See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318350677

Assessment of health of transformer: a survey

Article  in  International Journal of Pure and Applied Mathematics · April 2017

CITATIONS READS

0 477

2 authors, including:

Saptarshi Roy
National Institute of Technology, Warangal
24 PUBLICATIONS   18 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Condition Monitoring of Transformer View project

Wide Area Protection Schemes View project

All content following this page was uploaded by Saptarshi Roy on 11 July 2017.

The user has requested enhancement of the downloaded file.

International Journal of Pure and Applied Mathematics
Volume 114 No. 7 2017, 281-291
ISSN: 1311-8080 (printed version); ISSN: 1314-3395 (on-line version)
url: http://www.ijpam.eu
Special Issue
ijpam.eu

Assesment of health of
transformer: a survey
Ravi Kumar Jalli1
Saptarshi Roy2
1EEE Dept.,GMRIT ,

Andhrapradesh,
[email protected]
2EEE Dept., NIT Warangal ,

Telengana,
[email protected]

April 14-15, 2017

Abstract

Condition monitoring of transformer is the

process of acquisition and processing of data
related to various parameters of transformers
so as to predict and prevent the failure of a
transformer. This is done by observing the
deviation of the transformer parameters from
its expected values. Transformers are the most
critical assets electrical transmission and
distribution system. Transformer failures can
cause power outages, personal and
environmental hazards and expensive
rerouting or purchase of power from other
suppliers. Transformer failures can occur due
to various causes. Transformer in-service
interruptions and failures usually result from
dielectric breakdown, winding distortion caused
by short-circuit withstand, winding and
magnetic circuit hotspot, electrical
disturbances, deterioration of insulation,

281
International Journal of Pure and Applied Mathematics Special Issue

lightning, inadequate maintenance, loose

connections, overloading, failure of accessories
such as OLTC, Bushings etc. In this work it is
shown with case studies and examples how
health is assessed and precautions can be taken
to increase the life time of a transformer.
AMS Subject Classification: 00-02

Key Words and Phrases: Condition Monitoring, Tan

delta, Degree of Polymerization, Transformer
failures, dielectric breakdown, inadequate
maintenance of transformer.

1 Introduction
The life span of any equipment decreases with time[1]. So, to
increase the longevity of any equipment the condition of the
equipment should be assessed and preventive measures
should be taken during its operation time. Condition
Monitoring of Transformers is the process of collecting and
processing of data related to various parameters or
indicators of transformers so as to predict and prevent the
failure of a transformer. This is executed by observing the
anomaly of the parameters from their desired values.
Transformers are the very important components of power
system. Failures of transformers could cause power outages
various hazards or purchase of power from other suppliers.
Failures of transformer can happen due to several causes.
E.g-insulating medium breakdown, winding distortion, hot
spot in transformer windings , electrical disturbances,
degradation of insulation and paper insulators , lightning,
poor maintenance, bad-improper connections, overloading,
failure of supporting equipment’s such as OLTCs, bushings,
etc [2-3].
Several works has been done based on various condition
monitoring approaches of transformer from the past. [4]
describes distribution transformer’s condition monitoring of
mechanical parts by using Sweep frequency response
analysis.[5] describes a online condition monitoring approach
for power transformer’s bushing, tap changers and insulation
system.[6] demonstrates a technique for condition
monitoring of transformer oil using thermal analysis.[7]

282
International Journal of Pure and Applied Mathematics Special Issue

describes a method for condition monitoring of transformer’s

insulation system.[8] describes condition monitoring of
transformer insulation system by polarization and
depolarization of current measurement. In this work it is
shown with case studies and examples how health is
assessed and precautions can be taken to increase the life
time of a transformer.

2 Condition based monitoring methodologies

1)Tan delta on winding. 7) Partial Discharge Test.
2)Test of oil samples. 8)Acoustic Monitoring
3)Polarization Index. 9) High performance liquid
4)Dissolved Gas Analysis chromatography
5)Furan Analysis 10) Moisture Monitoring
6)Degree of Polymerisation

Fig 1: Automatic Ratio meter

Fig 1 shows an automatic ratio meter which is an associated
test kit during the work.

