Proceedings of

LIST OF SPONSORS

GOLD SPONSORS
SILVER SPONSORS
SILVER SPONSORS
 Plot no. 1503/4, GIDC estate, Vitthal
Udyognagar-388121(via-Anand),Gujarat, India
Phone no.: 02692-235023/237185
Web : www.atlantaelectricals.com

An ISO- 9001, ISO -14001 & OHSAS-18001 Company &

NABL Test laboratory

Manufacturer Of All Types Transformers

Product Range

Power Transformers (Up to and including 160MVA, 220 kV )

Distribution Transformers (Including Corrugated type)
Furnace Transformers (Up to and including 50MVA, 66 kV)
Inverter Duty Transformers & WTG transformers
Mobile Substation (Up to and including 60MVA, 132 kV )
Motor Starting Transformers (For L.T. & H.T. Motors)
Neutral Grounding Transformers
Rectifier Duty Transformers
Dynamic S.ckt Testing Transformer
Air Cooled Transformers
Booster transformers & Voltage Regulators
Unitized Substation
Air Core Reactors and Gapped Core Reactors

A progressive partner in power development…..

 Plot no. 1503/4, GIDC estate, Vitthal
Udyognagar-388121(via-Anand),Gujarat, India
Phone no.: 02692-235023/237185
Web : www.atlantaelectricals.com

An ISO- 9001, ISO -14001 & OHSAS-18001 Company &

NABL Test laboratory

Manufacturer Of All Types Transformers

Product Range

Power Transformers (Up to and including 160MVA, 220 kV )

Distribution Transformers (Including Corrugated type)
Furnace Transformers (Up to and including 50MVA, 66 kV)
Inverter Duty Transformers & WTG transformers
Mobile Substation (Up to and including 60MVA, 132 kV )
Motor Starting Transformers (For L.T. & H.T. Motors)
Neutral Grounding Transformers
Rectifier Duty Transformers
Dynamic S.ckt Testing Transformer
Air Cooled Transformers
Booster transformers & Voltage Regulators
Unitized Substation
Air Core Reactors and Gapped Core Reactors

A progressive partner in power development…..

 You specify. We do the rest.
Engineering | Procurement | Construc on | Project Management

Our experience of over four decades in execu ng electrical EPC projects,

has helped us build robust and efficient systems for design, engineering,
erec on and tes ng of substa ons. Our customers rely on us to execute
large turnkey contracts within the most demanding comple on schedules.

We are leaders in 220 KV class Power Transformers in India, and have

commissioned over 150 substa ons upto 220KV across a wide range of
industries. One of Asia’s largest Solar Power Plants of 132/33 KV was
commissioned by us.

To know more, log on to www.bharatbijlee.com

Motors | Transformers| Projects | Drives & Automation | Magnet Technology Machines

w w w. t b e a i n d i a . c o m

Manufacturer of Power
Power Transformers & Reactors
ranging upto 765 kV Class & HVDC Converter
Duty Transformers with an Annual plant
Capacity of 36,000 MVA
MVA

Special Features of TBEA India Factory at Karjan Gujarat

™ Fuly Airconditioned & Dust Free Factory
™ Special Air cleansing Shower for Man & Material
™ Low RH Area for final assembly
™ NABL Accredated Lab for Testing Transformers & Reactors
™ Exclusive VPD for Windings
™ Equiped with Horizontal & Vertical Winding Machines
™ Automatic Core Binding Machine
™ Short cirucuit Test done on following ratings
i) 315 MVA, 400/220/33 KV 3-Ph Auto Transformer @ KEMA
ii) 260 MVA, 420/20 KV 1-Ph, Generator Transformer @ KEMA
iii) 160 MVA, 220/66 KV 3-Ph Power Transformer @ CPRI

EPC DIVISION
TBEA India is possessing the in-hhouse capabilities of Project Design,
Engineering Procurement and Execution from Concept to commissioning for: 765kV 400 kV 220 kV Others
™ AIS up to 765kV AIS Package 4 5 4 2
™ GIS up to 765kV
GIS Package 0 3 3 1
™ Transmission Line projects
™ e-BOP projects. No. of Installations/ Projects

Major Customers
Power Grid Corportation of India Limited | Indian Railways | APTRANSCO | GETCO | MSETCL | HVPNL | UPPTCL | TSTRANSCO | KPTCL
Sterlite Grid | Sterling & Wilson | Larsen & Toubro | TATA Power | Techno Electric | KPTL | KEC International | CESC | Adani Transmission Limited

TBEA Energy (India) Private Limited

“TBEA Green Energy Park” National Highway No.8, Village Miyagam, Karjan, Vadodara 391240, Gujarat, INDIA.
Phone : (+91) 2666 660000 | Fax : (+91) 2666 660027
Regional Offices: Hyderabad Gurugram z Kolkata z Mumbai z Vadodara
 Manufacturers of Distribution
and
Power Transformers
up to 100MVA, 230/110/11KV

Established in the year 1978, M/s.Venkateswara Electrical Industries (p) ltd., has been manufacturing
technically superior Power & Distribution transformers till date. Our company is ISO 9001 certified. All
our transformers confirm to IS2026/Latest Revision, BS 171. Since our inception, we have gained
considerable experience and knowledge in designing and manufacturing transformers to suit any
working conditions.

The Company has manufactured 20,000 Nos. in Distribution Transformers and 900 Nos. in Power
Transformers up to and including 100 MVA / 230KV Class. The range of Power transformers extends
from a rating of 5 MVA to 100 MVA of 11kv to 230KV Class and Distribution Transformers from a rating
of 100KVA to 1800KVA to of 11, 22 and 33KV Class. The range of Power Transformer also includes 3
Phase ONAN/ONAF/OFAF Cooled 230KV Voltage Class Auto Transformer with maximum rating of
100MVA. In Tamil Nadu, we are supplying regularly to TANTRANSCO & TANGEDCO and we have
executed orders to MAHATRANSCO/Mumbai, APTRANSCO/Hyderabad, KSEB/Kerala and recently to
TSTRANSCO/Hyderabad.
 ONLINE MONITORING SYSTEMS OF POWER TRANSFORMERS

N.PRUTHVIKRISHNA, CH.PAVAN KUMAR, CH.SRIKANTH

TOSHIBA TRANSMISSION AND DISTRIBUTION SYSTEMS (INDIA) PVT.LTD.

040-3071 7213 (Ph)

040-3071 7263/2245 (FAX)

Pruthvikrishna Nakkani (pruthvikrishna.nakkani@toshiba-ttdi.com)

INTRODUCTION

Power Transformers are part of the Major Assets of

the substation and considering the same most of
the utilities are specifying the Online Monitoring
systems as part of Asset Management and to avoid
catastrophic failure and to avoid the Plant or Grid
shut downs.
Due to the increasing demand of Power the call for
reliability is increasing and it becomes more and
more mandatory to establish supervisory systems
which allows permanent monitoring of those high
valuable assets .Hence there is a demand for the
Intelligent and user friendly devices that are
capable of Remote control, monitoring, collecting
and sending the important data around the
transformer to the control room.
Online Monitoring systems like Dissolved Gas
Analyzer, Fiber optic temperature sensing system,
Oil drying & Moisture removal system, Digital RTCC
relays ,Bushing monitoring systems ,Online Partial
Discharge systems. etc. are increasingly used in
EHV transformers.
As these products and systems are developed by
different equipment manufacturer, their
compatibility, operation, installations in EHV
transformers .Their installations, commissioning in
EHV transformer possess great challenges to
Original Equipment Manufacturers (OEM) right from
design stage until their commissioning & operation
at site.
Aim of this paper is to summarize the major
utilities requirements of Online Monitoring
systems and to compare the similarities and
difference of different models and also the
challenges faced in implementation and
adoption of standard interface areas for
reliable operation of EHV transformers.
PREFACE
To monitor different critical parameters of
Transformers established vendors are
available with in India & Outside .They adopt
different technologies based on their own
research and developments .It is
transformer manufacturer responsibility to
review the product details and drawings,
understand its working philosophy, for right
implementation & mechanical fitment for
trouble free and reliable operating life of
transformers.
In this paper OEMs Implies Transformer
manufacturer and VENDOR implies
manufacturer of Online Monitoring systems
or authorized business partners and
UTILITIES implies end users of
Transformers.
Over the years, delivery lead time of OEMs
has drastically reduced from 18-24 months
to 8-10 months .Due to this every, OEM is
investing lot of time and energy on the
standardization of transformer design itself assembly installation of additional probes
considering design, drawing approval and material placed at an accessible position is
procurement, production and factory acceptance preferred. They are terminated at tank wall
tests cycle time .In such cases these monitoring plate with dust cap protection.
equipment’s implementation consumes Table -1
considerable lead time of OEM and create plenty of
varieties of mechanical fitments thus increasing its Vendor FOT PROBE construction
cost ,efforts and complexity to OEMs. In order to
save such efforts & cost, there is an urgent and
strong need for standardization of all these online A
monitoring systems. Considering the Transformer
expected life of 25-30 years and on an average
B
Online monitoring systems life span is around 10
years due to its electronic equipment’s. Due to this
utilities may opt to replace the monitoring systems
with another make or some advanced monitoring
systems. Considering such practical problems
standardization of Fitment of these ONLINE
Monitoring systems are required. c
ONLINE MONITORING SYSTEMS

1. FIBER OPTIC TEMPERATURE

MONTIORING SYSTEMS (FOTMS).
b) Placing the probes in the winding.
Comparison in terms Number of sensors and
FOT probes in windings requires specially
fitment of sensors in spacers and minimum bending
designed slot in radial spacers as per
radius, tank wall plate and its enclosure, portable
recommendation of each VENDOR. This
calibration devices are discussed in detail.
slot is critical for trouble free operation of
a) Specification Requirements sensor throughout its life. Each VENDOR
Utilities will specify the requirement of FOT probes have their own slot design (refer Figure 1)
to be installed for same rating of Transformer. It and associated special tool for probe
varies from 4 to 16 sensors. fitment; hence each OEM has to procure
Some typical requirements of the Utilities are as 8 and maintain such VENDORs specific slot
probe 4 channels, 8 probes 8 channels, and 16 cutting tools. To avoid such issue,
probe 16 channels. standardization of slot design is
In absence of Uniformity, it is recommended to recommended.
follow IEC 60076-2-2011 Annex E for selection of
Probes.
Each vendor has different patented technologies
and having varieties of FOT Probes constructions
and in general most will be the Kevlar Jacketed or
PTFE Teflon to add strength to the Probes.
Since Probes being delicate in nature, it requires
careful handling during different transformer
manufacturing process in order to avoid its
breakage or loss of signal .As replacement of
probes would result in a massive work of winding
 Table 2 : Dimensional variations in tank
wall plate for 24 feedthrough
(maximum)

Overall
Parameters► Mounding
diameter, PCD
holes
VENDOR ▼ mm

A 304 279 8x13mm

B 230 200 8x12.5mm
C 224 200 8x10mm

d) Probe bending radius

Each VENDOR also recommends

different minimum probe bending radius,
50mm bending radius shall be adhered as
manufacturing tolerance. Similarly tight
Figure 1: Types of slots in radial spacers for probe coiling of excess length of probes should
fitment. be avoided to avoid loss of signal. Spool
diameter should be at least 130mm.
c) Tank wall plate or Feed through plate
And enclosure e) FOT sensor portable instruments

Tank wall plate is essential part of FOT
system, since it calls for highly precise
dimensional threads to mount feed through.
This part acts as junction box for tank inside
probes and tank outside extension cable(s).
Each VENDOR has their own dimensions of
circular plates as tabulated in Table 2.
Due to difference in dimension of tank wall
plates its enclosure or hood are also of
different sizes, thus creating different conduit
arrangements for hoods for protection of
external fiber optic cables. Such conduits are
to be laid until local control panels.
These tank wall plates or feed through plates
are bolted to tank wall flange (rings). Each
VENDOR has their own tank ring design and
VENDOR supply such rings to OEMs or its
tank fabricators. In absence of standardization,
this involves huge co-ordination between
VENDORS, OEMs and Tank fabricators.
Largest size of feed through plates and
common tank wall flange (rings) is
recommended as “standard” in line with Table
2. This will enhance ease of operation and
handling.
FOT probes are quite delicate and it
requires special attention during
installation and routing. We need to check
signal strength of probes during different
manufacturing stages to prevent later
stage rework. To perform such signal
strength checks, special portable
instruments are needed which are
supplied by respective VENDORS. OEMs
should maintain such portable
instruments as manufacturing asset or
tool and associated cost of purchase and
its maintenance.
f) FOTs Monitoring systems
Monitoring systems are available from 4
channels to 24 channels.
g) Cable or hardware connectivity
Cable or hardware connectivity options
available are - copper LAN cable, FO
cable, RS-485 4-wire copper cables.
Copper LAN cables can be used
effectively up to 100m without any
signal loss. But it has been observed
that 100m cable length distance is very
small and hence FO cables are
preferred. With below basic criteria of
 cable selection, site issues such as cable
breakage, communication noise, signal
attenuation issues can be avoided.
h) FO cable
FO cables have following specification
as options. Preferred option is underlined.
Category: OM1, OM2, OM3, OM4
Size: 62.5/125μm, 50/125μm
Mode of operation: Single mode, Multi-mode
Above basic selection will cater application of
100BASE-FX Ethernet application up to
2000m cable distance i.e. from transformer
outdoor panel to substation control room
SCADA switch.
All outdoor cables must be rugged enough for
protection against water, oil-grease & flame
retardant, suitable for rodent free application
with steel armoring with UV protective
covering. Additionally copper LAN cables, if
used must be electromagnetically shielded for
industrial applications. Also some customers
ask GSM modems along with cable
communications. Important point here is OEMs
must discuss and clarify all these points during
drawing approval.
2 Online DGA monitoring system
Comparison in terms of number of gases, type,
size and location of valves, type of installation,
and types of communication protocols, cables
and hardware connectivity, additional signaling
contacts, power supply requirements, enclosure
and correction protection are discussed in detail.
Most of the DGAs work on Photo acoustic and
Gas chromatography principles.
a) Number of gases
VENDORS have different DGA models suitable
for selected gases to be analyzed starting from
Hydrogen (H2) up to 9 gases as follows.
Table 3:
b) Type & size of valves.
There are 2 types of valves (screwed
and flanged), Screwed: BSP / NPT /
BSPT
Flanged: DIN / BS 10 / IS 778 / ANSI 150
Most OEMs in India prefers flanged type
valves in accordance with IS 778
considering transformer life time
availability, lead time and Cost.
c) Single location or separate oil
inlet-outlet location
For some models of certain VENDORS,
single valves are adequate for their
operation(Refer Fig 2), Whereas few
other models or VENDORS needs
separate oil inlet and oil outlet valves on
tank for operation. No method is
superior or inferior, but UTILITIES must
specify make and model in specification
in order to avoid design delays.
Figure 2: Schematic representation
of the DGAs installation with
Single Valve.
d) Tank or Floor mounted.
Most UTILITIES don’t specify
clearly mounting arrangements
and t h e y leave the choice to
OEMs. Each UTILITY follows
different mounting pattern for
different projects considering other
aspects. UTILITIES concerns
should be clearly stated in project
technical specifications. OEM
shall preferably offer tank
 mounted type, subject to space
availability on tank. For tank mounted
arrangement anti-vibration mountings
considering lifelong outdoor application
must be supplied by VENDORs.
Figure 3: Schematic representation of the
DGAs installation.
e) Communication protocol
There are multiple protocols provided as a
options by VENDORS but its OEMs responsibility
to get it clarified (exact requirement) from
UTILITIES at the time of drawing approval stage
to avoid later stage rework and follow-ups with
multiple internal & external customers. Today
most of the VENDORS are offering IEC 61850 /
Modbus / IEC 60870 / DNP 3.0 protocols.
Considering future ready smart power system
IEC 61850 should be standardized.
f) Cable or hardware connectivity
As discussed in FOTs Monitoring system,
FO Port communications are preferable.
g) Additional input output contacts
Additional contacts are required by many
UTILITIES in addition of communication protocols
for direct hardwired connection to substation
SCADA. This additional contact also serves
the purpose of checking the healthiness of
equipment’s, taking instant readings in case
of on-site work.
Following few points must be considered
while selection.
Contact rating: All VENDORS provide
potential free output contacts of 230VAC
as default. But sometimes 230VAC
independent supply for emergency
situation may not be available at site.
However, 220VDC is made available at
site powered from battery backup. Hence
potential free contact rating should be
suitable for 220VDC.
Current breaking and making capacity of
contact depends on voltage magnitude,
AC or DC, load current and operating
power factor of load connected across the
contact. All such information is covered in
IEC 60947-5-1 [3].
As example, Table 4 contains “Utilization
category” defined for type of current
carried (AC or DC) and typical application
in which the switch is used.
● AC15 – Control of AC Electromagnetic
loads
● DC13 – Control of DC Electromagnets
“Designation” relates to the utilization
categories and defines conventional
thermal current Ith & rated operational
current Ie at rated operational voltages Ue
and VA rating.
● A400 – The “A” denotes AC thermal
current Ith 10 A and “400” denotes
maximum rated AC voltage.
● Q300 – The “Q” denotes DC thermal
current Ith 2.5 A and “300” denotes
maximum rated DC voltage.
● R300 – The “R” denotes DC thermal
current Ith 1 A and “300” denotes
maximum rated DC voltage.
 Utili 3) Online oil drying and moisture
zat Desi Rated current Ie at
VA Rating removal system
rated voltage Ue
ion gnati The insulation materials on all power
cat on transformers will degrade with time in
eg 120 240V 380V Make Break service. Insulation aging depends on
ory V
several different factors including the
AC15 A400 6 3 1.9 7200 720 original insulation material, oil
AC15 A300 6 3 - 7200 720 temperature, moisture content, oxygen
125 250V content, particle contamination and acids
V from oil and paper aging. Moisture
DC13 Q300 0.55 0.27 - 69 69 content has a significant impact on
transformer life-expectancy, especially on
DC13 R300 0.22 0.1 - 28 28
the aging of the solid insulation.

Conventional methods of removing

Table 4: Contact rating (example) as per IEC
60947-5-1
Considering all above, uniformity of potential
free contacts is recommended, though no such
standard exist now. Author recommend below
contact ratings for standardization.
S300 – DC13 – Ue: 240VDC, Ie: 0.5A, Ith: 0.5
A C300 – AC15 – Ue: 240VAC, Ie: 2A, Ith : 2.5A
h) Power supply
Power supply offered by VENDORS are 100-
240VAC, 90–264 VDC and Universal power
supply 90–264V AC/DC. Most UTILITY
specifies available AC & DC auxiliary supply at
site. Few UTILITIES are using only DC supply
for online DGA. For such cases, DC rated
MCB should be used for power supply, as AC
voltage rated MCB will not function properly in
DC at same magnitude of voltage. Therefore
UTILITIES must clearly specify the exact
supply to be used for monitoring system in
technical specification.
i) Enclosure and corrosion protection
All VENDORS are offering ingress protection class of at
least IP55 for their enclosures. Few VENDORS are
offering IP66 or NEMA 4X. Author recommends that
material of enclosures should be standardized as
Stainless Steel of grade SS316 with
thickness of 2mm minimum.
moisture, e.g. oil treatment plants, have
focused on rapidly drying out transformer
oil; however, 98% of the moisture is
actually stored in the cellulose insulation.
The rate of drying depends on the rate of
water diffusion from the insulation into the
insulating oil, which is a very slow
process.
Online oil dehydration system removes
moisture from the insulating oil through
disturbing the moisture equilibrium so that
moisture diffuses from the wet insulating
paper to the dried insulating oil, slowly
removing the moisture from the solid
insulation. This system is capable of
reducing transformer oil moisture levels
from 50 ppm to less than 5 ppm. This
process not only reduces the aging rate of
the cellulose, but it also improves the
dielectric strength of the insulation oil and
can increase transformer reliability and
life-expectancy.
a) Moisture extraction capacity
In new transformers total moisture content is
typically less than 0.5% of insulation weight.
Many of the VENDORs are offering 10 liters
of moisture extraction capacity model, while
few offer 20 liters capacity.
sheet
b) Type of mounting and enclosures.
In past vendors were offering ODS basic
models without Enclosures due to the
absence of SCADA and Control activities ,
However with Latest trend, smart sub-station
requires more control and SCADA
connectivity which demands enclosure type system.
technology this methodology to perform the
Vendors offering both Tank and Floor mounted
control and protection and Monitoring is bound
systems and Utilities shall specify their requirement
to change.
clearly in the Technical specifications and if there is
no stringent requirement of the customer, OEM
shall preferably offer tank mounted type, subject to
space availability on tank. For tank mounted
arrangement anti-vibration mountings considering
lifelong outdoor application must be supplied by
VENDORs
c) Power supplies
Vendors offering different type of power
supplies as listed in the below table
VENDOR Power supply requirement
A 415VAC, 3-Ø, 1.0A, 50Hz
B 240VAC, 1-Ø, 1.2A, 50Hz
C 240VAC, 1-Ø, 2A, 50-60Hz
D 240VAC, 1-Ø, 6A, 50Hz
Table 5: Table of voltages.
Most of the vendors offering, 1-Phase supply
can be standardize against the ODS systems.
However With the advent in the Processor Bus
IEC 61850 Standardizes the data types,
names and communications to be used in
a substation, it supports all substation
automation functions and their
engineering. The standard defines a
consistent methodology for
interconnecting IEDs & Online Monitoring
in substations using Ethernet Local
networks a set of communication set of
communication services and Applications
built on a standardized set of plant and
Equipment models. Most of the Utilities
are specifying for the same in their
specifications.
Architecture View Of Monitoring Systems
Communication To Control Room With
IEC61850 :
TRANSFORMER

d) Cable & Connectivity: FO connectivity is

preferred.

Testing of online monitoring systems

ONLINE
In absence of exclusive national or international FOTM
DRYOUT
Standards for routine, type testing of Online M SYSTEM
Monitoring systems, each VENDOR are
evaluating their products differently.
ONLINE
DGA
Communication of Monitoring systems to
SCADA using IEC 61850 Protocol

In the earlier days Transformer control & protection

of Power Transformers has been done with the ETHERNET SWITCH
extensive copper cables laid from Transformer
control panels to Substation Control Protection
IEC 61850
panels, even for the monitoring systems various COMMUNICATION
sensors will be installed near the Transformers and
data used to be transfer with the help of cables. CONTROL ROOM
 Manuals.
3) PTSS –VELCON Online Dry out
system Manuals.
4) IEC 60947-5-1 Low-voltage
switchgear and control gear - Part 5-
1: Control circuit devices and
switching elements -
Electromechanical control circuit
devices.

Figure 4: Monititoring systems

Communication to substation scada
system using IEC 61850 Protocal.

Conclusion

Modern online monitoring devices demands

knowledge up-gradation of OEM engineers in the
field of industrial communication and hardware right
from design concept to final commissioning stage.
Requirement of Different type of Online Monitoring
systems will be specified by UTILITES and also in
some cases OEM will suggest or provide.

Selection of right model of online equipment by

UTILITIES is very important. And also there is a
strong need to have a common national or
international standard which should be applicable
for all online monitoring systems based on
proposed recommendations highlighted in this
paper.

References

1) GE - Kelman TRANSFIX – family DGA

installation manual.
2) Qualitrol –DGA, FIBER OPTIC SENSORs

Outlook of Various fault in power transforms during manufacturing
and their analysis by low voltage and high voltage tests.
Case Studies:
Setu Vora1, Dr. Chirag Parekh2, Minesh Bhatt3 , Avni Parikh4

1 to 4 , Atlanta Electricals Pvt. Ltd., Gujarat, India

Abstract:

Transformer is more than 100 years old technology. Technology hasn’t changed drastically but the
challenges are continuous increase in size and rating, Limitation on weight and space, Global market –
competition, Accurate prediction of performance parameters,Increasing of power system
complexities,Increase in Voltage class, Up-gradation of manufacturing technology, Strict process control
and so on. Due to it’s certain limitation like complex geometry of winding and core, bulk wait of active
parts and tank etc , still entire process is manual. So at each and every stage of manufacturing manual
checks to be followed for best quality of product. If it is lapsed at any point, transformer may get failed
either in factory or at site. The testing of transformers is aimed at determining their suitability for the
application. The standards, such as Indian Standard IS2026, International Standard IEC 60076, ANSI, IEEE,
NEMA, etc., give the details of these tests. The objective of this article is to discuss methodology adopted for
fault finding of Power transformer by performing specific tests, when transformer doesn’t mitigates its
requirements and desired values specified by designer as well as by customer. Various case studies are
discussed with all practical approach and root cause analysis carried out for fault occurred during
manufacturing process of few EHV class power transformer. Corrective actions also applied and applicable
tests are performed to ensure healthiness of the Power transformer for long life performance. Such type of
analysis can be guideline for transformer manufacturer, consultants , utility engineers etc.

The routine tests are considered a minimum requirement and need to be conducted on each and every unit
manufactured. The other tests are performed either to prove a design and/or only when specified by the user
and usually conducted on one unit of a series.

KEYWORDS: No load loss Test, Magnetic Balance Test, Measurements, Di electric Test, Power Transformer.

Transformer Technical Data: Salient technical

I) INTRODUCTION: Power transformers are the most specification of the transformer designed with mineral
important and expensive assets of the transmission- oil is as under.
and distribution power networks. It acts as a vital link
between the generator and the load. Any failure / Rated Power 75/105/150 MVA Three
outage of the transformer will cripple the power Phase (Auto Transformer)
supply and hence the reliability of the entire power
supply system depends on the reliability of the Voltage Ratio 220/132/11 kV
transformer. Considering importance of the
transformer has a relatively long life approx. 25 years, Rated Frequency 50 Hz
each stage acceptance test and quality checked are
mandatory. Few case studies are discussed to find Vector Group YNa0d11
the fault of power transformer with various
methodology adopted to reach its root cause. Voltage Regulation +15% to -5% in step of 1.25
I. CASE STUDY: 150 MVA Auto Transformer failed % LV for LV variation
to withstand No load loss test.
Cooling Type ONAN/ONAF/OFAF Table 3: MBT from TV side

BIL HV -1050 kVP Interpretation of Results: From above results It was

LV - 650 kVP concluded that W phase have fault as it was drawing
TV - 90 kVP higher current. For identification of specific fault
location following activity carried out.
Before performing NLL Test, all other low voltage test
such as Ratio, Insulation Resistance, Magnetic Analysis for fault finding :
Balance, Vector, Winding Resistance, Oil BDV taken
and results found satisfactory. During No load test Oil drained up to top yoke frame level to see the
from TV winding 11 kV side, fault current observed at clearance of TV winding and Top side leads from
5 kV and particularly W phase was drawing higher Inspection Window which was found satisfactory. In
current. For confirmation of failure, It was tried to continuation, further oil drained from Transformer
release trapped gas at buchhloz relay but gas was tank and Inspection Window opened from OLTC side
not observed. Separate Source Voltage test was also to check clearances of all 3 phase leads and found
performed and found satisfactory. So Phase to earth satisfactory. Again MBT test checked but found same
clearance found assumed OK. After that Magnetic as previous one. Now as a step ahead in this fault
Balance test performed from HV, LV and TV side detection process, W phase OLTC connections
which were found as following. disconnected. Here ,MBT taken and results found as
following.
From HV Side
Voltage 1ɸ 3ɸ Wph OLTC Bypassed, Tap-13, Main Lead no 20-15
Applied 1U- 1V- 1W- current current From LV Side
N N N (mA) (mA) Voltag 2U-N 2V-N 2W-N 1ɸ 3ɸ
1U- e current current
N 450 450 444 6.78 3.2 10.8 Applied (mA) (mA)
1V- 2U- 451 451 412 36.2 8.1 4.1
N 448 426 448 22.9 3.3 8.5
N
1W
2V- 446 242 446 209 6.2 2.6
-N 448 30.2 418 448 46.3 19.3
N
Table 1: MBT from HV side 2W 449 50 404 449 8.2 4.4
-N
Table : 4 MBT from LV side
From LV Side
Voltag 2U- 2V- 2W- 1ɸ 3ɸ From TV Side
e N N N current current Voltag 1ɸ 3ɸ
e 3U- 3V- 3W- current current
Applied (mA) (mA)
Applied 3W 3U 3V (mA) (mA)
2U- 451 45 44 5.48 7.6 27.2
N 1 5 3U-
2V- 451 43 45 17.8 7.6 23.2 3W 447 447 383 66.7 197.7 173.5
N 4 1 2 3V- 208.
2W 449 46 41 449 22.2 50.5 3U 450 244.8 450 2 143.5 214
-N 1 3W
-3V 448 76.7 376 448 213.6 283.6
Table 2: MBT from LV side
From TV Side Table : 5 MBT from TV side
Voltage 1ɸ 3ɸ
Applied 3U- 3V- 3W current current Interpretation of Results: From above MBT results,
3W 3U -3V (mA) (mA) It was noticed that W phase magnetizing current
3U- result is improved.
3W 445 442 3.49 445 184.4 144.6
3V- 45 10.4 W ph tapping leads Connected to OLTC one by one
3U 450 0 2 441 183.2 3.7 A and Single phase MBT current pattern taken from TV
3W 38 side and results found as following.
-3V 446 3 446 73 4.7 A 3.6 A
 1 ɸ current (mA)
Tap:1 Tap:2 Tap:3 Tap:4 Tap:5 Tap:6 Table : 9 MBT from LV side
194.5 195.4 194.5 194.5 194.6 194.6
141.2 139.6 140 140 140.4 140.2
From TV Side
210.4 209.0 209.8 209.0 210 209.2 Voltage 1ɸ 3ɸ
Tap:7 Tap:8 Tap:9 Tap:10 Tap:11 Tap:12 Applied 3U- 3V- 3W- current current
194.6 194.7 194.7 195.6 194.9 194.5 3W 3U 3V (mA) (mA)
141 139.5 140.1 141 140.8 141.5
3U- 427 427 427 2.20 151.7 187
209.4 209.4 210.5 210.9 209.6 209.6
Tap:13 Tap:14 Tap:15 Tap:16 Tap:17
3W
194.3 194.4 194.2 194.4 194.5 3V- 433 423 433 10 4.18 184
140.9 140.8 140.6 140.6 140.3 3U (A)
209.6 210.0 209.6 209.5 209.5 3W- 430 47 406 430 4.22 4.16
3V (A) (A)

Table : 6 MBT from TV side Table : 10 MBT from TV side

With OLTC connection MBT taken from LV side data as New Diverter kept in W ph
below. From LV Side
From LV Side Voltage 2U- 2V- 2W-N 1ɸ 3ɸ
Voltage 2U- 2V- 2W 1ɸ 3ɸ Applied N N current current
Applied (mA) (mA)
N N -N current current
(mA) (mA) 2U- 444 444 440 4.85 7.8 2.79
2U- 445 445 416 29 8.5 4.7 N
N 2V- 442 423 442 20.70 8.0 23.1
2V- 447 238 447 210 6.5 2.8 N
N 2W- 443 21 424 443 121.9 50.9
2W 448 33 415 448 8.7 4.9 N
-N Table : 11 MBT from LV side

Table : 7 MBT from LV side From TV Side

Voltage 1ɸ 3ɸ
Applied 3U- 3V- 3W- current current
TV Side with oil filling and all connections are remain
3W 3U 3V (mA) (mA)
the same .
3U- 437 437 435 2.5 185 152
From TV Side
3W
Voltage 1ɸ 3ɸ
3U- 3V- 3W- 3V- 443 432 443 11.48 184 4.5 (A)
Applied current current
3W 3U 3V 3U
(mA) (mA)
3W- 440 50 416 440 4.2 (A) 4.3 (A)
3U- 451 440 449 12.36 195.5 4.18
3V
3W (A)
Table : 12 MBT from TV side
3V- 449 47.4 422 445 195.5 163.9
3U
Bottom Frame to Frame isolation removed and MBT
3W 446 445 443 2.370 4.60 4.05
-3V (A) (A) current pattern checked from TV side data as below.

