2019 IEEE International Conference on Innovative Research and Development (ICIRD)
Prediction of Power Transformers Lifetime Using
Thermal Modeling Analysis
Sarah Afifah1, Jannus Maurits Nainggolan1, Gunawan Wibisono1*, Chairul Hudaya1,2,*
1
Departement of Electrical Engineering, Faculty of Engineering, Universitas Indonesia, Depok, 16424, Indonesia
2
Energy System Engineering Master Study Program, Faculty of Engineering, Universitas Indonesia, Depok, 16424, Indonesia
*Corresponding author : gunawan@eng.ui.ac.id and c.hudaya@eng.ui.ac.id
Abstract — Transformer is one of the main components in
distribution system of electrical power system towards the
consumers, thereby any damage to the transformers will hinder
the distribution of electricity towards the consumers, and in
turn will make the SAIDI and SAIFI levels go up. High
temperature in transformers can cause degradation in the
insulation of transformers which in turn will cause failure in
transformers. When the temperature in winding reaches or goes
beyond the limit of 110°C, a degradation in insulation will start
happening and the remaining life of transformers will decrease.
Damage in transformers will cause disturbance in electrical
power system and result in a major economic loss. Before
damages occur, transformers need to be changed up until it is
deemed to be no longer efficient, this can be done by replacing
the transformer that is about to be damaged. To predict when a
transformer is about to break, a calculation is made based on
thermal modelling according to IEEE Std C57.91-1995 with its
most prominent variable being Hot Spot Temperature (HST).
HST is obtained by MATLAB programming using Annex G of
IEEE Std. C57.91-1995. By obtaining HST, thus the remaining
lifetime of transformers can be predicted. This study analysed
the effect of loading, hot-spot temperature, and ambient
temperature on the remaining lifetime of a transformer. The
higher those three factors, the quicker the transformer will
break, with loss of life percentage increasing exponentially. The
highest loading, hot spot temperature, and ambient temperature
given in this research gives percent loss of life of 0.0888332%,
0.0193394%, 0.020753% respectively.
Keywords — power transformer, thermal modelling, life
prediction, loss of life
I. INTRODUCTION
Viewed from a general perspective, an electric power
system is known as a large network that supplies power and
provides electricity from the generator to the load center. The
relationship between the generator and the load consists of
several subsystems, namely generation systems, transmission
systems, and distribution systems. Distribution system is a
subsystem of electric power which plays a role in the
distribution of electricity directly to the customers. So that any
damage to one component of the distribution system will also
have a direct effect on the customer.
The main component in the power system is the power
transformer. If there is a disruption to the transformer, the
distribution of electricity to the customer will stop, and in turn
the level of SAIDI and SAIFI will increase. Based on data
978-1-7281-2825-2 /19/$31.00 ©2019 IEEE
Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
collected from electric utility company in Indonesia (PLN
Lenteng Agung UP3) in 2017, the achievement of SAIDI and
SAIFI did not meet the target, where SAIDI had a target of
159.63 minutes/customer and the realization was 219.10
minutes/customer. While for SAIFI, the frequency target was
2.44 times/customer and the realization was 3.01
times/customer. The disturbances that occur in the transformer
are usually from the degradation and the aging of the
insulation system which consists of oil insulation and paper
insulation. These degradations are caused by deterioration
caused by electrical, mechanical, and thermal stresses
experienced during the operation of the transformer [2].
Therefore, the isolation on the transformer has an important
role in preventing disturbance.
Aging or deterioration of insulation is closely related to
temperature, humidity level and the amount of oxygen in the
air. Nowadays, transformer construction has eliminated the
effect of moisture and oxygen on the transformer oil, so that
the influence of these two factors on the isolation and
reduction of the life of the transformer can be ignored. The
remaining factor which greatly affects the life of the
transformer is temperature [3].
