CIRED 20th International Conference on Electricity Distribution
Paper 0686
TNB EXPERIENCE IN CONDITION ASSESSMENT AND LIFE MANAGEMENT OF
DISTRIBUTION POWER TRANSFORMERS
Young Zaidey YANG GHAZALI Mohd Aizam TALIB
Tenaga Nasional Berhad – Malaysia TNB Research Sdn. Bhd. – Malaysia
young@tnb.com.my mohdaizam@tnbr.com.my
ABSTRACT
Tenaga Nasional Berhad (TNB), an electric utility company
in Malaysia has embarked on condition assessment and life
management program for 707 units of in-service oil
immersed free breathing distribution power transformers
with on-load tap changer. The program consists of three-
tiered approach. The first tier involved transformer
population screening by means of on-line diagnostic
techniques through assessment of the insulating oil,
physical, thermal and operating performance. The second
tier focused on off-line diagnostic testing on few identified
transformers from the fleet using basic electrical testing
techniques. The third tier involved more advanced
diagnostic techniques namely frequency response analysis
and partial discharge test. Results of the condition
assessment on individual transformer were analyzed and
transformer health index was formulated to provide
engineering solutions and asset management direction.
INTRODUCTION
With the restructuring of the electricity sector into profit
oriented business entity, condition based maintenance
(CBM) strategies are now gaining importance. The main
goal is to reach a cost effective solution through effective
asset management. In the case for power transformers, an
effective CBM program is crucial as more transformers
have reached their service life. Diagnostic techniques able
to reveal their degradation state are a prerequisite to an
effective transformer life management program [1].
Diagnostic techniques such as Oil Quality Analysis (OQA),
Dissolved Gas Analysis (DGA) and Furfural Analysis
(FFA) are widely used to assess the insulation condition of
transformers. The Degree of Polymerization (DP)
measurement is also applied to predict the end-of-life of
transformers. However, direct DP measurement is not
always possible and diagnostic testing on insulating oil only
provides partial information about condition of the
transformers.
Various basic to more advanced on-site electrical diagnostic
techniques are also available. The combination of chemical
and electrical diagnostic testing can provide a better
assessment on the condition of the transformers. Any fault
in the transformer core, winding and tap changer for
example, can be diagnosed by excitation current, turns
ratios, winding resistance and insulation resistance
CIRED2009 Session 3 Paper No 0686
Prague, 8-11 June 2009
Hannah AHMAD ROSLI
Tenaga Nasional Berhad - Malaysia
hannaha@tnb.com.my
measurement. Partial discharge (PD) activities in the
transformer winding, core and other active parts can be
detected directly by on-line PD testing either by means of
acoustic signals detection or measurement of the electrical
signals. The mechanical integrity of the core and winding
assembly structure can be confirmed by Frequency
Response Analysis (FRA). Problems in insulation system
can be detected by dielectric dissipation factor or tan delta,
Recovery Voltage Measurement (RVM) and Frequency
Dielectric Spectroscopy (FDS) techniques. Finally, thermal
condition of the external surface temperature and problems
such as cooling system blockages, loose electrical
connection, unbalance loading effect and hotspots can be
detected by infrared emission testing.
This paper presents TNB experience in conducting
condition assessment and life management program on
distribution power transformers through diagnostic
approach and formulation of transformer health index.
BACKGROUND
Condition assessment program on power transformers in
TNB distribution system involved 707 units of in-service oil
immersed 3-phase free breathing power transformers with
on-load tap changer (OLTC). The voltage and capacity
rating of the transformers and the age distribution are shown
in Table 1 and Fig. 1 respectively. The program applied a
CBM methodology proposed by CIGRE Transformer
Committee (SC 12) [2]. The three main stages of the
program involved 1) transformer population screening
through assessment of the insulating oil, physical, thermal
and operating performance, 2) electrical diagnosis on few
transformers from the fleet which could not definitively be
classified as normal, and 3) evaluation on the remaining life
and ranking of the transformers to provide asset
management direction and engineering solutions.
CBM DIAGNOSTIC TECHNIQUES
The selection on the diagnostic techniques adopted for the
CBM program used a Failure Mode Effect and Criticality
Analysis (FMECA) and is discussed elsewhere [3]. These
techniques are divided into three practical levels or tiers as
described below. Even though various diagnostic techniques
are available, the ones selected were also subjected to
availability of the instruments and financial constraint.
 CIRED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0686