Table : 8 MBT from TV side From TV Side

After Diverter Removal MBT test as below. 3 ɸ current
From LV Side (mA)
Voltage 2U-N 2V-N 2W- 1ɸ 3ɸ 3U-3W 178.7
Applied N current current 3V-3U 190.4
(mA) (mA) 3W-3V 294.4
2U- 440 440 434 105 3.04 8.24 Table :13 MBT current pattern from TV side with
N frame shorting removed .
2V- 441 441 441 123 3.01 7.75
N From above results It was concluded that magnetizing
2W 445 0.140 0.145 445 0.113 0.925
current looked satisfactory when frame shorting
-N
removed while it was found abnormal while 2 frames
were shorted with link. Little smoke observed from
frame to frame isolation strip due to multiple time
voltage injection. Frame isolation copper strip was
burnt .

From TV Side
3 ɸ current
(mA)
3U-3W 177
3V-3U 1.5 (A)
3W-3V 1.8 (A)
Fig:3 , Core clamp tube damaged
Table :14 MBT current pattern from TV side with
Corrective Action :
frame connected.
All Core clamps ( Top and bottom Yoke binding)
checked by applying 2kV AC. Out of 16 clamps ,2
clamps of W phase side were taking higher leakage
current due to dead shorting with yoke clamps. Clamp
binding bolt steel washer touched with yoke clamp as
Insulation hole diameter was bigger than steel
washer hole.
Core bolts and yoke clamp bridging copper strips
replaced by new one. After all correction MBT which
was found as per table 15,16 and 17. After getting
satisfactory result transformer made ready will
complete oil filling to perform NLL. Now the No load
loss test performed. Voltage increased gradually from
10 % to 110 %. and kept charged for 110% for 4
hours. Here results found satisfactory which are as
following

From TV Side
Fig :1Bottom side frame earthing strip found burnt. Voltage 1ɸ 3ɸ
Applied 3U- 3V- 3W- current current
3W 3U 3V (mA) (mA)
3U- 445 451 406 52 178.7 207.5
3W
3V- 448 216 454 241 192.5 132.3
3U
3W- 444 49 409 453 296.1 193.1
3V
Table : 15 MBT from TV side

From HV Side
Voltage 1ɸ 3ɸ
Applied 1U- 1V- 1W- current current
N N N (mA) (mA)
1U-
N 454 454 443 20 3.1 2.1
1V-
N 454 221 454 228 2.3 1.0
1W-
N 455 17 444 455 3.0 2.0
Fig:2 , Top side CCA core bolts earthing strips. Table : 16 MBT from HV side
 From LV Side LV kV :4.1 kV and HV kV : 28.57 kV
Voltage 2U- 2V- 2W- 1ɸ 3ɸ PTR : 33000/110 =300 , TMF= 600
Applied N N N current current CTR : 10/5 =2 Freq : 50 Hz
(mA) (mA)
2U- 455 455 444 20 6.4 3.7
N VOLTAGE CURRENT WATT
2V- 456 241 456 226 4.6 1.8 U phase 26.097 0.5793 -2.825
N
V phase 20.312 1.5488 23.362
2W 455 21 439 455 5.9 3.4
-N W phase 29.082 0.8489 24.347
AVG 25.292 0.9882 45.62
Table : 17 MBT from LV side

II)CASE STUDY: 100 MVA Power Transformer failed LV kV :13.1 kV and HV kV : 91.59 kV
to withstand No load loss test. Pm = 27.372 kW

Transformer Technical Data: Salient technical Interpretation of Results : From above readings, It
specification of the transformer designed with mineral was observed that Magnetizing current and
oil is as under. measured losses both are abnormal. NLL was almost
100% just at 40 % of rated voltage. For further
analysis MBT performed which were found as
Rated Power 80/100 MVA Three
folloing.
Phase Power
Transformer)
Magnetic Balance & Magnetizing Current Test :
Voltage Ratio 230/33 kV From HV Side
Voltage 1U- 1V- 1W 1ɸ 3ɸ
Type of Winding 2 Winding Applied N N -N current current
(mA) (mA)
Rated Frequency 50 Hz 1U- 445 446 122 327 4.1 2.2
N
Vector Group YNyn0 1V- 446 226 446 219 9.0 3.5
N
% Impedance at Rated HV-LV 15 % (IS Tol) 1W 446 326 120 446 4.1 2.4
Tap -N
Voltage Regulation +10% to -10% in step of
1.25 % HV for HV From LV Side
variation Voltage 2U- 2V- 2W 1ɸ 3ɸ
Cooling Type ONAN/ONAF Applied N N -N current current
(mA) (mA)
2U- 445 445 63 381 101.7 81.4
Temperature Rise 0C 50 0C Oil N
55 0C Winding 2V- 446 220 446 225 383.5 149.3
Separate source voltage HV -460 kVrms N
test LV - 70 kVrms 2W 446 380 64 446 101.5 80.1
BIL HV -1050 kVP -N
LV - 170 kVP
Root Cause Analysis: All the required low viltage
No load loss Test : tests like Turns Ratio, Vector group, Insulation
PTR : 33000/110 =300 , TMF= 300 Resistance, MBT, Winding Resistance etc were
CTR : 5/5 =1 Freq : 50 Hz performed before starting NLL test and all found
satisfactory. Magnetic balance test result found
VOLTAGE CURRENT WATT abnormal. Separate source voltage test was also
U phase 8.288 0.5747 0.571 performed and found satisfactory. From results of
V phase 7.093 1.2226 5.999 MBT and NLL( Partial), fault suspected in V phase
W phase 8.334 0.6597 5.464 either HV or LV winding.
AVG 7.905 0.8190 12.034
For further step investigation Oil drained and Core
coil assemble taken outside. Partial voltage applied to
V phase only and kept under observation.
Temperature of each disc of regulating winding was
monitoring. After certain time, one or two discs of
regulating winding found heated. Here both HV main
and regulating winding was having 2 parallel winding
design( top and Bottom in each limb). From V phase
Tap winding each tap lead ( Top and bottom parallel
bridging link) was dis connected gradually and every
time magnetizing current recorded. There was
suspecting the unbalance turns in regulating winding
during manufacturing process. This was creating
circulating current between winding parallel circuit.
Here 1 virtual turn was added in heated disc on top
winding and Magnetizing current measured. Results
for the same are presented in tables.

1 and 2 no leads are cut and voltage applied and

results as below.
Fig : 4, CCA of 100 MVA

V I W
R 9.992 182.15 0.0510 Observation and Corrective Action: On top side of
Y 10.100 338.39 2.3321 regulating winding , disc no 8 identified with one turn
9.294 176.66 1.6376
lesser than bottom side disc no 8 turns. Same was
B
complied by addition of subtraction of turns from
∑ 9.795 232.40 4.021 respective discs. After rectification, again MBT and
Magnetizing current test were performed. All tests
Magnetic Balance & Magnetizing Current Test results were found satisfactory. Results fore the same
are as following.
From HV Side From HV Side
Voltage 1U- 1V- 1W- 1ɸ 3ɸ Voltage 1U- 1V- 1W 1ɸ 3ɸ
Applied N N N current current Applied N N -N current current
(mA) (mA) (mA) (mA)
1U- 454 454 398 56 2.5 1.3 1U- 449 448 395 53 2.4 1.4
N N
1V- 454 230 454 225 1.9 0.8 1V- 448 225 449 224 1.8 0.7
N N
1W- 456 58 399 456 2.5 1.3 1W 449 52 396 449 2.4 1.2
N -N

From LV Side From LV Side

Voltage 2U-2V- 2W 1ɸ 3ɸ Voltage 2U- 2V- 2W- 1ɸ 3ɸ
Applied N N -N current current Applied N N N current current
(mA) (mA) (mA) (mA)
2U- 455 455 363 92 71.9 33.5 2U- 449 449 359 89 71.5 35.1
N N
2V- 456 227 456 229 51.6 20.8 2V- 448 223 448 226 51.4 20.8
N N
2W 458 90 365 458 71.3 34.7 2W- 449 89 359 449 70.8 33.7
-N N

All the required process like VPD drying, Pressing ,

Oil filling ect were performed before tanking and final
testing. All the test on ready transformered as per
customer requirement were performed and results for
the same found satisfactory. Specially, NNL test
results after rectification are as following.
 tripped. After required checking and review of
No load loss test : parameters ,again the transformer energized but
PTR : 33000/110 =300 , TMF= 600 same phenomena of tripping observed.This time
CTR : 5/5 =1 Freq : 50 Hz instead of 100% it was tripped at 90% of rated
voltage. Results for the same are as following.
VOLTAGE CURRENT WATT
U phase 64.973 1.1065 65.07 No load loss Test :
V phase 64.405 0.6920 36.32 PTR : 11000/110 =100 , TMF= 100
W phase 65.188 0.9684 53.26 CTR : 5/5 =1 Freq : 50 Hz
AVG 64.855 0.9223 154.65
VOLTAGE CURRENT WATT
LV kV :33 kV and HV kV : 230 kV R 63.127 0.7229 38.346
Y 63.265 0.6012 28.392
Pm = 46.395 kW B 64.287 0.6435 33.097
AVG 63.560 0.6559 99.83
III. CASE STUDY: 10 MVA Power Transformer failed
to withstand No load loss test.
Pm – 9983
Transformer Technical Data: Salient technical
specification of the transformer designed with mineral oil Analysis of fault identification and Corrective
is as under. Action :

Rated Power 10 MVA To investigate the fault more precisely, Induced over
voltage test carried out and at 75 % of test voltage
Voltage Ratio 110/11 abnormal sound observed nearby OLTC and Voltage
of W phase dropped instantly. Here air trapped in
Type of Winding 2 Winding
Buchholz relay, which was released and that found
Rated Frequency 50 Hz inflammable. It was concluded the burning of some
insulating component inside the tank. It was also
Vector Group Dyn11 concluded that as it failed during NLL and IOV test,
fault might be relevant to Phase to phase lead
% Impedance at Rated Tap 10% (IS tolerance) clearance. To justify it ,Separate source voltage test
No 5 carried out and satisfactory.
Voltage Regulation +15% to -5% in step of
1.25 % HV for LV variation Magnetic Balance test also again carried out and
Cooling Type ONAN result found as following.

Temperature Rise 0C Oil- 50 Winding- 55

Separate source voltage HV -230 kVrms From HV Side

test LV - 28 kVrms Voltage 1ɸ 3ɸ
Applied 1U- 1V- 1W current current
BIL HV -550 kVP 1W 1U -1V (mA) (mA)
LV - 75 kVP 1U- 449 449 442 6.8 .0.9 15
1W
1V- 447 433 447 16 1.0 0.7
Before No load Test, all low voltage test such as 1U
Turns Ratio, Insulation Resistance, Magnetic 1W 444 102 430 444 15 12
Balance, Vector Group, Winding Resistance, Oil BDV -1V
and High voltage test carried out and results found
satisfactory. Table : After Failure Magnetic Balance
During No load test ( LV winding supplied 11 kV and Interpretation of Results: Result of MBT and
HV winding 110 kV kept open) Voltage and current Magnetizing current were found abnormal.
read fluctuation observed after energizing transformer
at rated voltage and within few seconds it was
Corrective Action: Oil drained from transformer and
OLTC side window opened to review the fault point.
110 kV main line leads which were routing through
OLTC for making delta connection and line lead too.
V and W phase main delta leads clearance
maintained with Permawood insulation which was
wrapped by insulating tape. May be due to impurity in
this this bridging material , voltage tracking happened
between V & W phase. It was dead shorted due to thick
carbon line in wooden piece.
Corrective Action : Instead of permawood insulation
fiber tube used for clearance. All work done during
positive pressure applied in transformer. After
rectification and oil filling transformer re offered for
Testing.
All low voltage and high voltage tests including NLL
and IOV carried out and found satisfactory. No load
Transformer kept on shock charged for 110% for 4
hours after completion of all Di electric test.

After Rectification Magnetic balance

From LV Side
Voltage 2U- 2V- 2W 1ɸ 3ɸ
Applied N N -N current current
(mA) (mA)
2U- 447 448 377 70 43.8 39.30
N
2V- 447 225 448 223 32.0 18.70
N
2W 449 68 380 448 42.2 43.30
-N
From HV Side
Voltage 1ɸ 3ɸ
Applied 1U- 1V- 1W current current
1W 1U -1V (mA) (mA)
1U- 447 447 425 23 1.2 1.1
1W
1V- 448 228 445 218 0.8 0.7
1U
1W 446 20 424 440 0.7 0.7
-1V

Figure shows clearance by permawood insulation. No load loss Test :

PTR : 11000/110 =100 , TMF= 100
CTR : 5/5 =1 Freq : 50 Hz

VOLTAGE CURRENT WATT

R 63.741 0.7583 37.75
Y 63.392 0.6154 27.93
B 64.782 0.6937 32.53
AVG 63.972 0.6891 98.21

Pm – 9821 Watt

IV. CASE STUDY: 50 MVA Power Transformer failed

While testing .

Transformer Technical Data: Salient technical

specification of the transformer designed with mineral oil
is as under.

Fig shows Permawood insulation punctured

Rated Power 50 MVA Interpretation of Results:

Voltage Ratio 132/33 kV In Magnetic Balance test , 3 phase current pattern of

U phase found erratic but in single phase result were
Type of Winding 2 Winding found satisfactory.
Further HV SSV and LV SSV performed and
Rated Frequency 50 Hz withstood ok. While performing NLL , typical results
observed which were completely different from
Vector Group YNyn0 conventional result pattern. Test results are as below.

% Impedance at Rated At tap no-5- 8% (-/+10% No load loss Test :

Tap tolerance)
Voltage Regulation +15% to -5% in step of PTR : 33000/110 =100 , TMF= 300
1.25 % HV for LV CTR : 5/5 =1 Freq : 50 Hz
variation
Cooling Type ONAN VOLTAGE CURRENT(A) WATT
R 62.377 1.6713 -91.87
BIL HV -650 kVP Y 35.335 0.9889 17.69
LV - 170 kVP B 38.024 0.7335 20.99
AVG 45.245 1.1313 130.55

Analysis of Fault identification and Corrective Pm = 39165 watt

Action :
Interpretation of Results: From above results It can
During low voltage test Ratio, be said that U phase is taking double voltage as
compared to V and W phase as well current and watt
Insulation Resistance, Vector Group, Winding pattern also erratic. This may be due to unequal Flux
Resistance and Oil BDV carried out and results found distribution.
satisfactory. Magnetic Balance test results are as
below. U phase HV winding was reverse compared to V & W
phase. It was happened inadvertently by Forman. It
was observed during CCA process and decided to
change only physically the start and finish lead
From LV Side connection of LV winding of U phase to make star
Voltage 2U- 2V- 2W 1ɸ 3 ɸ current point. All the low voltage test including polarity was
Applied 2N 2N -2N current (mA) found satisfactory and on that based it was proceed
(mA) for final process of drying, Tanking and testing. Now
2U- 434 434 337 97 37.4 124.7 during final testing all low voltage tests as well as
N SSV found satisfactory but during NNL test
2V- 439 214 439 224 26.2 70.6 abnormality was identified.
N
2W 438 96 344 438 36.5 53.8 Corrective Action:
-N
As it was known from manufacturing stage, CCA
From HV Side taken out from tank after draining of oil. Top yoke and
Voltage 1U- 1V- 1W- 1 ɸ 3ɸ U phase HV/regulating winding was dismantled. U
Applied 1N 1N 1N current current phase Regulating winding separated from main
(mA) (mA) winding and again fixed concentrically with correct
1U- direction. All low voltage performed during CCA
1N 442 437 362 77 3.4 10.4 process and found satisfactory. Further process of
1V- Drying and tanking carried out and send the ready
1N 441 202 438 237 2.4 5.7 transformer for testing.
1W
-1N 439 68 377 438 3.1 4.1
 Magnetic Balance test after rectification : presented in this paper with detailed analysis which can
* From LV Side be useful to everybody who deals with transformer
Voltage 2U- 2V-2N 2W- 3ɸ business. Here 3-4 different instances are discussed but
Applied 2N 2N current many of other faults can be come across while
(mA) manufacturing or handling transformer. Same to be
2U- 434 439 337 101 12.9 analysed and documented for further course of
N implementation while fault happen with transformer.
2V- 439 224 439 215 8.6
N V. ACKNOWLEDGEMENT:
2W 438 105 335 441 11.5
-N Authors are thankful to the Management of Atlanta
Electricals Pvt Ltd for allowing publication of this paper.
From HV Side Author is thankful to the Management of CPRI for
Voltage 1U- 1V- 1W- 1ɸ allowing publication of this paper at this platform of
Applied current National conference on “Transformer and allied
1N 1N 1N
(mA) Equipment”.
1U-
1N 442 440 376 65 1.5
1V-
1N 441 228 440 213 0.8 VI. REFERENCES: -
1W
-1N 439 72 371 442 1.6 Atlanta Electricals NABL Accredited Test lab data and
documents.
No load loss Test :
PTR : 33000/110 =100 , TMF= 300 "Power Transformers—Part 1: General, IEC Standard
CTR : 5/5 =1 Freq : 50 Hz 60076-1:2011".

"Power Transformers—Part 3: Insulation Levels,

100% CURRENT(A
Dielectric Tests and External Clearances in Air, IEC
(63.508) VOLTAGE ) WATT Standard 60076-3:2013".
R 64.974 0.8796 47.52
Y 64.412 0.5867 21.41
B 64.328 0.9471 35.50
AVG 64.572 0.8045 104.43 VII. AUTHOR’S BIOGRAPHIES

SETU VORA
Pm =31329 watt She has received
Bachelor of engineering
degree in year 2008. She
Interpretation of Results : After corrective action No
has 11 Years of working
load loss results found satisfactory. experience for Power
Transformer in various
disciplines like Testing,
V. CONCLUSION: Quality, Design, Sales
and SAS. Currently
Transformer is the simple object but as it comprises Working with ATLANTA
various field of science, Engineering and Technology, ELECTRICALS P LTD
care must be taken while designing manufacturing and as Head Testing.
installing. While analysing failure for almost 70 to 80 %
cases , it is observed that transformer got failed due to
manufacturing stage discrepancies. In this paper many
of case studies presented with practical data and
respective root cause and corrective actions. With the
help of MBT and magnetizing current test, we can clarify
the flux distribution of transformer. It can be greatly
influenced due to Interturn fault in winding, Core related
faults like earthig, Phase to phase insulation failure,
Winding direction and so on. Such case studies are
DR.CHIRAG PAREKH
He has received PhD,
M.E and B.E. degrees in
electrical engineering
from R.K.University,India
in 2017, S.P. University,
India in 2002 Saurastra
University, India in 2000
and respectively. He is
working with Atlanta
Electricals Pvt. Ltd., India
from 2002. His areas of
interest are transformer
Engineering ,design and
development ,Testing and
commissioning, electrical
power system and green-
energy sources. Also he
delivered many expert
lectures. He has
presented/published
many technical papers in
national/international
journals and conference.
Many UG and PG
students worked under
the guidance of him.
MINESH BHATT
He is having 16 yers
of experience in
transformer design
and perusing Master
study in electrical
engineering.
Currently Working
with Atlanta
Electricals Pvt. Ltd as
DGM-Design.

AVNI PARIKH
She has received
Bachelor of engineering
degree in year 2005. She
has 14 Years of working
experience for Power
Transformer in various
disciplines like , Quality,
Design, System
analysis,PPC. Currently
Working with ATLANTA
ELECTRICALS P LTD
as Sr.Manager(Mech.
Design & CAD.
 260 MVA, 420/20KV GENERATING TRANSFORMER PROTOTYPE
SHORT-CIRCUIT WITHSTAND TEST AT KEMA, NETHERLANDS

Author: Hitarth Gohel, Xu Chunmiao, Satyam Dewangan

TBEA Energy (India) Private Limited, Karjan
hitarth.gohel@tbeaindia.com

1. Introduction: 3. Methodology:

High Voltage Transformer that holds a Technical Specification of National

good market, is an important equipment Thermal and Power Corporation Limited
that drives power sectors and related (NTPC) was adopted to manufacture
260MVA, 420/20kV 1-ph, Generating
utilities in a global set-up. According to a
Transformer unit,
report by the International Agency of
Energy (IEA), India shall achieve a General Parameters/Characteristics of the
comprehensive electrification as the power transformer are stated below:-
demand is likely to become double by the
1 Voltage 420/√3kV±
year 2030. In a Power System it is (2x2.5%)/ 20kV
asserted that any short-circuit damage 2 Power 260MVA
shall lead to large area power outage, long 3 Current 1072.2A/13000A
overhaul period and heavy losses. If the 4 Short-Circuit 15.0%
main transformer is directly connected to impedance
the generator then any short-circuit 5 Vector Group Ii0
impairment shall seriously affect the 6 Cooling ODAF
reliability of the power supply and cause Method
7 Frequency 50Hz
huge economic losses. Moreover, it is
8 Insulation
difficult to repair the winding at site and Level
under strict climatic and seasonal a HV SI 1175, LI 1425,
conditions, appropriate maintenance is AC 630
challenging. b HV Neutral AC 38
c LV LI 170, AC 70
TBEA Energy has manufactured High 9 Winding Core-LV-HV-TAP
Voltage Transformer rated 260MVA, Arrangement
420/20kV, 1-ph, Generating Transformer
which has been successfully Short-Circuit As per IEC 60076-5, this two-winding
Tested at DNV-GL KEMA Laboratories, transformer falls under category III. The
Netherlands in Year 2015. Oil-immersed short time current rating of Transformer ‘s
Generating Transformer, with all the test accessories like Bushings, Bushing
procedures and inspection protocols Current Transformer, Tap Changer, etc
observed at factory premises and at were selected to withstand the thermal
KEMA Laboratory along with all the and dynamic effects due to short circuit
important technical points considered in condition especially observed in
the design and manufacturing of the Generating Transformers due to
transformer, before the unit was affirmed nonsynchronous operation. Central
to sustain Short-Circuit tests successfully. Power Research Institute, Bangalore was
also invited to witness the tests as a virtual
2. Keywords: customer at KEMA Laboratories.
High Voltage Transformer, Generating
Transformer, Short-circuit Test, KEMA

Page 1|5
 260 MVA, 420/20KV GENERATING TRANSFORMER PROTOTYPE
SHORT-CIRCUIT WITHSTAND TEST AT KEMA, NETHERLANDS

Author: Hitarth Gohel, Xu Chunmiao, Satyam Dewangan

TBEA Energy (India) Private Limited, Karjan
hitarth.gohel@tbeaindia.com

5. Important points taken care during of 5 micron. As per ISO – 9, the limiting
manufacturing: value of dust particle is within 293,000.
Hence, dust control method is
a. For winding radial stability, radial support implemented quite effectively in the
and winding conductor proof stress values factory.
are important. LV winding is provided with
sufficient number of inner strips and a Prior to commencement of Short-Circuit
thick pressboard cylinder to avoid radial tests at KEMA laboratory, routine tests
buckling. including applied voltage test, induced
b. The conductor buckling depends upon the voltage withstand test and also HV tests of
amplitude of stresses and the thickness of impulse and partial discharge
copper conductor. The Low Voltage coil is measurement tests were performed on the
designed with CTC conductor bearing low transformer unit.
current density with epoxy-coating to
enhance strength of winding. 6. Short-circuit impedance and load
c. The axial compression force has a huge loss Result before short circuit test are
effect on the axial stability of the winding. as follows
All windings have proper axial alignment. Pk
Tap Power Voltage Current Freq Z (t)
The block width are decided in such a way (75⁰C)
No. (kW) (kV) (A) (Hz) (%)
(kW)
to give axial stability, not affecting the 1 384.6 40.40 1018.0 49.7 15.9 399.4
cooling of the winding. 2 380.9 37.64 1071.8 49.7 15.6 396.0
d. The copper conductor stresses are used 3 396.8 35.09 1119.2 49.8 15.4 422.3
within yield strength zone considering
additional safety margin. Result: The measured short-circuit and
e. Pre-compressed high density board is load loss are within limit
used for the blocks and all the blocks are
firmly fixed with Pressboard washer to The measured values of applied voltage
avoid any further movement. test, induced voltage test with PD
f. Controlling Manufacturing process with measurement (IVPD), frequency response
strict tolerances. analysis, losses, AC withstand voltage test
g. Tighten winding lead takeout zones and between core and clamps, Insulation
designing of lead support structure to Characteristics - insulation resistance,
resist any movement of lead during short absorption ratio, polarization index before
circuit conditions. and after electrical test, capacitance and
h. The transformer is manufactured in a tan δ of winding insulation and bushing
controlled RH value atmosphere. The and BDV of oil and chromatographic
packing services are carried out in a low analysis were in well permissible limits.
RH value room where the humidity is
After this, the unit was dismantled and
within the acceptable limits and an
sent to KEMA for further short-circuit
extended exposure time does not affect
testing. (Note: peak current is ±5% to the
the products.
rated current; fault current is ±10% to
i. The dust control in the factory is within a
rated current)
range of 30,000-60,000 with particle size

Page 2|5
 260 MVA, 420/20KV GENERATING TRANSFORMER PROTOTYPE
SHORT-CIRCUIT WITHSTAND TEST AT KEMA, NETHERLANDS

Author: Hitarth Gohel, Xu Chunmiao, Satyam Dewangan

TBEA Energy (India) Private Limited, Karjan
hitarth.gohel@tbeaindia.com

7. Short-circuit testing procedure at

KEMA:-

a. IR ratio and winding resistance tests are

done. Once these were found to be
acceptable, short circuit test are
performed on the transformer.
b. After the test results are within the
permissible limits reactance are measured
and recorded.
c. At the beginning of short-circuit tests, the
average temperature of the winding shall
preferably be between 10 °C and 40 °C as
per IEC 60076-5.
d. The test shall be performed with current
holding maximum asymmetry as regards
the phase under test.
e. The amplitude î of the first peak of the
asymmetrical test current is calculated as
follows: i = I x k x √ 2
f. In this case X/R>14; therefore the factor k
√ 2 is assumed to be equal to 2.69 for
transformers of category III in accordance
to IEC 60076-5.
Test Circuit:
a. A calibration shot of 50-70% of HV
voltage is injected on the HV winding
by grounding the LV winding for
250msec at maximum tap.
b. Once the calibration shot is sustained
by the unit, a shot bearing 100% of
fault current, which is the first short-
circuit shot is applied to the winding at
maximum tap position for 250ms.
c. Reactance between the windings is
then measured. The reactance
measured at this stage shall not be
greater than or equal to the reactance
measured earlier before the application
of the calibration shot by ±1%.
d. The same process is be followed for
the second and third short-circuit shot
at normal and minimum tap position.
e. In no case, the reactance value
measured after the application of short-
circuit shot should exceed the
reactance value measured at the time
by ±1% when calibration shot is
injected.
f. In this way, short circuit shots were
applied on a transformer terminals at,
maximum, normal and minimum tap
positions each.

Page 3|5
 260 MVA, 420/20KV GENERATING TRANSFORMER PROTOTYPE
SHORT-CIRCUIT WITHSTAND TEST AT KEMA, NETHERLANDS

Author: Hitarth Gohel, Xu Chunmiao, Satyam Dewangan

TBEA Energy (India) Private Limited, Karjan
hitarth.gohel@tbeaindia.com

8. Reactance measurement overview:-

Reactance
Measured between Change
phases measured
Ω %
Tap
1 3 5 1 3 5
Position
Before
39.74 35.37 31.64 - - -
Tests
At
150227- 39.84 - - 0.2 - -
4004
At
150227- - 35.49 - - 0.3 -
4006
At
150227- - - 31.76 - - 0.4
4008
After
39.89 35.51 31.76 0.4 0.4 0.4
tests

System Power: 36370MVA; System

Voltage: 420kV

3 5
Tap Position 1 Max.
Nom. Min
254.6 242.5 230.4
Tap Voltage
kV kV kV
Impedance 15.99 15.60 15.42
Voltage % % %
Resistance 0.16
0.15% 0.15%
(75⁰C) %

The maximum deviation in reactance

measured was 0.4%. The maximum
deviation allowed in accordance with IEC
60076-5 is 1%.

Page 4|5
 260 MVA, 420/20KV GENERATING TRANSFORMER PROTOTYPE
SHORT-CIRCUIT WITHSTAND TEST AT KEMA, NETHERLANDS

Author: Hitarth Gohel, Xu Chunmiao, Satyam Dewangan

TBEA Energy (India) Private Limited, Karjan
hitarth.gohel@tbeaindia.com

is also performed on the transformer unit

before and after the short-circuit test.

The out-of-tank inspection with respect to

displacement, deformation of core and
windings, connection and supporting
structures or traces of discharges did not
reveal any apparent defect.

After all the results of the tests at the

factory premises were under allowable
Externally no visible change or limits, the transformer unit was set for the
disturbance was observed. physical inspection. During the physical
inspection it was observed that the
windings and block-washer alignment was
intact and all the leads to tank clearances
were maintained as before the unit was
short-circuit tested.

10. Conclusion:

It is seen that to successfully solve the

From the above result, it is clear that the
Generating Transformer Unit did not
display any abnormality during the short-
circuit testing. The transformer was then
dismantled and dispatched to TBEA for
further testing and inspections.
9. Testing and Inspection at factory
premises:
All the routine tests as specified in the IEC
and as required by the customer were
conducted. The impedance measured at
this stage as well shall not vary more than
±1% before and after the short-circuit test.
HV tests like impulse and partial discharge
measurements are done at 100% Volt.
Sweep Frequency Response Analysis test
problem of power transformer short-circuit
withstand capacity, proper design and
development along with controlled and
well-monitored manufacturing processes
is a must. In the case of the described
260MVA Generating Transformer, the unit
withstood all the routine tests and short-
circuit test and physical inspection was
ascertained with the established norms
mentioned in the international standards
within the permissible limits. Hence, the
transformer was declared to pass the
short-circuit test successfully.
10. References:
a. Technical Specification of National
Thermal and Power Corporation
Limited (NTPC)
b. International Standard IEC 60076
Part 5: Ability to withstand shirt circuit
Page 5|5
 AN APPROACH TO DETECT ELECTRICAL ARC INSIDE TRANSFORMER USING ARC
SENSING TECHNOLOGY

S.Linga Karthik, Easun – MR Tap Changers Private Limited

s.lingakarthik@easunmr.com

Abstract

Electrical arc is the source for fire in Transformer or any other Electrical devices. Detecting arc
early and cut down the source for arc could be the better way of mitigating arc related hazards.
Fiber Optic and Point Sensors are the Arc detecting Sensors available in the market which are
capable of detecting Electrical arc with intensity starting from 8000Lux.These Sensors are
effectively used in Switchgear industries where Sensor operating medium is Air. When it comes
to Oil medium, both Fiber optic and Point Sensors are found to be least effective and therefore it
induces the need of identifying an efficient Sensing mechanism for Oil medium which so far not
being available nowhere in the Market. We from Easun – MR Tap changers Private Limited has
done research in this area and identified an effective Sensing mechanism with the concept
called Light Sensing, which is a patented concept for Arc detection in Oil medium.

