Dissolved Gas in Oil Analysis – Part 2 (refer to newsletter No. 4 for introductory info)

It is never good practice to look at one Dissolved Gas in Oil Analysis (DGA) from a laboratory and make an assessment on transformer health. History, as in all maintenance endeavours, is an important part of effectively interpreting DGA results. Ideally we are able to assess several DGA results in combination with standard oil tests and power/dissipation factor data.

Threshold Values

The following are published threshold values of dissolved gas in oil, gas concentrations.  Values below those indicated are deemed to be “normal” for transformers that have seen some years of service.  You will note the variation between the authorities, and there are other threshold guidelines, with many user organizations establishing their own levels.  This data is shown for reference only.  The best analytical comparisons for DGA come from transformers of similar size and rating, exposed to similar loading and stress conditions.

 
IEEE
Dorenberg & Strittmatter
US Bureau of Reclamation
Hydrogen (H2)
100 ppm
200 ppm
500 ppm
Oxygen (O2)
N/A
N/A
N/A
Nitrogen (N2)
N/A
N/A
N/A
Carbon Monoxide (CO)
350
500
750
Methane (CH4)
120
50
125
Carbon Dioxide (CO2)
2500
6000
11000
Ethylene (C2H4)
50
80
175
Ethane (C2H6)
65
35
75
Acetylene (C2H2)
35
5
15

Ratio Methods for Dissolved Gas in Oil Result Interpretation

There are well established ratio comparison methods that look at pairs of gases, and develop a coding system to help define potential fault conditions.  Two of these methods are Rogers Ratios, a 4 ratio system and IEC Ratios which employ a 3 ratio approach.  This newsletter does not purport to provide a complete background on ratio analysis buts offers an overview of the methodology.

In general, care must be taken when interpreting results from the ratio methods.  A ratio method will return the same result if the compared values are 2/1 or 2000/1000, referring to threshold levels will help weed out “good” results.  There are a number of combinations that do not result in an analysis

(see page 2).  It is also important to note that the extensive data that went into establishing the ratio methods comes primarily from Utility transformers. 

Overview of IEC Ratio Method

This method looks at the dissolved gas in oil ppm ratio of C2H2 / C2H4 , CH4 / H2 , C2 H2 / C2H4 and assigns a code for each gas pair.  For some of the code combinations, an analysis/probable cause is defined.

 
35
resulting
C2H2
CH4
C2H4
ratio
C2H4
H2
C2H6
< 0.1
0
1
0
0.1 –<1.0
1
0
0
1.0 – <=3.0
1
2
1
> 3.0
2
2
2
 
IEC DGA Ratios
 
 
C2H2
CH4
C2H4
 
Case
C2H4
H2
C2H6
 
0
0
0
0
No Fault, Normal
1
0
1
0
Partial discharges of low energy
2
1
1
0
Partial discharges of high energy density
3
1
0
1
Discharges of low energy, Arcing
3
2
0
1
Discharges of low energy, Arcing
3
2
0
2
Discharges of low energy, Arcing
3
2
0
2
Discharges of high energy, Arcing
5
0
0
1
Thermal Fault, 150 C, Conductor Overheating
6
0
2
0
Thermal Fault, 150 – 300 C, Oil Overheating, Mild
7
0
2
1
Thermal Fault, 300 – 700 C, Oil Overheating, Moderate
8
0
2
2
Thermal Fault, 700 C, Oil Overheating, Severe

Summary:

Establish some threshold levels to weed out the “normal” readings and focus on the potential problem transformers. Apply ratio methods with care. Compare apples to apples. Solicit advice from an experienced Transformer Consultant (self serving recommendation).