The temperature of the transformer surface is not uniform,
so to find out the most influential temperature which reduces
the life of the transformer, hotspot or the hottest point on the
transformer is used. Because of the many factors that affect
temperature, the reduction in transformer service life cannot
be predicted with high accuracy even if it is under controlled
and constant conditions, and it will be even more difficult in
uncontrolled conditions [3]. Therefore, to get the most
accurate prediction of the remaining life of the transformer, it
is imperative to use internationally defined standards, namely
the IEEE and IEC standards.
In this study, in order to improve the reliability of SAIDI
and SAIFI, to prevent failures, and to extend the operational
life of transformers so that blackouts for customers can be
avoided, an approach is executed on predicting the remaining
life of transformers based on hotspot temperature. The
implementation of the following prediction strategies can
make the transformer operate with high efficiency at the
appropriate time period before any damage can occur to
transformers and cause blackouts.
 II. THERMAL MODELING
Thermal modeling is a study used to estimate top-oil
temperature (TOT) and hot-spot temperature (HST) in a
transformer. TOT is defined as the average oil temperature at
the tank outlet and oil pocket temperature. Whereas, HST
consists of ambient temperature (AT), increase in TOT, and
increase in HST. One approach to implement thermal
modeling transformers is to use a numerical network thermal
model. At present there are two standards that use this
approach, namely IEEE C57.91-1995 and IEC 60076-7.
A. Effects of HST and TOT on the Towards Aging of
Transformers
The aging of the transformer can be evaluated using HST.
The value of HST depends on ambient temperature (AT),
increase in TOT against AT, and increase in HST against
TOT. The increase in the TOT value which is also an increase
in HST has an effect that can reduce the insulation life of the
transformer [4], [5], [6]. Abnormal conditions, such as
overloading, supplying non-sinusoidal loads, and the
influence of high ambient temperatures, can accelerate the
aging of the transformer, which will eventually speed up the
use of transformers. So it can be concluded that the increase
in TOT and HST can shorten the life of the transformer.
The main contributor to the temperature of insulation is
heat generated from load losses. Because the deterioration of
isolation is related to the insulation temperature and
temperature spread due to uneven load losses in the
transformer, it can be said that the greatest deterioration in
insulation will occur in parts operating at the highest
temperature conditions [7]. Therefore, the study of
transformer aging factor is often considered as the effect of
aging caused by HST.
B. Thermal Modeling in Transformers
When the transformer is energized and loaded with
ambient temperature (A), the dissipation caused by core loss,
winding loss, and stray losses in the tank, as well as metal
support structure, as sources of heat, will cause oil temperature
and winding temperature to increase. Transformer oil is
cooled by a radiator arrangement which flows to the bottom
of the cooling ducts so that it reaches bottom oil temperature
(bo). Transformer oil flows vertically over the winding ducts
and exits through the upper winding ducts. Then the
transformer oil enters the radiator on the top oil temperature
in the main tank (top) [3].
This research used IEEE std. C57.91-1995 to calculate
hotspot temperature, bottom oil temperature, top oil
temperature, and top of duct oil temperature. Fig.1 shows
transformer thermal diagram [3].
Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
Fig. 1 Transformer Thermal Diagram
In this model, the hot-spot temperature is the sum of
ambient temperature, top oil temperature rise (Ʌ୲୭୮ ), and
hot-spot to top oil temperature gradient (Ʌୌ = H.g), where H
is hot-spot factor and ‘g’ is thermal gradient between winding
and oil average temperatures [8].
This diagram is based on the following assumptions:
• The change in the oil temperature inside and
along the winding is linearly increasing from
bottom to top.
• The increase in the winding temperature from
bottom to top is linear with a constant
temperature difference ‘g’.
• At the top of the winding HST is higher than the
average temperature rises of the winding. The
difference in the temperature between the hot
spot and the oil at the top of the winding is
defined as H x g, where H is a hot spot factor. It
may be varied from 1.1 to 1.5, depending on short
circuit impedance, winding design and
transformer size[3].