Key words: Electrical Arc, Transformer Fire, Fiber optic and Point Sensors, Arc Detector, Lux
level, Numerical Protection Relays

Introduction researches and methodologies are being

Transformer catching fire is the worst threat
any Substations, Industries or buildings can
experience since the damages caused by
such fire is huge in terms of money and at
times could lead fatality too. Fire Protections
systems are introduced to minimize such
damages caused by Transformer fire and
explosion to some degree but still significant
damages has been done at least for the
Transformer once it catches fire.
Theory of Transformer Fire
Electrical arc created inside Transformer
because of any internal fault causing gas
formation inside the Transformer which
leads to the pressure buildup and
subsequently Transformer explosion.
Transformer explosion allows the oil to react
with Oxygen to catch fire. There are several
proposed and being tried to effectively
encounter and suppress the possibilities of
a Transformer catching fire. Arc Sensing
technology, whereby isolating the
Transformer as quick as possible from the
occurrence of Arc shall be more reliable
technology in Transformer Fire prevention.
Role of Existing Fire Protection System
Fire Protection Systems available in the
market are designed to activate on the basis
of Transformer Protective Relays activation.
There are incidents where, even though
Transformer Fire Protection systems are
installed with Transformer and the
associated Protection Relays are activated,
still Transformer catches fire. This is
because, the initiated Internal Arc is allowed
to a Self sustainable fault because of
relatively slow acting nature of the existing
Transformer Protection Relay inputs. On the
other hand, if the internal fault has been
sensed quickly and Transformer has been
isolated much quickly so that the generated
fault not allowed becoming a self
sustainable fault, then the probability of
Transformer fire is very minimal.
Possible ways to mitigate Transformer
catching Fire
If we analyze the primary cause for the fire
due to Transformer Internal faults, it will
boils down to Electrical arc created due to
Transformer Internal faults. Therefore, if we
could sense the arc and quench it (as quick
as possible) by not feeding further power to
Transformer, then the chances of pressure
buildup and Transformer explosion can be
greatly reduced.
As depicted in Figure 1 – Fire Triangle, for
the fire to catch and sustain there has to be
a continuous feeding of Fuel, Heat and
Oxygen.
Fig. 1 – Fire Triangle
If we cut down any one of these three
elements, then the fire won’t be there or in
case of fire, fire will not sustain. In the case
of Transformers, during faulty scenario, the
created fault will release enormous amount
of heat energy which will increase the tank
pressure and at the same time it will reduce
the dielectric strength of the Transformer oil
and thus, transform the Transformer oil as
fuel for the fire to sustain. But if we could
detect the fault and isolate the Transformer
as early as possible, then the fault cannot
sustain by considering volume of the
Transformer Oil.
Fiber and Point Sensors are extensively
used in Switchgear Industries to detect the
arc but it is suitable only for air medium and
therefore there is a need for Development of
Arc Sensor which can detect the arc in oil
medium and that is when the development
of Arc Sensor initiated.
IEEE has established a formula for
calculating the Arc fault current Ifc (which is
estimated to be substantially less than the
corresponding Short Circuit Current Isc).
Fig. 2 shows us few Ifc and Isc values
derived from that formula.
Fig. 2 – Short Circuit Fault Current Vs
Arc Fault Current
This comparison is giving us the data about
the anticipated Arc Fault Current during
Transformer Internal fault and of course in
the case of Phase to Phase fault inside a
Transformer.
Arc Sensor
When an Electrical Arc initiated, then the
same can be seen in the form of visible
light. Several experiments were conducted
to ascertain the spectrum of arc and the
resultant spectrum has been shown in the
Fig. 3.
 Fig. 3 – Arc Spectrum

Arc Sensor is capable of detecting any such Fig. 5 – Test Arrangement

arc created inside the Transformer. Arc
Sensors are also tested for its suitability to
detect Electrical arc under oil filled
conditions. Figure 4, represents the test
scenario wherein the Arc Sensor immersed
in used transformer Oil to see its
performance.
The test Transformer used in this
experiment is with a height of 3 M and the
Arc Sensors were fitted on top cover of this
tank.
Response time
Arc Sensor Response time for an Electrical
arc measured to be in the range of milli
seconds (Average time measured to be less
than 10mS)

Fig. 4 – Tank filled with Used

Transformer Oil

Since Arc Sensor is working on the principle

of Arc intensity, it can sense the Electrical
arc irrespective of the location of the fault or
the geometry of the Transformer/winding
arrangement or the place where it is
installed.

Below figure (5) the arrangement to test the

sensitivity and response of the Arc Sensor
for the created Electrical Arc.
Fig. 6- Arc Sensor Response time
Intensity of Electrical Arc
During fault, produced Electrical arc will
liberate enormous amount of heat energy
and Photo energy.
There were experiments conducted to
understand the amount of Arc light created
during real Electric Arc @ 2000A short
circuit current and 17000A Short circuit
Current. The amount of Arc light recorded is
being 2,000,000 Lux and 12,000,000 Lux
respectively at 1M distance from the arc
source.
Arc Sensor mounting location
As per CIGRE study, 60% of Transformer
failure and fire happened due to Phase to
ground fault and 50% of Transformer fire
happened due to the flashover happened
between HV lead to Bushing turret.
Therefore, Arc Sensor is designated to be
installed around HV and LV Bushings.
Fig.7 – Arc Sensors (Marked as Green
Circle) mounted around Bushing Turret
Conclusion
With the help of Arc Sensing technology,
fault can be cleared much quickly in
Transformer. As mentioned in CIGRE 537,
probable Bushing related failures can be
easily sensed by Arc Sensors and therefore
subsequent fire hazards can be mitigated.
Short circuit and Arc flash events are
capable to produce very high photo
emission in the order of few crore; this will
enable to Arc Sensor to be effectively sense
the arc created inside oil medium as well.
Biography
Linga Karthik received his Electrical and
Electronics Engineering from Anna
University in 2010. He is a Deputy Manager
– Product Development in Easun-MR Tap
Changers Private Limited. Karthik has 10
years of experience in OLTCs, Transformer
Fire Protection System.
 SEISMIC QUALIFICATION OF TRANSFORMER
BUSHING
R Panneer Selvam1, Joint Director and Yamini Gupta2, Engineering Officer
Earthquake Engineering and Vibration Research Centre
Central Power Research Institute
Bengaluru, India
1selvam@cpri.in
2yaminigupta@cpri.in
Abstract - High voltage substation equipment are
highly vulnerable to seismic event. The poor
seismic performance of high voltage substation
equipment during pervious earthquakes and
need for power supply after earthquake disaster
led to focus on seismic qualification. Transformer
is the most important equipment in electrical
power transmission and distribution network.
Functioning of transformer under various
environmental conditions including earthquake
is paramount in addition to basic electrical
functional performance. Power transformers are
highly vulnerable to earthquakes. Bushings are
the most fragile component in the high voltage
power transformer. Failure of high voltage
transformer bushing had been noticed even for
the mild earthquake with 0.1g zero period
acceleration. Hence seismic qualification of high
voltage transformer bushing is recommended in
standards. Seismic loading based on substation
location should be taken into consideration while
designing transformers. This will ensure
uninterrupted and safe power supply in the event
of earthquake. Central Power Research Institute
(CPRI) is equipped with state-of-the-art tri-axial
shake table facility for simulating true
earthquake vibration. Seismic qualification
methods, codal provisions and seismic
qualification of transformer bushing by shake
table method are discussed in this paper.
Keywords - seismic qualification, natural
frequency, transformer, transformer bushing,
damping
I. INTRODUCTION
Earthquake is a natural disaster causing
damage to structures and equipment. Ground
vibration due to earthquakes is complex multi-
frequency in nature having both horizontal and
vertical components. Frequency analysis of
past earthquake data’s have shown that the
frequency of earthquake vibration is usually
less than 33Hz and most of the energy is in the
frequency band of less than 10 Hz. Earthquake
vibration is low frequency vibration, which
means vibration with large displacement. Low
frequency vibration with large displacement is
the major concern for tall and slender high
voltage equipment.
Expected magnitude of earthquake at any
substation location can be computed based on
geographical region, local site and soil
conditions, and historical seismic data. The
magnitude of earthquake can also be obtained
based on country specific seismic zones.
Seismic qualification of critical facilities like
Nuclear power generating station is carried out
using site specific seismic levels. Generalised
response spectrums are used for seismic
qualification of substation equipment. Normally
critical equipment and structures are designed
to resist major earthquakes and others to the
intensity of moderate earthquakes.
High voltage substation equipment are highly
vulnerable to seismic loading. Increase in
substation voltage rating increases fragility. Tall
and slender substation equipment usually has
low natural frequency, in the high energy band
of earthquake motion. When the natural
frequency of equipment matches input motion
frequency, resonance occurs thus amplifying
dynamic motion. Resonance in high energy
band of earthquake motion results in
amplification of low frequency components and
may result in failure of equipment. Based on
past earthquake records, major physical
failures of high voltage substation equipment
are observed, whereas in the case of low
voltage equipment functional failures are mainly
noticed.
Electrical equipment should be designed for
expected level of earthquake loading based on
mounting location. Design validation and
functional performance should be checked by
actual testing. Substation equipment are
interconnected, hence in addition to seismic
design of individual components
interconnection should also be designed to
avoid failure. This will secure substation
equipment and ensure uninterrupted and safe
power supply.
 Demonstration of equipment’s ability to
perform its function during and/or after
subjected to the tri-axial earthquake dynamic
force is seismic qualification. In addition to
functional capability, mechanical integrity of
equipment shall also be demonstrated.
In order to meet the basic requirements
regarding seismic qualification of equipment
and thereby to ensure reliable electrical power
system, Earthquake Engineering Laboratory
capable of performing a diverse range of
seismic qualification requirements on
equipment has been established. Seismic
qualification methods, codal recommendation
for seismic qualification of electrical equipment
along with seismic qualification of 420 kV
Transformer bushing by shake table method
are discussed in this paper.
qualification by shake table method is
recommended in standards. Seismic ground
acceleration is amplified by transformer and
acceleration at bushing mounting flange
depends on dynamic behaviour of transformer.
Computing acceleration level at bushing flange
is the primary requirement for seismic
qualification of bushing. The severity level at
the bushing flange can be computed by finite
element modelling of transformer and exciting it
with the required ground acceleration. If details
regarding ground motion amplification by
transformer are not available, required
response spectrum at bushing flange can be
obtained by multiplying with superelevation
factor (K). Amplification of ground acceleration
in transformer is shown in Fig. 1.

II. TRI-AXIAL SHAKER SYSTEM

The tri-axial shaker system with six

degrees of freedom is capable of performing a
diverse range of seismic qualification test on
equipment, sub-assemblies and components
as per National / International standards. The
shake table is 3 m x 3 m in size and test
specimen weighing up to 10000 Kg can be
seismically qualified. Simultaneous three-
dimensional random vibrations that are
statistically independent can be simulated using
shaker system. Frequency range of interest for
earthquake vibration is usually less than 33 Hz
and frequency below 5 Hz is very important.
CPRI tri-axial shaker system can simulate
earthquake vibration from 0.1 Hz to 50 Hz
frequency range. Test specimen of 10 m height
can be tested using this unique facility. The
seismic qualification tests on various equipment
like control panel, switchgear cubical,
instrument transformer, outdoor circuit breaker,
isolator, transformer, transformer bushing etc.
are being conducted using the tri-axial shaker
system.

III. SEISMIC QUALIFICATION OF BUSHING

The Power transformers are vulnerable to

seismic events. Major structural failure of high
voltage bushings had been observed during
past earthquakes. Due to payload and volume
limitation of shake table test facilities, it is
practically impossible to test high rating power
transformer by shake table test. Hence power
transformers are seismically qualified by
analysis. In addition to seismic qualification of
power transformers by analysis, transformer
bushing being most fragile component, seismic Fig.1. Ground acceleration amplification
A. Codal Recommendations testing can reduce failure in the event of
IEEE standard 693:2018 – “IEEE earthquakes.
Recommended Practice for Seismic Design of B. Bushing Details
Substations”. Seismic qualification by shake
table method is recommended for transformer 420 kV, 800A Oil impregnated paper
bushings having voltage rating more than 138 condenser type bushing with porcelain insulator
kV. Seismic qualification is carried out by tri- was subjected to seismic qualification as per
axial multi-frequency time history test with zero IEC TS 61463: 2016. Bushing is 4265 mm
period acceleration of 0.5g for moderate height from mounting flange and weight is 1360
performance level and 1.0g for high Kg.
performance level. Required Response Photograph of bushing mounted on tri-axial
Spectrum (RRS) at bushing mounting flange is shake table is shown in Fig.2.
obtained by multiplying required acceleration Response accelerations were mounted at
level by superelevation factor of 2.0 and plateau both ends of bushing and at mounting flange.
of spectral acceleration is also modified to take Strain gauges were mounted at critical
into account the amplification of ground motion locations to measure strain induced during
by transformers. Bushings mounted at a slope seismic vibration. Strain gauges were mounted
will experience more dynamic load when near porcelain and metal flange cementing
compared to vertically mounted condition. joint, both on porcelain and metal.
Hence seismic qualification by time history test C. Recommended tests
is carried out by mounting bushings at a slope.
It is recommended that the bushing be tested at The test sequence is as follows:
20 degrees measured from vertical. a) Vibration response investigation
b) Seismic qualification test
c) Electrical routine test
IEC TS 61463:2016 – “Bushings seismic
qualification”. Seismic qualification of bushing
by static calculation, dynamic analysis and by
shake table testing are mentioned in this
standard. Seismic qualification by static method
is valid for rigid equipment, bushing being
flexible equipment, dynamic amplification of
zero period acceleration due to resonance is
taken into consideration. Seismic qualification
by static method is simpler evaluation with
increased conservatism. Static method is
usually used at the initial design stage.
Dynamic analysis is performed for whole
transformer including bushing and mounting
condition. Electrical equipment being complex
in nature, it is difficult to model and predict the
equipment performance by finite element
analysis precisely. Also by finite element
analysis method, only physical conditions like
deflection and stress due to seismic loading can
be evaluated. In case of electrical equipment, in
addition to physical conditions it is required to
check functioning during and/or after seismic
event, which can be checked only by shake
table testing. Seismic qualification by shake
Fig.2. 420 kV, Transformer Bushing
table method is used for testing bushing due to
test facility limitation. If RRS at the flange of
bushing is not available, testing can be carried D. Vibration Response Investigation
out by multiplying the ground RRS with Dynamic characteristics of bushing can be
superelevation factor of 1.5 as per this found by vibration response investigation.
standard. Natural frequency and damping of bushing can
Thus power utilities with service areas in be found by Free oscillation method (Snap back
seismic Zone III, IV and V as per BIS 1893, Part test) and by base excitation method. Base
1 should have their equipment seismically
excitation method is more accurate and
qualified. Designing of substation equipment for
reliable.
seismic loading and validation of design by
 Free oscillation method TABLE 1
RESONANCE FREQUENCIES
Bushing is made to oscillate freely and Axis Location Resonance frequency, Hz
oscillations are recorded using transducers.
Bushing filled with the insulating medium was
installed vertically on stand. Test was done by 3.50
Horizontal-1 Top
applying 2500 N at bushing terminal and 12.00
Flange
released suddenly. Response of accelerometer
mounted at the top of bushing recorded is 3.56 & 12.00
Bottom
shown in Fig.3. 3.69
Horizontal-2 Top
Approximate values of natural frequency and 17.00
Flange
damping values are evaluated from
accelerometer response signal and are as 3.75 & 17.00
Bottom
follows: 28.25
Vertical Top
a. Natural frequency – 4.39 Hz Flange
28.25
b. Damping – 3.11% 28.25
Bottom

3.5 Hz

Transmissibility (Acceleration)
Acceleration

Time, sec
Fig.3. Free oscillation test

Frequency, Hz
 Base Excitation method
Fig.4. Resonance search test

Vibration response investigation is usually

carried out by low level base excitation uni-
axially in each three axes individually.
Sinusoidal sweep vibration from 0.1 to 33 Hz
at frequency sweep rate of 1 octave/minute with
constant acceleration of 1m/s2 was used for
vibration investigation.
During this test table excitation and response
accelerometer signals were recorded to
compute resonance frequency and damping.
The Resonance frequencies of bushing are
shown in Table 1.
Natural frequency and damping values are
evaluated from accelerometer response signal
and are as follows:
a. Natural frequency – 3.50 Hz
b. Damping – 2.81%
Typical resonance search test graph is
shown in Fig.4. Value of damping calculated by
half power method from resonance search test
data.
E. Seismic Test
Seismic qualification test was carried out for
0.3g zero period acceleration. Details of RRS
amplification by transformer was not known,
hence RRS is multiplied by superelevation
factor of 1.5. Vibration levels in vertical axis are
80% of horizontal axes. Statistically
independent spectrum compatible time
histories were generated for two horizontal and
vertical directions for 2% damping spectrum,
value computed from vibration investigation
test. Time history test was carried out by
simultaneous three-dimensional random
vibration. Required Response Spectrum (RRS)
and Test Response Spectrum (TRS) for
horizontal axis are shown in Fig.5.
During time history test, response
acceleration on test specimen and strain values
at critical locations are recorded. Maximum
strain of 448 microstrain was developed close
to mounting flange as shown in Fig.6. Time
history of strain recorded close to mounting
flange is shown in Fig.7, based on this data
stress developed due to seismic loading can be F. Acceptance Criteria
computed. Acceptance criteria as per IEC TS 61463
standard:
TRS  The bushing shall insulate and carry current
FRS: Acceleration, m/s2

after seismic test. This can be ensured by

RRS carrying out electrical routine test after
seismic test.
 No crack, leakage, permanent deflection or
relative movement of parts shall occur.
 Stress due to seismic loading shall be added
to stress due to;
Operating load, wind pressure, mass of
bushing and internal pressure. Combined
Frequency, Hz
stress should be within limit.
Fig. 5. RRS and TRS
 Bushing was meeting all the requirement of
standard and is seismically qualified.

Fig. 6. Maximum strain location
IV. CONCLUSIONS
The expected Magnitude of the earthquake
for any region can be computed based on the
geographical conditions, local site conditions,
soil conditions and historical seismic data. But
time of occurrence of earthquake cannot be
predicted. Dynamic loading due to
earthquakes should be taken into consideration
during design.
Seismic design should be validated by
shake table test. Loss of time and finance to
restore substation to normal working condition
after earthquake can be avoided by seismic
qualification. This will ensure uninterrupted,
safe power supply during crucial post-
earthquake rescue operation. Power utilities
may utilize the state-of-the-art facilities
available at CPRI to ensure reliable power
supply to their customers.

V. ACKNOWLEDGEMENT

The authors gratefully acknowledge the

encouragement of Shri. V S Nandakumar,
Microstrain

Director General, Central Power Research

Institute in bringing out this technical paper.

VI. REFERENCES

[1] IS 1893 (Part 1) : 2016 – “Criteria for Earthquake Resistant

Time, Sec
Fig.7. Time history of strain
Peak value = +368 & -448 microstrain
Design of Structures”.
[2] IEEE standard 693-2005 – “IEEE Recommended Practice
for Seismic Design of Substations”.
[3] IEC TS 61439 : 2016 – “IEC standard for Bushings –
Seismic qualification.”

SMART REPLACEMENT OF LARGE POWER TRANSFORMERS
Komelabbas Lakhani, SIEMENS, India
Contact no. +91-9820697197 / 022-39608796 Email: Komelabbas.lakhani@siemens.com
ABSTRACT
Large power transformer, being one of the
most expensive, complex, custom-built static
equipment in the grid, it is a complicated
process to replace or interchange with a new
transformer. Though interchangeability is in
practice for the ages, the challenges and
guidelines & recommendations of
replacement are found in scattered form.
The paper covers on major challenges or
potential risks which can be mitigated in
design and execution phases of a transformer
replacement. The need for harmonization of
critical accessories such as bushings, tap
changer, terminations is highlighted.
Being a global OEM with vast technological
expertise & knowledge, an attempt is made in
this paper to provide a holistic view of
electrical, mechanical & control
characteristics with regards to
interchangeability and replacing power
transformers and its accessories.
Recommended checklist is added to aid
purchasers for SMART transformer
replacement.
INTRODUCTION:
SMART replacement in power transformers is
defined as the ability to replace an existing
unit with minimum changes in existing
substation, minimum possible time & incurring
lowest cost, provided that, the corresponding
electrical & mechanical aspects either
matching with existing units or technically
superior.
Replacement needs are ever increasing due
to ageing of old fleets of transformers in
substations due to introduction of renewable
generation.
Expectations for a quick & efficient
replacement is beneficial to both purchasers
and OEMs.
National Conference 2019, CPRI
Replacement of transformers in an existing
substation layout with other equipment are
tricky, since custom made set ups would
necessitate changes during removal of old
transformer & installation of new.
Disposal of oil filled non-working transformers
poses high risk due to the flammable nature
of the transformer oil & its possibility of
seeping into soil on the site – creating the
earthing resistance reduction on the soil &
contamination. Environment friendly disposal
of the PCB oil filled equipment is mandatory is
local regulations.
Due to different power transmission
philosophies, different terminations interfaces
(Oil to outdoor air, Oil to SF6, Oil to Oil, Oil to
Busduct) are required on transformers leading
lack of harmonization in terms of mediums,
construction, accessories & layouts. In case
of change of interface as technological
improvement, critical review electrical
clearances and creepages should be done in
existing substation layout to mitigate risk
arising out interface change during
transformer replacements.
As per CIGRE brochure 528, the enquiry
document should include the following details,
where applicable:
 Outline and foundation drawings of
existing transformers, if
interchangeability is required
 proposed drawing of substation
showing general position of
transformer specified
 details of primary connections, such as
LV phase isolated connections and gas
insulated connections
 list of applicable purchasers, statutory
or regulatory requirements
 if parallel operation with existing
transformers is required, information in

accordance with IEC 60076-1 should
be provided
Timely availability of critical site information is
essential for mutual benefit of purchaser and
OEMs in projects with short delivery
expectations and milestones should be clearly
defined.
Design & Manufacturing challenges in
SMART replacement
The challenges are grouped into
a) Electrical Design Aspects
b) Control Design Aspects
c) Mechanical Design Aspects
These needs to be assessed and evaluated
before the replacement transformers are
designed & executed.
1 Electrical Design Aspects
Most critical aspect of any transformer design
is to align the key electrical parameters of the
customer like vector group, voltage, MVA,
regulation, impedance. It should be carefully
evaluated if a replacement transformer is to
run in parallel operation - in line at an existing
substation in the same installation. IEC-60076
Part 8 contains all critical electrical aspects of
transformers for successful parallel operation
which states, the transformer would require:
 The same phase-angle relation –
clock-hour number
 The same ratio with some tolerance
and similar tapping range
 the same relative short-circuit
impedance – percentage impedance –
with some tolerance.
 Transformers can exhibit different %
Impedance across tapping range, care
required in parallel operation
(reason: different design concepts e.g.
winding relative position, regulation
type)
 Consequence of minor mismatches of
data should not be overestimated.
(e.g. insisting precisely same tapping
voltage in parallel operation).
National Conference 2019, CPRI
 Mismatch in relative loading of ≤10%
between non-identical design is
reasonable.
As per CIGRE brochure 528, the tender
specification for a transformer which is
intended for parallel operation should contain
details of existing transformer.
This also means that the variation of relative
impedance across the tapping range should
be similar for the two transformers. It is not
recommended to combine transformers of
widely different power rating.
As the natural relative impedance for optimal
designs varies with size of transformer.
Transformers built according to different
design concepts are likely to present different
impedance levels and different variation
trends across the tapping range. In practice, a
mismatch of relative loading of no more than
about 10 % between two transformers of non-
identical designs should be regarded as
reasonable.
2 Control Design Aspects
Control design philosophy for the new
transformers is to be line with the existing
transformers & substations. One of the
challenges for control systems is that, design
of accessories varies for different makes is not
harmonized & its correlation is key to
seamless operation. Central Board for
Irrigation and Power Publication 317:2013
covers the general schematics of parallel
operation. Below aspects of schematics are
recommended for parallel operation.
 Suitable selector switch be provided,
so that anyone transformer of the
group can at a time be selected as
‘Master’, ‘Follower’ or ‘Independent’.
 The scheme will be such that only one
transformer of a group can be selected
as ‘Master’
 An out-of-step device shall be provided
for each transformer which shall be
arranged to prevent further tap-
changing when transformers in a group

operating in ‘Parallel control are one
tap out-of-step.
It is important to focus on following aspects,
if the customer requires parallel operation.
 Control schematics of OLTC Drive
Mechanism (DM), Remote Tap
changer Control Cubicle (RTCC),
Marshalling and Cooler control
cubicles of transformer which is to be
run in parallel should form the basis of
reference.
 Accessibility, mounting arrangement
(on-tank or ground supported), door
opening sides of control cubicles shall
be furnished.
 For OLTC with automatic parallel
operations, make & models of AVR
shall be furnished.
 Contacts rating should be selected
based on numbers of transformers to
be run in parallel as per
recommendation of OLTC
manufacturers.
3 Mechanical Design Aspect
The Mechanical Design challenges are
grouped into:
1. Overall dimensions
2. Foundation matching
3. Coolers (on tank mounting or separate
cooler bank)
4. Termination interface
5. Transportation
6. Removal & re-installation.
IEC 60076-22 published in 2019, contains
specification requirement of various
accessories. However, India specific
harmonization of accessories is pending,
leading to different practices among
accessories OEMs. This results in additional
challenges during transformer replacements.
Areas of harmonization was compiled by
author’s organization for various accessories,
had been submitted to BIS technical
committee (TC 16) are,
 Ingress Protection class
 Paint type
National Conference 2019, CPRI
 Paint shade
 Contact ratings
 Type and no. of contacts
 Cable gland size
 Cable size and type
Lack of harmonization in various accessories
& standardization & estimation of all inputs
results in delays and plug and play type
arrangement will reduce the replacement shut
down time.
According to author’s judgement, Additional
engg time in evaluation of replacement units
and associated constrains are approx. ~20%
more than that of new designs.
3.1 Challenges in overall dimensions
Providing a replacement transformer within
the existing footprint of “higher capacity” are
technically more challenging than
replacement of identical capacity, with due to
consideration to minimum spacing
requirements to adjacent objects specified in
local regulatory requirement.
With addition of new protective accessories,
sensors and monitoring devices along with
newer termination arrangements in
replacement units without exceeding
permissible footprint needs special attention.
Sometimes purchaser levy penalty for
excessive footprint especially urban areas
with high real estate values and can form a
basis for vendor selection.
3.2 Challenges in Foundation
Site foundation of transformer are made for
specific loading and pull-out directions having
standard rail gauge tracks (e.g. 1676mm). For
replacements units, pull-out directions with
due consideration to other substation
equipment in vicinity should be reviewed.
Condition assessment of existing civil
foundation (of main units and cooler banks
etc.,) in terms of load carrying capacity for
expected life span needs to be reviewed &
 National Conference 2019, CPRI

becomes critical during replacement of higher interchangeability of identical units with

capacity upgrades. minimal mechanical assembly changes, as
shown in below picture.
Geometrical matching with existing jacking
points and rail track gauge are key in Retrofitting with Ester fluids: Due to different
replacement and changes should be mutually viscosity of Esters fluid compared to mineral
agreed between purchaser and OEMs. oil, entire thermal design including cooler
surface, thermal head, frictional flow, oil
3.3 Challenges in Coolers velocity etc. should be properly evaluated with
due consideration to specified temperature
On tank mounted coolers are preferred for rise limits
replacement units, to reduce site foundation
mismatch or modified foundation. On-tank
coolers in place of separate cooler bank
conversions are more challenging since it
involves due considerations to thermal
efficiency without exceed guaranteed
temperature rise limits, vibration and sound
pressure.

Based on actual measurement of temperature

rises in top oil, winding hot spot, core
temperature and tank surface temperature,
capacity enhancement proposal are reviewed
subject to availability of design margins &
mechanical design feasibility for increased
cooler capacity (e.g. increased radiators,
fans) with consideration of foot print and
temperature rise limits.
For capacity enhancement of replacement
transformer, change in cooling type (e.g. unit
cooler or heat exchanger) can be proposed to
limit the footprint coupled with changes to
active part construction and cooler bank
interfaces.
Among the replacement units, for increased
reliability, reduced maintenance effort, robust
and standardized cooler modules should be
preferred, which are easy to replace features
along with low noise. The “modular” cooler
should contain exchangeable heat exchange
parts (cooler body). The modular unit cooler
should permit easy and safe cleating with
water jets with larger pitch distance and fixed
fins. Considering the interchangeability of
transformers at various locations, purchaser
can opt for cooler bank mounting provision for
both sides of transformers tank as practiced
by few utilities. This will enable faster
Fig.1 Transformer with separately
mounted radiator bank
3.4 Challenges in terminal interfaces.
Based on CIGRE brochure 755: 2019, 21% of
failure are attributed due bushing and 19%
are associated with fire. Hence during
replacement, most purchasers prefer RIP or
RIS type of bushing in place of OIP condenser
bushings. The RIP and OIP bushings have
different oil end and air end dimensions.
OEMs needs to review of such termination
arrangements within existing footprint and
required air clearances.
Due consideration to terminal connecting
interface is required to avoid undue
mechanical cantilever stress on bushings.
Bushing suitability of Seismic acceleration in
case of substation relocation or change in
geological seismic data over the years.

In case of cable box (air-filled or oil-filled)
attention to be paid to healthiness of cables,
termination kits and cable glands should be
reviewed.
The special testing arrangements, if any (e.g.
AIS testing of SF6 & Oil-Oil terminals) should
be evaluated during bid stage including the
supply. With proper upkeep of special test
arrangement at purchaser end, it will render
flexibility to purchaser for change of
termination. Such special testing arrangement
can be utilized during retesting, thus saving
time, efforts and cost.
For quick interchange of cable box in
replacement unit, plug-in terminations shall be
preferred in place of traditional cable box
arrangement, as seen in figure:
Fig 2: Conventional Cable box arrangement:
Fig 3: Plug & play Cable box arrangement
permanent & test set up:
3.5 Challenges for Transportation
Transportation of transformer (factory to site
& installation to repair location) can be a major
National Conference 2019, CPRI
cause of concern, considering inland
transportation dimensions restriction of
respective countries and mode of transport.
Missing records of transportation in (from
factory to site) can lead to incorrect
assumption during transit from site to repair
location. Medium of filling (Oil filled / dry air)
needs to be evaluated based on the size &
transportation conditions.
Road surveys should be performed to check
the transportation constraints to prevent
wrong assumptions, risk and delay.
For guidance on transformer transportation,
kindly refer CIGRE brochure No: 673: 2016.
3.6 Removal & Re-installation
In some instances, removal of transformers
can become quite challenging as it may
involve removal of other equipment’s in
vicinity, arising from improper substation
layout and unplanned bay extensions. Hence
holistic planning with adequate emphasis on
the transformer interchangeability is needed
on substation layout, to minimize down time.
DROP-IN TYPE ARRANGEMENT
Drop in type arrangement is considered as the
removal of one transformer unit at the site
without disturbing the live loaded grid. A
transformer should be designed in a such a
way that, it will face minimum movements and
should be installed safely at desired location.
A suitable crane & access needs to be
checked for such an arrangement.
Fig.4 Drop In transformer at site
 National Conference 2019, CPRI

Recommendation to purchaser for SMART 3. CIGRE Brochure 523 –2016 Guide on

transformer replacements:
1. Purchaser specification should include the
interchangeability requirement in their
specification.
2. Purchaser shall furnish specific make,
models of accessories meant for replacement
with geometrical dimensions, foundations.
Transformer transportation.
4. CIGRE Brochure 755 -2019 Failure
analysis of Power Transformers
5. Transformer Manual Central Board for
Irrigation and Power Publication 317:2013
About the author:

3. Purchase can opt for plug in arrangement

in place of conventional cable box.

4. Purchaser are advised to procure the

special testing arrangements.

5. Purchaser can insist for installation of latest

monitoring devices and associated sensors.
Komelabbas Lakhani is a Gold medalist for
6. Purchaser should insist for harmonized B. E mechanical Engineering & received this
accessories as applicable to country of usage. M Tech & B.E degrees from the Mumbai
University. He has been working for Siemens
Transformer factory since 2007 in the
Conclusion :
Engineering Design department. He is
currently positioned as Manager Design of
Author has made a small attempt to provide a
Mechanical Design of Power transformers.
holistic picture about the challenges and risks
associated with the replacement of the
transformer. Interchangeability and ease of
replacement need to be inherited at
contractual stages to have long term vision of
SMART execution.

The annexure attached contains a checklist

that can be a guide to mitigate the risks
discussed during this paper.
The attempts for harmonization of
accessories need to be pursued enabling the
customers for a smooth replacement
experience.

