation: fault-type classification and severity quantification answer different questions, and both are needed for a complete diagnostic picture. CIGRE TB 296 [3] notes that modern DGA interpretation practice increasingly moves toward multi-method assessments that combine fault classification (what is happening?) with severity quantification (how serious is it?) rather than relying on any single method.

Practical Complementarity in a Multi-Method Framework

The Duval Triangle [2] has largely superseded Rogers Ratios in modern DGA guidance, including in IEC 60599:2022 [4], because it avoids the unclassifiable zone problem. By using the proportion of CH₄, C₂H₄, and C₂H₂ as coordinates in a triangular composition space, the Duval Triangle classifies every gas profile into one of seven zones without undefined boundaries. The graphical representation also makes trends visible: movement from one zone toward another over successive samples can indicate fault type evolution.

The most comprehensive DGA assessment uses all available methods within a structured analytical workflow:

1. Concentration evaluation against IEC 60599 [4] or IEEE C57.104 [5] background values: establishes whether elevated gas is present and which gases are involved
2. Fault-type classification via the Duval Triangle [2]: identifies what type of fault or condition is most consistent with the gas profile
3. R-DGA severity assessment via CSEV and HF [7]: quantifies how serious the condition is relative to the fleet and whether it is progressing

A case where all methods converge, with the Duval Triangle in the D2 (arcing) zone, IEC 60599 ratios consistent with high-energy discharge, and high CSEV and rising HF, presents an unambiguous assessment requiring urgent attention. Where methods diverge, the divergence is itself informative: a transformer with Duval Triangle in the T3 (high-temperature thermal) zone but low CSEV may have had a historical high-temperature event that has since resolved, rather than an active progressing fault.

TOA software applies the Duval Triangle, IEC ratio analysis, and R-DGA metrics within a single analytical environment [6], updating the complete assessment for every transformer in the database each time new DGA results are entered. This integration eliminates the inconsistency that arises when different methods are applied through separate tools or by different analysts.

For technical background on the R-DGA methodology, visit the Science page. For product information on TOA, visit the TOA page, or contact us to discuss applying these methods to your DGA programme.