Int. J. Chem. Sci.

: 13(4), 2015, 1777-1786

ISSN 0972-768X
www.sadgurupublications.com

FORECASTING THE NATURE OF POWER TRANSFORMER

INSULATION OIL USING CHEMICAL PROPERTIES OF
DISSOLVED GASES
K. IYSWARYA ANNAPOORANI*

VIT University Chennai Campus, CHENNAI (T.N.) INDIA

ABSTRACT
The life expectancy of transformers at various operating conditions is not accurately known.
Deterioration of power transformer insulation is a function of time and temperature. Since in power
transformer, the temperature distribution is not uniform, the part which is operating at the highest
temperature will ordinarily undergo the greatest deterioration. Therefore, it is usual to consider the effects
produced by the highest temperature hottest spot. The hottest-spot winding temperature is the principal
factor in determining life due to loading. The temperature cannot be measured directly because of the
hazards in placing a temperature detector at the proper location because of voltage. In this paper a neural
network based regression analysis was done to forecast the nature of power transformer insulation oil and
life time of transformer including loading conditions.

Key words: Dissolved gas analysis, Levenberg algorithm, Regression analysis.

INTRODUCTION

The largest portion of capital investment in transmission and distribution substations

represented by power transformers. A lose in a single unit can have a multimillion dollar
impact on financial consequences. But a failing transformer removed from service in time
can usually be economically reconditioned. Fault conditions in a power transformer are
detected in several ways. One of the methods is based on detection of the degradation
products in the insulating oil. The degradation of insulating oil occurs due to abnormal
dissipation of energy within the transformer. This causes carbon by products dissolved in the
insulating oil1. However, the energy released through fault processes such as overheating,
partial discharge and arcing, is often sufficient to generate the fault gases initially in the
form of bubbles. Also high moisture conditions and sudden overloads can cause the
inception of moisture vapor bubbles released from conductor insulation. When abnormal
________________________________________
*
Author for correspondence; E-mail: iyswarya.annapoorani@vit.ac.in; Mo.: 9884077358
1778 K. I. Annapoorani: Forecasting the Nature of Power….

gassing from dissolved gas analysis results, the transformer is subjected to frequent testing
weekly or even daily to understand, what is happening inside the transformer and to prevent
a catastrophic failure.

Table 1: Fault with dissolved gases

H2 (Hydrogen) Exposed to partial discharge or corona in the transformer also created
with excessive moisture
CH4 (Methane) Exposed to excessive heat from an intimate contact with a hot metal
C2H6 (Ethane) Exposed to excessive heat from a hot metal. Heat required is greater
than heat required for methane
C2H4 (Ethylene) Exposed to excessive heat from a hot metal. Heat required is greater
than heat required for ethane
C2H2 (Acetylene) Typically associated with electrical arcing.
CO2 (Carbon Exposed, when cellulose insulation is exposed to excessive heat, such as
dioxide) during period of over loading.

When a transformer’s insulating oil exposed to excessive heating under normal or

abnormal operating conditions, the heat is transformed to the oil and if sufficient amount of
heat is present, combustible gases are created. Since the decomposition of oil indicates a
threat to the operational safety of expensive machines, electrical engineers focus attention on
the amount and nature of the gas evolved. Based on the result of DGA and the potential
cause of such differences, appropriate preventive measures are taken to protect the
transformer. Measuring the levels of the light gases found in an electrical transformers
insulating oil is one way to monitor the health of a transformer1,2. A gas sample is extracted
from the oil and nine key gas components-namely hydrogen, oxygen, nitrogen, methane,
carbon monoxide, carbon dioxide, ethane, ethylene and acetylene are analyzed. When a
transformer is failing, the chemical compounds in the oil break down to give off these gases.
If analyzed on a routine basis, a failing transformer can be identified and replaced without a
power loss or the potential of a serious explosion.
Interpretation of dissolved gas analysis