REFERENCES:
1. IEC_60076-8.6, Power Transformers,
application guide.
2. IEC 60076-22: 2019 IEC 60076-22
published in 2019, Specification requirement
of various accessories.
 National Conference 2019, CPRI

Annexure:
Check List for SMART Replacement
Sr. Existing Offered
no. Parameters Unit Unit Remarks
1 Parallel Operation ( Yes/No)
2 MVA Rating
3 Voltage ( KV) HV/LV/MV
4 Vector Group
5 Frequency ( Hz)
6 % Impedance @ MVA Base ( Min / Normal/Max Tap)
7 Voltage Variation
Voltage Variation HV (+/-Variation % in xx steps) or
Voltage Variation LV (+/-Variation % in xx steps) or
Voltage Variation MV (+/-Variation % in xx steps)
8 Overall Dimensions ( Lx B x H) (m)
Busduct Height from Rail Level (m) HV/LV/MV
Centre Line of CBbox/BD from Rail Gauge ( Longitudinal
Dimension) HV/LV/MV
Marshalling Box (Ground Mounted/ Tank Mounted)
CBBox (Ground Mounted/ Tank Mounted)
DGA (Ground Mounted/ Tank Mounted)
Online Drying System (Ground Mounted/ Tank Mounted)
9 Make & Model of the AVR
10 Alternate Solutions can be proposed to customer for approval Yes/No
Unit Coolers Allowed?
Ester Oil Design Allowed?
Plug in Bushings for cable terminations allowed?
11 Restrictions Comments
Overall Dimensions ( Lx B x H) (m)
Transport ( Lx B x H) (m)
Existing Coolers can be used?
Drop In arrangement available?
12 Documents Required Comments
Rating & Diagram Plate
%Impedance at main extreme taps
General Arrangement & Foundation
Transport drawing
Cable Box / Bus Duct Matching Dimensions
Bushing Details
 Mineral oil based nanofluids for transformer
insulation
P. Thomas
Dielectric Materials Division
Central Power Research Institute
Bangalore, India
thomas@cpri.in
Abstract - Mineral oil (transformer oil), a petroleum based
bi-product has been used as an insulating oil for power
transformers till date. In recent years, in order to have
improved critical parameters and heat transfer characteristics,
nanofluids were considered the next generation insulting fluids.
In this work, Mineral oil based nanofluids were developed by
employing various nanoceramics such as CaCu3Ti4O12 (CCTO),
Ba0.85Ca0.15Zr0.1Ti0.9O3 (BCZT), and Ca2Fe2O5. The effect of
these nanoparticles on the critical parameters like AC
breakdown voltage (BDV), Thermal conductivity, Flashpoint,
Resistivity and Viscosity of mineral oil studied in detail. The
mineral oil based CCTO nanofluids, besides showing enhanced
electric breakdown voltage, improvement in the thermal
conductivity also has been observed.
Keywords—nanofluids, mineral oil, nanoceramics, BDV,
thermal conductivity.
I. INTRODUCTION
Transformers occupy prominent positions in the power
system, being the vital links between generating stations and
points of utilization. Transformers in electric power delivery
systems around the world are filled with liquid. The liquid
functions as an electrical insulation and heat transfer fluid.
The quality of oil in transformer plays an important role in
performing this function. There is an tremendous increase in
the power generation and due to this, there is increasing
awareness on improving insulation both solid as well as liquid
[1]. Monitoring the transformer insulation is very important
since the failure of the insulation directly affects the reliability
and economy of the power system. The failures can be
overcome by proper designing of the equipment as well as
employing insulating materials with improved characteristics
[2]. Traditionally, mineral oil is used as a liquid insulation in
transformers. Mineral oil (transformer oil), a petroleum based
bi- product has been used as an insulating oil for power
transformers till date [3]. This is due to the fact that, mineral
oil is low cost, exhibit better dielectric strength, and possess
low viscosity. Mineral oil is a liquid by-product of refining
crude oil to make gasoline and other petroleum products. This
type of mineral oil is transparent, colorless oil, composed
mainly of alkanes and cycloalkanes, related to petroleum jelly.
Mineral oil contains hydrocarbons which are broadly
classified as paraffinic, naphthenic and aromatics. The
paraffinic molecules can be either straight chain or branched.
The naphthenic molecules are also known as cycloalkanes
[4,5]. The development of nanotechnology provides an
effective way to improve the performance of insulating oils.
The field of nanodielectrics is the future for the development
of insulating oils with improved critical parameters that can
change the design aspects of high voltage power apparatus.
Nano fluids exhibit more efficient heat transfer properties
compared to the host fluid [6]. The addition of micro sized
particles into base fluids is not efficient due to its heavier
weight and gravitational force which causes sedimentation of
particles leading to degradation in thermal properties [7,8].
This problem is significantly reduced by using nano sized
particles which are lighter and smaller in size and thus
enhancing the thermal properties of the fluid. Especially for
the transformer oil insulation, many nanoceramics were
employed for the development of nanofluids and tremendous
improvement in the critical characteristics has been achieved
[9-10]. In this work, various nanofluids were prepared by
employing few nanoceramics and the results obtained were
discussed in this work.
II. EXPERIMENTAL
A. Preparation of Nanofluids
Mineral oil available in the local market has been used in
this study. Nano ceramics such as CaCu3Ti4O12 (CCTO),
Ca2Fe2O5 (CFO) and Ba0.85Ca0.15Zr0.1Ti0.9O3 (BCZT) were
prepared in the laboratory and used for preparing nanofluids.
Series of nanofluids were prepared by mixing nanopowders at
various concentrations: 0, 0.001, 0.0025, and 0.005 by weight
percent in mineral oil. To achieve homogeneous sample with
less agglomeration, the samples were ultra-sonicated using
probe ultrasonicator.
B. Characterization techniques.
The electric strength was measured using a 100 kV
Automatic Breakdown voltage tester as per IEC 60156
standard [11]. The standard spherical electrode with the gap
spacing of 2.5 mm was used. The average of five values were
taken and computed as breakdown voltage. The Resistivity
has been measured as per ASTM D 1169 [12]. The applied
voltage is set as 500 V/mm (rms) in the three-electrode test
cell. The testing has been carried out at 27°C and 90°C. The
Viscosity of the samples were measured using a U-tube
calibrated glass capillary viscometer at various temperatures
like 27, 45 and 70°C as per ASTM D 445 standard [13]. The
Flash point tests were carried out in Automated Pensky-
Martens closed cup apparatus as per ASTM D 93 standard
[14]. The thermal conductivity test measures the temperature
gradient produced across the liquid by a known amount of
energy introduced into the test cell by an electrically heated
platinum element. The test method is carried as according to
ASTM D2717.
III. RESULTS AND DISCUSSION
The nanofluids were characterized for parameters like
Resistivity, AC breakdown voltage (BDV), Viscosity,
Flashpoint, and Thermal conductivity as per guidelines given
in their respective ASTM/IEC standards.
A. AC Breeakdown Voltage
Electric strength values obtained for the mineral oil based
Nanofluids are shown in Fig.1. The electric strength values
obtained for the CCTO nanofluid shows enhanced BDV
results than base sample. A linear increase in the BDV laues
has been observed, however, beyond 0.055%, nanoparticles
aglomeration has been observed.The nanofluid with 0.005wt
% of CCTO exhibited better results, indicating the optimum
level of nanoparticle loading. Similar trend has been observed
in the case of BCZT nanofluid and around 15% enhancement
in the AC breakdown voltage was noticed. In the case of
calcium ferrite decrement in the BDV has been observed,
which needs to be studied in detail.
Fig.1. Effect of nano particles on breakdown
voltage of mineral oil samples
B. Resistivity
Mineral oil having the resistivity 1.58X1014 at room
temperature and it is very clear that with increasing the
concentrations of CCTO nanomaterial, the volume resistivity
has increased as shown in Figures.2. The other nanofluids had
showed decreasing trend.
Fig.2. Resistivity measured at 27°C for all the
nanofluids
C. Thermal condcutivity
Fig. 3. shows thermal conductivity values obtained for the
nanofluids. Transformer fluids are expected to act as an
electrically insulating medium and heat transfer agent. It is
thus desirable for the fluid to have a higher thermal
conductivity. Thermal conductivities are directly related to
aromatic content and the thermal conductivity increases with
increase in viscosity. Thermal conductivity also has an
influence on size and shape of the additives that are used for
the development of nanofluids. The room temperature
thermal conductivity of the base oil is around 0.138 W/mK.
By employing the Nanoceramics into the base oil the
increased thermal conductivity was achieved for all the
nanofluids.
Fig.3. Thermal conductivity behavior of
nanofluids.
D. Flash Point
The flash point characteristics of the nanofluids are shown
in the Figure.4. Flash point of the base oil is 150°C which has
increased as the weight percent of CCTO ceramics increased
in the mineral oil. The BCZT nanofluid does not show any
variation and the Calcium ferrite addition had decreased the
flash point characteristics.
Fig.4. Flash point characteristics of nanofluids.
E. Viscosity
It is observed that the viscosity for the nanofluids does not
show much variation with the increasing concentration of
nanomaterial in transformer oil as shown in the Figure.5. The
viscosity decreases with increase in the temperature for all the
nanofluids. These results indicated that the nanoparticles
influence on the viscosity is minimal. However, nanoparticles
have greater influence on the electrical properties. Similar
trend has been observed for all the other nanofluids.
Fig.5. Viscosity measured at different temperatures
for the CCTO nanofluids.
IV. CONCLUSIONS
The mineral oil based nanofluids were successfully
prepared by employing various nanoceramics under
ultrasonication. The electric strength values increased as the
weight percent of the CCTO nanoparticles increased in the
mineral oil. Though there is no variation in the viscosity and
acidity of the nanofluids, improvement in the thermal
conductivity were observed as the weight percent of the
nanoparticles increased in the Mineral oil. The stability of the
nanofluids needs to be examined thoroughly by employing a
suitable dispersant or surfactant. These results indicated that
CCTO nanoparticles is a good candidate material for the
development of Mineral oil based nanofluids for transformer
insulation applications.
ACKNOWLEDGMENT
The management of Central Power Research Institute is
acknowledged for the financial support (CPRI Project No.
IHRD/2015/TR/7/15122015).
REFERENCES
[1] Transformers, Second edition, Bharat Heavy Electricals
Limited, Tata Mc-Graw Hill Publishing Limited, 2003.
[2] A.A. Halacsy and G.H. von Fuchs, “Transformer
Invented 75 Years Ago,” Trans. Amer. IEE Power
Apparatus Syst., Part-III, Vol.80, No. 3, pp. 121-125,
1961.
[3] V. Saurabh and V. Prashant, “Transformers History and
its Insulating Oil”, 5th National Conf., Computing for
National Development, India, pp.1-4, 2011.
[4] C. M. Wilson, “Insulating Liquids, their use,
manufacture and properties” Peter peregrinus Ltd.
(1980).
[5] Frank M. Clark, “Insulating Materials for Design and
Engineering Practice” Wiley and Sons, Inc., Publishers,
New York (1962).
[6] Das, Sarit K, Stephen U. S Choi, Wenhua Yu, and T.
Pradeep, “Nanofluids: Science and Technology”,
Wiley-Interscience, 2008.
[7] Wei Yu andHuaqing Xie, “Review Article on A Review
on Nanofluids: Preparation, Stability Mechanisms and
Applications”, Journal of Nanomaterials, Volume 2012,
Article ID 435873, 2012.
[8] V. Trisaksri and S.Wongwises, “Critical review of heat
transfer characteristics of nanofluids,” Renewable and
Sustainable Energy Reviews, vol. 11, no. 3, pp. 512–523,
2007.
[9] R. T. A. R. Prasath, N. K. Roy, S. N. Mahato, P. Thomas,
“Mineral oil based high permittivity CaCu3Ti4O12
nanofluids for power transformer application,” IEEE
Trans. Dielectr. Electr. Insul., vol. 24, no. 4, pp. 2344–
2353, August 2017.
[10] P. Thomas, N. E. Hudedmani, R. T. A. R. Prasath, N. K.
Roy, S. N. Mahato, “Synthetic ester oil based high
permittivity CaCu3Ti4O12 nanofluids an alternative
insulating medium for power transformer,” IEEE Trans.
Dielectr. Electr. Insul., vol. 26, no. 1, pp. 314–321,
February 2019.
[11] IEC 60156 Third Edition, “Insulating Liquids-
Determination of Breakdown voltage at Power
Frequency- Test method, 2003-11.
[12] ASTM D 1169, “Standard Method of Test for Specific
Resistance (resistivity) of electrical Insulating Liquids”,
2011.
[13] ASTM D 445, “Standard Test Method for Viscosity of
Transparent and Opaque Liquids (Kinematic and
Dynamic Viscosities), 2011.
[14] ASTM D 93, “Standard Test Methods for Flash Point by
Pensky-Marten Closed Cup Tester”, 2012.
EXPERIENCE OF TOSHIBA (INDIA) IN DESIGNING SHORT CIRCUIT
PROOF TRANSFORMERS
V.S.KRISHNA PAVAN. S, R.GOPALA KRISHNA, G.SREELATHA,

Y.SUDHAKAR, A.VIKRAM

TOSHIBA TRANSMISSION AND DISTRIBUTION SYSTEMS (INDIA) PVT. LTD.

INDIA

040-3071 7206 (Ph)

040-3071 7263/2245 (Fax)

krishnapavan.sanagala@toshiba-ttdi.com

Abstract: the three-phase fault and three phase-

The reliability of large power transformers
is of paramount importance for the bulk
transmission of electric power.
Transformers experience a variety of
electrical, mechanical and thermal
stresses during its operation. To assess
the inherent quality of transformers
numerous routine, type and special tests
are performed. Short circuit withstand
capability of power transformer is the
ability to withstand the full asymmetrical
short circuit currents in each phase and in
each representative tap position without
impairing its suitability for normal service
conditions. Successful completion of this
test proves the transformer’s robustness
to withstand various types of short circuit
currents in the field, validation of design
excellence and flawless manufacturing.
This paper deals about Toshiba (India)
experience in designing and
manufacturing short circuit proof
transformers.
ground fault (symmetrical fault) are the
most severe one. Hence, it is usual
practice to design a transformer to
withstand a three-phase short circuit at its
terminals, the other windings being
assumed to be connected to infinite
systems/sources.
Short Circuit Forces:
The basic equation for the calculation of
electromagnetic force is
F=BxIxL
Where, B is leakage flux density vector,
I is current vector and L is winding length.
The leakage flux density at any point can
be resolved into two components, viz. one
in the radial direction (Bx) and other in the
axial direction (By). Therefore, there is
radial force in the x direction due to the
axial leakage flux and axial force in the y
direction due to the radial leakage flux, as
shown in figure 1.0.

INTRODUCTION:

Short circuit currents:

There are different types of faults, which

result into high over currents, viz. line-
ground, line-line, line-line-ground, three-
phase and three phase-ground. Usually Fig. 1.0 Radial and Axial Forces
Radial forces:

The radial forces act outwards on the

outer winding tending to stretch the
winding conductor, producing a tensile
stress (also called as hoop stress).
Whereas the inner winding experiences
radial forces acting inwards tending to
collapse, producing a compressive stress. Fig. 3.0 Bending between radial spacers
Conductors of inner windings, which are
subjected to the radial compressive load, When these forces are more than a
may fail due to bending between supports certain limit, a failure can occur due to
or buckling. There are two types of tilting of conductors in a zig-zag fashion as
buckling phenomenon, one is forced shown in figure 4.0
buckling & other is free buckling. Forced
buckling is bending of conductors between
radial supports all along the circumference
as shown in figure 2.0. Free buckling is
bending of conductor in the span of radial
supports as shown in figure 2.0.

Fig. 4.0 Tilting phenomena

Fig. 2.0 Buckling phenomena

There are many factors which may lead to

the buckling phenomenon, viz. winding
looseness, inferior material
characteristics, eccentricities in windings,
lower stiffness of supporting structures as
compared to the conductor, etc…
Axial forces:
Under the action of axial forces, the
winding conductor can bend between the
radially placed insulation spacers as
shown in figure 3.0.
The failure due to tilting under the action
of axial compressive forces is one of the
principal modes of failures in large power
transformers.
i) Ampere-Turn balancing of all windings
to reduce axial forces.
ii)
copper yield strength to
withstand stresses (both radial and axial).
Additionally epoxy bonding of conductors
is used to enhance the mechanical
strength.
iii) Selection of radial coil supports to
withstand buckling.
iv) Alignment of radial support spacers
with the contact point between winding
support structure and core.
V) Dimensioning of spacer area
withstand axial forces.
vi) Dimensioning of clamping ring to
withstand axial forces.
vii) Design of core clamps and coil
pressing structure to withstand axial
forces.
viii) Coil spacers and runners are made
burr free to eliminate insulation damage.
ix) Individual windings are stabilized to
achieve designed heights.
x) Locking of end blocks to avoid
dislocation.
xi) Securing all winding exit leads and
connections to bushings & tap changers.
xii) Proper locking of active part with tank.
Toshiba (India) design facilities:
Toshiba (India) has in house validated
sophisticated analytical tools to check
dynamic short circuit withstand capability
of all types of power transformers. With
the help of in house sophisticated
analytical tools, following are the
forces/stresses calculated along with
safety factors:
i) Radial compressive stress & bending
stress of inner windings.
ii) Radial tensile stress of outer windings.
iii) Spiralling forces on helical winding lead
exits
iv) Axial compressive force and bending
stress in windings.
v) Compressive stresses on the paper of
the winding conductors & spacers.
vi) Tilting stresses of the winding
conductors.
to vii) Axial forces on end supports,
compression and bending stress in
mechanical structure.
Toshiba (India)’s Success Story in
producing Short Circuit proof
transformers:
With existing design facilities, Toshiba
(India) has experience of dynamic short
circuit testing of different voltage class of
power transformers up to 400kV for
different kind of applications like
Generator transformer, Auto transformer,
Split winding transformer, Scott
connection transformer, etc…Toshiba
(India) success rate in dynamic short
circuit testing is more than 90% in first go.
Toshiba (India) has done dynamic short
circuit testing of power transformers at
high power testing laboratories in India as
well as outside India i.e. KEMA-
Netherlands. Toshiba (India) is credited
with first time success of Short Circuit
testing of more than 40 transformers.
Some of the large rating transformers are
as under:
Sl. No. Rating
82MVA,15/420kV
1
1-Ph Generator Transformer
315MVA,400/220/33kV
2
3-Ph Auto Transformer
105MVA,400/220/33kV
3
1-Phase Auto transformer
70MVA,220/6.9-6.9/12kV
4
3-Ph Start Up Transformer
70MVA,21/6.9-6.9 kV
5
3-Ph Unit Transformer
110MVA,132/11.5/11.5 kV
6
3-Ph Station Transformer
 Sl. No. Rating iv) Insulation Levels:
131MVA,220/66 kV
7 3-Ph Power Transformer HV-630kVrms/1425kVp/1050kVp

45MVA, 27/11.5 kV PD Levels-U1:510kVrms & U2:460kVrms

8 3-Ph Unit Transformer
LV-70kVrms/170kVp
115MVA, 11/220 kV, HVN-38kVrms/95kVp
9 3-Ph Generator Transformer
v) % Impedance: 15% ± 7.5% tolerance
100 MVA, 220/55-55kV
(at principal tap)
10 Scott Transformer
Design features:
270MVA, 20/420 kV,
11 1-Ph Generator Transformer i) Winding Arrangement:
315MVA, 400/220/33 kV,
12 3-Ph Auto Transformer

275MVA, 23.5/420 kV,

13 1-Ph Generator Transformer

Some of case studies of dynamic short

circuit tested transformers of Toshiba
(India)

Case-I:

275MVA, 23.5/420kV, 1-Ph Generator

Transformer:-

This is the highest rating 1-Phase ii) Type of windings:

Generator Transformer which Toshiba
(India) was designed, manufactured and LV- Helical, HV-Disc, Tap-Disc
under gone dynamic short circuit test
successfully.This transformer was iii) Type of winding conductor:
supplied to one of the thermal power
LV, HV & Tap: Epoxy bonded CTC
generation stations in India. Brief
description of the transformer is given Design Considerations:
below.
i) Care taken to avoid dislocation of collars
Technical particulars: of LV winding.

i) Rating: 275 MVA 1-Ph Generator
Transformer (3-Ph Bank Capacity:
825MVA)
ii) Voltage Class: 23.5/420/√3 kV
ii) Turns are supported and locked to
avoid spiralling of LV winding.
iii) Epoxy bonded CTC used in LV winding
to avoid buckling.

iii) Cooling: ONAN/ONAF/ODAF

iv) Dummy runners are provided to avoid c) Axial Displacement Vs Height
waviness of the LV winding.

v) Close manufacturing tolerances are

given for stack height to reduce axial
forces.

vi) Lead exits are locked between top and

bottom pressure rings.

Vii) Care taken to avoid damage to top

disc insulation of HV winding.
d) Axial Force Vs Height
viii) Top and Bottom blocks of tap winding
ends are locked.

Design Analysis:

Some of the typical graphs for different

windings are as below

LV Winding:

a) Axial Displacement Vs Time

HV Winding:

a) Axial Displacement Vs Time

b) Axial Internal Force Vs Time

b) Axial Internal Force Vs Time

 c) Axial Displacement Vs Height c) Axial Displacement Vs Height

d) Axial Force Vs Height d) Axial Force Vs Height

Tap Winding: Testing Experience:

a) Axial Displacement Vs Time i) 275MVA, 23.5/420/√3 kV 1-Ph

generator transformer is the largest rating
1-Ph generator transformer successfully
tested indigenously.

ii) Number of shots during SC test:

Maximum tap – 1 Shot

Principal tap – 1 Shot

Minimum tap – 1 Shot

b) Axial Internal Force Vs Time
iii) After application of 3 shots of short
circuit current, variation in inductance
observed is -0.375% against the
acceptable limit of ± 1%.

Case-II:

315MVA, 400/220/33kV, 3-Ph Auto

Transformer:-

315MVA, 400/220/33kV, 3-Phase auto

transformer was successfully short circuit
tested. Brief description of the transformer
is given below.
Technical particulars: iii) Type of winding conductor:

i) Rating: 315MVA, 3-Ph Auto Transformer
ii) Voltage Class: 400/220/33kV
iii) Cooling: ONAN/ONAF/ODAF
iv) Insulation Levels:
HV-570kVrms/1300kVp/1050kVp
PD Levels-U1:510kVrms & U2:460kVrms
IV-395kVrms/950kVp
LV-95kVrms/250kVp
N-38kVrms/95kVp
LV, Tap, COM & SER: Epoxy bonded
CTC
Design Considerations:
i) Turns are supported and locked to avoid
spiralling of TAP winding.
ii) Close manufacturing tolerance are
given for stack height to reduce axial
forces.
Design Analysis:
Some of the typical graphs for different
windings are as below

v) % Impedance: Tertiary Winding:

HV-IV: 12.5% ± 7.5% tolerance (at a) Axial Displacement Vs Time

principal tap)

HV-LV: 60% Minimum (at principal tap)

IV-LV: 45% Minimum

Design features:

i) Winding Arrangement:

b) Axial Internal Force Vs Time

Fig. 6.0

ii) Type of windings:

TER- Disc, Tap- Multi start helical, COM-

Disc, SER- Disc
 c) Axial Displacement Vs Height c) Axial Displacement Vs Height

d) Axial Force Vs Height d) Axial Force Vs Height

Tap Winding: Common Winding:

a) Axial Displacement Vs Time a) Axial Displacement Vs Time

b) Axial Internal Force Vs Time b) Axial Internal Force Vs Time

 c) Axial Displacement Vs Height c) Axial Displacement Vs Height

d) Axial Force Vs Height d) Axial Force Vs Height

Series Winding: Testing Experience:

a) Axial Displacement Vs Time i) Number of shots during SC test:

HV-IV mode:

Maximum tap – 3 Shots

Principal tap – 3 Shots

Minimum tap – 3 Shots

IV-LV mode:

9 Shots ( 3 shots in each phase)

b) Axial Internal Force Vs Time
ii) After application of total 18 shots of
short circuit current, variation in
inductance is as follows

HV-IV: +0.36%

IV-LV: +0.594%

The acceptable limit is ± 1%.

Case-III:

70MVA, 220/6.9-6.9/12kV, 3-Ph Split

Winding Transformer:-

It is proud moment for Toshiba (India) that

70MVA, 220/6.9-6.9/12 kV, 3-Ph split
winding transformer successfully tested in
first attempt. After this success Toshiba
(India) has successfully completed Short
Circuit test on similar type of transformers
(total 3 No’s) and supplied to various
utilities. We have also designed & tested
such split winding transformers of 400kV Fig. 7.0
class. A brief description of 70MVA
transformer is given below. ii) Type of windings:

Technical particulars: LV1 & LV2- Helical, HV- Disc, Tap- Disc

i) Rating: 70MVA, 3-Ph Split Winding iii) Type of winding conductor:

Transformer
LV1 & LV2- Epoxy bonded CTC, HV &
ii) Voltage Class: 220/6.9-6.9/12kV Tap- PICC

iii) Cooling: ONAN/ONAF Design Considerations:

iv) Insulation Levels: i) LV1 & LV2 turns are locked & tightened
to avoid spiralling effect.
HV-460kVrms/1050kVp
ii) Epoxy bonded CTC used in LV1 & LV2
LV1, LV2, LV1N & LV2N-28kVrms/75kVp winding to avoid buckling.
TER-28kVrms/75kVp iii) Lead exits are locked between top and
HVN-50kVrms/125kVp bottom pressure rings.

v) % Impedance: iv) Due care to be taken while designing

HV-LV1 & HV-LV2: 12% ± 5% tolerance
(at principal tap) on 35MVA base
LV1-LV2: 22 ± 15% tolerance on 35MVA
base
Design features:
i) Winding Arrangement:
bus bar supporting structure to restrict
electromagnetic forces between them.
v) Top & Bottom blocks of tap winding are
locked.
Design Analysis:
Some of the typical graphs for different
windings are as below
LV1 Winding: LV2 Winding:

a) Axial Displacement Vs Time a) Axial Displacement Vs Time

b) Axial Internal Force Vs Time b) Axial Internal Force Vs Time

c) Axial Displacement Vs Height c) Axial Displacement Vs Height

d) Axial Force Vs Height d) Axial Force Vs Height

HV Winding: TAP Winding:

a) Axial Displacement Vs Time a) Axial Displacement Vs Time

b) Axial Internal Force Vs Time b) Axial Internal Force Vs Time

c) Axial Displacement Vs Height c) Axial Displacement Vs Height

d) Axial Force Vs Height d) Axial Force Vs Height

Testing Experience: success rate of full scale Dynamic Short
Circuit test (approx. 40%), we Toshiba
i) Number of shots during SC test: (India) have achieved a success rate of
HV-LV1 (Maximum tap) – 3 shots more than 90%.

Acknowledgement:
HV-LV2 (Principal tap) – 3 shots

HV-(LV1+LV2) (Minimum tap) – 3 shots The authors are thankful to Toshiba

(India) management for encouragement
ii) After application of 9 shots of short given at different stages of transformer
circuit current, variation in reactance is design learning. The authors are also
+0.29% against acceptable limit of ± 2%. thankful for the permission given to us to
share our success story.
Toshiba (India) Failure experience:
References:
In addition to success story, Toshiba
(India) experienced certain problems 1)Transformer Engineering (Design and
during short circuit test which are listed Practice) by S.V.Kulakarni &
below: S.A.Khaparde

i) Free buckling of LV winding 2)M.V.Prabhakar,T.K.Ganguli,

S.C.Bhageria, R.C.Agarwal, S.K.Gupta
ii) Breaking of top pressure ring due to
disposition of the coil pressing screw. Short Circuit Withstand capability of
Power Transformers
iii) Collapse of winding due to dislocation
of bottom coil supports. 3)V.K.Lakhiani, Dr.S.V.Kulakarni

Keeping in view above failures, due care Short Circuit Withstand of Power
is taken in all subsequent designs and Transformers – A Perspective
success story of Toshiba (India)
continued.

Conclusion:

We at Toshiba (India) design each and

every transformer to cater to Short Circuit
withstand in the network. The design is
conceptualised by taking into account all
type of short circuit forces that the
transformer may experience during its life
time. During design stage, designer
makes all design elements with care and
represents the same properly. After which,
the main task starts at the manufacturing
stage where proper materials are used
and manufacturing is done in line with
design. The translation of design to the
manufacturing process is the main
challenge to produce Short Circuit proof
transformers. As against the world
 Pre-Magnetizing of Transformer Core for
Inrush Current Restraint-New Proposals
1] Kumar Virvani 3] 4]Santosh
2]Gaurang Chauhan Ram Krishna Mishra C. Vora
Raychem RPG Pvt Ltd Raychem RPG Pvt Ltd Raychem RPG Pvt Ltd Institute of Technology,
kumar_virvani@rayche gaurang_chauhan@rayche ram_mishra@raychemrp Nirma University
mrpg.com mrpg.com g.com santoshvora@yahoo.com

Abstract - Transformer is a key component in factors such as circuit complexity, cost effectiveness,
electrical power system and its reliability is of practical feasibility, extent of inrush current mitigation (%
utmost importance for uninterrupted functioning of reduction of inrush current magnitude) etc.
electrical system. When a transformer is energized Conventionally, point on wave method.
at no-load, large amount of inrush current is drawn
because of core saturation. The magnitude of this
current may be as high as 50-60 % of terminal short
circuit current. The effect of Inrush Current on the
transformer windings is quite severe as it lasts for
longer duration (0.2 to 0.5 sec) compared to short
circuit current (0.08 to 0.1 sec) and it can lead to
reduction in lifespan of transformer, mal-operation
of protection system if it experiences switching
surges over a short span. In this paper, analytical
and FEM technique to compute inrush current has
been described and few methods for inrush current
mitigation have been explored.