3 Polarization Index Test

The insulation resistance readings change w.r.t time
because of alignment of dielectrics . Actually, the dielectric
system contains polarized atoms with dipoles and the
polarized atoms aligns (polarize) with the applied dc voltage .
As they polarized , the dielectric resistance will increase and
after sometimes it gets saturated . The direct ratio of two
certain times can indicate the probable health condition of
the insulation in the system. The ratio is called the
“Polarization Index” .

283
International Journal of Pure and Applied Mathematics Special Issue

R (t 2 )
Therefore ,P.I =
R(t1 )
R(t1) = Insulation resistance after t1 times of applying d.c
voltage .
R(t2) = Insulation resistance after t2 times of applying d.c
voltage .
Generally we use t1 =10 sec t2 =60 sec
P.I gives an idea of the cleanliness of the insulation
system. The several interpretations obtained from the
polarization indexed test are described as follows :
P.I < 1.3 – not good.
2 < P.I <5 - permissible
P.I < 2- dirty or moist insulation .
P.I > 5- Very dry, brittle insulation system.
Exceptions : HV Vacuum pressure impregnated system may
yield a P.I between 1&2,for a perfectly good winding .

3.1Case Study :

The results shown are tested on a 16 MVA make

transformer, Voltage Class 6.6 KV /433V in Table-1
Table-1 : P.I test results
Connection I.R at 10 I.R at 10 sec(Mega P.I
sec(Mega Ohm)
Ohm)
Ph- Earth 2.3 2.58 1.12
Ph-Earth 1.2 2.25 1.96
Ph-Ph 1.04 2.08 2.00
Recommended Action :
1)The insulation is okay .
2)But little bit of oil purification will make its condition
better. If even after oil purification the condition of the
insulation is not improved then it is suggested to take the
help of other methods like nitrogen purging or induction
heating etc.
If P.I value is very high, even if >5 ,then it is not a matter
of concern .But it is really harmful , if its value is very low
,specially below 1 . It indicates the insulation contains
moisture . Moisture content make it conducting and thus
degrade its insulation property .
4 DGA (Dissolved Gas Analysis)

284
International Journal of Pure and Applied Mathematics Special Issue

It is very important for assessment of health of transformer.

Generally oil samples from Buchlozz relay is collected , gas
extracted , then they are separated , then analyzed , then we
make interpretations from collected data. Some general
interpretationsof DGA analysis is described below :
4.1 Some interpretations from DGA
Free Gas–
i)Colourless and odourless : Trapped air .
ii)Greyish with pungent smell: non-flamable overheating of
insulation and insulation board .
iii) Yellowish inflammable : decomposition of wood .
iv)Dark Grey : Flash over in oil / Overheat .
The results are shown on a 50MVA Transformer , 132/33 KV
in Table-2
Table-2: DGA Test results
Test Values Top Sample Bottom Sample-I Bottom Sample-II
BDV(KV) 54 55 57
Water content(ppm) 13 12 12
DGA Results
TGC(%) 7.1 7.2 7.2
N2(%) 5.11 5.23 5.19
O2(%) 1.51 1.49 1.53
H2(ppm) 1111 1159 1164
CH4(ppm) 450 455 455

4.2 Recommended Action :

i)Content of H2,CH4 indicates partial discharge,C2H2
indicates arcing ,Content of C2H4 and C2H6 indicates
decomposition of oil (Kitone formation).
ii) The DGA shows high temperature arcing problem of the
transformer . Some part of oil is also decomposed . Need
proper internal inspection of the transformer , whether
internal condition of winding insulation is o.k or not .