The main difficulty in making use of dissolved gas analysis results is that it is not
easy to draw the line between normal and abnormal results. So in4 interpretation schemes
include a normal condition as one of the diagnostic outcomes, but have not been particularly
effective in reliably identifying a normal condition. There are three most important
discussions needed from user point of view in power transformer.
Int. J. Chem. Sci.: 13(4), 2015 1779

(i) Normal condition

(ii) Transformer needs immediate breakdown maintenance
(iii) Condition demands continuous monitoring with period

If there is no appreciable rise in concentration of various gases then transformer is

healthy as in first two cases and if the rise in concentration is in the order of hundreds or
thousands of ppm every week or at lesser intervals, then transformer needs immediate
breakdown maintenance5. For the third question, the interpretation given in Table 1 will
have to be further simplified and made applicable to those cases of condition monitoring,
where rise in concentration of one or more individual gases is being observed. The
simplified approach forms the new method of interpretation, applicable to all the cases of
condition monitoring as in Table 2.

Table 2: Interpretation result of dissolved gas analysis

Rise in gas Key gas

Ratio method As per IS : 10593
concentration method
Nil Normal aging Normal aging Normal aging
CH4 ---- Thermal fault below 150oC Thermal fault from 150oC
to 300oC
C2H4 Over heating General conductor over Thermal fault of low
heating temperature upto 150oC
CH4 & C2H6 --- Fault from 150oC to 300oC ----
CH4 & C2H4 ---- Circulating currents and/or Thermal fault of 300oC to
over heated joints 700oC or above
C2H2 ---- Flash over without power ----
flow through
C2H2 & C2H6 ---- Tap changer selector ----
breaking current

Drawbacks from the conventional DGA interpretation schemes are –

• Mainly developed based on human judgment and no systematic attempt has

been made. High degree of inconsistency and ambiguity.
• Unable to detect with high confidence multiple faults that occur concurrently
within the transformer
1780 K. I. Annapoorani: Forecasting the Nature of Power….

• Unable to detect new or unknown faults owing to lack of expert knowledge in

them.
Proposed work

In this paper, we use regression method of Artificial Neural Network to overcome

the above drawbacks. The problem of neural network learning can be seen as a function
optimization problem, where we are trying to determine the best network parameters
(weights and biases) in order to minimize network error6. This said several function
optimization techniques from numerical linear algebra can be directly applied to network
learning, one of these techniques being the Levenberg-Marquardt algorithm. The Levenberg
Marquardt algorithm provides a numerical solution to the problem of minimizing a nonlinear
function over a space of parameters for the function. It is a popular alternative to the Gauss-
Newton method of finding the minimum of a function. As our problem related to Fault
Diagnosis by Dissolved Gas Analysis of a Power Transformer, the neural can be viewed as
highly nonlinear functions7. From this perspective, the training problem can be considered as
a general function optimization problem8,9, with the adjustable parameters being the weights
and biases of the network, and the Levenberg-Marquardt can be straight forward applied in
this case. For training the neural network, we consider the DGA data with corresponding
load for a power transformer in an electrical utility in India rated as 105 MVA, 400/230KV,
BHEL make transformer in an substation with serial No. 6006653.

Table 3: DGA data of a 400/230 KV Transformer

Carbon
Date of Hydrogen Methane Ethane Ethylene Acetylene Load Current
dioxide
test (H2) (CH4) (C2H6) (C2H4) (C2H2) (MW) (A)
(CO2)
26.7.11 4 1 1 0.1 0.1 947 100 439.17
30.9.11 5 2 1 1 1 1150 120 527
30.11.11 6 2 1 2 1 1120 110 483.091
30.1.12 11 1 2 0.1 0.1 1478 100 439.17
29.3.12 12 1 1 0.1 0.1 1560 120 527
21.7.12 5 1 2 0.1 1 1152 120 527
20.9.12 3 1 0.1 0.1 0.1 441 110 483.09
27.2.13 34 10 5 5 2 2345 170 746.59
29.4.13 62 32 46 15 10 2654 170 746.59
21.6.13 71 43 49 32 5 2763 160 702.67
Int. J. Chem. Sci.: 13(4), 2015 1781