Keywords— Finite Element Analysis (FEA), Inrush (IR)

Current, Multi-Winding Transformer, Electromagnetics.
I. INTRODUCTION
Pre-magnetizing of transformer is an evolving to
reduce inrush current in transformer. Magnetizing inrush
current is an unsolved phenomenon since the inception
of transformer, and it needs ample study to reduce the
damaging effects of inrush current during no-load
switching of the power transformer.
When transformer is energized at no-load, its
winding acts as an inductive circuit. Magnetic flux
generated will be lagging the source voltage by 90
degree electrical [1]. Due to absence of initial magnetic
flux in the core, supply voltage and induced flux start
from zero instance. Due to this effect, more than twice
the magnetic flux is induced in the core during first cycle
which can result in transformer core saturation and large
amount of magnetizing inrush current flows through
windings. This phenomenon is shown in figure 1 with i0
waveform[1].
There are a few methods available in literature to
mitigate the inrush current of transformer, but each
method has its own limitation depending on various
Figure 1 Initial core magnetizing current waveform (with residual flux)
[2] is used to limit inrush current to less than the no-load
current. It is basically the controlled switching of circuit
breaker at the time of transformer energization.
Limitation of this method is that it fails in giving
accurate residual flux present in core.
During this phenomenon, rate of decay of inrush
current depends on fluctuation of air-core saturated
reactance value [3]. This inrush current has significant
impact on transformer life and results in protection
system mal-operation due to significant electromagnetic
forces experienced by the windings[4].
There is also one method namely pre-insertion of
resistor (PIR) [5] which regulates the input voltage and
reduces the inrush current peak. But this method is
limited for low short circuit capacity system. For marine
application, additional auxiliary transformer [6] of 0.5-1 %
rating is included with main transformer. It pre-
magnetizes the transformer core before connecting main
supply to transformer and hence reduces the core
saturation. Cost of auxiliary transformer and relay

`
coordination are major concern in implementation of this
method.
Soft start circuit [7] can also be used to limit inrush
current but, the availability and cost of these anti-parallel
phase-controlled thyristor are limiting factor for
commercial use. In this paper, two modified methods
namely pre-magnetization of transformer by in-built
auxiliary winding and combined control of inrush current
by pre-insertion resistors and phase-controlled thyristor
are analyzed in detail.
II. ANALYTICAL COMPUTATION
Inrush current calculation is done for transformer
to compare the magnitudes of both short circuit and
inrush current condition. The formula to calculate inrush
current is referred from “T. Specht” [8]:
A (1)
Where, is constant (1.15) up to 20 MVA rating,
is air-core reactance during saturation (in ohms), E is
RMS phase voltage (in volts), θ is switching angle at
which circuit breaker is closed which is given by;
) (2)
Where, is saturation flux density (in Tesla), is
rated flux density (in Tesla) and is residual flux
density (in Tesla).
Dry type vacuum pressure impregnated converter
duty transformer of 4.9 MVA rating is used for analysis. It
is a special purpose star/ex-delta configuration
transformer which has 2 high voltage (HV) winding in
parallel which forms single primary winding as well as
there are 6 secondary low voltage (LV) windings.
Windings are wound in concentric manner with HV as
inner winding and LV as outer windings. Using equation
(1), magnitude of inrush current is calculated as 5.91 kA
for this transformer.
Figure 2: Window diagram
`
To reduce the simulation time, certain
assumptions are considered like edge strip and electrical
insulation are not modelled, only ‘Normal Tap’ simulation
and electrical height of the windings is considered.
Transformer is designed with 102 turns in primary
winding.
Table 1. Transformer Specifications.
____________________________________________________
Specifications Value
____________________________________________________
Rated power [MVA] 4.9
Voltage [kV] 6.6/1.920
Current [A] 428.6/1473
No. of windings [HV/LV] 1(split windings)/6
Impedance 9.79 %
____________________________________________________
Fig. 2 shows window diagram of multi-winding dry
type transformer (all dimensions are in mm) and
transformer specifications are mentioned in Table 1.
III. EFFECT OF INRUSH CURRENT ON TRANSFORMER AND
DIFFERENT MITIGATION MATHODS:
i. Effect of inrush current on transformer:
Inrush current effect on transformer can be
analyzed by comparing with short circuit condition. Finite
element analysis has been done for both the conditions
and results are discussed. It is observed by analysis that
although the magnitude of inrush current (5.91 kA using
equation-1) computed as 55 % of short circuit current
(Magnitude of short-circuit current is computed in [9] for
same transformer), duration of inrush current is 4 times
the short circuit current. Also, the frequency of
occurrence of inrush current is more compared to short
circuit current. The magnitude of axial force experienced
between high voltage windings in inrush current
condition is 26.5 N compared to short circuit force of
22.4 N [9]. These results emphasize that inrush current
condition has significant impact on transformer windings
compared to short circuit current especially in terms of
magnitude of axial electromagnetic forces.
ii. Mitigation Methods:
Method – 1: Combined control of inrush current by
resistors and phase-controlled thyristors:
The proposed method for inrush current mitigation
is presented in fig 3. The main function of this circuit is to
energize the transformer core with both pre-insertion of
resistor and regulated voltage source by controlling the
firing angle of anti-parallel thyristors from 180 to 0
degree electrical. An optional switch could be used to
bypass combined circuit from the main circuit in steady
state operation. The proposed method could be applied
to restraint large amount of inrush currents in medium-
voltage transformer.
Major cost contributing components are pre-
insertion resistor, anti-parallel thyristor, phase lock loop
(PLL) and bypass contactor of medium voltage. Benefit Table 2 Design calculation for auxiliary winding
of using combined resistors and thyristors method is __________________________________________________
reduction in voltage ratings of thyristor and heat Parameter Value Remarks
generation from resistors is less as only magnetizing Power capacity 37 kVA S=VxI
current flows through primary winding. Also, the method Voltage supply 240 V star configuration
is very simple and does not require magnetic flux Current rating 52 A as per no-load loss
calculation or specific point on wave voltage to energize Volts/turn 24 V -
Total no. of turns 10 formula, 240/24 =10
the transformer. The same method is suitable for either
Current density 2.5 -
single-phase or three-phase transformers. Area in sq.mm 20.8 A = I/J
Conductor size 11.5 x 2.5 as per winding area
Winding resistance (R) 22.24 mΩ formula R
*Remarks: formula ohm

IV. SIMULATION STUDY:

i. MATLAB circuit of combined control by resistor

and anti-parallel thyristors:

Figure 3: Block diagram of combine circuit method To investigate the proposed inrush current mitigation
technique, MATLAB simulation of transformer is done on
Method – 2: Pre-magnetizing of transformer by in-built real time basis for (time period). A whole circuit with
auxiliary windings: provision of bypass circuit breaker is given in fig. 6.
The second proposed method of in-built winding
energization is presented in window diagram shown in
Fig. 4. The main function of this method is to magnetize
the transformer core at starting with applying regulated
voltage source to auxiliary winding before connecting
with main supply and prevent it from saturation.

Figure 5: Firing circuit of phase A for phase-controlled thyristor

4.9 MVA, 6.6 kV/1.920 kV, two windings transformer

Figure 4: Window diagram with auxiliary winding
Design of auxiliary winding is done based on
percentage of rated current which magnetizes the core
and other required data is formulated in Table 2. After
energization of transformer, auxiliary winding is
disconnected using bypass contactor.
is fed by 6.6 kV three phase voltage source. Resistor of
100 ohm and a pair of anti-parallel thyristor of 2.8 kV
across each phase are placed in between source and
transformer. These both components provide the ramp
input at its terminal from 0 to 0.5 sec. There is also a
provision of contactor to bypass this control circuit at 0.5
sec instance. Firing circuit for anti-parallel thyristors is
mentioned in Fig. 5. Total simulation run-time is kept at 1
sec and results are discussed in next section.

In the next section, combined controlled circuit

method is analyzed by MATLAB simulation and in-built
winding method is analyzed by using finite element (FE)
analysis. The simulation results for each case is
discussed afterwards.

`
 Figure 8: External circuit diagram for voltage excitation

Figure 6: MATLAB block diagram of combined control method

ii. FE analysis of in-built winding:
To validate the proposed method of mitigation, FE
analysis of transformer under transient solver is done.
Fig.7 shows the 2-D axisymmetric model of transformer
with in-built auxiliary winding. As discussed in previous
section, design of auxiliary winding is done considering
location, distance from yoke, percentage of rated
current, etc. Boundary condition and excitation is
mentioned in table 3.
Excitation for the auxiliary winding is given by
external circuit. 240 V rms single phase source with
22.24 mΩ (refer table 2) is given to the auxiliary winding.
Remaining both high voltage and low voltage windings
are kept open circuited with higher impedance.
Simulation run-time is 0.5s.

Table 3 Input parameters for FE analysis

_______________________________________________
Parameter Value
Solver 2D Transient
Mesh Extreme fine
Boundary condition Vector potential
Excitation External circuit
Simulation time 0.5 sec
Time step 0.5 msec
Figure 8: External circuit diagram for voltage excitation

V. RESULTS AND DISCUSSION:

Figure 7: 2-D Axisymmetric model and mesh model
i. Combined control by resistor and anti-parallel
thyristors:
There are two results shown in this section which
comprises of both ‘without and with mitigation method’.
The source voltage, inrush current and magnetic flux
waveform without any mitigation is shown in Fig 9. First
peak of inrush current in phase A is 2229 A at worst
residual flux condition of (0.8, -0.4, -0.4) pu. Flux in
phase A is 2.1 pu and decay with respect to change in
inrush current. The source voltage, inrush current and

`

Figure 9: Waveform of voltage, inrush current and flux without mitigation
flux waveform with mitigation is shown in Fig 10. From
the duration of 0 to 0.5 sec, presence of both resistors
and anti-parallel thyristors gives ramp voltage input at
transformer input terminals.
Figure 10: Waveforms of source voltage, inrush current and flux
with mitigation
Only 20 A (0.27 pu) of peak current has sailed in
the circuit and magnetized the core up to 0.7 pu. At 0.5
sec, contactor has been operated to bypass this control
circuit. The first peak of inrush current after bypassing is
58 A with rich harmonic content. Due to this current,
magnetic flux also shoots up to the value of 1.1 pu with
rated supply voltage, but it does not saturate the core
and transformer core is magnetized as desired.
ii. In-built winding:
After simulation, it is observed that:
`
Figure 11: Magnetic flux density and flux distribution plot in transformer
core
 Axial force of 6.5 N is experienced by auxiliary
winding.
 Auxiliary winding is supplied with 220 V/phase
voltage source which provides 87.5 A peak current in
1st cycle and 52.5 A peak current in 25th cycle (0.5
sec).
 Transformer core is magnetized up to 1.93 T and it
will not get saturated.
Main advantage of this method is that it is simple
and cost-effective compared to other mitigation method
as low supply voltage is needed during pre-magnetizing
period.
VI. CONCLUSION:
In this paper, transformer core behavior under
energizing condition is analyzed and explained. Inrush
current has been calculated and verified for an actual
transformer. Effect of inrush current on transformer is
also studied by finite element analysis. Two methods are
identified namely ‘pre-magnetization of transformer by
auxiliary winding’ and ‘combined control by resistor and
phase-controlled thyristors’. Both selected methods are
analyzed through finite element analysis and MATLAB
circuit respectively. In auxiliary winding method, it is
observed that current flow from the winding does not
saturate the core. In combined control by resistor and
phase-controlled thyristors method, only magnetizing
current is flowing from the winding. Also, there are
certain benefits like low voltage rating of thyristors are
required as compared to primary voltage, low power is
dissipated from resistors. Therefore, it is concluded that
both methods do not saturate the core hence restrains
inrush current.
 VII. REFERENCES:
[1] S. V. Kulkarni and S. A. Khaparde, Transformer Engineering:
Design and Practice, 2nd ed. Published, New York: Marcel
Dekker, May 2004, Pg no. 56-61.
[2] Urmil Parikh, Bhavesh R. Bhalja, "Mitigation of magnetic inrush
current during controlled energization of coupled un-loaded power
transformers in presence of residual flux without load side voltage
measurements", International journal of Electrical power and
energy system,2015
[3] Allan Greenwood, Power System Transients, 2nd ed. Published.
New York: Wiley Inter-science, 1991, pg no. 68-72.
[4] L F. Blume,G. Camilli,S. B. Farnham, H. A. Peterson,
"Transformer Magnetizing Inrush Currents and Influence on
System Operation", AIEE Transaction, Vol.63, March 1944.
[5] China Nat O_shore Oil Corp.,Cui Rong Zhang Wanbing , "Pre-
insertion Resistor with switching control", CN103594222, March
2016.
`
[6] Wuhan Guide Electric Co Ltd., Xiang li Zheng Tu, "Transformer
Pre-magnetizing De-vice",CN201732648, Feb 2011.
[7] Igor A. Pires et al., “Mitigation of Electric Arc Furnace Transformer
Inrush Current using Soft-Starter-Based Controlled
Energization” IEEE Transaction on Industrial Application, March
2017O. J. Bedrij, “Carry-select adder,” IRE Trans. Electron.
Comput., pp. 340–344, 1962.
[8] T. R. Specht, "Transformer Magnetizing inrush current", AIEE, Vol
70, NO. 3, March 1951.
[9] G. Chauhan, and K. Virvani, “FE Analysis of Short-Circuit &
Inrush Current Electromagnetic Force in Multi-Winding
Transformer” National Conference on Innovation & Best Practices
in Transformer Design, Testing and Maintenance at CPRI, Bhopal
2019.
 LIFE CYCLE MANAGEMENT OF POWER TRANSFORMERS-CPRI EXPERIENCES

Dillip Kumar Puhan, Rajat Sharma, K. P. Meena, Arunjothi R, Thirumurthy

Central Power Research Institute (CPRI), Bengaluru, India
Emails: dillip2009@cpri.in, rsharma@cpri.in, meena@cpri.in, arunjothi@cpri.in, thiru@cpri.in

ABSTRACT
Electric power plants as well as transmission and distribution grids include a large number of
equipment like high voltage rotating machines viz Generators, motors, power cables, and sub-
station equipment like Transformers, CTs, PTs, Switch gears etc of different age, manufactured and
installed during decades. These equipment are quite expensive and form a significant portion of
plant assets and more importantly vital components in the reliable delivery of electric power.
Insulation systems for power equipment are a complex combination of materials and have
undergone changes in the last few decades. Insulating materials do comply with the required
performance at the beginning of their life, though during their course of operation ageing and
deterioration may occur to the effects of various stresses. Failure of the insulation directly or
indirectly will result in failure of power equipment which in turn results in forced outages, reduced
reliability and increased maintenance and repair costs. Quality requirements are increasing as
outages of electric power distribution are expensive. In this paper various condition assessment
methodology and diagnostic techniques for life cycle management of power transformers are
discussed and few case studies are presented.

1. INTRODUCTION

Transformer life cycle is best understood form traditional “Bathtub Curve”. Transformer failure
rate has been found to follow bathtub curve. The bathtub curve depicts transformer life cycle in
three stages. The first stage is known as infant mortality period and it has a decreasing failure
rate. The infant mortality is followed by a nearly constant failure rate period, which is usually
long. Finally the curve ends in the third stage, a period of wear out with increasing failure rate,
which is mostly related to equipment aging. Sometimes the first stage has increasing failure rate
that indicates failures arising from inadequacy of material used or imperfect manufacturing,
handling or control processes. In such cases failure rate in the first stage peaks quickly and is
followed by a period of decreasing failure rate. In some other cases decreasing failure rates is
followed by an increasing failure rate in the first stage. The decreasing failure rate corresponds to
infant mortality of the traditional bathtub curve. The increasing failure rate which follows is
mostly due to problems similar to those mentioned above but which requires a longer exposure
time for manifestation. The infant mortality period of life cycle results from failures in weak
subsamples. Infant mortality period includes all failures prior to steady state. International
experience shows global failure rate is less than 2% per year. Highest failure rates are observed at
power transformers at upper voltage levels.

In order that infant mortality is low, design acceptance tests are performed on transformers
before they are put in service. Design acceptance test is defined as a test carried out to determine
the adequacy of the design of a particular type, style or model of equipment and its component
parts to meet the assigned ratings and to operate satisfactorily under normal service conditions if
specified, and to determine compliance with appropriate industry standards.
When the transformers are at the third stage of the bathtub curve, the failures increase rapidly
overtime. In the final quartile of transformer life cycle, the failure probability can increase by 5
times than the normal system failure rate.
The transformer insulation system in service is subjected to a combination of electrical,
mechanical, thermal and environmental stress. Neither all the factors those affect the life
expectancy of transformers are known, nor are properly understood. However in general, the
following facts are significant.

(a) The initial thermal, electrical and mechanical properties of insulating materials used
affect life expectancy. Various materials lose their properties at different rate and during the
aging process the same material may lose different properties at different rates.

(b) Thermal, electrical and mechanical stresses also affect life of transformer. Thermal stress
arise due to overload currents, local overheating, leakage fluxes as well as well as malfunctioning
of cooling system.

(c) Electrical stresses are caused by system as well as transient over-voltages, winding
resonances etc.

(d) Mechanical stress between leads, conductors and windings are produced by short-circuit
inrush currents.

(e) Normal load cycle along with environmental factors like ambient temperature and
humidity affect life expectancy. Both mechanical and dielectric withstand strength of the
transformer is reduced by aging of its insulation. During the service life of transformer, it is
subjected to faults that result in high radial and compressive forces. With system growth the
operating stress on transformer increases as the load increases. In an aged transformer, typically
the conductor insulation is weakened to the point where it can no longer sustain mechanical
stresses of the fault. Then dielectric failure of turn to turn insulation occurs or loosening of
winding clamping pressure takes place, which reduces the transformer’s ability to withstand
future short circuit forces [1, 2].

Central Power Research Institute (CPRI), Bengaluru, a premier institute for Indian Power Sector
has been conducting diagnostic testing on power transformers in service to assess the condition
and their useful life.

2. CAUSES OF INSULATION FAILURE

Transformer failure can be broadly classified as electrical, thermal and mechanical. The failures
are also classified in a different manner as internal or external. Failure due to insulation
degradation, PD, increased moisture content, overheating, winding resonance etc. fall under the
internal category, whereas the faults due to lightning strikes, switching over-voltages, system
faults, system overload etc. fall under the external category. So far as the location of failure is
concerned, it could be in the main tank, bushings, tap changers or in the transformer accessories.
Notable defects in the transformer major insulation are (a) moisture in cellulose insulation, (b)
contamination of oil with water, particles and insulation aging product, (c) Insulation surface
contamination, which occurs mainly due to absorption of aging by products on a cellulose
surface or due to deposition of conducting particles and insoluble aging products and (d) Partial
Discharges in weaker parts of insulation. The presence of moisture and impurities changes
dielectric parameters of deteriorated components, viz. their conductivity, permittivity and
dissipation factor, particularly with temperature. This in turn results in related changes in
dielectric characteristics of the whole transformer. Defects related to excessive moisture, oil
contamination or surface contaminations are reversible defects, while damage created by partial
discharge activities is usually irreversible in nature [1,2].

3. CONDITION ASSEEEMENT METHODOLOGY

Though there are several criteria proposed by the researchers over the years there is no unanimity
in the global technical community about the best criterion for remaining life assessment or End
of Life prediction. Residual Life Assessment [4], [5] in true sense involves following steps:
1. Collection of O&M history of the equipment.
2. Visual inspection & examination of the equipment.
3. Conducting appropriate diagnostic tests to assess the present status of the equipment.
4. Analysis of the data to detect extent of deterioration or to detect defective components.
5. Recommendation of appropriate remedial measures to
a. avoid forced outages
b. extend residual life

The diagnostic tests need to be conducted periodically [every (2 / 3 / 4 years)] to monitor the
trend in the parameters. Data logged over the years and trend analysis provides useful
information to initiate appropriate remedial measures to extend life of the equipment.
CPRI has carried out extensive condition assessment studies of power transformers. No
quantitative assessment of life in terms of years is possible with the data generated. However, on
the basis of analysis of the data, appropriate remedial measures such as run, repair or replacement
to extend the remaining life of the equipment is recommended. Following diagnostic tests have
been proved to be effective in condition monitoring power transformers in service.

Tests on Generator Transformer:

1. Insulation Resistance / Polarisation Index Test
2. Tan delta and Capacitance Test on Transformer windings
3. Tan delta and Capacitance Test on HV Bushings
4. Moisture Estimation by Recovery Voltage Measurement (RVM) or Dielectric
Spectroscopy or Polarization/Depolarization Current (PDC) method
5. Winding Resistance Measurement
6. Transformer Turns ratio (TTR) Measurement.
7. Short Circuit Impedance Measurement
8. Magnetising / Excitation Current Measurement at 10kV.
9. Sweep Frequency Response Analysis (SFRA) Test

Tests on Transformer Oil:

 1. Tests on Transformer Oil sample as per IS:1866:2000
2. Dissolved Gas Analysis
3. Furan Analysis

Tests on Paper Sample: (If required)

1. Degree of polymerization test.

5.0 CASE STUDIES

CPRI has been conducting diagnostic testing on power transformers for utilities, process
industries and power plants for condition assessment of their insulation system by conducting
appropriate diagnostic tests. Analysis of the diagnostic test data has helped the plant managers in
life cycle/asset management. Few case studies are discussed as given below.

CASE-I 10.5/220 kV, 120 MVA, 3- Φ, 50 Hz GENERATOR TRANSFORMERS 35 YEARS

OLD (COMMISSIONED IN 1984), DIAGNOSTIC TESTS CONDUCTED IN 2019:

O&M History:
 Date of commissioning 10.04.1984.
 Last Capital maintenance carried out during end of 1995.
 Oil replaced with new oil in April2007.
 The corresponding Generator is under Renovation, Modernization and Uprating (RMU).
The Generator is getting uprated by 16%.
 The Unit is under shutdown since April 2019 for RMU of the Generator Unit.
 The transformer has been subjected to vacuum filtration and hot oil circulation two
months before testing.

Visual Inspection and Examination:

 Oil leakage observed from off circuit tap switch near the handle shaft. Oil leakage was
arrested by wrapping cotton tape with resin around off circuit tap switch handle. Oil was
continuously dripping after the cotton and resin was removed from the shaft for operating
the tap switch for turn ratio/winding resistance measurement.
 There was minor oil leakage near LV side bushings. Few petticoats of two LV bushings
were damaged/chipped off.
 When B phase HV bushing test tap/power factor terminal was opened for conducting tan
delta measurement, continuous oil leakage was observed from the test tap/power factor
terminal.
 When R phase bushing test tap/power factor terminal was opened for conducting tan delta
measurement, carbonizing inside the test tap observed. There was no continuity between
power factor terminal and ground, indicating power factor terminal is not getting
grounded.

HV & LV windings:
Results of the Insulation Resistance / Polarization Index and Tan Delta tests obtained on the
transformer are presented in Table – 1A.
Table – 1A
 Capacitance
Insulation section PI Tan (%)
IR (G) (pF)

HV winding versus LV winding 7.05 1.73 0.221 10869.36

connected to Grounded tank

HV winding versus Ground with -- -- 0.241 3837.50

LV winding Guarded

HV winding versus LV winding 11.1 1.60 0.214 7029.79

(Ungrounded)

LV winding versus HV winding 3.79 2.42 0.277 25945.59

connected to Grounded tank

LV winding versus Ground with -- -- 0.302 18913.17

HV winding Guarded

 The estimated moisture content in the solid insulation is less than 1.75%.

 The measured TTR value lies in the permissible range.

 The measured short circuit impedance lies in the permissible range.

 The measured winding resistance values lie in the permissible range.

 In absence of previous SFRA data comparison has been made between phases. SFRA results
indicate that there is no deformation in the HV and LV windings of the transformer.

 The oil test results are normal indicating healthy condition of the transformer oil. D.G.A
results are normal indicating healthy internal condition of the transformer. Furan analysis
shows no furan content detected indicating healthy of solid insulation.

HV Bushings
Results of the tan delta tests obtained on the transformer HV Bushings are presented in
Table-1B.
 Table–1 B
Tan  @ 10 kV Capacitance @ 10 kV
Phase
(%) (pF)

R 0.249 243.45
Y 0.408 243.73

B 0.304 244.89

 The tan delta values are low indicating low dielectric losses in the bushing insulation.

Recommendations:
From the diagnostic test data, it can be inferred that the condition of the transformer is healthy.

As the Generator Unit is getting uprated 16%, the transformer has to be operated at 150 MVA
continuous overload by running both the coolers simultaneously as per OEM data given in the
name plate. In view of this recommendation of the OEM designer engineers may be sought for
additional cooler and modification in cooling control system.

HV bushings are in healthy condition. In view of the findings of visual inspection and
examination of R and B phase bushings, it is recommended for ensuring grounding of the power
factor terminal of the R phase bushing test tap and arresting leakage in B phase bushing test tap
before putting the transformer in operation. However for high system reliability, it is
recommended for considering R and B phase HV bushings for replacement.

Oil leakage from off circuit tap switch and on LV side of the transformer must be arrested to
avoid moisture ingress into the transformer.

As diagnostic testing is a continuous and condition assessment is based on trend analysis, it

commended to conduct diagnostic testing after three years. Tests on transformer oil as per IS
1866:2017 including DGA and Furan Analysis may conducted after one year.

Risk Assessment:
Without additional cooler reliability and availability of power is questionable. If the hot spot
temperature increases beyond the limiting value, plant managers will be forced to derate the
transformer and operate which defeats whole effort of uprating the Generator Unit.

CASE-II 11/220 kV, 99 MVA, 3- Φ, 50 Hz GENERATOR TRANSFORMERS 40 YEARS

OLD (COMMISSIONED IN 1975) DIAGNOSTIC TESTS CONDUCTED IN 2015:
History:
Transformer was in operation prior to testing. Transformer is running under 16% continuous
overload (Considering Generator rating 103.5 MW & Transformer rating 99MVA).
HV & LV windings:
Results of the Insulation Resistance / Polarization Index and Tan Delta tests obtained on the

transformer are presented in Table –2.

Table – 2
 Tan  Capacitance
Insulation section IR PI (%) (pF)
(M)

HV winding versus LV winding connected to 76.8 1.17 2.527 20258.52

Grounded tank

HV winding versus Ground with LV winding -- -- 2.975 3886.08

Guarded

HV winding versus LV winding 111 1.26 2.424 16370.16

(Ungrounded)

LV winding versus HV winding connected to 53.3 1.21 2.891 30819.93

Grounded tank

LV winding versus Ground with HV winding -- -- 3.419 14447.28

Guarded

 The IR and PI values low.

 The tan delta values obtained on the three insulation sections of the transformer are high
indicating high dielectric losses in the transformer insulation system.

 The estimated moisture content in the insulation system is 4.69%.

 The measured TTR values lie in the permissible range.

 The The measured short circuit impedance lies in the permissible range.

 The measured winding resistance values lie in the permissible range.

 The present context because of absence of baseline SFRA data comparison is made between
the three phases of the generator transformer. SFRA results indicate that there is no
deformation in the HV and LV windings of the transformer.

 The transformer oil test results are furnished in Section IV. The oil test results show higher
concentration (114ppm) of ethylene gas even after recent filtration. The presence of the key
gases methane, ethane and ethylene is an indication of probable overheating in the
insulation. Furan analysis shows higher level (1510ppb) of furan content indicating
deterioration of solid insulation. However, the values are not alarming. The maximum
permissible limit of furan content is 2500 ppb.

Recommendations:
The diagnostic test data indicate that the transformer is not healthy. The dielectric losses and the
moisture content are high in the transformer insulation system. The transformer oil indicates
symptom of overheating. The furan analysis indicates deterioration of the solid insulation.
In view of this, it is recommended to withdraw the transformer from service and subject for
thorough internal inspection. It is also recommended to extract three paper samples from the HV
windings for conducting Degree of Polymerization test which is the confirmatory test to take
decision to replace the transformer or not.

Risk Assessment:
Continuous overloading of the transformer without additional cooler and modification in cooling
control system will lead to accelerated aging of the transformer insulation system and reduces
remaining life of the transformer.

CASE-III 11/33 kV, 25 MVA, 3- Φ, 50 Hz GENERATOR TRANSFORMERS 58 YEARS

OLD (COMMISSIONED IN 1958), DIAGNOSTIC TESTS CONDUCTED IN 2016:
History: The transformer has been overhauled. Winding insulation system and Core have
been cleaned by oil jet. Gaskets have been changed and transformer was dried out by
applying vacuum and nitrogen cycle during 31.05.2016 to 09.06.2016. New hot circulated
transformer oil was filled under vacuum on 09.06.2016 before CPRI testing. DVC has
provided CPRI Furan Analysis Report conducted during February 2016. The Furan Analysis
shows higher level of furan content as high as 2380ppb.

HV & LV windings:

Results of the Dielectric Spectroscopy and DP tests obtained on the transformer are presented
in Table – 3.

Table – 3
10 kV Equivalent Power Frequency Value
Insulation Section DP Value
Tan  (%) Capacitance (pF)

HV winding versus LV HV Winding -

winding connected to 2.586 3874
562.5
Grounded tank

HV winding versus LV 3.836 8562 --

winding (Ungrounded)

LV winding versus HV LV Winding-

winding connected to 3.029 10980
562.3
Grounded tank

 The estimated moisture content in the HV winding insulation, barrier insulation & LV
winding insulation sections are 4.2 %, 4.1% & 4.1% respectively. These moisture levels are
slightly higher than the maximum permissible level of 4.0% for an in-service 33kV class
transformer.
 The DP represents the average number of glucose chains in the cellulosic paper. As the paper
undergoes ageing, these polymer chains breakdown into smaller units. As a result of this, the
DP of the paper decreases [Normally a fresh (new) unaged Kraft paper sample exhibits a DP
around 1100-1200]. After factory drying tests, this value tends to drop to 800-1000 and then
decreases exponentially with ageing. The aged paper with a DP of 150-200 exhibits poor
mechanical strength and makes the transformer windings more susceptible to failure during
system extreme conditions. The DP value of 150-200 can be used to indicate end of the
useful service life of the cellulose.

In the present context the DP values obtained on the subject transformer (HV Winding-
562.5,LV Winding-562.3) indicate moderate deterioration of the paper insulation. The
mechanical strength of the paper insulation is reasonably good.

Recommendation:
 From the diagnostic test data it can be inferred that the transformer is generally healthy. As
condition monitoring is a continuous process and condition monitoring is based on trend analysis,
it is recommended to conduct diagnostic testing after one year.

Risk Involved:
Normally Generator Transformer rarely sees short circuit faults unlike power transformers
operating in substations. Though the mechanical strength of the paper insulation is reasonably
good, an unlikely event of heavy short circuit fault can lead to failure.

CASE-IV 11/220 kV, 130 MVA, 3- Φ, 50 Hz GENERATOR TRANSFORMERS 34 YEARS

OLD (COMMISSIONED IN 1985) DIAGNOSTIC TESTS CONDUCTED IN 2019:
History:
 Generator Transformer had flash over at DETC Taps in June-2015.
 Oil had been replaced with new Oil (27,000 Ltrs) after Vacuum for 24 hrs.
 All Gaskets had been replaced.
 DETC had been permanently fixed to Tap-4.
 KPCL, Varahi has provided DP Test Results conducted during November 2015. The
mean value of Degree of Polymerization is 404.0.
 Previous oil results (2017 to 2019) have been reviewed. Oil tests results indicate low
dielectric strength (BDV) and high moisture content.

a) HV & LV Windings:
Results of the Insulation Resistance / Polarization Index and Tan Delta tests obtained on the
transformer are presented in Table – 4A.

Table – 4A

Tan  Capacitan
Insulation section PI (%) ce (pF)
IR (G)

HV winding versus LV winding connected to 2.39 1.36 1.287 10951.35

Grounded tank

HV winding versus Ground with LV winding -- -- 0.914 3975.67

Guarded

HV winding versus LV winding (Ungrounded) 2.11 1.48 1.508 6971.74

LV winding versus HV winding connected to 0.0495 1.05 1.426 24424.50

Grounded tank

LV winding versus Ground with HV winding -- -- 1.426 17439.64

Guarded

 The estimated moisture content is 1.9%. Oil conductivity 29 ps/m.

 The measured TTR value lies in the permissible range.

 The measured short circuit impedance value lies in the permissible range.
 The measured winding resistance values lie in the permissible range.

 The present SFRA data have been compared with the previous SFRA data. The comparison
graphs indicate that there is no deformation / displacement of the HV & LV windings.

b) HV BUSHINGS:
Results of the tan delta tests obtained on the transformer HV Bushings are presented in
Table-4B.
Table–4 B
Insulation Tan 
Phase Resistance 60sec Capacitance (pF)
(G) (%)

R 6.17 0.407 226.63

Y 6.88 0.416 227.26
B 6.54 0.384 226.56
The tan delta values are low indicating low dielectric losses in the bushing insulation.

Recommendations:
From the diagnostic test data, it can be inferred that the transformer is generally healthy. In view
of low BDV (39.7 kV-July 2019) and high water content (27 ppm-July 2019) it is recommended
for vacuum filtration and hot oil circulation.

As diagnostic testing is a continuous process and condition assessment is based on trend analysis,
it is recommended to repeat the tests after Two years.
Risk Assessment:
If vacuum filtration and hot oil circulation if not carried during no load/shutdown condition
moisture will migrate from oil to paper insulation and will accelerate the aging of paper
insulation.

CASE-V 132/6.9 kV, 10 MVA, 3- Φ, 50 Hz STATION TRANSFORMER 34 YEARS OLD

(COMMISSIONED IN 1965), DIAGNOSTIC TESTS CONDUCTED IN 2018:

HISTORY:

On dated: 29.06.2017 at 00:00 hrs, Station Transformer- 10T tripped on differential with heavy
sound, flash over, vibration and 6.6 kV voltage dip. When reached at site, it was found that its
‘B’ phase jumper with clamp broke from bushing and oil was leaking from top of the transformer
main tank. It was also observed that its turret CTs displaced from its position.

After enquiring it was known that the Ash Pump (6.6 kV Motor) of ‘B’ Power House, Which was
completely merged in water, was run to evacuate the water from its pit causing three phases to
ground fault. Being a high potential fault, the Transformer felt a heavy vibration displacing its
turret CTs, Radiators etc causing the breaking of B phase bushing Jumper resulting tripping of
Station Transformer- 10T on differential protection.

After Repair/ Maintenance, this Transformer was energized on 07.08.2017. The Transformer was
in service during 07.08.2017 to 23.05.2018.
On dated: 23.05.2018 at 10:00 hrs, Station Transformer-10T tripped on Differential Protection
during changing Tap position of OLTC. After enquiring, The Main/Common shaft of Tap
Changing mechanism found broken at R phase. This caused uneven distribution of tap positions
between all three phases and hence tripped the Transformer. Also, the reactance coil of the
Transformer found damaged during inspection. R & B phase reactance coils are damaged
whereas R phase reactance coil is severely damaged.

(a) HV & LV Windings :

Results of IR and Tan delta tests obtained on the transformer are presented in Table–5A.

