Use of the DGA Matrix

Camlin DGA Matrix processes and analyzes DGA data input using 11 diagnostic methods.
Each row of the matrix corresponds to a specific method (IEC, Rogers, Dornernburg, Key Gas, Duval Triangles and Pentagons, ETRA, CO2 / CO) while each column corresponds to the diagnostic result (No Diagnosis, OK, PD, Stray Gassing, T1, Overheating, Carbonization, T2, T3, DT, D2, D1). By looking at the results of each method on the same graph, you can eliminate outliners, reduce uncertainty and focus on the predominant degradation phenomenon.

 

Limits

The limits above which each method must be used are not clearly defined in the IEC, IEEE and ETRA standards. Some methods have built-in limits, others need to define a threshold. DGA Matrix uses the limits of IEEE C57.108: 2008 condition 1. Therefore, if all gases are below the limits mentioned above, the result is OK for each method. As soon as a gas exceeds the limit, each method is run and a result is obtained.

Methods

Dornenburg (1974)

It introduced the differentiation between types of electrical and thermal failure using the analysis of 4 significant gas ratios (CH4 / H2, C2H2 / C2H4, C2H2 / CH4, C2H6 / C2H2).

Gas : H2, CH4, C2H6, C2H4, C2H2

It identifies:

  • No diagnosis
  • No fault
  • Corona (PD low energy)
  • Thermal fault
  • Arcing (PD high intensity)

Rogers (1975)

It originally used 4 gas ratios (CH4 / H2, C2H6 / CH4, C2H4 / C2H6, C2H2 / C2H4), 2 are the same as in Donenburg. It is able to distinguish different temperature ranges of faults. The method was subsequently reduced to 3 gas ratios and the description of the faults is more generalized. Both the Dornenburg and Roger methods are capable of defining “normal states” for the DGA state.

Gas : H2, CH4, C2H6, C2H4, C2H2

It dentifies:

  • No diagnosis
  • No fault
  • Low Energy Density Arcing
  • Low temperature Thermal
  • Thermal <700 ⁰C
  • Thermal > 700 ⁰C
  • Arcing – High temperature discharge

IEC ratios (1978)

It is based on the 3 ratios of Rogers with a slight difference in the ratio intervals. It provides a general description of the types of faults.

Gas : H2, CH4, C2H6, C2H4, C2H2

It identifies:

  • No Diagnosis
  • PD – Partial Discharges
  • D1 – Discharges of low energy
  • D2 – Discharges of high energy
  • T1 – Thermal fault, t < 300 ⁰C
  • T2 – Thermal fault, 300 ⁰C < t < 700 ⁰C
  • T3 – Thermal fault, t > 700 ⁰C

Duval Triangle T1 (1974)

Graphic model of the fault using 3 gas concentrations in a ternary with CH4, C2H2 and C2H4.
The concentrations of the three gases are expressed as a percentage of their sum. The Duval Triangle method always gives a diagnosis, ie the method is unable to define a “normal” DGA condition. That said, it should therefore only be used if the gases have abnormal concentrations. The method introduced DT – discharges plus the type of thermal defect.

Gas : CH4, C2H4, C2H2

It identifies:

  • PD – Partial Discharges
  • D1 – Discharges of low energy
  • D2 – Discharges of high energy
  • T1 – Thermal fault, t < 300 ⁰C
  • T2 – Thermal fault, 300 ⁰C < t < 700 ⁰C
  • T3 – Thermal fault, t > 700 ⁰C
  • DT – Discharges + Thermal fault

Duval Triangles T4, T5 (2008)

These additional triangles were introduced to provide additional information on fault types, depending on the results of the Duval triangle (triangle DT1).
Triangle Duval 4 should be user if DT1 defines the defect as PD, T1 or T2.
DT5 to be used if DT1 defines the defect as T2 or T3.
These new triangles are more sensitive to the PD defect type and introduce new defect subtypes: O, S, C.

Gas : H2, CH4, C2H4, C2H6

It identifies:

  • ND – No diagnosis
  • S – Stray gases
  • O – Overheating <250°C without carbonisation of paper
  • C – Possible carbonisation of paper
  • T2 – Thermal fault, 300 ⁰C < t < 700 ⁰C
  • T3 – Thermal fault termico, t > 700 ⁰C

Duval Pentagons (2014)

The Duval Pentagons represent a further development of the Duval triangles, using 2 additional gases.
The methods provide a diagnosis based on the simultaneous use of H2 and 4 hydrocarbon gases.  

Gas : H2, CH4, C2H6, C2H4, C2H2

It identifies:

  • PD – Partial Discharges
  • S – Stray gases
  • D1 – Discharges of low energy
  • D2 – Discharges of high energy
  • T1 –  Thermal fault, t < 300 ⁰C
  • T2- Thermal fault, 300 ⁰C < t < 700 ⁰C
  • T3 – Thermal fault, t > 700 ⁰C
  • O – Overheating < 250°C without paper carbonisation
  • C -Possible carbonisation of paper
  • T3-H – Thermal fault T3 in mineral oil only

Key Gas Method (1974)

It is based on the percentage of each fault gas to the sum of all the fault gases. Four failure models are provided depending on the “gas model”. In general, the method is too sensitive to CO concentration, resulting in false positives for older transformers with normally high CO concentrations.

Gas: H2, CH4, C2H6, C2H4, C2H2, CO

It identifies:

  • Overheated Oil
  • Overheated Cellulose
  • PD
  • Arcing in oil

ETRA (1999)

Developed by the Electric Technology Research Association (ETRA), it uses only two ratios С2Н2 / С2Н6 and С2Н4 / С2Н6.

Gas : C2H6, C2H4, C2H2

It identifies:

  • PD – Partial Discharges
  • D1 – Discharges of low energy
  • D2 – Discharges of high energy
  • T1 – Thermal fault, t < 300 ⁰C
  • T2 – Thermal fault, 300 ⁰C < t < 700 ⁰C
  • T3 – Thermal fault, t > 700 ⁰C
  • DT – Discharges + Thermal fault

CO2/CO Ratio

This interpretation was aimed at detecting a critical aging situation of the overall paper insulation. It was first introduced by Roger, then changed over the years. Depenidng on the level of CO and CO2 it can help to determine if it is an overall overheating or if it is mainly involving the paper.
The recent IEEE C57.107-2018 introduced the use of CO2 / CO only together with the “above normal” presence of hydrocarbon gases. Using this report it is possible to detect both the overall insulation hearing and the overall charring of large quantities of paper.

 

Gas : H2, CH4, C2H6, CO, CO2

It identifies:

  • No diagnosis
  • Insulation Overheating
  • Cabonisation of paper