Tables 3 and 4 show the dissolved gas value in ppm with variable load for a 400/230
KV transformer. Figs. 1 and Fig. 2 show that variation of acetylene with respect to load and
variation of hydrogen with respect to load. The inference from the Fig. 1 is, when sudden
increase in acetylene gas level in transformer oil from 2 ppm to 10 ppm shows that there is
increase in load hence load, current. Also from Fig. 2, we conclude that as load increases,
due to thermal effect the hydrogen value in ppm also get increased10. As per the standard,
sudden increase in hydrogen leads thermal fault. Due to this thermal effect, cellulose
degradation will occur and acetylene gas level also get increased, which leads to arcing fault.
This data is used to train the neural network using Levenberg algorithm and the fault of that
transformer was predicted.

800

600
C 2H 2
400
Load (MW)
Current (A)
200

0
.11

.11

.11

13

13

13
12

12

12

12
.2.

.4.

.6.
.1.

.3.

.7.

.9.
.7

.9

.11
26

30

27

29

21
30

29

21

20
30

Fig. 1: Load influence in acetylene formation in insulating oil

800
700
600
500
H2
400 Load (MW)
300 Current (A)
200
100
0
11

11

.11

13

13

13
12

12

12

12
.7.

.9.

.1.

.3.

.7.

.9.

.2.

.4.

.6.
.11
26

30

27

29

21
30

29

21

20
30

Fig. 2: Load influence in hydrogen formation in insulating oil

Levenberg’s main contribution to the method was the introduction of the damping
factor λ. This value is summed to every member of the approximate Hessian diagonal before
1782 K. I. Annapoorani: Forecasting the Nature of Power….

the system is solved for the gradient. Typically λ would start as a small value such as 0.1.
Then the Levenberg-Marquardt equation is solved commonly by using LU decomposition.
After the equation is solved, the weights w are updated using δ and network errors for each
entry in the training set are recalculated. If the new sum of squared errors has decreased, λ is
decreased and the iteration ends. If it has not, then the new weights are discarded and the
method is repeated with the higher value of λ.

DGA data

Initialize network
weight w   and δ

Compute Jacobian J
 

Compute error
t
gradient g = J E

Solve (H+λI) δ = g
and find δ
 

Update weights
w and
  δ

Recalculate sum  squared error Increase  λ

If
Sum squared error <
          
initial value
 

Stop

Fig. 3: Algorithm flow chart

Int. J. Chem. Sci.: 13(4), 2015 1783

Variations of the algorithm may include different values for v, one for decreasing λ
and other for increasing it. Others may solve (H + λ diag (H)) δ = g instead of (H + λI) δ = g,
while others may select the initial λ according to the size of the elements on H, by setting
λo = tmax (diag (H)), where t is a value chosen by the user. We have chosen the identity
matrix equation because it is the same method implemented internally by the Neural
Network Tool Box in MATLAB.

Mean squared error (mse) 106

105

104

103
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
2 Epochs

Fig. 4: Performance plot of developed NN

The above performance plot shows, how the network performance improved during
training. It shows the value of the performance function verses the iteration number. The
network performance is measured in terms of mean squared error, shown in log scale. It
rapidly decreased as the network was trained. The performance plot doesn’t indicate any
major problems with training because the validation and test curves are very similar at both
ends. If the test curve increased, then it is possible that some over fitting might have
occurred. Performance is shown for each of the training, validation and test sets. The version
of the network that did best on the validation set was after training. The mean squared error
of the trained neural network can now be measured with respect to the training samples. This
will give us a sense of how will the networks will do, when data applied from real world.
The average squared error is the difference between the network outputs (a) and the target
outputs (t). It is defind as –

N
mse = 1 (Ci ) 2 = 1 ∑ (t i − a i ) 2 …(1)
N N i=1
1784 K. I. Annapoorani: Forecasting the Nature of Power….