III. RESEARCH METHODOLOGY
A. Calculation of Hotspot Temperature using Annex G
To calculate the hot spots, this study used Annex G from
IEEE Std C57.91-1995. The mechanical and dielectric
properties of an insulation will decrease at a temperature that
exceeds the normal limit. Based on the IEEE Std C57.91-1995
standard, if the hot-spot temperature of a transformer exceeds
110°C, the insulation will be aging rapidly and even exceeding
its normal limit. Therefore, the highest temperature on the
transformer can be helpful in calculating the isolation state.
In this study, the temperature obtained at‫ݐ‬ଵ was used for
the next time (‫ݐ‬ଵ + t) = ‫ݐ‬ଶ . This study used an increment,
where the last calculated temperature was used to calculate the
temperature at the time after. Formula for the hottest-spot
temperature of a transformer is:
ߠு = ߠ஺ + ߠ஻ை + ߠௐைȀ஻ை + ߠுȀௐை (1)
Where ߠ஺  is the ambient temperature (°C), ߠ஻ை is the
bottom oil temperature (°C), ߠௐைȀ஻ை is the oil temperature
rise over the bottom oil at the hotspot location (°C), and
ߠுȀௐை is the winding hotspot temperature rise over the oil at
the hotspot location (°C). The calculation for each of those
temperature is given in detail in Annex G of IEEE std. C57.91-
1995.
B. Calculaltion of Loss of Life Percentage in a Transformer results was then used to calculate the aging acceleration factor
Using Annex I (‫ܨ‬஺஺ ), the equivalent aging acceleration factor (‫ܨ‬ா஺஺ ), and the
To calculate the loss of life percentage in a transformer, loss of life percentage in transformers. These three
this study used the Annex I approach based on IEEE std components were calculated using Annex I from IEEE Std
C57.91-1995.
C57.91-1995. Based on these standards the age of isolation is
an exponential function of HST [1]. Measurements were made on three different days, with
different loads. The first day measurement, on Monday, 18
ా March 2019, shows the highest load with the average hourly
% of Insulation life = A.݁ ഇ೓ శమళయ
(2) loading of 0.79162 p.u; the second day, on Friday, 22 March
2019, exhibits an average load of 0.71721 p.u, and the third
Where h is HST (°C) that has been obtained from the day of 15 June 2018 postulates an average load of 0.44908 p.u.
calculation of Annex G, A and B which are constants and These three types of load conditions were used as validation
determined based on the insulation material and HST which in analysing the effect of loading, room temperature and hot-
is defined as the age of initial isolation. Equation (2) can be spot temperature on the life of the transformer.
used for distribution transformers or power transformers These three load conditions used uniform room
because both are made using the same cellulose isolation. For temperature. The room temperature in this study used the
example, the HST for initial isolation is 110°C. means that if temperature of the weather forecast. The reason was because
the transformer works on this HST, then its life will be 1p.u the weather forecast is quite accurate considering that the
[3]. Using the above assumptions, equation (2) will give: substation being measured is an outdoor substation so that it
does not require further measurements when compared to the
భఱబబబ
indoor substation.
Per unit life = 9.8 x ͳͲିଵ଼ ݁ ഇ೓ శమళయ (3) The transformer used in this study has a 30-year lifetime
based on the information provided by the factory. Tables 1-
The above equation is used to calculate the aging Table 3 shows the results obtained after the MATLAB
acceleration factor (‫ܨ‬஺஺ ሻ. ‫ܨ‬஺஺ is the speed at which the life of performed the Annex G and Annex I calculation.