Table 1: Voltage and capacity ratings of power transformers oil or FFA was performed. By measuring the quantity and
Voltage Capacity No. of Units types of furans present in a transformer oil sample, the
30 MVA 474 insulation’s overall DP and remaining life estimation can
33/11 kV 15 MVA 138 also be inferred with a high degree of confidence since it is
7.5 MVA 23 not practical to obtain a paper sample from de-energized
22/11 kV 12.5 MVA 24
12.5 MVA 39 distribution power transformers [7].
7.5 MVA 3
22/6.6 kV
5 MVA 3 Inspection on Physical Conditions and Operating
3.5 MVA 2 Performance
11/6.6 kV 3 MVA 1
Transformer life is affected by electrical loading profiles
and thermal effect. Other factors such as mechanical
299
300
damage, corrosion, oil leaks also have strong effects. In
addition, the transformer maintenance and performance
250 historical records such as OLTC maintenance records and
No. of Transformers

200
records of transformer internal and external tripping are also
essential in assessing the condition of the transformers.
150
117

100 91 Thermography
60 62 Infrared thermography is a technique that can provide the
50
42
16 11
first indication of thermal problems due to overheating of
7 2
0
conductor or between bad contacts. Infrared scanning is
≤5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 ≥ 46 applied on the external components such as tanks, bushings,
Age (Year)
radiators and cooling systems. The hotspot temperature
Fig. 1: Age distribution of distribution power transformers difference was used as main criterion for the assessment.

Tier 1 Diagnostic Techniques

Tier 1 diagnostic techniques were applied on all 707
energised units as baseline audit to assess for presence of
faults, quality of the insulating oil, degradation level of
insulating paper as well as physical, thermal and operating
performance of the transformers.
Tier 2 Diagnostic Techniques
Tier 2 diagnostic techniques were applied on few
problematic transformers from the fleet which Tier 1
diagnosis could not definitively classify these transformers
as normal.

Dissolved Gas Analysis (DGA) Transformer Turns-Ratio Measurement

DGA is an establised technique to detect presence of fault in
the transformers. DGA results for transformer main tank
were analysed for key gases and total dissolved combustable
gases (TDCG) concentrations as well as key gases and
TDCG rate of generation in accordance with IEEE Std
C57.104 [4] and IEC 60599 [5] limits and interpretation. To
enable the assessment of rate of gas generation, diagnosis
was performed based on minimum two oil samples at
different time intervals. Results were also analysed and
interpreted for presence and type of faults using IEC ratio
and Duval’s Triangle methods.
Oil Quality Analysis (OQA)
In OQA, oil samples were tested for breakdown voltage,
water content, acidity and power factor. These tests are the
basic routine tests for mineral insulating oil in accordance
with IEC 60422:2005 [6], and are sufficient to indicate the
condition of the insulating oil. Limits and interpretation of
test results in accordance with IEC 60422:2005 were used.
Furfural Analysis (FFA)
As the life of the cellulosic material is directly related to the
life of the transformer, analysis of the furanic compound in
Transformer turns-ratio was measured at every tap position
of each phase against calculated nameplate values [1]. The
deviation between the measured and calculated ratios at
each tap position was captured to detect shorted turns or
open circuit in the transformer main and tapping windings.
Winding Resistance Measurement
Winding resistance test is a dc test performed at every tap
position to detect broken conductor strands, loose
connections, and bad contacts in the tap changer. Test
results were compared between phases or with factory
results at corrected temperature and deviations were
measured at all tap positions for primary and secondary
windings [1]. Assessment was carried out by considering the
maximum deviation between any phases at each tap.
Dielectric Dissipation Factor/Tan Delta Measurement
Paper insulation in transformer winding forms an effective
capacitance network and it has a measurable amount of
dielectric power loss regardless of age [8]. The dielectric
loss will develop heat in the insulation during transformer
operation that can cause deterioration of the insulation. In
this test, the tangent of the insulation dielectric loss angle is

CIRED2009 Session 3 Paper No 0686

CIRED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0686

measured between primary and secondary windings (CHL), Start

primary winding and ground (CH), as well as secondary

winding and ground (CL) at different test voltages Tier 1 Tests

depending on the rated operating voltages of the winding. Poor/Very Poor

Oil Oil Quality All Good/Fair
The tangent delta measured for each CHL, CH and CL was reconditioning
Health
Index
taken from the mean values at different test voltages. or regeneration.
Poor/Very Poor