Fig. 5: Regression plot of developed NN

The regression plot represents the training, validation, testing and all data. The
dashed line in each axis represents the perfect result. i.e. Output = Target. The solid line
represents the best fit linear regression line between outputs and targets. The R value is an
indication of the relationship between the outputs and targets. If R = 1, this indicates that
there is an exact linear relationship between outputs and targets. If R is close to zero, then
there is no linear relationship between outputs and targets. From the above regression, plot
training data indicates a good fit. The scatter plot is helpful in showing that certain data
points have poor fit. If it is, it would investigate this data point to determine, if it represents
extrapolation. If so, then it should be included in the training set and additional data should
be collected to be used in the test data.

Table 4: Predicted diagnosis

Predicted testing Approximated load

Predicted fault
year in MW
2014 Normal ageing 132.8
2016 Arc or flashover along with over heating 165.9
2019 Arc or flashover along with over heating 160.3
2020 Arc or flashover along with over heating 162.9

Table 4 shows that when sudden increase is there in load, beyond the rating due to
utility, the thermal stress get increased, which lead degradation of insulation and possibility
Int. J. Chem. Sci.: 13(4), 2015 1785

of arcing fault. When the loading reaches 50% of nameplate, the hot metal gases ethylene,
ethane and methane starts increases. If it was unpredicted, it leads to thermal fault of 300oC
to 700oC with high and low energy discharges. When the temperature is greater than 700oC,
the breakdown of oil occurs and produces acetylene. It causes a sustained arcing, a more
serious operational issue that can lead transformer failure, if left unpredicted. So by
predicting the dissolved gas data by regression method using Neural Network, we can
deduce the health of the transformer by finding the type of fault will occur in future.

CONCLUSION

The important need for condition evaluation is that there is an aging problem of
transformers installed in industries and transmission and distribution of power system
network. Most of the transformers were installed 20 to 30 years ago, when large investments
were made in expanding electrical power system network. These transformers have been
exposed to various accumulative worsening stresses and these are in high risk of failure. So
there is increase in need of reburshment, repair and replacement. Based on this view, the
predictive analysis of transformer will enable the power system network to give reliable
availability of power supply.

REFERENCES

1. M. Duval, IEEE Electr. Insul. Mag., 5(6), 22-27 (1980).

2. R. R. Rogers, Elec. Insul., 13(5), 349-354 (1978).
3. Indian Standard Method of Evaluating of Gases in Oil Filled Electrical Equipments,
IS: 10593 (1983).
4. IEEE Guide for the Interpretation of Gases Generated in Oil Immersed Transformers,
ANSI/IEEE Standard, C57.104 (1991).
5. IEEE Guide for the Detection and Determination of Gases in Oil Immersed
Transformers and their Relationship to the Serviceability of the Equipment, ANSI/
IEEE Standard C57.104 (1978).
6. K. F. Thang, R. K. Aggarwal, D. G. ESP and A. J. McGrail, Dielectric Materials,
Measurement and Applications, Conference Publication No. 473, IEE (2000).
7. E. Y. Kang, H. Kim and J. Cho, LNAI 4251, Springer-Verlag (2006).
8. A. Shintemirov, W. Tang and Q. H. Wu, IEEE Transactions on Systems, Man, and
Cybernetics – Part C.: Appl. Rev., 39(1) (2009).
1786 K. I. Annapoorani: Forecasting the Nature of Power….

9. Y. Jinhong, X. Huang, C. Wu and P. Jiang, IEEE Transactions on Dielectrics and

Electrical Insulation, 18(2), 478-484 (2011).
10. K. Bacha, S. Souahlia and M. Gossa, Electric Power Systems Research, 83, 73-79
(2012).

Revised : 03.08.2015 Accepted : 05.08.2015