the transformer insulation is accelerated compared to the life Table 1 Simulation Result on Monday March 18th, 2019
at 110°C [3]. Parameters Value
భఱబబబ భఱబబబ 96.699 °C at 18.775 hours
൬ ି ൰ Maximum hot-spot temperature
‫ܨ‬஺஺ = ݁ యఴయ ഇ೓శమళయ
(4) Maximum top-oil temperature 75.504°C at 19.008 hours
Final hot-spot temperature 86.1142°C
‫ܨ‬஺஺  will be greater than 1 when HST is greater than Average winding temperature 48.6882°C
110°C and less than 1 when HST is smaller than 110°C [3]. Final top oil temperature 72.0777°C
According to Annex G. HST varies depending on load and Final top of duct temperature 42.1735°C
Final bottom oil temperature 42.1203°C
room temperature therefore ‫ܨ‬஺஺  is used to determine the Equivalent aging 2.47632 Hours
equivalent aging acceleration factor of the transformer. The Equivalent aging factor (۴۳‫)ۯۯ‬ 0.10318 per unit
equivalent aging acceleration factor at the reference % Loss of life 0.0038097%
temperature, at a predetermined time which is over a 24-hour
loading period, is define as
Table 2 Simulation Result on Friday March 22nd, 2019
σಿ
೙సభ ிಲಲǡ೙ ௱௧೙
‫ܨ‬ா஺஺ = (5) Parameters Value
σಿ
೙సభ ௱௧೙ Maximum hot-spot temperature 88.8398°C at 18.6083 hours
Maximum top-oil temperature 70.714°C at 18.825 hours
Where ‫ܨ‬ா஺஺ is the equivalent aging acceleration factor Final hot-spot temperature 79.4768°C
Average winding temperature 45.3137°C
over a period of time. N is the total value of the time 67.1172°C
Final top oil temperature
period, ߂‫ ݊ݐ‬is the nth time interval and ‫ܨ‬஺஺ǡ௡ is the aging Final top of duct temperature 39.5281°C
acceleration factor for the temperature at the time interval Final bottom oil temperature 39.4838°C
߂‫ݐ‬௡ [3]. Equivalent aging 1.03351 Hours
Equivalent aging factor (۴۳‫)ۯۯ‬ 0.0430629 per unit
Lastly, loss of life percentage of a transformer within 24- % Loss of life 0.0015900%
hour period can be determined by multiplying ‫ܨ‬ா஺஺ by the
time period in hours, where in this case is 24 hours, so that
Table 3 Simulation Result on Eid Fitri June 15th, 2018
the equation of loss of life percentage is: Parameters Value
Maximum hot-spot temperature 64.3394°C at 18.2917 hours
ிಶಲಲ ୶௧୶ଵ଴଴ Maximum top-oil temperature 57.2608°C at 18.0083 huors
% Loss of lifetime = (6)
୒୭୰୫ୟ୪୧୬ୱ୳୪ୟ୲୧୭୬୪୧୤ୣ Final hot-spot temperature 60.514°C
Average winding temperature 36.7511°C
IV. RESULTS & DISCUSSION Final top oil temperature 54.1556°C
Final top of duct temperature 33.2787°C
MATLAB program was used to simulate the input loading Final bottom oil temperature 33.2559°C
and transformers data so that the hot spot temperature, top-oil Equivalent aging 0.0653022 Hours
temperature, top of duct temperature, and bottom oil Equivalent aging factor (۴۳‫)ۯۯ‬ 0.00272092 per unit
temperature could be obtained. Calculations used to obtain % Loss of life 0.0001005%
these temperatures used Annex G of IEEE Std C57.91-1995.
The acquisition of the hot-spot temperature from the Annex G

Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
A. Analysis of the Effects of Loading on Loss of Life
Percentage
In this study, five types of scenarios were carried out on
transformer loading where the p.u value was varied into five
variations for the three days listed above. There were some
variations of p.u which were reduced, and some are added to
see the effect of loading on the percent reduction in
transformer’s life. The variations of the p.u were; p.u-0.1, p.u
normal, p.u+0.1, p.u+0.2, and p.u+0.3. The correlation
between those varied loading can be seen in the graph on fig
2.