5 Tan Delta Test

Tan delta test indicates the dielectric strength , dielectric
loss and the moisture content of the various insulations.
Delta is the angle between ideal capacitive current and lossy
capacitive current . Tan delta test is very important to know

285
International Journal of Pure and Applied Mathematics Special Issue

the condition of insulation of transformer bushing and other

cables. The various interpretations obtained from the Tan
Delta values are described in section 4.1 .
5.1Limit of Tan Delta Values:
For good transformer ==< 0.5%
Acceptable ==0.5%-2.0%
Unacceptable ==2.0%
5.2Case Study :
Results shown on a 138/11.5 KV Transformer , 16.5/33/82.5
MVA ,Ynd11 in Table-3 and Table-4.

Table3:Tan Delta measurement of the winding :

Test Condition Voltage(KV) % Tan Delta Capacitance(Pf)
HV-LV+Gnd(GST) 10 0.415 11433.48
LV-Gnd (GSTg) 5 1.187 15325.46
HV-Gnd(GSTg) 10 0.401 4133.06
HV-LV(UST) 10 0.402 7301.83

UST indicates Ungrounded specimen test .

GST indicates Grounded specimen test .
GSTg indicates Grounded Specimen test with guard.

Table-4:Tan Delta measurement of the Bushing :

Bushing Voltage(KV) % Tan Delta Capacitance(Pf)
1U 10 0.48 252.25
1V 10 0.60 254.53
1W 10 0.36 256.34
Diagnosis :
1)Tan Delta value of LV-Gnd indicates wet insulations .
2) Tan Delta value of HV bushing ( 1V)slightly exceeds the
limits . It may contain little moisture but within acceptable
limit.
6.SFRA ( Sweep Frequency Response Analysis )
Test :
It assesses the structural health of core and winding . It
indicates the response of transformer winding to variable
supply frequency . The frequency response from a newly
made transformer given by original equipment
manufacturer( OEM) and the frequency response at the time
of the condition monitoring of a transformer is compared and

286
International Journal of Pure and Applied Mathematics Special Issue

then analyzed by seeing the anomaly. Then desired

prevention is taken.
Fig 2 and Fig 3 shows the results.

Fig 2: Frequency Response of newly made transformer

Fig 3: Frequency Response at the time of condition based monitoring

Recommendations :
The 2nd picture shows distortion with original figure ,
indicates displacement of winding with respect to original
place of core – coil assembly at first.It indicates that it needs
liftingof core coil assembly from the structure and refit .

7 Furan Analysis
Furan analysis is the paper insulation degradation
test.Furan is derivative of an aromatic compound formed
due to paper degradation .Derivatives of Furan soluble in
oil are :
Furaldehyde , Acetyl Furans,Methyl Furaldehyde ,
Furfurylalchohol ,Hydroxy Methyl Furaldehyde .
Limit:
Acceptable : Total Furan < 100 ppb
Questionable : 100ppb < Total Furan < 250
7.1Case Study :
Transformer Details : 138/11.5 KV ,16.5/33/82.5 MVA. Test

287
International Journal of Pure and Applied Mathematics Special Issue

results are shown on Table 5

Table 5: Test Report
Analysis of Furanic Compounds Obtained Value (ppb)
5-Hydroxymethyl-2-Furfural ND
2-Furfuryl Alchohol ND
2-Furfural 240.8
2-Acetyl Furan ND
5-Methyl-2-Furfural ND
Total Furan content in Oil 240.8
ND: Not Detected.
Recommendations :
The transformer is having furan level 240.8ppb , which
indicates normal aging of transformer. It needs proper
attention before effective life is getting expired.

8 Degree of Polymerisation Test

It requires paper samples. Aging of paper leads to loss of
tensile strength and reduced life of electrical equipments.
Insulation is long chain of hydro-carbons . With respect to
time , it will deteriorates . At the beginning of life the degree
of polymerization is around 1100-1500 and at the end of life
it is around 100-120 .
Transformer Details : 138 /11.5 KV MVA Class : 16.5 /33
/82.5 MVA Vector Group: Ynd11.
8.2 Test Results :
The insulation paper samples were taken for measurement
of D.P from various locations .The sample details and result
as below (ref. Table-6).