Table-5A

Tan  Capacitance
IR (M) PI (%) (pF)
Insulation section
HV Winding versus LV Winding
1.06 17.626 10500.26
shorted to grounded tank 4.06

HV winding versus LV winding Guarded -- 30.206 3311.78

--
HV Winding versus LV Winding
4.23 1.03 11.234 7190.79
Ungrounded.
LV Winding versus HV Winding
12.7 1.04 12.638 15182.98
shorted to grounded tank.

LV winding versus HV winding Guarded -- -- 14.049 7992.13

 The estimated moisture content is 5.4%, 5.5% and 5.4% for HV, Barrier and LV insulation
section respectively.
 The measured TTR values obtained on the Y and B phases are marginally exceeding the
permissible range and TTR deviation on R phase is very high.
 The measured values are normal and lie in the permissible range as per IEEE-62.

 The measured short circuit impedance value is normal and the value is comparable to the name
plate value within ± 3%.
 The magnetic balance test results are abnormal indicating unhealthy condition of the core
(magnetic circuit) of the transformer.
The exciting (magnetizing) current test is useful in locating problems such as defects in the
magnetic core structure, shifting of the windings, failures in the turn to turn insulation and
problems in the tap changers. These defects result in a change in the effective reluctance of
magnetic circuit which affects the current required to produce the expected flux through the
core. The magnetizing current obtained between R phase and HV Neutral is abnormally high
indicating defect in magnetic core structure.

 Phase to phase magnitude response comparison indicate that there is core deformation in the
magnetic structure and axial winding movement and deformation in the R phase windings.
Minor movement and deformation in the Y and B phase windings suspected.
 Transformer oil results indicate very low BDV (23 kV), high water content (63 ppm) and
very low specific resistance. Furan analysis shows very high level (1570 ppb) of furan content
indicating deterioration of solid insulation. The maximum permissible limit of furan content is
2500 ppb.
 DP values obtained on the subject transformer (3 Samples of Insulating Paper from HV
Winding-211.32, 170.7 and 165.7) indicate end of life of the paper insulation and it has
reached the maximum of its service life.

Recommendations:

From the analysis of the diagnostic test data, it can be inferred that insulation condition of the
transformer is not healthy.

Transformer core is not in healthy condition. In view of core deformation and axial winding
movement and deformation, it is recommended for removal and inspection of the core and
winding assembly to confirm the finding of diagnostic testing.

In view of the Degree of Polymerization Test results, It is recommended for replacement.

6.0 CONCLUSIONS
Based on field experiences the following conclusions are drawn.

Though there are various diagnostic tests for diagnosis of power transformer, one single
diagnostic test does not give complete information about the condition of the transformer.

For meaningful analysis a comprehensive diagnostic program to be evolved and appropriate

diagnostic tests to be conducted to assess the condition of the transformers.

Diagnostic testing and condition monitoring is a continuous process and based on trend analysis.
The diagnostic tests need to be conducted periodically to monitor the trend in the parameters.

Diagnostic test results can be used for asset management decision support for different
maintenance activities, repair, replacement and condition assessment steps.
ACKNOWLEDGEMENTS
The authors are thankful to Management of Central Power Research Institute, Bangalore for
giving permission to present the paper.

REFERENCES

[1] Sivaji Chakravorti, Debangshu Dey, Biswendu Chatterjee, A Book on “Recent Trends in the
Condition Monitoring of Transformers-Theory, Implementation and Analysis”

[2] IEEE Std. 62-1995 IEEE Guide for Diagnostic Field Testing of Electric Power Apparatus-
Part1 Oil Filled Transformers, Regulators and Reactors.
SOLID STATE TRANSFORMER TOPOLOGIES –
A REVIEW
N Maheswara Rao, B R Vasudevamurthy, Swaraj Kumar Das, R A Deshpande
Short-circuit Lab, CPRI-Bengaluru

ABSTRACT
Conversely, turning it off can block voltage of
In order to cater future renewable system either polarity.
architectures flexible and desired technologies
development are required for reliable
operation of electricity grids. As a part of this,
solid state transformer (SST) have been
identified as one of the potential equipment.
SST is a static equipment that transfers
electrical energy from one circuit to another
through power electronic conversion
technologies. With the latest breakthrough and
maturity in power electronic modules
development, creates interest to design SSTs.
Massachusetts Institute of Technology has
listed SSTs as one of the ten breakthrough Fig. 1. AC/ac Buck Converter
technologies that will influence the future of
electricity grids. Hence SST gained
importance and continual growth is happening In Fig. 1, the two devices tied in series indicate
in this area. Since input and output voltages an arrangement of some number of
are ac, the switches in an SST are required to semiconductors that would be needed to
block voltage in both polarities as well as achieve the desired voltage ratings. (The
switches need to conduct current in both number of devices required would be a
directions to realize bi-directional power flow. function of voltage level and the type of
Unlike the conventional transformer, the SST semiconductor). Bidirectional switches are
can be used not only to transform ac but also alternately turned on at a high frequency
to transform dc. Further SST is a powerful
relatively to the fundamental component.
power-electronics-based apparatus covering
the functions of the transformer including Under these conditions, it can be shown that
flexible ac transmission system. Infact, it
should be regarded as a new technology
involving flexible transmission for distribution
systems. This paper aimed to review various
topologies proposed in the latest literature where d is the duty cycle (the time the upper
survey. switch is on relative to the switching period)
Keywords: solid state transformer (SST), Electricity Grid, and designates the phasor representation
Flexible AC transmission, Topology, Flexible of the fundamental component.
Transmission
Although this is the most straightforward
INTRODUCTION approach to ac–ac power conversion, it is
problematic in several respects. First, each
The earliest AC–AC power electronic switch must be able to block full primary
transformer was proposed by W. McMurray in voltage and also be capable of conducting full
1970, followed by the AC/AC buck converter secondary current. The fact that both the
developed by the United States Navy in 1980. blocking voltage and peak current are large
However, due to slow switching and low power
rating of semiconductors, application of SSTs implies that this design would be very costly.
in power systems could not be implemented This disadvantage could be partially mitigated
during this era. by performing the power conversion in several
stages, although such a tactic complicates the
The ac/ac buck converter is perhaps the most design. Other drawbacks of this approach
direct approach to single phase ac power include the use of series tied devices, which
conversion. In this arrangement, switches are are often difficult to control, the lack of
bidirectional. Turning a switch on permits it to magnetic isolation, inability to correct load
conduct current in either direction. power factor, and inability to prevent load
harmonics from propagating into the primary-
voltage system. In view of these
disadvantages, the ac/ac buck converter is not
practical for distribution transformer
applications.

SINGLE STAGE CONVERSION

Fig. 4. Dual Active Bridge (DAB) configuration

The relatively high switching frequency and

Fig. 2. Block Diagram of Single stage SST
As a solid state transformer, galvanic isolation
is required at some stage of power conversion.
A high frequency transformer is preferred to
reduce the weight and volume of the magnetic
core. For the direct ac-ac power conversion
described here, there is one switching network
to convert the source ac voltage to an
intermediate high frequency ac voltage and
another switching network to convert the high
frequency voltage back to grid frequency.
the low number of switching devices and
absence of electrolytic capacitors make it
suitable for size-critical applications. An ac-ac
DAB converter shown in Fig. 4 is controlled by
phase-shift modulation (PSM). The amount of
power transferred is controlled by the phase
shift angle between two bridges φ and the
input/output voltages Vi and Vo, as described
in (1),
where Xpu is the lumped transformer leakage
reactance in per unit and φ is in radians.
TWO STAGE CONVERSION

For topology shown in Fig. 3, the incoming ac

waveform is modulated by a power-electronic
converter to a high-frequency square wave
and passed through a small high-frequency
transformer. Another converter, synchronous
with the high-voltage side but at a lower
voltage, demodulates it.
Fig. 5. Block Diagram of Two stage SST
Fig. 6 shows a detailed view of two stage SST.
The secondary side converter consists of
single-phase full bridge active rectifier and
conventional voltage source inverter structure.
The step-upped secondary side HFAC signal
is rectified by active rectifier and establishes a
desired magnitude of fluctuated dc link voltage
Vdc by a proper selection of transformer turn
ratio n.

Fig. 3. Single stage SST Topology

Many improvements happened in this topology

as this topology is of single stage and mainly
provides galvanic isolation. Futures like
regulation of power flow, control of voltage
magnitude, harmonic content & power factor
and efficiency by implementing soft switching.
while the soft switching is achieved only for a Fig. 6. Two stage SST Topology
limited range around the rated load. Thus, this
design would be most suitable for constant
THREE STAGE CONVERSION
load applications.
A three-stage design consists of high-voltage
stage, isolation stage, and low-voltage stage.
Fig. 7 shows a block diagram representation of
three stage conversion.
Fig. 7. Block Diagram of Three stage SST
It contains input full-bridge active rectifier
(FBAR), full-bridge DC-link inverter
(FBDCI), high-frequency isolating
transformer (HFIT), cyclo-converter and
output filter as depicted in Fig. 8.
Fig. 8. Detailed multi stage SST Topology-1
Fig. 9. Detailed multi stage SST Topology-2
The three-stage conversion topology shown in
Fig. 9. The first stage utilizes an AC/DC
rectifier aimed at regulating the voltage across
the HV DC link, shaping the input current and
achieving bi-directional power flow and
reactive power compensation as well as
harmonic elimination. While the second stage
includes a high frequency, the DAB is required
to regulate the active power flow, provide
galvanic isolation, and control the voltage at
the LV DC bus. During this stage, the voltage
is stepped up by a high-frequency transformer
and transferred to LVDC link. Finally, the
voltage is shaped in to 50 Hz waveform on the
load side.
MULTI LEVEL TOPOLOGIES
High-voltage part is composed of several
modules (half-bridge ac/dc converters)
connected in series through high-frequency
transformers to cope with high input voltage,
while the low-voltage part is composed of
several modules connected in parallel
(bidirectional half-bridge dc/dc converter and
dc/ac PWM inverter) to cater high current
requirement.
The high-voltage stage is realized using an
interfacing inductor and several cascaded
voltage source converters (VSCs) to meet the
high voltage level requirement. The isolation
stage consists of front-end VSCs, medium-
frequency isolation transformers(MFITs), and
back-end VSCs.The front end VSCs are back-
to-back connected with the cascaded VSCs in
the high-voltage stage. The back-end VSCs
are back-to-back connected with the
converters in the low-voltage stage. The low-
voltage stage is achieved by several
converters with parallel connection to meet the
high current requirement. So, the single phase
subunit is a typical series-input-parallel-output
type of connection.
CONCLUSIONS
A review of various topologies proposed in the
existing literature survey is presented in this
paper. Unlike traditional magnetic core
transformers, SST is flexible enough to be of
modular construction, enabling bi-directional
power flow which can be employed for AC and
DC grids. Moreover, SSTs can control the
voltage level and modulate both active and
reactive power at the point of common
coupling without the need of external flexible
ac transmission system as per the current
practice in conventional electricity grids. The
rapid advancement in power semiconductors
switching speed and power handling capacity
will soon allow for the commercialization of
grid-rated SSTs.
Electronics Specialists Conference, Jeju,
Korea, 18–22 June 2006; pp. 1–6.
[8] Casarin, J.; Ladoux, P.; Martin, J.;
Chauchat, B. AC/DC converter with medium
frequency link for railway traction application.
Evaluation of semiconductor losses and
operating limits. In Proceedings of the
SPEEDAM 2010, Pisa, Italy, 14–16 June
2010; pp. 1706–1711.
[9] Sabahi, M.; Hosseini, S.H.; Sharifian, M.B.;
Goharrizi, A.Y.; Gharehpetian, G.B. Zero-
voltage switching bi-directional power
electronic transformer. IET Power Electron.
2010, 3, 818–828.
Authors Biodata

N Maheswara Rao
REFERENCES Joined CPRI Short-Circuit
Laboratory in 2009. He is
[1] Ronan, E.R.; Sudhoff, S.D.; Glover, S.F.; a gold medalist in Power
Galloway, D.L. A power electronic-based electronics from National
distribution transformer. IEEE Trans. Power Institute of Technology
Deliv. 2002, 17, 537–543. [NITT], Tiruchirappalli. He
is having experience of
[2] Hengsi, Q.; Kimball, J.W. AC-AC dual more than 10 years in the
active bridge converter for solid state field of short-circuit testing and evaluation of
transformer. In Proceedings of the 2009 IEEE various electrical equipments as per National
Energy Conversion Congress and Exposition, and International standards. He is actively
San Jose, CA, USA, 20–24 September 2009; involving in various R&D projects of CPRI SC
pp. 3039–3044. Lab. His areas of interest are Short-circuit
testing, Design and developments of power
[3] Qin, H.; Kimball, J.W. Solid-State electronic modules for short-circuit
Transformer Architecture Using AC–AC Dual- applications and Short-circuit studies of
Active-Bridge Converter. IEEE Trans. Ind. distributed generation systems and algorithm
Electron. 2013, 60, 3720–3730. developments.

[4] Chen, H.; Prasai, A.; Divan, D. Dyna-C: A B Vasudeva Murthy is

Minimal Topology for Bidirectional Solid-State presently working as Test
Transformers. IEEE Trans. Power Electron. In-charge for the short-
2017, 32, 995–1005. circuit lab and has more
than 20 years of
[5] Solid State Transformers Topologies, experience in testing &
Controllers, and Applications: State-of-the-Art performance evaluation of
Literature Review by Ahmed Abu-Siada, Jad LV switchgear &
Budiri and Ahmed F. Abdou controlgear equipments. Also specialized in
creation of new test facilities, automation of
[6] Cha, H.J.; Enjeti, P.N. A three-phase tests & measuring instruments. ASTA
AC/AC high-frequency link matrix converter for Observer for witnessing of tests for ASTA
VSCF applications. In Proceedings of the 2003 Certification (Intertek) for LV switchgear &
IEEE 34th Annual Power Electronics Specialist controlgear equipments. Obtained Bachelor of
Conference, Acapulco, Mexico, 15–19 June Engineering from Bangalore University.
2003; Volume 1974, pp. 1971–1976.

[7]. Jin, A.; Li, H.; Li, S.ANewMatrixTypeThree-

PhaseFour-WirePowerElectronicTransformer.
InProceedings of the 2006 37th IEEE Power
 Mr. Swaraj Kumar Das
was graduated in
Electronics &
Communication
Engineering from N.I.T
(R.E.C) Durgapur, West
Bengal in 1988. He
worked as Engineer in R &
D Centre of M/s.
Hindustan Cables Ltd.,
Hyderabad during 1989 – 1991 then joined
CPRI, Short Circuit Laboratory, Bangalore at
the end of 1991. Currently holding the post of
Joint Director, heading Short Circuit
Laboratory in CPRI with more than 20 years of
wide experience in the field of short circuit
testing and performance evaluation of LV &
HV switchgear and control gear equipments.
He is a member of Bureau of Indian Standards
– ET 34 & ET 07 committee for CT & PT and
Low voltage switchgear & control gear
assemblies respectively. He has publications
in the area of Distribution Transformers, CT’s
& LV switch gear and controlgear assemblies

R A Deshpande obtained
his B.E. in Electronics and
Power (1981) and M.Tech
in Integrated Power
System (1983) from
Visvesvaraya Regional
College of Engineering,
Nagpur. He joined CPRI
in 1984 and continues to
work there. Currently he is Additional Director
in charge of Short Circuit Division. His area of
interest are Power System Planning and
operational Studies, Earthing Systems design
and evaluation, Design and operation of
Distribution Systems, failure investigation of
Power System Components, Simulation
Studies, Inspection and Evaluation of
Distribution Projects, Solar Power generation
and integration with grid, . He is a senior
member of IEEE.
 Partial Discharge measurement and Dielectric testing of
High Voltage Transformer Bushings

P Rajamani, K A Aravind, K Sandhya, K Rajeshwara Rao, K Urukundu,

A Ramulu, B Krishna and Pradeep Nirgude
Central Power Research Institute, Ultra High Voltage Research Laboratory, Hyderabad 500 098
rajamani@cpri.in

Abstract transformer bushing, is to be ensured before putting

Ultra High Voltage Research Laboratory (UHVRL)
of Central Power Research Institute (CPRI),
Hyderabad is having state-of-the-art facility to
perform electrical type and routine test on bushings
as per National and International Standards. In this
manuscript, some of the experiences and
precautionary measures taken while performing
partial discharge quantity measurement test,
dielectric tests, electromagnetic compatibility
emission test and other test on 420 kV class oil
impregnated paper and resin impregnated paper
transformer bushings as per IEC 60137:2017 are
elaborated.
1. Introduction
Transformer is one among the vital and expensive
equipment in electric supply system. Failure of in-
service transformer may cause expensive outages
and unscheduled costly repairs. Hence, to provide
uninterrupted power supply to the customers,
reliable operation of all essential parts of
transformer is mandatory. Bushing is one of the
essential part of transformer. Bushing-failure, is
identified as one of the major causes of transformer
breakdown in different reported studies [1].
Moreover, failure analysis of damaged transformer
bushing revealed that 44% of forced outages of
power transformer is related to winding and
bushings [1]. As per CIGRE survey, 10% of
transformer failures were caused by bushing
damage, followed by catastrophic consequence [2].
Though, identification of initial breakdown process
of bushing structure is really challenging, if
unattended, it may lead to sudden failure of bushing,
sometimes, damage the whole transformer
completely. Therefore, in addition to regular
diagnosis, integrity of such vital equipment,
into service. Eventually, integrity of new or in
service bushing is confirmed by testing in
accordance with relevant international and national
standards. In this manuscript, along with facilities
available in CPRI UHVRL to perform testing as per
IEC 60137:2017[3], in particular, experiences on
partial discharge measurement and other type
testing of transformer bushing, viz. Oil Impregnated
Paper Bushing (OIP) and Resin Impregnated Paper
Bushing (RIP) are discussed.
2. Facilities available in UHVRL
UHV indoor double shielded laboratory is equipped
with 1200 kV (2X600 kV PD free testing
transformers in cascaded mode), 2 A, 2400 kVA
cascaded AC generator to perform partial discharge
(PD) measurement test and other power frequency
voltage tests. The indoor laboratory is specially
designed and constructed to perform partial
discharge test on various electrical equipment, in
particular, bushings and instrument transformer.
The background ambient partial discharge observed
at UHV equipment test voltage is around 3 pC. The
sensitivity of the PD measuring system is in
compliance with IEC 60270:2015[4]. Photograph of
1200 kV HVAC test system along with calibrated
divider is shown in Fig.1. To cater the need of
impulse testing, UHVRL is also having 5 MV, 500
kJ, 25 stage outdoor impulse voltage test facility to
perform lightning, switching and chopped impulse
voltage test on various electrical equipment.
Photograph of impulse test facility is shown in
Fig.2. In addition to indoor facility, the lab having
1600 kV (2 X 800 kV testing transformers in
cascaded), 6 A and 9600 kVA outdoor HVAC test
facility to cater the source requirement of artificial
pollution test on various electrical equipment of
UHV rating. Photograph of outdoor HVAC test
facility along with artificial pollution test chamber
of 24 m diameter and 27 m height and 850 kV wall
mountable bushing is shown in Fig.3. This outdoor
HVAC source is in compliance with requirement of
test system according to IEC 60507:2013[5].
UHVRL also have facilities to perform accuracy
test, electro-magnetic compatibility test,
capacitance and dielectric dissipation factor
measurement test, temperature rise test, transient
response test, ferro-resonance test and short circuit
withstand capability test on instrument transformer.
Fig.1. 1200 kV (2X600 kV), 2 A and 2400 kVA Indoor HV
AC test system with measuring divider
Fig.2. Outdoor 25 stage 5 MV, 500 kJ impulse generator
Fig.3. 1600 kV (2X800 kV), 6 A and 9600 kVA outdoor HV
AC test facility along with artificial pollution test chamber
3. Capacitance (C) and Dielectric Dissipation
factor (tan δ) Measurement test
To assess the quality of insulations and check the
uniformity of production of capacitance graded
bushings this test is performed. Dissipation factor is
an indication of drying and impregnation process of
capacitors. During the test power frequency AC
voltage is applied to high voltage terminal of
bushing and the capacitance is measured at test tap
is connected to capacitance and tan δ measuring
bridge. Flange of bushing is connected to earth
along with body turret tank. As dielectric dissipation
factor is vital assessment parameter to judge the
quality of insulation, the HV connection lead is
maintained at-least half the size of bushing head to
avoid increase in dielectric dissipation factor value
due to ionic current emission from HV connecting
leads. During type tests on capacitance graded
bushing, this test is particularly carried out before
and after series of types test to check whether
damage has occured on any insulation. To ensure no
puncture has occurred during dielectric tests,
measured value of capacitance before and after tests
are compared, if no notable difference in measured
capacitance values, all the layers of bushing are
intact. On the other hand, if there is change in
capacitance value compared to before measured
value, the raise in capacitance value is attributable
to the capacitance of single layer, or the change in
the measured capacitance shall be less than
capacitance of one layer of capacitance graded
bushing.
As both capacitance and dissipation factor are
subjected to temperature and frequency, value of
ambient temperate and frequency at the time of
testing is also reported. Harmonics present in the
supply may affect the measurement, hence, a
 Schering bridge method, which excludes error due Table 2. Capacitance and Dielectric Dissipation Factor - after
to harmonics are used for measurement. An Dielectric Tests of 420 kV transformer bushing
automatic high precision C, L and R measuring Sl. Test OIP Bushing RIP Bushing
bridge is used for measurement along with 1200 kV No. Voltage C (pF) tan δ C (pF) tan δ
(kV)
SF6 gas filled 20 pF standard capacitor. The 01 10 469.00 0.00374 686.70 0.00305
photograph of standard capacitor and test object is 02 255 469.04 0.00396 686.86 0.00319
shown in Fig.4. According to standard the C and tan 03 420 469.98 0.00424 686.99 0.00352

δ value is measured at 1.05 𝑈𝑚 ⁄√3 and at 𝑈𝑚 for Ambient Td Tw P Td Tw P (kPa)

Conditions (ºC) (ºC) (kPa) (ºC) (ºC)
bushings having highest voltage of equipment, 𝑈𝑚 During Test 29 24 96.53 35 26 96.53
≥ 52 kV. Measured value of C and tan δ before and
after power frequency test is given in Tables 1 and
4. Partial Discharge (PD) Measurement Test
2 for 420 kV OIP and RIP transformer bushings. It
is observed that there is no noticeable difference in
capacitance value before and after dielectric tests. Partial discharge quantity measurement is
Moreover, the value and raise in dissipation factor applicable to all types of bushings, except, gas
measured at 1.05 𝑈𝑚 ⁄√3 and 𝑈𝑚 , i.e., 255 kV insulated and cast or moulded resin insulated
and 420 kV, for this case, is also in compliance with bushings. Partial discharge is evidence of degrading
IEC 60137:2018, i.e., less than 0.007 and 0.001, insulation system, which could or may lead to costly
respectively. The values at lower voltages, e.g., 10 repairs. The presence and magnitude of PD are
kV is also measured and recorded for later routine important criteria to measure for early detection of
maintenance and diagnosis purposes insulation and assessment of manufactured or
repaired product quantity. Schematic representation
of and picture of PD measuring system of UHVRL
is shown in Figs.5 and 6. The measuring system is
in accordance with IEC 60270:2015.

As recommended in standard, the sensitivity of PD

detection of UHVRL measuring circuit is 5 pC. In
combination with coupling device, the frequencies
at which the transfer impedance (ratio of output
voltage amplitude to a constant input current
amplitude) has fallen by 6 dB from band pass values
is 93 kHz and 413 kHz, the lower and upper cut of
frequencies respectively. In suitably designed and
constructed double shielded indoor laboratory along
with HVAC test system, the background noise,
signal detected during PD tests, which are not
originated in the test object, has observed around
3.4 pC at UHV power equipment test voltage, i.e.,
Fig.4. 1200 kV SF6 gas filled standard capacitor at 850 kV. Hence, the sensitivity of PD measuring
system is in compliance with IEC 60270:2015. PD
Table 1. Capacitance and Dielectric Dissipation Factor - Before
measurement is made after dry power frequency
Dielectric Tests of 420 kV transformer bushing
Sl. Test OIP Bushing RIP Bushing voltage withstand test at stipulated value in IEC
No. Voltage C (pF) tan δ C (pF) tan δ 60137:2017 during decrease of the voltage from that
(kV)
01 10 470.12 0.00344 686.10 0.00300 level at
02 255 470.21 0.00378 686.42 0.00306 𝑈𝑚 , 1.5 𝑈𝑚 ⁄√3 , 1.1 𝑈𝑚 ⁄√3 𝑎𝑛𝑑 1.05 𝑈𝑚 ⁄√3 ,
03 420 470.62 0.00402 686.73 0.00341
i.e., at 750 kV is the pre-stress voltage followed by
Ambient Td Tw P (kPa) Td Tw P (kPa) PD measured at 420 kV, 363.7 kV, 266.7 kV and
Conditions (ºC) (ºC) (ºC) (ºC)
During Test 28 23 96.12 34 27 96.13 254.6 kV for 420 kV rated transformer bushings.
Measured value of PD of OIP and RIP transformer
bushing is given in Tables 3 and 4 and are in
compliance with requirements of IEC
60137:2017for OIP and RIP bushings. In addition to
this, PD quantity measured after lightning and
power frequency voltages test as given in Table 6.
After all dielectric tests, the measured PD value is
less than 10 pC and is in compliance with IEC
60137:2017. As a substitute for the measurement of
PD quantity, the radio interference voltage (RIV), is
measured by RIV meter by calibrating the meter in
compliance with relevant standard.

Table 3. PD quantity measured before after dielectric test of 420 kV OIP

transformer bushing

Sl. Test Before After Max. permissible Fig.6. PD / RIV measuring system along with control panel
No. Voltage Dielectric Dielectric limit as per IEC
(kV) test test 60137:2017
(pC) (pC) (pC) 5. Electromagnetic Compatibility Test
1 420 4.6 4.8 10
2 363.7 4.2 4.4 10
3 266.7 3.9 4.1 5 This radio interference emission test is applicable to
4 254.6 3.8 3.9 5
all indoor and outdoor bushings having highest
Table 4. PD quantity measured before after dielectric test of 420 kV RIP system voltage ≥ 123 kV. During this test, the
transformer bushing
bushing is erected as in service conditions with
Sl. Test Before After Max. permissible flange and other normally earthed parts connected
No. Voltage Dielectric Dielectric limit as per IEC
(kV) test test 60137:2017 to earth. The bushing is in dry and clean condition
(pC) (pC) (pC) and also approximately maintained at the same
1 420 3.4 4.6 10 temperature as in test room, moreover, nearby
2 363.7 3.2 4.2 10
3 266.7 3.1 4.0 5 earthed object, which may influence the
4 254.6 3.1 4.0 5 measurement, are removed. Ability of bushings to
Table 6. PD quantity measured before and after lightning and power frequency function satisfactorily in an electromagnetic
test RIP Bushing environment without introducing intolerable
Test Before After After Max.
electromagnetic disturbance to other equipment in
Voltage Lightning Lightning power permissible the system is assessed during this test. Radio
(kV) Impulse impulse frequency limit as per
Voltage test voltage test withstand IEC
interference voltage (RIV) level is measured at test
test 60137:2017 object at 1.1 𝑈𝑚 ⁄√3, shall be less than 2500 µV for
(pC) (pC) (pC) (pC)
420 3.4 4.4 4.6 10 compliance to the relevant standard. The measuring
circuit of UHVRL, Hyderabad tuned to measuring
frequency of 1 MHz, measured across the measuring
impedance of 300 Ω and is in compliance with
CISPR 18-2:2017 [6]. During testing test voltage of
1.1 𝑈𝑚 ⁄√3 is applied and maintained for 5 minutes.
The test voltage then decreased to 0.3 𝑈𝑚 ⁄√3,
raised again in steps to 1.1 𝑈𝑚 ⁄√3 and finally
decreased to 0.3 𝑈𝑚 ⁄√3 in steps. The amplitude of
voltage steps is maintained as 0.1 𝑈𝑚 ⁄√3. For every
steps the radio interference level emitted from the
Fig.5. Schematic of Laboratory PD measurement arrangement
test object is recorded, the value observed in last run
is plotted against voltage and is referred as RIV
characteristics. The RIV level measured across
300 Ω and the RIV characteristics of 420 kV OIP
transformer bushing is given in Table 7 and Fig. 7, 6. Dielectric Testes
respectively. Though it is not supported by standard
care is taken to perform the test below 80% Dielectric tests are carried out on complete bushings
humidity level. Further, during RIV measurement with their necessary fittings and all other accessories
the high voltage connection is extended in line with fitted when in use. During all tests, the temperature
bushing axis at least for 1.5 m, which is around 0.2 of ambient air and immersing medium shall be
times of arcing length of the bushing. Arrangement between 10 °C and 40 °C. Suitable correction is
of OIP bushing for radio interference emission test applied to the test voltage according to IEC 60060-
is shown in Fig.8 1:2010[7] and IEC 60137:2017 for ambient
conditions deviated from standard atmospheric
Table 7. Measured RIV level of 420 kV OIP transformer Bushing
Test Voltage conditions. Test taps is maintained at earth potential
RIV level across 300 Ω (µV)
(kV) during dielectric tests. Sufficient clearance from
266.7 478
242.5 421
surrounding earthed objects is maintained to avoid
218.5 364 direct flashover between them either in ambient air
194.0 316 or in immersion medium. IEC 60137 has not
169.75 273 specified the order of the tests performed, except the
145.5 204
121.3 134
impulse voltage test shall be made before dry power
97.0 80 frequency voltage withstand test. Withstand
72.8 42 capability of bushing insulation against various
internal and external over voltage stress are verified
during dielectric tests. Series of tests, comprising of
RIV Characteristics dry power frequency voltage withstand test, long
600 duration power frequency voltage withstand test,
500 dry lightning impulse voltage withstand test and
Measured RIV (µV)

400 wet or dry switching impulse voltage withstand test

300 are referred as dielectric tests. Before and after this
200 series of type tests, measurement of C and tan δ,
100 partial discharge quantity is measured to check the
0 integrity of capacitor graded insulation.
300 200 100 0
6.1. Dry Power frequency voltage withstand test
Test Voltage (kVrms)
Fig.7. RIV characteristics of 420 kV OIP transformer bushing
This test is applicable to all bushing types which are
not subjected to routine tests. Power frequency
voltage of magnitude as in IEC 60137:2017 is
applied to the high voltage terminal of bushings and
maintained for 60 sec. If no puncture or flashover
observed, then the bushing have considered as pass.
For 420 kV capacitor graded transformer bushing,
this test performed at 750 kV (test voltage corrected
to ambient atmospheric conditions as per IEC
60060-1 is applied during the test) test voltage and
no flashover is observed. The value of capacitance
is measured after this test and compared with
previously measured value and given in Tables 1
and 2. No major deviations observed, except minor
changes due to change in ambient atmospheric
Fig.8. 420 kV OIP transformer busing arrangement for Radio interference
conditions.
emission test
6.2. Long duration power frequency voltage 6.3. Lightning Impulse Voltage (LIV) Withstand
withstand test (ACLD) Test (type test)

ACLD test is applicable to all transformer bushing
having highest system voltage ≥ 170 kV. UHVRL
has performed ACLD test of 420 kV OIP
transformer bushing. The power frequency test
voltage (without atmospheric correction) is applied
to HV terminal of bushing in the as shown in Fig.9
and partial discharge quantity is measured after
every application. Measured value of partial
discharge of 420 kV OIP transformer bushing for
applied voltage as per Fig.9 is given in Table 8.
During the application no flashover or puncture is
observed, in addition, observed value of PD is less
than 10 pC and 5 pC at Um,
1.5 𝑈𝑚 ⁄√3 𝑎𝑛𝑑 1.1 𝑈𝑚 ⁄√3 of test voltage. Hence
the tested bushing is in compliance with the
requirement of the test conducted as per IEC 60437.
This test is applicable to all types of bushings.
Transformer bushings having Um >72.5 kV
subjected to 15 full LIV of positive polarity, 01 full
LIV of negative polarity at 110 % of rated LI
withstand voltage, 05 chopped impulse voltage of
negative polarity at 121 % of rated LI withstand
voltage and 14 full LIV of negative polarity at
110 % of rated LI withstand voltage. Test voltage
corrected to atmospheric ambient conditions as in
IEC 60060-1 and IEC 60137 and corrected voltage
is applied. No flashover on liquid immersed parts,
not more than two flashovers in air at positive
polarity and no flashover in air at negative polarity
are the requirement to pass the test. During testing
of 420 kV OIP transformer bushing no flashover or
puncture was observed. Representative Oscillogram
of LIV of 420 kV bushing is shown in Fig.10