Fig. 2 Effects of Transformer Loading on Equivalent Aging
It can be seen from figure 2 that the greater the value of
loading, the greater the value of equivalent aging will be, in
which the result looks exponential because the formula of ‫ܨ‬஺஺
is exponential.
On Monday, a scenario was carried out in which the
addition of 0.3 to p.u the load caused a load exceeding the limit
of the rating nameplate that reached 1.22 p.u. This situation
may occur when the load demand from the customer has
exceeded the capacity of the existing transformer, while the
mitigation process such as feeder rupture is still under control,
so to avoid the occurrence of blackouts the operator is forced
to continue to let the transformer exceed the rating nameplate.
In the case of p.u + 0.3, it can be seen that the value of
aging results reached 58 hours, which means that within 24
hours of operation the transformer is the same as it has been
operating for 58 hours. This certainly indicates that there will
be a significant decrease in the life of the transformer.
For other days, the addition of load did not adversely affect
the value of the transformer aging results, because even if the
load was added both on Friday and Eid, the value of aging
results remained below 24 hours, which means that the
transformer will not experience a reduction in the proper life.
Decrease in service life due to increasing load will make
the current increase as well. When the current operating
exceeds the rated current, the copper losses that arise will
make the winding temperature goes up. So that the isolation
from the winding will be degraded, the greater the degradation
experienced by the insulation, the faster the transformer will
be damaged. The extreme thing that might happen if the
transformer overloads for a long period of time while the
insulation is in a bad condition is the possibility of a short
circuit or earth fault will increase. But if the overload occurs
only in a short time, significant damage will not occur.
Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
Fig. 3 Effect of Loading Towards Percent Loss of Life
The results of the effect of loading on the loss of life
percentage of the transformer in Figure 3 are in line with the
effect of loading on equivalent aging results in figure 2, the
greater the value of loading, the greater the loss of life
percentage will be, where the result is exponential, because
the formula of ‫ܨ‬஺஺ is a function of exponential.
As can be seen in the graph of figure 3, the addition of 0.2
and 0.3 to p.u loading on Monday produced a reduction
percentage in the life of the transformer by 0.0305867% and
0.0888332%. Based on IEEE Std C57.91-1995, the standard
reduction in life per 24 hours is 0.0133%. Results on Monday
showed that it exceeded this value, means that the transformer
would have a shorter life than it should. The same thing
happened to the addition of 0.3 to p.u loading on Friday, where
the result of the reduction in transformer life was 0.0352362%.
Figure 3 and 2 above prove that the greater the load given
to the transformer, the greater the life reduction of the
transformer would be. Keep in mind that the purpose of life
here is the age of the insulating material that protects the
transformer, because isolation is a component of the
transformer which is affected by temperatures arising from the
current. Life here is not the functional age of the transformer.
Given the increase in loading, the current will be even greater
so that heat will arise in the transformer winding. If the current
exceeds the capacity of the transformer then the heat that
arises will be even greater, this heat will cause the isolation
system of the transformer to degrade that will accelerate the
life of the transformer eventually.
Based on SPLN D3.002-1: 2007 and PLN 017: 2014
Board of Directors Letters, the transformer will be within safe
limits and function optimally if it is used at 50% -60%, the
load in that interval will result in a small reduction in the life
of the transformer. Therefore, many transformers are still in
good condition even though their ages have passed the
manufacturer's prediction, this is because the operating
conditions are below the limits specified by the manufacturer
and of course the conditions of the transformer loading.
B. Analysis of the Effect of Hot-spot Temperature on Loss
of Life Percentage
In this study, five types of scenarios were carried out at the
transformer hot-spot temperature where the temperature value
was varied into five variations for Monday (18 March 2018),
Friday, (22 March 2018), and Eid, (15 June 2018). Variations
in temperature were reduced and some were added to see the
influence of the hot-spot on the reduction percentage in
transformer life. Variations in the temperature of the hot spots
were; -5°C, normal, +5°C, +10 °C, and +15°C.