Table6: D.P Results

Paper Samples taken from D.P Value
HV lead ,top 650
HV lead,bottom 600
LV lead,center 563
LV lead ,bottom 576
LV lead , top 580
HV lead , Center 669

288
International Journal of Pure and Applied Mathematics Special Issue

The above results indicates the normal aging of transformer

is already started. With the degree of polymerization the
residual life of transformer is also can be assessed. For
assessing the residual life of a transformer the degree of
polymerization vs residual life curve is required which is to
be provided by original equipment manufacturer ( OEM) at
the time of purchasing of the transformer.
9 Partial Discharge Test
Local dielectric breakdown of a nominal area of a solid
/liquid electrical insulation system under HV stress. Normal
range of partial discharge from a transformer is in terms of
pico-coulomb. Partial discharge creates a hissing sound.
Corona is a kind of partial discharge.
10 Acoustic Monitoring
With the load of the transformer increases noise will also
increase .Hence several measurement of noise needed to
monitor the transformer performances .e.g-
i) The noise level of new transformer .
ii)Changes of noise level over time at given load.
iii)relative level of noise with the background noise.

11 Conclusion
Finally, the condition monitoring of transformer is very
effective in case of large size power transformers which is
having rating in terms of 100,200,300 MVA or even more. A
standard 315 MVA transformer costs in terms of crores, may
be 5-6 crores with on an average life span 15-20 yrs. Instead
of replacing the whole transformer with erection , testing
and commissioning costs after the effective life time, it is
suggested to do condition monitoring and increase its
effective life time up to its optimum usage.Prone to damage
spares are needed to be replaced and obsolete spares are
needed to be replaced with spares with updated features. It
is seen from the experiences if the user invests just 10% of
the transformer’s cost on the account of its condition
monitoring the effective life time increment of transformer
will be of about 25%, 30% or even more. So, condition
monitoring of transformer is a very very important aspect

289
 International Journal of Pure and Applied Mathematics Special Issue

w.r.t its maintenance and it is expected that this article will

be very helpful for all industry practitioners to create a
general awareness about its different methodologies with
practical examples.

References
[1]Saptarshi Roy,” Performance assesment of SCADA based
wind turbine : condition monitoring approaches “,
International Journal of Electrical Power System and
Technology,vol.1, issue 2,pp: 1-9,September 2015 .
[2] Johnson &Philips ,” J & P Handbook of Transformers “.
Transformer , BHEL
[3]Ghani S, Thayoob Y.H.M and Hanum Y ,”Condition
Monitoring of Distribution Transformer’s Mechanical Parts
using Sweep Frequency Response Analysis ( SFRA)” ,2013,
The Malaysian International Tribology Conference
[MITC2013],Elsevier, pp:469-476
[4] Stirl T, Vilaithong R and Tenbohelen S. ,”On-line Condition
Monitoring and Diagnosis for Power Transformers their
Bushings,Tap Changer and Insulation
System”,http://www.unistuttgart.de/ieh/forschung/veroeffentl
ichungen/CMD2006Tobias.pdf
[5]Degeratu S, Rotaru P and Rizescu S ,” Condition Monitoring of
Transformer Oil using Thermal Analysis and Other
Techniques “, Journal of Thermal Analysis and Caliometry
,Vol.119, Issue : 3 ,pp. 1679-1692 ,March 2015 .
[6] Mehta A.K, Sharma R.N and Chauhan S,” Condition
Monitoring of Insulation System in Power Transformers “,
International Journal of Computer and Electrical
Engineering ,Vol.4 ,No. 2 ,pp. 149-152 , April 2012 .
[7] Saha T.K, Purakait P and Yao Z.T ,” Condition Monitoring of
Transformer Insulation by Polarisation and Depolarisation
Current Measurements “, January 2002
,https://espace.library.uq.edu.au/view/UQ:9800/Condition_M
onito.pdf

290
291
 292

View publication stats