Fig.9. Test voltage profile of ACLD test

Table 8.Long Duration Power Frequency Voltage Test – OIP Bushing

Sl. Test Test duration / PD measured
Fig.10. Oscillogram of LIV withstand test on 420 kV transformer bushing
No. voltage Number of Disruptive (pC)
(kV) discharges observed
1 266.7 5 min / 0 3.9 6.4. Dry or Wet switching impulse voltage (SIV)
2 363.7 5 min / 0 4.1 withstand test
3 420.0 1 min / 0 4.8
4 3.8
5 3.8
SIV withstand test in general applicable to all
6 3.8 bushings of Um ≥ 300 kV. Dry SIV withstand test
7 3.8 is applicable to indoor, indoor-immersed, and
8 3.8 completely immersed bushings. In addition, dry SIV
9 363.7 60 min / 0 3.6
10 3.6
withstand test is applicable to all transformer
11 3.6 bushings of Um ≥ 245 kV. However, wet SIV
12 3.6 withstand test is applicable to outdoor bushings.
13 3.8 During the testing, it was ensured that the bushing is
14 3.8
15 3.8
mounted on the earthed plane, radially extended in
16 266.7 5 min / 0 3.4 all direction of 0.4 times of arcing length of bushing.
High voltage connection is made on top of bushing
at the height of at-least 0.4 times of arcing length in
the axis of bushing. Wetting arrangement and other
relevant parameters are made in compliance with
IEC 60060-1 for wet SIV test. During the test, the
bushing is subjected to 15 pulses each of positive
and negative polarity sequentially at the magnitude
as given in IEC 60137:2017. For transformer
bushing having Um of 245 kV and above, additional
dry test with 15 impulse of negative polarity at
110 % of rated withstand voltage is applied.
Representative oscillogram of switching impulse
test on 420 kV transformer bushing is shown in
Fig.11.
Fig.11. Oscillogram of SIV withstand test on 420 kV transformer bushing
7. Conclusions
CPRI-UHVRL having facility to perform dielectric
tests, partial discharges (PD) measurement test,
electromagnetic compatibility (EMC) test and
capacitance (C) and dielectric dissipation factor
measurement (tan δ) test according to relevant
standards on transformer bushing upto ultra high
voltage rating.
8. Acknowledgement
The authors wish to thank the CPRI authorities for
giving permission to publish this paper.
9. References
[1]. Christina AJ, M A Salam, Q M Rahman, and et al,
“Causes of transformer failures and diagnostic methods
– A review”, Renewable and Sustainable Energy
Reviews (2017).
[2]. CIGRE-WG 12-05: An international survey on failures
in large power transformers in service, Electra No. 88;
1983. S. 21–48.
[3]. IEC 60137, “Insulated bushings for alternative voltages
above 1000 V”
[4]. IEC 60270, “High-voltage test techniques – Partial
Discharge measurements”
[5]. IEC 60507, “Artificial pollution tests on high-voltage
ceramic and glass insulators to be used on a.c. systems”
[6]. CISPR 18-2, Radio interference characteristics of
overhead power lines and high-voltage equipment Part-
2: Methods of measurements and procedure for
determining limits.
[7]. IEC 60060-1, “High-voltage test techniques - Part 1:
General definitions and test requirements
 Gaseous Contents in Transformer oil and Dissolved Gas Analysis
Introduction:
Transformers are the critical elements that step up
and down the voltages at each end of the
transmission line. This devices used in electrical
circuits changes the voltage of electricity flowing in
the circuit. Transformers can be used either to
increase the voltage (called "stepping up") or
decrease the voltage ("step down"). The power
transmitted along the line is equal to the voltage times
the current. The higher the voltage the lower the
current that must flow within the transmission lines to
deliver the same power. The power is distributed
through different rating of power transformer based
upon the requirement of utilities and applications.
During power transmission there is power loss which
is imperative and our aim should be of minimizing this
power loss. The devices like transmission lines,
transformers, Insulators etc.; the material quality and
equipment design will decide the degree of power
loss. Each device is unique and important in a system
and its proper satisfactory functioning is expected.
Here our focus is on the transformer and its
component specially transformer oil. In this paper we
will discuss about the quality of oil that affects the
performance of a power transformer.
A satisfactory functioning and long service life is
always expected which is dependent on many factors
like its quality of constituent components, design and
maintenance. The Oil in power transformer acts as
insulant and coolant media between primary and
secondary windings of a transformer. There are
various types of liquid insulants like Mineral oil,
Silicone Oil, Natural and Synthetic esters that are
available in field and market depends upon the
application cost, quality and availability; above all the
effect of oil on the performance of equipment
transformer. The mineral oil is common liquid insulant
that finds place in power industry. This oil for use in
power transformer needs to be tested before its
application in and in-service conditions its behavior to
be monitored regularly for the satisfactory functioning
of transformer. The inherent impurities of oil with
added and or generated impurities during the service
….Ramjeet Singh, Additional Director, CPRI
life will reduce the working life of power transformers.
The power transformers under thermal and electrical
stress with overloading conditions due to power
demand and continuous use while poorly maintained
environment are the catalyst for early degradation of
oil and failure of power transformer in service.
The quality of intended oil for use must qualify the
tests as per National / International standards of IS
335-2018 / IEC 60296-2013. Before energization of
power transformer also the oil should be checked to
meet the requirement as per IS 1866-2017/ IEC
60422 and maintenance in service.
As we know Mineral Oil is a mixture of Hydrocarbon
liquids; in-spite impurities of in-organic, organic in
nature, free or dissolved, added or generated during
service conditions are inherent part in oil and these
impurities affects the quality of oil. We need to
minimize the impurities that come from refining stage,
transportation and service conditions. The common
physical impurities that gets included in oil are : air,
moisture, dust, particles , excess additives, acids etc
that needs removal to improve and maintain the oil
quality. During thermal and electrical stress on
transformer these impurities can be of catastrophic in
nature.
In this paper we will study the causes of gaseous
presence or generation in power transformer oil and
their role in ascertaining the oil quality and gaseous
limits under Factory testing of transformers as well as
transformers in operational or field service from
maintenance perspective using Dissolved Gases
Analysis technique of testing for monitoring of these
power transformers.
GASEOUS PRESENCE IN NEW TRANSFORMER
OIL: A gaseous presence in oil is unfavorable
condition and oil should not contain gas in it; hence
an oil before use in transformer is degassed. A typical
air presence limit for different ratings of power
transformer like large transformer >100 MVA is <0.5
%, for Medium power transformer (between 20 MVA
to 100 MVA) is < 1% and for Small and distribution
transformers; it is upto 2%. An new oil that have not CO2 152 184 289 66
gone under any thermal or electrical tests generally H2 0 0 0 0
also has air dissolved in it and this air can be removed O2 10406 13050 14595 6233
by degassing/ filtration of oil. DGA of fresh new oil N2 19063 23861 26850 13260
shows the presence of air constituents like Oxygen From above we see that air ( O2+N2) is less in filtered
and Nitrogen mainly and very trace amount of oil. Multiple filtration/ high vacuum can further bring
down the air content in oil. The Solubility of air in oil
Methane. This comes because of air is dissolved in
at saturation depends on the atmospheric pressure
Oil and oil is not properly degassed. A properly and temperature and type of oil used.
filtered oil should have no or minimum air constituent.
Effect of Air Content on BDV test :
We carried out DGA of atmospheric air samples and
air extracted from new oil samples -given below which Many electrical apparatus use air as the insulating
indicates that air dissolution in oil is very less. This medium. A dry air acts as a good insulant however
humidity and moist affects negatively on BDV value.
DGA result of oil should be taken as reference point
In few cases other gases such as N2, CO2, CCl2F2
for further increase of any other gas generated during (Freon) and SF6 (hexafluoride) are also used. BDV
the testing of transformer due to thermal or electrical value of Air ( 50% RH) is around 8 kV/2.5 mm gap; for
stresses. Sometime we also have found the so called mineral oil it may be up-to 80-90 kV/2.5 mm gap. A
or claimed new oil has initial presence of hydrocarbon paper/ press board has value less than of oil but an
gases which indicates the oil is not new or it is oil impregnated pressboard has higher value of its
own. Decrease of the average breakdown voltage
blended with other used oil; however if there is no
under the influence of air bubbles are noticeable; an
further increase of gas formation, then this does not early flash-over is seen in presence of air bubble in
indicate the effect of thermal of electrical stress on oil. That is why as per IS 6792 and IEC 60157 an
transformer due to test conditions. initial stand time of 5 minutes are given before testing
of first break down. Whereas for natural ester oil, 15-
30 minutes of initial stand time is given so that
GASEOUS CONSTITUENTS OF AIR :
dissolved air bubbles comes to surfaces before first
Gases AIR 1 AIR 2 AIR 3 AIR 4 In PPM BDV test. The moist air with carbon or dust
contaminants in oil enhances early break down of oil
CH4 1.97 1.88 1.66 1.68
value.
C2H6 0 0 0 0 0
Effect of Gas Content on Flash Point:
C2H4 0 0 0 0 0
C2H2 0 0 0 0 0 Flash point of oil is a temperature at which produced
combustible gas mixture catches momentary fire. The
CO 0.78 2 0 1.31 breakdown of oil caused by electrical discharges or
prolonged exposure to very high temperature may
CO2 685 227 454 474 4
produce sufficient quantities of low molecular weight
H2 0 0 0 0 0 hydrocarbons lowering the flash point of the oil.
Lowering of flash point is the indication of presence of
O2 199261 196825 163628 164470 1 volatile combustible products in oil may be due to
N2 698851 693586 558491 561130 0 contamination of solvent or dissolved gases that were
generated due to electrical sparking discharge
leading overheating of oil.
GASEOUS CONSTITUENTS IN NEW MINERAL
TRANSFORMER OIL (Filtered) : A data analysis of 70 different transformers oil sample
Gases NEW 1 NEW 2 NEW 3 NEW 4 In PPM were carried out for Flash Point and Dissolved gas
CH4 0.23 0.29 0.39 0 0 content correlation was established. It was found that
19 samples in range of Flash Point 161-168 .C had
C2H6 0 0 0 0 0 no Dissolved gas content in it; 24 nos. had gas
C2H4 0 0 0 0 0 content and 20 with no gas content found in range of
C2H2 0 0 0 0 0 151- 159 .C flash point; while all the 7 samples with
CO 0.57 1.3 1.28 3 143 – 150 .C flash point had gas content in it. From
this experimental data analysis we come to know that these gases and their quantum we can identify the
presence of generated gases (combustible in degree and nature of fault inside the transformer. As
nature) decrease the flash point of oil. As per IS these gases are soluble in oil, the fault imprints goes
1866 / IEC 60422 a decrease of 10 % of Flash point into oil; and after extraction of these gases from oil
is allowed for safe working of oil. It may be noted and Gas Chromatographic method of analysis
that each sample may have different initial flash point. prediction of the fault taking place inside the
transformers is possible and it is well documented test
procedures.
Effect of Gas content on Flash
In case of transient fault very less quantity of
Point hydrocarbon gases are generated ( as happens in
30 Factory tests) and hence DGA accuracy of analysis
is less; to overcome this problem some time oil can
25
be oxygenated to a given range of concentration such
20 as 8000 – 12000 ppm in initial new oil to get
15 hydrocarbon gas results in Factory test ( like Impulse
and temperature rise test) samples as they have low
10 formations of gassing rates i.e to sensitize the DGA
5 results by Gas chromatograph. Here the air acts as a
carrier for diffusion of generated hydrocarbon gases
0
in oil and makes DGA with better accuracy.
161-168 .C 151-160 .C 142-150 .C

No. of Trf. Without Gas No. of Trf. With gas In degassed oil, oxygen concentration is very low.
Variation in Oxygen and Nitrogen gas content is also
seen sample to sample the solubility of Oxygen
Effect of presence of Non-Combustible gas decreases with increase of temperature while for
impurities on Flash Point Nitrogen solubility increases with increase of sample
temperature.
The thermal or electrical stress on mineral oil filled
transformer evolves the gases like Hydrogen, GAS CONTENT / GAS ANALYSIS (DGA) TO FACTORY
Methane, Ethane, Ethylene, Acetylene, Carbon Mono
oxide which are combustible in nature, while air ( O2 TESTS ON ELECTRICAL EQUIPMENTS:
+ N2 ) : O2 helps in burning / oxidation. The N2 , CO2
and Moisture are non combustible in nature; that is Before commissioning of a power transformer in field
why we get slight higher flash point of a sample in service it is put under factory testing to ensure its
comparison to what we get on consequitive test of proper working in actual service life conditions. A
same tested sample. transformer is put under Short Circuit test, Lightening
Impulse tests, Heat Run Test, HV Tests like
Sample 1 Sample 2 Sample 3
conditions wherein electrical and thermal stresses
154 .0 165.5 157.5
152 .0 163.5 155.5 develop and there are chances of gaseous formations
152.0 161.5 155.5 in oil. The gaseous formation beyond limits indicates
the abnormal condition of transformer under these
Sample 4 Sample 5 Sample 6 testing. However these tests are for short durations
153 .5 160.0 156.0 and gaseous formations are very less giving
151 .5 159.5 155.5 challenges for a chemist / scientist to measure these
152.0 159.5 156.0 small quality of gases in precision.
“ The type of volatile Impurities affects the flash
point value.” Factors affecting Gases rate during Factory
Tests:
Gaseous Presence in Oil : The mineral oil is
hydrocarbon liquid and under thermal heating and Gas measurements are used to detect the effect of
arcing conditionals it breaks into smaller constituents abnormal temperature in windings, leads, magnetic
of gases like CH4, C2H6, C2H2, C2H4, C3H6,C3H8 circuits, structural elements or abnormal leakage flux.
etc. The Paper which is cellulosic in nature may also The design of these transformers thus imparts
break into CO, CO2, H2O. By nature of presence of influence in gas formations. Beside the transformer
design there are other aspects that may effect on gas
production rate :-
-Oil to Cellulose mass ratio : If there is less oil to
absorb the gases produced, higher gassing rate will
be observed.
- Paper/ Cellulose type or quality : Thermally
upgraded or not
- Oil type or Brand : Stray gassing tendency in oil
will produce high rate of gas formation
- Cooling method and cooling efficiency
- In some transformers paints, glues/ resins, stainless
steels and other materials may evolve gas formations
- Test duration : It is an important aspect for how long
and at what stress rate the transformer has gone
under testing
Due to stray gassing tendency of some oils they
produce gases such as hydrogen , hydrocarbon
gases and carbon oxides at low temperature ( < 120
.C) without thermal and electrical faults and even
without operation stress, resulting high production of
gases and misinterpretation of DGA results. Inhibited
oil typically produce less stray gassing than
uninhibited ones. To measure stray gassing method
ASTM D7150 or CIGRE Brochure 296 is adopted.
Gassing Tendency : Gassing Tendency of oil differs
from oil to oil. It is the gas absorbing tendency of oil
when subjected to corona partial discharge and it is
necessary and important to measure in case of HV
transformers and bushing. It is a measure of rate of
absorption of gases into oil under prescribed
conditions. The method IEC 60628 -1985 is adopted
to carry out this test. A gassing tendency of 5
mm3/Minute is used for => 400kV equipment.
RECOMMONDED METHODS OF GAS
EXTRACTION FOR FACTORY TEST:
The gases dissolved in oil should be extracted and
analysed by gas chromatography. As in the factory
test samples the gas content are very low, special
precautions are to be taken for gas extraction,
selection of GC equipment and highly trained,
qualified and experienced hand to work and analyse
the DGA test results is required. Oil samples should
be analysed as soon as possible after being taken
and in no case later than seven days. The detection
limit of GC system for overall determination is crucial
and indicated as below to qualify for selection of
equipment.
Detection Limit (S) for factory tests
Gases_ Concentration _ul/l
CH4 0.1
C2H6 0.1
C2H4 0.1
C2H2 0.1
CO 5
CO2 10
H2 2
O2 500
N2 2000
A highly sensitive capillary column is recommended
for GC system to detect the above gases. In addition
to that a very good repeatability ( r ) of test results is
necessary to prevent the misinterpretation of test
results. It is also recommended the same laboratory
is used and laboratory repeatability is monitored
regularly. The required criteria of repeatability at low
gas concentrations as per IEC 60567 is : r<S. The
CPRI Oil Test Lab follows the required guideline for
DGA testing.
The gas extraction efficiency is crucial for end DGA
results. There are various gas extraction methods like
Multi-cycle vacuum extraction using vacuum pump
apparatus, , Vacuum extraction by partial degassing
method applying gas correction factors, Stripping
extraction methods and head Space methods. In
CPRI we are adopting Multi-cycle vacuum extraction
using vacuum pump apparatus with Gas
chromatograph and latest Agilent make Head Space
Gas chromatograph for DGA analysis. The accuracy
of extraction by different methods differ and also
dependent on the concentration level of gases
present in oil. A sample is considered medium gas
concentration level when hydrocarbons are between
9 – 60 ppm, CO and CO2 between 100 to 500 ppm;
and known as low gas concentration level when
hydrocarbons are between 1 – 10 ppm, CO and CO2
between 30 to 100 ppm. In factory tests gases
generated are generally in low gas concentration
level.
Limits of DGA after Temperature Rise Test : As per
IEC 61181 reference standard, in 90% typical cases
the increase in gases found ( _ul/l/h) as blow and Similarly after lightening impulse test on
problem noticed in no. of cases are also tabulated transformers, the gaseous limiting values are as
below: follows :
Transformers H2 Cn H2+Cn CO CO2
Gases H2 CH4 C2H4 C2H6 C2H2 CO
0.1-0.3 0.04-0.3 0.1-1.6 0.4- 2 5-18
_ul/l 15 4 1 1.5 0.5 15
Special cases: 1.7 0.5 2.2 5 20 (
As factory tests are for short duration, the gas
in ppm) where Cn = CH4+C2H6+C2H4+C2H2
formation are very less, but in route operation in
service over the period of years and overloading
Here special cases refer to such oil with higher
gassing tendency. conditions the oil remains at agitated temperature, the
natural degradation of oil takes place and the gases
Generally in temp. rise test C2H2 is not found or gets accumulated.
below the detection level.
Remark : From the above it is seen that slight
( ul/l/h) increase of gases is seen after temp. rise test
H2+Cn Problem cases however in some cases significant amount of gases
are noticed indicating the concern to be
<0.5 1/215
investigated.
0.5 -1 1/36
1 -2 4/21 Case Study : DGA of Power Transformer in
2-5 4/12 Service conditions: -
5-10 2/4 The relative quantities of above gases vary
>10 3/3 depending upon how the energy available to decompose
the oil is released at fault and hence depends on
At low concentration gas content the DGA result characteristic of the type of fault, if a fault is present. The
accuracy varies a lot and also dependent on type of most significant gases generated by decomposition of oil
gas extraction methods. As per standard the effect of are Carbon monoxide (CO), Carbon dioxide (CO2),
gas extraction method is tabulated as below : hydrogen (H2), methane (CH4), ethane (C2H6), ethylene
(C2H4), and acetylene (C2H2).
Accuracy of Extraction Methods :
During ordinary operation, gases particularly
Extraction Medium Conc. Low Conc. CO2 and CO are also formed by normal ageing
Procedure % Accuracy % Accuracy processes in the transformer. For cellulose degradation
Toepler 13 35 by heat alone, at current operating temperatures,
Partial Degassing 13 30 statistical analysis for normally operating conservator
Stripping 18 23 transformers gives a CO2/CO ratio of about 7 although
Headspace 18 37 with a widespread of values. Some cases in which
CO2/CO is below 3 or above 11 should be regarded as
perhaps indicating a fault involving cellulose. When
The accuracy of DGA results of low gas content excessive paper degradation is suspected (CO2/CO<3),
samples in oil (after gas extraction) further goes it is advisable to ask for a furanic compounds analysis or
down. At very low gas content level DGA accuracy as a measurement of the degree of polymerization of paper
per IEC 61181 for 1-3 ppm of hydrocarbon, 2.5 ppm
sample.
of H2, 5 ppm of CO, 40 PPm of CO2 level is approx..
+44%.
Partial discharges occur in case of fault of low
level energy (breakdown in gas filled voids surrounded by
oil or oil impregnated material); the main cause of
decomposition in this case is ionic bombardment of the
oil molecules and the major gas produced in hydrogen. Based on actual test on various samples carried out there
In other cases, decomposition of the oil is mainly caused are examples that indicate the presence of different
by heat, with variations in the types of hydrocarbons amount of gases and types of fault inside the transformer.
produced as the temperature rises. Thus a little
decomposition occurs at normal operating temperature, Example No. 1:-
produced mainly hydrogen and methane. Higher
S.No. Test
temperatures and higher energies are caused by hot
Test values
spots or conductor overheating; temperatures from a little
DISSOLVED GAS obtained
above normal operating temperature, say 1500C, to as 1
ANALYSIS (DGA)
high as 10000C may occur in such cases to decompose
Name of Gases
the oil. The principal gas produced by low temperature
hot spots is CH4, but as the fault temperature rises C2H6
1. Methane, ppm 18517
and C2H4 appear in increasing quantities. In the much
2. Ethane, ppm 3804
higher temperatures occurring at sparking flashovers, or
3. Ethylene, ppm 35223
in power arcing where temperatures of over 3000 0C
,C2H2 becomes significant product. At this temp. oil 4. Acetylene, ppm 1236
decomposes to elementary components. C3 5. Carbon dioxide, ppm 352
hydrocarbons are also gives complementary 6. Carbon mono-oxide,
73
information which is useful to make the diagnosis more ppm
precise. Moreover, experience has shown that, in most 7. Hydrogen, ppm 154
cases, a satisfactory diagnosis can be made without 8. Oxygen 12646
taking into account C3 hydrocarbons, they have been 9. Nitrogen 48041
omitted from the interpretation method. Date of commissioning
2008
of transformer
Dissolved O2 and N2 may be found in oil. The Date of Filtration of Oil July 2008
O2/N2 ratio in oil reflects air composition and is close to
0.5. In service, this ratio may decreases as a result of oil
oxidation and/or paper aging. Factors such as the load As per Roger’s ratio method of analysis the Ratio
and preservation system used may also effect the ratio, Code is 0 2 2 indicating Thermal fault of high temp.
but ratio less than 0.3 are generally considered to indicate O2/ N2 ratio less than 0.3 also indicates that severe
excessive consumption of oxygen. oxidation of oil has taken place. As per Duval’s
Triangle Method of Analysis, we also find the
In general the DGA gas content is dependent “Thermal Fault > 700 .C”.
upon the transformer design, load and temperature etc;
there is no standard limit for the gas content. The gas
content is equipment/ transformer specific and its fault
analysis is done based on “Gas increasing Trend
Analysis” ; Rogers ratio Method (C2H2/C2H4, CH4/H2,
C2H4/C2H6), Duval’s Triangle Method etc. before
coming to conclusion.

The persisting faulty conditions thermal or

electrical leads high concentration of gas formations
making gas safety relay operational and raising
alarm. The DGA of such oil samples are carried out
at regular intervals for maintenance of power
transformers.
Example No. 2:- As per Roger’s ratio method of analysis the Ratio
Code is 1 0 2 indicating Discharge of High energy.
S.No. Test
DISSOLVED GAS Test values As per Duval’s Triangle Method of Analysis, we also
2 obtained find the “ Electrical and Thermal fault.”
ANALYSIS (DGA)
Name of Gases Example No. 4:-

1. Methane, ppm 1008

S.No. Test
2. Ethane, ppm 163 Test values
DISSOLVED GAS
3. Ethylene, ppm 1566 4 obtained
ANALYSIS (DGA)
4. Acetylene, ppm 393 Name of Gases
5. Carbon dioxide, ppm 1182
6. Carbon mono-oxide, 1. Methane, ppm 5
126
ppm 2. Ethane, ppm 2
7. Hydrogen, ppm 775 3. Ethylene, ppm 10
8. Oxygen 28650 4. Acetylene, ppm 9
9. Nitrogen 47999 5. Carbon dioxide, ppm 432
Date of commissioning 6. Carbon mono-oxide,
2005 46
of transformer ppm
Date of Filtration of Oil Feb-2008 7. Hydrogen, ppm 0
As per Roger’s ratio method of analysis the Ratio 8. Oxygen 11928
Code is 1 2 2 indicating Discharge of High energy
9. Nitrogen 26172
and Thermal Fault of high temp. As per Duval’s
Triangle Method of Analysis, we also find the “ Date of Filtration of Oil 2014
Electrical and Thermal fault.”

Example No. 3:- As per Roger’s ratio method of analysis the Ratio
Code is 1 2 2 indicating Discharge of High energy
S.No. Test Thermal Fault of high temp. As per Duval’s Triangle
DISSOLVED GAS Test values Method of Analysis, it denotes the beginning of “
3 obtained
ANALYSIS (DGA) Electrical Arcing fault.”
Name of Gases
CONCLUSION:
1. Methane, ppm 23
2. Ethane, ppm 11 1. The Volatile impurities like solvents, low molecular
wt. Hydrocarbons, Gassing tendency of oil, Moisture
3. Ethylene, ppm 140
content play significant role in BDV and Flash Point.
4. Acetylene, ppm 91
5. Carbon dioxide, ppm 269 2. Data Analysis of Dissolved gases ( DGA) is a
6. Carbon mono-oxide, significant and powerful tool for diagnostic study of
1
ppm nature of faults in the electrical power equipment.
7. Hydrogen, ppm 9
8. Oxygen 10033 3. High precision equipment like Gas Chromatograph
9. Nitrogen 18666 is needed for DGA testing and analysis of gases in
Date of commissioning low gas content level. of equipment. DGA of even
New new oil may be prescribed to ensure the initial level
of transformer
of presence of gases that to be taken as reference
Date of Filtration of Oil ---- point and to see the further rise of gas concentration
if any during factory tests of power transformer or
in-service operational condition for fault analysis.

4. The Gas concentration level, Gas Ratio indicates

the type of faults inside the transformer.

References:

1. National and International Standards like IS

10593, 9394, IEC 60422, 60567, 60599 ,
61181 etc.
2. Operation of Gas Analyser for transformer
oil with Headspace sampler – M/s Agilent
Technologies, Inc.
3. Guide for the Interpretation of gases
generated in Oil- Immersed Transformers –
IEEE C57.104

Acknowledgement : The author is thankful for the

technical lab resources and management of CPRI for
providing opportunity to work and gain experience in
the field. He is also thankful to the colleagues and
staffs for their valuable help and support.

About Author :