Fig. 4 Effect of Hotspot Temperature Towards Equivalent Aging
It can be seen from figure 4 that the greater the value of
the hot spot, the greater the equivalent aging per day will be,
where the result looks exponential because the formula of ‫ܨ‬஺஺
is an exponential function.
On Monday, there was a scenario where the addition of
15°C in the hot-spot produced aging for 12,5706 hours, which
means that with 24-hour operation in the temperature
conditions in Figure 4 the transformer was experiencing the
operation of 12,5706 hours. For the addition of +10° C and
+5°C hotspots, the aging results showed smaller results,
namely 7.41854°C and 4.31734°C. The value is considered
safe because it is still under 24 hours, so the transformer did
not experience aging more than it should. Although the
temperature that had been added is quite significant, namely
15°C, but because the original temperature was indeed small
due to a small load, there was still no excessive aging.
On Friday, with a smaller load than Monday, it can be seen
in graph of Figure 4 that with the addition of 15 °C, the aging
results obtained was 5.6428 hours. This value is also still far
below the maximum, while on Eid day, with a load that was
much smaller than Monday and Friday, the value was even
smaller, plus the loading on the transformer did not reach 50%
which indicated that the load was very small .
From graph 4 it can be concluded that by increasing the
temperature of the hot spot, the value of the aging results of a
transformer will increase as well. This happens because the
heat from the temperature of the hot spot will affect the
insulation of the transformer. The heat that arises due to the
increase in hot spots makes the oil and paper insulation from
the transformer degrade. Good maintenance of transformer oil
can reduce excessive aging.
Fig. 5 Effect of Hotspot Temperature Towards Percent Loss of Life
The results on the effect of hot-spot temperature on the
loss of life percentage in Figure 5 are in line with the effect
of hot-spot temperature on equivalent aging results in figure
Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
4, the greater the hotspot value, the greater the reduction in
transformer life will be, where the result is exponential, this
is because the formula of ‫ܨ‬஺஺ is an exponential function.
It can be seen in figure 5 that the addition of 15 °C in
hot spots on Monday caused a large reduction percentage in
transformer life to 0.0193394%, this value exceeded the
normal average reduction in age per 24 hours based on IEEE
Std C57.91-1995, which is 0.0133%. The results on Monday
that had exceeded this value still could be tolerated because
the results of aging were still under 24 hours. While for other
days they were still within normal limits, so the transformer
could survive according to the manufacturer's life or even
more.
Figure 4 and 5 above prove that the greater the
temperature of the hotspot given to the transformer, the
greater the age reduction of the transformer will be. The
temperature of the hotspot is very influential on the isolation
of the transformer, if it is too hot, the insulation will be
degraded so the life of the transformer will decrease.
C. Analysis of the Effect of Ambient Temperature on the
Loss of Life Percentage of Transformers
In this study, five types of scenarios were carried out at
transformer room temperature where the temperature value
was varied into five variations for Monday, (18 March 2019),
Friday, (22 March 2019), and Eid, (15 June 2018), all with the
same temperature. Variations in temperature were reduced
and some were added to see the effect of room temperature on
the reduction percentage in the life of the transformer.
Variations in the temperature of the room were; -5°C, normal,
+5°C, +10°C, and +15°C.
Fig. 6 Effect of Ambient Temperature Towards Equivalent Aging
From figure 6, it can be concluded that with increasing
ambient temperature the value of the equivalent aging of a
transformer will increase as well. This happens because the
room temperature value affects the hot-spot temperature
where the hot-spot temperature value according to formula (1)
depends on the temperature of the top-oil and the average
winding temperature. So that by increasing the ambient
temperature the isolation side of the transformer will be
affected. The heat arising from the increase in the following
ambient temperature deteriorates the oil and paper insulation
of the transformer, thence the life of the transformer will
decrease. Good maintenance of transformer oil can reduce
excessive aging.