Excitation Current Measurement Minor repair/calibration

of external components
The purpose of this test is to detect short-circuited turns,
core and winding problems and poor tap changer contacts Tier 2 Tests
[8]. On a 3-phase core type delta-star or star-delta
connection, the excitation current pattern will have two Health
Good/Fair Maintain/Revise

phases on the outer limbs of a 3-legged core, higher than the Index Frequency of
Tier 1 Tests
phase on the center limb. Test results of the two higher Poor/Very Poor

currents at each tap were compared between phases. Tier 3 Tests

Insulation Resistance and Polarization Index (PI) Health

The insulation resistance measurement can be of value in Index Good/Fair

Poor/Very Poor
determining the presence or absence of harmful
Reduce Loading and prepare
contamination, degradation as well as failure of the for major repair, upgrade or
insulation. Since insulation varies inversely with replacement/relocation

temperature for most insulating materials, PI was applied as

a method to assess the transformer insulation condition [8]. End

Tier 3 Diagnostic Techniques Fig. 2: CBM workflow for in-service power transformers

Tier 3 diagnostic techniques are advanced techniques

applied on few problematic transformers after Tier 2 and
Tier 1 diagnosis could not definitively classify these
transformers as normal.
Frequency Response Analysis (FRA)
This test is used to detect faults such as short-circuited
turns, movement, mechanical deformation or displacement
to windings or core and loose turns. It provides a picture of
the frequency transfer function of the transformer. Results in
the form of the cross correlation index (CCI) were assessed
either by comparison to previous measurements, or
comparison to transformers of similar design and
construction, or between phases. The CCI limits
recommended by the instrument manufacturer were used.
Partial Discharge (PD) Measurement
Insulation failure inside transformers is preceded by PD
activity which is commonly related to moisture in the
insulation, cavities in solid insulation, metallic particles, and
gas bubbles generated due to some fault condition. A
significant increase in the PD level can provide an early
indication of a failure. PD measurement with acoustic
signals detection method was applied in the assessment.
TRANSFORMER HEALTH INDEX
With so much data provided by various diagnostic
techniques, asset managers are having difficulties in
evaluating the risks and prioritizing the corrective solutions
on problematic transformers in a consistent manner. Hence,
health index is formulated for capturing and quantifying the
test results to provide a qualitative indication of the
condition of individual transformer. In order to provide the
overall health condition, the transformer age was also taken
into account. The process for arriving at the Transformer
Health Index (THI) is as shown in Table 3. The transformer
health index is ranked based on the Transformer Condition-
Based Ranking in Table 2, to determine the next
recommended mitigating actions.
Table 2: Transformer condition based ranking
THI Indication Recommended Mitigating Action
Maintain normal frequency of current tier
85 ≤ THI ≤ 100 Good
tests
Revise frequency of current tier tests to 6
55 ≤ THI < 85 Fair months interval or as recommended by
the expert
10 ≤ THI < 55 Poor Perform next tier tests (2nd priority)
THI < 10 Very Poor Perform next tier tests (1st priority)

CBM IMPLEMENTATION RESULTS

From the assessment, 63 units out of 707 power
The workflow practised by TNB for implementing CBM for
transformers from the fleet required Tier 2 tests and out 63,
in-service free breathing power transformers is as illustrated
50 units identified for Tier 3 tests. The Final THI was then
in Fig. 2.
determined for all the 707 power transformers. Based on the

CIRED2009 Session 3 Paper No 0686

 CIRED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0686

final THI, the transformer condition were ranked as Fig. 3 shows that, as the transformer aged, the condition of
“Good”, “Fair”, “Poor” and “Very Poor”. The conditions of the transformers deteriorated. However, a number of in-
the transformers according to age and the entire population service transformers aged 5 years and below, were also
are illustrated in Fig. 3 and Fig. 4 respectively. diagnosed as in “Poor” conditions. Most of these
transformers were found to have considerably high
Table 3: Transformer health index calculation concentration of gases that relate to overheating. On the
TIER 1
contrary, most common problem found for transformers
Amplified Total aged above 30 years were paper degradation and
Condition Indicator
Weighting Ranking ranking Ranking deterioration in physical conditions. Based on the problems
factor number Score found from the assessment, the corrective actions that were
(a) (b) (c) (d)
DGA 1.2 3,2,1,0 20,12,-18, -20 (a) x (c)
recommended and carried out are as shown in Fig. 5.
OQA 1.2 3,2,1,0 20,12,-18, -20 (a) x (c)
133
FFA 1.2 3,2,1,0 20,12,-18, -20 (a) x (c) 140