In addition, the ambient temperature also affects the
transformer activity. Ambient temperature can be absorbed
into the transformer which results in an increase in
temperature in the transformer. The increase in temperature
can be increased by heating due to continuous use of the
transformer.
Fig. 5 Effect of Ambient Temperature Towards Percent Loss of Life
Same as the effects of loading and hot spots, the loss of
life percentage of transformer at different room temperatures
also has the same correlation with the results of aging. The
higher the ambient temperature, the higher the reduction in
the life of the transformer will be.
V. CONCLUSION
The effect of loading on the loss of life percentage of
transformer is exponential, means that the higher the load on
the transformer, the greater the value of its loss of life
percentage will be, so that the transformer will be damaged
quickly. The transformer used in this study experienced the
greatest life reduction on Monday's load, 18 March 2019, with
the addition of 0.3 p.u which provided 58 hours of equivalent
aging results and percent loss of life of 0.0888332%..
The effect of the hotspot temperature on the loss of life
percentage is exponential, because the higher the hotspot
temperature, the greater the percent loss of life will be. The
heat generated from the hot spots will cause the insulation to
degrade, so the transformer will be damaged quickly. The
transformer in this study experienced the largest age reduction
on Monday, 18 March 2019, with the addition of a hot-spot
temperature of 15°C which gave equivalent aging results of
12,5706 hours and a percent loss of life by 0,0193394%.
The ambient temperature in this study has affected the hot-
spot temperature, thus the higher the ambient temperature, the
Authorized licensed use limited to: University of Glasgow. Downloaded on July 21,2020 at 00:39:08 UTC from IEEE Xplore. Restrictions apply.
greater the loss of life percentage would be. This study
experienced the largest age reduction on Monday, 18 March
2019, with the addition of ambient temperature of 15°C which
gave equivalent aging results of 13.4895 hours and a percent
loss of life by 0.020753%.
ACKNOWLEDGMENT
The authors would like to thank the financial support
provided by Universitas Indonesia through PIT 9 Funding
Scheme managed by the Directorate for Research and
Community Engagement (DRPM) Universitas Indonesia.
REFERENCES
1. IEEE Guide for Loading Mineral-Oil-Immersed
Tansformers, IEEE Standard C57.91-1995, 2004.
2. Forouhari, S. and Abu-Siada A. Remnant life
estimation f power transformer based on IFT and
acidity numbere of transformer oil. 2015.
3. Najdenkoski, K. and Rafajlovski G. Thermal aging
of distribution transformaers according to IEEE
and IEC Standards. 2007.
4. G.Swift, T. S. Molinski, and W. Lehn. A
fundamental approach to transformer thermal
modeling-Part I: Theory and equivalent circuit,
IEEE Trans. Power Delivery, Vol. 16, No. 2, p.171–
175. 2001.
5. Susa.D, Lehtonen.M, and Nordman.H, Dynamic
thermal modelling of power transformers, IEEE
Trans. Power Delivery. 197–204. 2005.
6. Jauregui-Rivera.L and Tylavsky.D.J. Acceptability
of four transformer top-oil thermal models-Part 1:
Defining metrics, IEEE Trans. Power Delivery, Vol.
23, No.2, P.860–865. 2008.
7. Mohamed Yazdani-asrami, Mohammed Mirzaie,
A.S.Akmal, S.A.Gholamian. Life estimation of
distribution transformers under non-linear loads
using calculated loss by 2D FEM, J. of Electrical
systems, Vol.7, No.1,p. 12-24. 2011.
8. M. Srinivasan, A. Krishnan. Prediction of
Transformer Insulation Life with an Effect of
Environmental Variables, International Journal of
Computer Applications, vol 55- No.5. 2012.