Thermography 0.6 3,2,1,0 20,12,-18, -20 (a) x (c)

120
Physical & operating 102
0.4 3,2,1,0 20,12,-18, -20 (a) x (c)
condition 100
Age 0.4 3,2,1,0 20,12,-18, -20 (a) x (c)

No. of Cases
Tier 1 THI (Sum of individual ranking score) THI 1 80 66
Subtract % estimated life used above 100% (-10,-15,-20,-30) A
Final Tier 1 Transformer Health Index THI 1-A 60

Tier 1 Weighting Factor 0.5 36

40
TIER 2 19
Turns-ratio 1.2 3,2,1,0 20,12,-18, -20 (a) x (c) 20 11
Winding resistance 1.2 3,2,1,0 20,12,-18, -20 (a) x (c)
Tangent delta 1.2 3,2,1,0 20,12,-18, -20 (a) x (c) 0
Replace Refurbish Monitor Monitor gas Monitor Recondition
Excitation current 0.6 3,2,1,0 20,12,-18, -20 (a) x (c) transformer transformer furanic generation thermal & oil
content loading
I. Resistance & PI 0.8 3,2,1,0 20,12,-18, -20 (a) x (c) conditions
Final Tier 2 THI (Sum of individual ranking score) THI 2 Mitigating Actions
Tier 2 Weighting Factor 0.5
Total THI (Sum of individual indices) {(THI 1-A) x 0.5} +{THI 2 x 0.5}
Fig. 5: Mitigating actions on problematic transformers
TIER 3
Frequency Response Analysis (FRA) 0,-5,-10,-15 X CONCLUSION
Partial Discharge (PD) 0,-5,-10,-15 Y
Final Transformer Health Index Total THI – X – Y Through condition assessment based on CBM methodology,
TNB was able to plan and implement a life management
100% 1 5 2 3 program for its distribution power transformers. With the
9
22 9
5 3 use of transformer health index (THI), TNB was also able to
89
80% 8 33 22 perform a more effective and consistent assessment on the
Relative Proportions

1
19 4
entire transformer population. THI has become a tool not
60% 53
9 4
3 only in determining the correct remedial actions but also in
21
12 prioritizing the most appropriate engineering solutions.
40%
43
75
152 47
2
1
REFERENCES
5 7 5
20% 25
[1] M. Wang, A.J Vandermaar, K.D Srivastava, 2002, “Review of
7 1 Condition Assessment of Power Transformers in Service”, IEEE
0% Electrical Insulation Magazine, Vol. 18, No. 6
≤5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 ≥ 46 [2] CIGRE W.G 12.18 “Life Management Techniques for Power
Age (Year) Transformers”, CIGRE SC 12
[3] Y.Z Yang Ghazali, M.A Talib, H.A Rosli, 2008, “Condition
Good Fair Poor Very Poor
Assessment of Power Transformers in TNB Distribution System and
Fig. 3: Transformer health index according to age Determination of Transformer Condition Index” CEPSI
[4] IEEE Std. C57.104, 1991, “IEEE Guide for the Interpretation of
Gases Generated in Oil-Immersed Transformers”
Very Poor (33) [5] IEC 60599:1999, “Guide to The Interpretation of Dissolved and Free
5%
Poor (206) Gases Analysis”, 2nd Ed.
29% Good (349) [6] IEC 60422:2005, “Mineral Insulating Oil in Electrical Equipment –
49% Supervision and Maintenance Guidance”, 3rd Ed.
[7] R.D Stebbins, D.S Myres, A.B Shkolnik, 2003, “Furanic
Compounds in Dielectric Liquid Samples: Review & Update of
Diagnostic Interpretation & Estimation of Insulation Ageing”, IEEE
Proceeding on Properties and Applications of Dielectric Materials
Fair (119)
17% [8] M. Horning, J. Kelly, S. Myers, 2001, ”Transformer Maintenance
Guide”, 2nd Ed.
Fig. 4: Condition of transformers population based on THI

CIRED2009 Session 3 Paper No 0686