Ramjeet Singh, M.Sc. First Class Chemistry pass

out from Meerut University Meerut, has worked in
Indian Railways, Rail Coach Factory Kapurthala as
Chemist and Metallurgical Assistant ( Quality
Control), in the beginning of his carrier for about two
years and then joined CPRI in 1989 as a Scientific
Officer. He has working experience of past 30 years
in field of testing and certification in CPRI. He has
served in the institute at various Divisions/ Units in
different capacities and presently working as
Additional Director, HOD - Oil Test Lab, CPRI
Bhopal. Email: ramjeet@cpri.in; Mob. 09754864113
 A STUDY ON THERMAL BEHAVOUR OF LOW VOLTAGE IRON CORE SERIES
REACTORS USED WITH CAPACITOR BANKS
V.Vaidhyanathan, T.Bhavani Shanker, R.Shyam, A.Sheik Mohamed
Central Power Research Institute, Bangalore
Abstract
In Distribution sector increased emphasis is
on reduction in power losses and improved
system efficiency. With more and more
non-conventional sources being added to
the grid the concept of distributed
generation and smart grid are gaining
credence. The importance of power factor
improvement is a significant contributor to
the scheme of things in the smart grid
conceptualization. Fixed capacitor banks
are used for correction of power factor in LV
distribution networks. As nature of industrial
loads has become very dynamic, power
factor compensation have also become
dynamic with the installation of
automatically switchable capacitor banks in
compact metal enclosed assemblies.
These are used to supplement the fixed
shunt capacitor banks to cater to the system
requirement. With increased usage of
variable frequency drives and power
electronic components the nature of loads
have changed over a period of time and
power quality is becoming a major issue. In
addition, increased harmonic loading is a
major concern as capacitor banks being low
impedance devices are subjected to not
only excessive overloads, but also face
resonance phenomenon. To minimise the
effect of harmonics series reactors which is
also known as detuned reactors are used.
Air core and iron core reactors can be used
as de-tuning reactors in the LV systems.
But the later is preferred as it is compact
and can be accommodated in constrained
spaces of the metal enclosed LV
assemblies.
This paper discusses the thermal
performance of these Iron core reactors
which are used in the LV assemblies. Two
case studies of tests performed on these
type of reactors are presented in this paper.
Key words: Detuned reactor, Iron core
reactors, series reactors, LV capacitors,
APFC panels.
1. Introduction
Transformer and Reactors are both wound
apparatus. But while transformers are used
only for transformation, reactors are more
application specific. Based on specific
system requirements reactors are designed.
Reactors are classified in IEC 60076-6 [1]
as shunt reactors. series current limiting
reactors, neutral earthing reactors, power
flow control reactor. motor starting reactors
filter (tuning)reactors, capacitor damping
reactors, discharge reactors, neutral
coupled arc suppression reactors,
smoothing reactors, etc., depending upon
the application.
In LV distribution systems due to the
increased presence of industrial loads like
variable frequency drives, power electronics
devices, inverters and converters, induction
furnaces, etc., the problem of power quality
particularly harmonics are very high. In
absence of proper harmonic filters these
have a serious impact on other devices
connected in the network. Since shunt
capacitor banks used in network offer low
impedance path at increased frequencies
they are susceptible to the harmonic
overloading and could also lead to some
unwarranted resonance phenomenon. To
minimise this problem in the LV network,
series reactors also called as de-tuned
reactors are introduced between the LV
system and the shunt capacitor banks. This
increases the series impedance of the
capacitor bank and also minimises the
problem of transient inrush current due to
capacitor switching. De-tuned reactors
used for this purpose are designed for
specific location based needs instead of
being generic. They are also indicated by a
percentage along with the rated voltage and
rated current.
In LV networks widely used de-tuned
reactors are dry type and designed with
Iron core of CRGO or CRNGO core, Air
natural (AN) cooled, aluminium or copper
wound with class F or class H insulation.
Distributed multiple core gaps are provided
to reduce the stray flux losses and to
enhance the linear performance of the
reactor.
Generally, tests on these reactors are
carried out in accordance with IEC 60076-6
[1] and IS 5553 part 5 [2]. Among which,
Temperature rise test is one of the type
tests to verify the thermal performance of
reactors. Temperature rise test performed
on two different types of iron core reactors
of similar rating are discussed in this paper
along with some of the important
observations.
2. Assessment of Thermal performance
An important test for assessing the thermal
performance of reactors is performing a
temperature rise test. The test reveals the
combined effects of electrical and thermal
stresses.
During the test the reactor will be energised
continuously to the rated equivalent current
(Iequ). This value of current is generally
higher than the rated current as in service
the reactor is expected to get loaded with a
certain level of harmonic currents. This will
be close to simulating the service conditions
under which the reactor is expected to
perform.
Performing temperature rise test on LV de-
tuned reactors is a technical challenge due
to the following factors:
1. Test procedures for de-tuned reactors
are not specifically covered in
standards.
2. Maintaining the currents at the required
level as the excitation voltage are higher
depending on the tuning order.
3. Measurement of temperature in
energised condition of the reactor.
Power Capacitors Lab of CPRI, Bangalore
has carried out temperature rise test on LV
detuned reactors of rated reactive power of
upto 100 kvar, in the tuning order of 2.7 to
4.2.
The temperature rise test is carried out
generally as per cl. 9.10.8 of IEC 60076-6
[1] which cross refers to IEC 60076-11[3] &
IEC 60076-2[4].
However IEC 60076-2 deals with
temperature rise test on oil filled power
transformers and IEC 60076-11-2004[3]
deals with dry type transformers. In the
above IEC standards and equivalent Indian
standards there is no specific reference for
temperature rise test method on reactors,
for both air core and iron core types.
However test method generally prescribed
in IEC 60076-2 and 60076-11 is adopted for
carrying out temperature rise test on LV
detuned series reactors
3. Application of current.
Temperature rise test is performed at rated
equivalent current (Iequ) and rated frequency
of the reactor. In exceptional cases, for
example in extremely large rated power it
will be generally difficult to maintain these
test conditions and in such cases the test is
performed at reduced current, but not less
than 0.9 Iequ. The test level shall be stated
by the manufacturer and agreed upon
between manufacturer and purchaser
before commencement of the test. The
temperature rise shall be corrected to the
rated equivalent current at power frequency.
The winding temperature rise above
ambient temperatures shall be multiplied by
(Ir/Itest)y, with y=1.6 of AN cooling reactor
and y= 1.8 for AF cooling reactors.
Appropriate connectors and electrical leads backwards in time to the instant of
shall be used to connect the reactor to the shutdown of ac power source.
power supply. Since reactors used could be either copper
wound or aluminium wound, the following
3.1 External cooling air temperature formulae are used for determining the
Precautions shall have to be taken to average winding temperature.
minimise variations of cooling air
temperature particularly during the later part The ratio between resistance value R2 at
of the test when steady state is temperature θ2 and R1 at θ1 is taken as:
approaching. At-least three temperature
sensors with suitable time constant shall be 𝑅2 (235+𝜃2 )
used for measuring cooling air temperature = : for copper
𝑅1 (235+𝜃1 )
at 1 to 2 m away from the reactor. The
average of the readings shall be used for
𝑅2 (225+𝜃2 )
evaluation of the test. Readings shall be = : for Aluminium
taken at regular intervals or automatic 𝑅1 (225+𝜃1 )
continuous recording shall be used.
Where, R1 is cold resistance at external
3.2 Determination of steady state cooling temperature (θ1) and R2 is hot
condition resistance at instant of shutdown (θ2)
To monitor the attainment of steady state
condition, temperature sensors are fixed at From the above equations, the temperature
various locations on the reactor surface. value θ2 is obtained as follows:
The ultimate steady state condition of
temperature rise is reached when the 𝑅2 (235+ 𝜃2 )
𝜃2 = − 235 : for copper
temperature rise becomes constant: This is 𝑅1
considered to have been achieved when the
𝑅2 (225+𝜃2 )
temperature rise does not vary by more 𝜃2 = 𝑅1
− 225 : for Aluminium
than 1K per hour for three consecutive
hours. The external cooling medium at the time of
shutdown is θa. The winding temperature
3.3 Temperature rise determination rise (∆θw) is calculated as ∆θw = θ2 - θa.
The temperature of the winding at the end
of the test will normally be determined by
measurement of winding resistance 3.4 Limits specified
immediately within three minutes. The The temperature of components used in
measurement is commenced after reactor, metallic parts and adjacent
shutdown of test power and reconnection of materials shall not reach a value that will
the reactor terminals from ac test power cause damage to any part of the reactor.
source to the dc measuring current source.
Table 1; Winding temperature limits
The winding temperature and its resistance for reactors as per IEC 60076-6/11[3]
vary with time and the problem is to Class of Insulation Average
extrapolate backwards in time to the instant Insulation system winding
of shutdown. Resistance measurement is temperature temperature
commenced as soon as possible within rise (K)
three minutes after the connection of the
A 105 60
reactor terminals to the measuring
E 120 75
equipment. A considerable number of
B 130 80
discrete resistance measurements are
obtained within a time span of 20 minutes F 155 100
from the time of disconnection. The H 180 125
measured values are extrapolated
4. Case Studies
Several LV iron core type reactors have
been tested at the laboratories of CPRI for
temperature rise test. The ratings of the
reactors tested ranges from 5.67%, to 14 %
in the voltage range 380 V to 525V. 1 kvar
to 100 kvar and tuning order of 2.7 to 4.2.
The test has been carried out at a system
frequency of 50 Hz and the prevailing
ambient temperature of 25°C to 30°C.
In this paper, temperature rise test on two
LV detuned reactors of similar rating and
different makes are presented as case
studies. While the first reactor is aluminium
wound the second is copper wound.
Case study-1
Details of the reactor tested are as follows:
Rated Inductance LN/mH : 0.784
Rated Voltage / P% UN/V: 415, 12.5%
Rated current IN/A : 139.1
Rated frequency fN/Hz : 50
Rated equivalent Iequ A: 154.6
Qk/kvar : 100
Conductor: Aluminium wound
Insulation class: Class H
The reactor was mounted on a wooden
platform and was energized to the rated
equivalent current of the reactor with the
help of a 3Ø AC power source. The rated
equivalent current measured was 154.6A.
Energization was continued until a steady
state temperature condition was reached
and maintained for further three hours. A
view of the tested reactor is as shown in
Figure-1.
Figure- 1: A view of reactor used for the Case
study-1.
Temperature was monitored around the
outer surface of the reactor at nine locations
until a steady state temperature condition
was reached and maintained for three
hours. Since the middle phase encounters
the lowest thermal dissipation, the first
measurements were carried on the middle
phase. The reactor attained stabilization of
first (middle) phase after nine hours
Thereafter measurement were carried out at
the each of the other phases with an interval
of 4 hours. The temperature rise was
measured by resistance extrapolation
method. Resistance measurements were
carried out using an automatic resistance
meter. The calculated temperature rise for
individual phases of the reactor at steady
state condition is given below:
The absolute temperature rise ∆θw was
112K for R-phase, 122K for Y-phase and
111K for B phase against the specified
value of 125 K for class H insulation.
The reactor successfully withstood the
temperature rise test ensuring satisfactory
thermal and electrical performance of the
reactor.
Case study-2
Details of the reactor tested are as follows:
Rated Inductance LN/mH : 0.46
Rated Voltage / P% UN/V: 440, 7%
Rated current IN/A : 132
Rated frequency fN/Hz : 50
Rated equivalent Iequ A: 154.6
Qk/kvar : 100
Conductor: Copper wound
Insulation class: Class H
The reactor was energized to the rated
equivalent current of the reactor with the
help of a 3Ø AC power source. The rated
equivalent current measured was 144.4A.
Energization was continued until a steady
state temperature condition was reached
and maintained for further three hours. A
view of the reactor under test is as shown in
Figure-2.

Figure- 2: A view of reactor used for the Case
study-2.
Temperature was monitored around the
outer surface of the reactor at nine locations
until a steady state temperature condition
was reached and maintained for three
hours. Since the middle phase encounters
the lowest thermal dissipation, the first
measurements were carried on the middle
phase. The reactor attained stabilization of
first (middle) phase after eight hours
Thereafter measurement were carried out at
the each of the other phases with an interval
of 4 hours. The temperature rise was
measured by resistance extrapolation
method. The calculated temperature rise for
individual phases of the reactor at steady
state condition is given below:
The absolute temperature rise ∆θw was 80K
for R-phase, 89 K for Y-phase and 78K for
B phase against the specified value of 125
K for class H insulation.
The reactor successfully withstood the
temperature rise test ensuring satisfactory
thermal and electrical performance of the
reactor.
5. Analysis and Conclusion
The first reactor stabilized after nine hours
and withstood the test successfully as the
temperature rise was within the specified
limits and no deformity was observed during
temperature rise test. However the rise in
temperature was close to the specified limits
indicating very little scope for overloads.
The second reactor stabilized after eight
hours and withstood the test successfully as
the temperature rise was within the
specified limits and no deformity was
observed during temperature rise test. The
margin of rise in temperature was
significantly below the specified limits.
Both the reactors tested stabilized after
eight to nine hours for stabilisation of the
first phase. In both the reactors tested the
middle phase exhibited the maximum
temperature rise.
Power Capacitors Laboratory of CPRI has
necessary facilities to carry out the
temperature rise test which is one of the
important type tests on iron core detuned
reactors up to 3 phase 200 kvar of different
tuning order from 2.7 to 4.2 at power
frequency of 50 Hz to check the behaviour
of the reactor under the combined electrical
and thermal stresses.
6. Acknowledgement
Authors sincerely thank the management of
CPRI for according permission to publish
the paper.
References
[1] IEC 60076-6-2007 – Power transformers-
Reactors
[2] IS 5553 (part-5)- 1989 Tuning Reactors –
[3] IEC 60076-11-2004 – Dry Type
Transformers-
[4] IEC 60076-2-2011 – Power transformers-
Temperature rise Test on oil immersed
power transformers.
 INVERTER DUTY TRANSFORMER APPLICATION IN GRID-TIED
SOLAR PHOTOVOLTAIC POWER GENERATION PLANT IN INDIA

M. Chandra Sekhar, GovindaRao G, S. Sudhakara Reddy, Anupam Awasthi

Central Power Research Institute

Abstract: Due to climatic change concerns, the future additions to the power generation
capacity in the world will be predominantly renewable energy. Renewable energy generation
major contribution from solar. The special operating and performance requirements of
Inverter duty transformer in Grid-tied solar photovoltaic power generation plant call for
different design and analysis techniques. This paper covers the design requirements of solar
inverter duty transformer as per IEEE Std. C57.159 and also High Power Laboratory short
circuit testing experience.

Key words: Inverter duty transformer, Distributed Photovoltaic (DPV),

1. INTRODUCTION:

The climate change across the global, the more priority is given to Renewable energy
generation. India is aiming the total target of 175 GW of renewable energy by 2022 that
India announced at Paris. The renewable energy sector to meet the targets indicated below
by 2022:

1. Solar (utility-scale, distributed, off-grid/mini-grid – 100 GW)

2. Wind (utility-scale – 60 GW)
3. Small hydro (5 GW)
4. Bioenergy (10 GW)

In order to handle huge potential of solar power, inverter duty transformer will plays an
important role and it will transfer power to MV/HV power transmission network. The
transformer industry is rising to this challenge and has developed special transformers for the
DPV generation. This paper focuses on the design challenges of these transformer and
growth of transformer technology in India.

2. TYPICAL GRID TIED SOLAR PV POWER GENERATION PLANT AND

INVERTER DUTYTRANSFORMERS:

A single line configuration of grid tied solar PV power generation plant is shown in fig.(1),
the major components are Solar panels, String combiner box, Solar inverter, LV switchgear,
Inverter duty transformer, MV switchgear.
 Fig. (1): A single line configuration of grid tied solar PV power generation plant

In the case of photovoltaic solar power generation plant, electrical poweris generated by
converting solar radiation into direct current electricity using semiconductors that exhibit the
photovoltaic effect. Photovoltaic power generation employs solar panels made up of a
number of cells containing photovoltaic material. The output of solar panels DC voltage is
then converted to three-phase AC voltage using inverter. The inverter is subsequently
connected to a DPV system inverter transformer. This DPV system inverter transformer is
further connected to a collector bus, which can feed a suitable load on the distribution system
or will be connected to the substation collector transformer at the utility substation that
combines the power output of the multiple distributed sources of generation and steps the
voltage up to the local load, distribution, or transmission level.

The inverter duty transformer can work as a step-up transformer and step-down transformer
depending upon flow of energy. In daytime as long as sun light exists the inverter
transformer can work as a step-up transformer. At night it works as a step-down transformer
feeding an auxiliary transformer from the medium voltage grid, the flow of energy in the
inverter transformer is reversed in this case compared to the daytime PV power generation
process. For economic reasons, associated capital expenditures and limited space
requirements, the inverter duty transformers are designed with multi-winding. The existing
inverter duty transformer designsare like three winding, four winding and five winding.

Existing multi-winding inverter duty transformer

Three
winding Four Five winding Five winding
inverter winding inverter duty inverter duty
duty inverter duty transformer transformer with 2
transformer transformer core coil assembly
3. MAJOR DIFFERENCES BETWEEN POWER TRANSFORMER AND
INVERTER DUTY TRANSFORMER
Based on the application and operation of inverter duty transformer, the following are major
differences in Inverter duty transformer over conventional power transformer;
a. It works as step-up transformer and step-down transformer depending upon flow of
energy. In daytime as long as sun light exists the inverter transformer can work as a
step-up transformer. At night it works as a step-down transformer feeding an
auxiliary transformer from the medium voltage grid. Hence, the transformer remains
energized always, the no-load losses are present 24 h per day.
b. The output of solar inverters is fed to LV windings. The solar Inverters of the PV
system convert DC to AC through a semiconductor switching circuit, but the AC
wave obtained from these devices will not be a perfect sinusoidal wave. The AC
waveform of voltage to ground has a pulsed nature; some harmonics and spikes are
also present. Hence, the inverter transformer needs to be designed to work with the
pulse voltage shape of the inverter.
c. The solar irradiation and ambient temperatureimpacts the inverter output ac power.
The inverter transformer is loaded according to the inverter output ac power. During
the peak time of the annual solar activity, the transformer may be fully loaded for 6
hours perday, but it can be unloaded for close to 12 hours when only no-load losses
are present. The day-time fluctuation of the load in conjunction with the unloading of
the transformer at nighttime can cause a considerable thermal and mechanical force
cycling and may adversely affect thetransformer.
d. Inverter transformers can experience a high current inrush when energized, i.e. when
it is working in step-up duty.

4. KEY INPUTS FOR INVERTER DUTY TRANSFORMER DESIGN:

In addition to the standard transformer parameters, the following characteristics may be

required to be specified for the Inverter duty transformer design:
a) Rated output KVA:
The transformer is rated based on the inverter output, load diagram, and specific
consideration (harmonic content, weather patterns, unusual service conditions, etc.)
and is suitable for the operation with the expected continuous maximum required
load. Additional spare capacity for future load growth may also be considered if
necessary information is specified. It is recommended to rate inverter transformer at
least equal to an expected maximum of the inverter output rating.
b) Voltage regulation:
The output voltage of a transformer can be adjusted by typically, a de-energized tap
changer on the secondary (HV) winding of the inverter transformer in order to match
its secondary (HV) winding voltage level with the grid voltage. A de-energized tap
changer (DETC) with a range of +/- 5% of the HV winding nominal voltage is
recommended.
c) Voltage class and test voltages, including LV impulse test, if any.
d) Current harmonic content needs to be specified or reference needs to be made to
IEEE Std C57.12.00 or IEEE Std C57.12.01.
e) Winding configurations and Impedance requirements
Generally, solar inverter duty transformer is a multi-winding transformer. It is
important to have a close coupling and equal impedances of each LV winding to HV
winding in all configurations of the windings and also it is recommended to have high
enough impedance between two LV windings in order to decouple these windings.
Based on the number of windings and Impedance requirements, the designer has
several possible winding configurations. For example, difference possible winding
configurations of the three winding transformer as shown in Fig(2).

Fig. (2): Possible winding configurations of the three winding transformer

f) Losses, eddy current and stray losses

The total load losses are calculated based on the maximum load profile losses and
losses contributed by harmonics.
Eddy currents and stray losses are present in every transformer. The majority of stray
and eddy losses are dueto the fundamental frequency currents, however, the eddy and
stray losses can increase in the presence of thesignificant harmonic content. If the
inverter feeding the power into the inverter transformer is producing morethan a
standard level of harmonics, the stray and eddy losses will increase. The effect of
these increased loadlosses on the efficiency is not typically a concern. Of much
greater concern is a possible increase in the windings’ hot spot temperature and hot
spots in metallic parts. The increased hot spots can reduce the transformerlife.
Localized extreme overheating can result in a catastrophic transformer failure.
Increasing the transformerpower rating reduces the load losses at operating current
and may compensate for the higher operating temperatures that occur in this situation.
Special attention is needed to the eddy and stray losses in multi-winding
configurations in case of any appreciable difference in the load between LV windings.
 The stray losses in the electrostatic shield can lead to thetemperature increase above
insulation thermal class limits.

g) Magnetizing inrush current

Inverter transformers can experience a high current inrush when energized, because
inverter transformers can perform step-up or step-down duty, both cases require
consideration of expected inrush currents. If the source winding is the inner winding
of a transformer, located next to the core, the inrush current will be higher than if it is
the outer winding of the transformer.

h) Electrostatic shield
In inverter transformers of the solar power generation systems, it is recommended to
design an electrostatic ground shield between the primary and secondary windings.
The user needs to specify the requirements for the shield. Typically, the winding
connected to the inverter circuit is ungrounded. The electrostatic ground shield
between the primary and secondary windings intended to prevent capacitive coupling
of these windings. The main advantage is that the electrostatic ground shield
minimizes possible transfer of the high frequency voltage disturbances (harmonics,
pulsations, surges that are created in the voltage inverting process) from the primary
(LV) winding to the secondary (HV) winding and the power system. The other
advantage of the electrostatic ground shield is to reduce a transfer of the high voltage
transients (overvoltage) to the primary (LV) windings and the inverter system
connected to the primary winding. Without the electrostatic ground shield, a very
high percentage of the high voltage transients on the secondary (HV) side may
transfer to the primary (LV) side of the transformer. The electrostatic shield also
filters the voltage gradient of the pulsed primary (LV) voltage.
It is recommended that the design of the electrostatic shield considers the effect of the
eddy losses due to the magnetic field. Either aluminium or copper shields can be
used. Normally, the thinner the shield metal the less are the eddy losses. Higher
conductivity shielding (copper foil) has less eddy losses than lower conductivity
(aluminium strip) material.

i) Short-circuit consideration
For the multi-winding inverter duty transformer, the short-circuit considerations in the
design of the transformer are fairly standard. Faults may occur on either primary (LV)
winding(s) or on secondary (HV) winding. Winding configuration of the multi-
winding transformers affects the magnitude and distribution of the short-circuit
forces. These forces and specifics of the mechanical integrity of various winding
configurations (e.g., stacked windings case with a short-circuit on one of the LV
windings) are important aspects of the reliable design.

Inverter duty transformer short-circuit rating is defined based on the maximum short-
circuit current under all operating conditions and takes into account transformer
voltage regulation.
 The following are possible short-circuit situations in the DPV power generation
system:
1) Short-circuit on the high voltage side of the inverter transformer between the
transformer and a power grid (Fig. 3).

Fig.3—a short-circuit on the power grid side of the transformer

In this case, the main short-circuit current comes from the power grid side of the
system. During theshort-circuit, the current of the inverter is controlled by the
semiconductors, and it is independent ofthe inverter transformer impedance; initial
symmetrical and peak short-circuit currents are equal to orslightly greater (up to
150%) than the inverter nominal current.

2) Short circuit on the low voltage side of the inverter transformer between an
inverter and a transformer (Fig. ).

Fig. 4—a short-circuit on the low voltage side of the transformer

In this case, the main short-circuit current comes from the power grid side and will be
limited with the transformer impedance. The inverter short-circuit current is equal to
the inverter nominal current.
In this case, for the three-winding transformer faults may occur on either one or both
of the input windings, these situations are required to be considered in the design as
per short circuit current.

5. CPRI CONTRIBUTION TO SOLAR POWER PARK

High Power Laboratory, CPRI Bangalore and STDS, CPRI Bhopal conducted special
tests on various ratings of Inverter duty transformer as per National or International
standards for delivering quality product into solar power generation system. The
following Inverter duty transformer ratings are tested at High Power Laboratory,
CPRI Bangalore;
A. Three winding inverter duty transformer
6.25MVA, 33/(0.8-0.8)kV, 50Hz, Three phase inverter duty transformer
5.0MVA, 33/(0.38-0.38)kV, 50Hz, Three phase inverter duty transformer
3.50MVA, 33/(0.4-0.4)kV, 50Hz, Three phase inverter duty transformer
2.0MVA, 33/(0.38-0.38)kV, 50Hz, Three phase inverter duty transformer
 B. Five winding inverter duty transformer
12.5MVA, 33/(0.6-0.6-0.6-0.6)kV, 50Hz, Three phase inverter duty
transformer
10.0MVA, 33/(0.63-0.63-0.63-0.63)kV, 50Hz, Three phase inverter duty
transformer
5.0MVA, 33/(0.35-0.35-0.35-0.35)kV, 50Hz, Three phase inverter duty
transformer

6. CONCLUSIONS:

Solar power will dominate the source of electric power in future. This will reduce the CO2
emissions substantially. The transformers required for solar DPV require special design
features to meet the challenging operating conditions. The industry today has developed
different types of transformers for the application.

REFERENCES:

a) IEEE C.57. 159-2016; “IEEE Guide on Transformers for Application in Distributed

Photovoltaic (DPV) Power Generation Systems”
 Effects of production errors on short circuit testing of transformers

Rajaramamohanarao Chennu, S Sudhakara Reddy, Anupam Awasthi, Gurudev T

Maroti, Arun Kumar S & Sreeram V
Central Power Research Institute, Bangalore

Abstract: Over the last decade there has

been, rightly, a renewed recognition of the Although it is customary to lump these
importance of the human factor in different failings under the catch-all heading
manufacturing. To a large extent, this was of "human error," the distinctions between
due to the advent of new technology and a them are important because different error
growing awareness of latest manufacturing types have different underlying effects on the
techniques. In the case of the former, not short circuit performance evaluation, occur in
only did the new manufacturing technologies different parts of the transformer. The key to
raise questions concerning skill, but also all error management is to target limited
many studies showed that human factors resources on those things most amenable to
were crucial to their successful withstand correction and improvement. In short, it
capability of short circuit. Transformer is one means managing the manageable.
of the most expensive equipment in power
generation station, power transmission
system. The design validation is confirmed
by subjecting a series of tests on the
transformers, which will reduces the outage
of power system and increase the life of the
transformer. In these the effects of human
errors for the failure of the transformers is
very significant.
CPRI is having the decades of
experience in short circuit testing of
transformers. In this paper the short circuit
withstanding forces and its effects on the
transformer is discussed. Different case Fig. 1 Trends in accident causation
studies were presented depicting human
errors for different types of failures on The number of power and distribution
transformers during the short circuit transformers in the power system. These
withstand test. transformers are in service for many years
and should sustain for different
I. INTRODUCTION environmental, electrical and mechanical
stresses (Between conductors, leads and
windings due to over currents or fault
In common with most other complex
currents caused by short circuits and inrush
technical activities, human error is implicated
currents).
in the majority of transformer short circuit
Failure rate depends on some factors.
related quality lapses in the production.
In service winding electro dynamic
General estimates of this human error
deformations after short circuit currents can
contribution have increased over the years,
result in insulation disruption and further to
from a low of around 20% in the 1960s to
turn-to-turn internal short-circuit immediately.
values in excess of 80% in the 1990s. This is
However, in other cases, the insulation
shown in Fig. 1
weakness center can probably appear in the
winding deformation point. This insulation
weakness center can exist in the winding for
few years. And an increase of partial
discharge (PD) intensity, which will result in
insulation disruption, is registered.
This paper outlines the varieties of
human error and some of the ways in which
they can be managed for improving the short
circuit performance of the transformers.
II. SHORT CIRCUIT CURRENTS AND ITS
EFFECTS
The most commonly occurring short
circuit event is a single-line to earth fault in
which one phase is short circuited to ground.
This type of fault may occur due to lightning
strikes, debris, pollution effects, animals and
vegetation. Other types of faults such as
three-phase to ground, double earth fault are
also considered in short circuit current
calculations. In general the short circuit
current is calculated using symmetrical
components for various situations taking into
account:
 Tapping arrangement
 Fault position (e.g. low voltage or high
voltage winding)
 Short circuit power combination
(network and transformer)
 Short circuit type (e.g. single phase to
earth, three phase symmetrical, etc.)
Fig. 2 Short Circuit current
𝑡
𝑖(𝑡) = 𝐼𝑚 [sin(𝜔𝑡 − 𝛼 − ∅) − 𝑒 𝜏 sin(𝛼 − ∅)]
I(t) = instantaneous short circuit current
Im = peak short circuit current
ω = angular frequency in rad/s
α = voltage angle at which short circuit
occurs
Φ = impedance phase angle
The short circuit current in above
equation consists of two components: a
steady state component at power frequency,
and an exponentially decreasing
unidirectional component as shown in Fig. 1
[2].
Fig. 2 shows that the first current peak
is the maximum peak, becoming
progressively smaller as the unidirectional
(green line in Fig. 2) component decays. For
this reason the first peak is usually used to
calculate maximum occurring forces. To aid
the understanding of short circuit current and
its relationship to the power transformer, the
simplest fault scenario, viz. three-phase
symmetrical short circuit fault, is discussed.
This allows the use of an equivalent single
phase circuit model, as all phases remain
balanced and nominal or rated values of the
transformer current can be used. The steady
state component of short circuit current for
this simple case can be expressed as a
multiple of nominal current by the overcurrent
factor, which depends on the transformer
impedance.
To account for the initial direct current
(DC) offset and therefore the first peak, a
further factor called the asymmetry factor "k"
is used [2]. This results in a more accurate
representation of the current to precisely
calculate the peak electromagnetic force.
Values of k are specified in the applicable
standards such as the IEC 60076-5 [3]. The
asymmetry factor is dependent on the ratio of
the transformer reactance (X) and resistance
(R). Values of k increase for increasing X/R
ratios. Values of 1,8 for transformers up to
100 MVA and 1,9 for transformers in excess
of 100 MVA are indicated when the X/R ratio
is unavailable [3].
In the case of the appearance of short
circuit as a result of the internal damage of
transformer windings in service or during
electrodynamic testing for short-circuit
withstand with the artificial short-circuiting of
the outlets of windings, the value of the
greatest steady short-circuit current in the
double winding regime without taking into
account the intermediate network elements
will comprise
𝑈𝑛𝑜𝑚−𝑡𝑟
𝐼𝑠𝑐 =
√3(𝑍𝑡 + 𝑍𝑠 )
Where Unom-tr is nominal voltage of
transformer’s tap changer; then short-circuit
impedance of transformer comprises (U in
kV, S in MVA):
2
%𝑍 ∗ 𝑈𝑛𝑜𝑚−𝑡𝑟
𝑍𝑡 = 100∗𝑆𝑛𝑜𝑚−𝑡𝑟
where; percentage impedance of transformer
(from formular of transformer); Snom-tr nominal
capacity of transformer, then short-circuit
impedance of network will be:
𝑈𝑆2
𝑍𝑠 =
𝑆𝑆
Where: US is the highest system voltage and
Ss is the short circuit capacity of the system,
which is determined by the network capacity
Effects of short circuit currents:
During short-circuit, the copper wires
of transformer winding were deformed under
the influence of electromagnetic forces by
Biot-Savart’s law, which will be in differential
form [10-15]:
𝑑𝑓 = [𝐵. 𝑗)𝑑𝑣
where: df – force’s vector, which is
influenced on the element of current with
volume dv in magnetic field with magnetic
induction В and current density j. The vector
product in the right part of equality showed
that electromagnetic force is perpendicular to
the direction of magnetic induction and
current density (by left-hand rule).
Fig.3 Current and induced forces waveforms
The force, which acts on the winding
or its part, can be calculated by integrating
the above equation:
𝐹 = ∫ [𝐵. 𝑗]𝑑𝑉
𝑉
If В and j are perpendicular to each other and
they are constant throughout entire volume,
then electromagnetic force, which has
influence on the element of current in
magnetic field, is:
𝐹 = 𝐵·𝑙·𝑖
where l–length of wire or winding; i–value of
current into wire or winding; B–value of
magnetic induction.
Fig. 4 directional relationship between fields
Short circuit current will influence both
flux density B and current density J the
above equation implying that force is
proportional to the square of the current. Fig.
3 shows the relationship between current
and force and the nature of electromagnetic
forces. The electromagnetic force pulsates at
approximately twice the power frequency and
is unidirectional. Considering that the
overcurrent factor is typically between 8 and
10, the force generated can easily be in the
region of a hundred times the nominal force
magnitude and may reach several thousand
kN [2].
The direction of the forces are
stipulated by the vector product indicating
that the force will act perpendicular to the
plane formed by the magnetic field intensity
and current direction. The force direction can
be determined using the left hand rule, as
indicated in Fig. 4 which describes the field
relationships for the cross-section of a single
conductor. Electromagnetic force tends to
minimise the magnetic energy density in the
volume. In this way the forces tend to:
 Reduce the radius of inner windings
 Increase the radius of outer windings
 Reduce the height of windings –
windings are compressed toward the
median horizontal
Fig. 5 Magnetic fields and associated field
directions
Viewing the transformer from a cross
sectional view (Fig. 5), and taking into
account the direction of the current as being
constant, it is noted that the force is
perpendicular to, and follows the bending of
the flux lines. In the middle of the windings
this results in a radial force pushing outward,
and toward the ends of the windings this
results in an axial force pushing onto the
windings.
Electromagnetic forces in transformer
windings can be sub-divided into axial and
radial force by means of associated axial and
radial modes of failure. Axial forces occur in
a direction parallel to the winding height.
Radial forces occur perpendicular to the
winding height. Axial and radial forces,
although sharing a common origin, can for
the most part, be treated as mutually
exclusive modes [2].
Radial forces:
The flux at mid winding height is for
the most part parallel to the winding height.
According to the left hand rule, the resulting
force therefore acts perpendicular to the
winding height. For windings on the inside of
the main flux field (situated between primary
and secondary windings) the force acts
inward, and for outside windings, outward.
Forces acting on the inside windings
result in a compressive stress whereas on
the outer winding this force leads to a tensile
stress acting to elongate the winding turn –
as shown in Fig. 6 [2]. This is as a result of
the cylindrical profile of the windings where
forces, perpendicular to the winding
circumference, create tangential stresses.
Fig. 6 Conductor force directions and
associated stress
There radial forces resulted in the
forced buckling, free buckling and hoop
forces as shown in fig. 7
Forded buckling Free buckling
Fig. 7 Effects of radial forces
Axial forces:
Axial forces are generated parallel
with the winding height. It results in
conductor tilting and collapse of end
supporting structure. Due to the pattern of
the magnetic leakage field, the windings
experience opposing forces at the winding
ends, leading to compressive forces. Axial
force is generated when the transformer's
magnetic field lines are radially orientated. point of missing ampere-turns or where the
Fig. 4. shows that the highest ‘bending' of the displacement occurs. The forces then work
magnetic field occurs at the winding ends, to increase the displacement i.e. windings
consequently maximum axial force is are pushed further apart, resulting in
generated here. The local force generation iteratively worse displacements and therefore
accumulates toward the middle of the higher forces [2].
windings resulting in a maximum occurring
force at mid-winding height. This force is
compressive towards the winding centre as
the circulation direction of the magnetic field
leads to opposing forces at each end of the
winding. A schematic representation of the
axial force distribution, in a typical
transformer winding, is shown in Fig. 8.

Fig. 9 Winding displacements and induced

forces
III CASE STUDIES OF TRANSFORMER
FAILURES
The impact of the human errors on the
short circuit test on the transformer is
discussed. Some important case studies of
Fig. 8 Cumulative and local force distribution short circuit evaluations carried out in CPRI
along winding height is presented as case studies indicating the
In addition to the compressive force human errors.
occurring in the winding the effect of recoil is Case Study 1: when a 300 kVA, 33kV class
also considered as transformer windings are transformer was tested for dynamic short
essentially springs from a mechanical point circuit withstand test by three phase.
of view. This spring action can be explained
by noting the nature of the electromagnetic
wave in Fig. 2. The force magnitude follows a
sinusoidal pattern and therefore compressive
tension stored in the winding is released as
the force approaches zero. This force will be
exerted against the core yokes and end
insulation structures, and will have a
magnitude of less than the compressive
force generating it.

Winding displacement from the centre

line result in excessive axial forces, and
provision for this possibility is necessary –
shown in Fig. 9. This is due to the difference Fig. 10: Improper crimping.
in the ampere-turns (i.e. mmf) distribution
along the winding heights resulting in more During the short circuit test the PRV
pronounced bending of the field lines at the operated in the second shot. During physical
inspection it is found that the crimping was
not properly done on the neutral side of B-
phase coil. This is clearly shown in the Fig.10
that, the other two crimping joints were found
in good condition, even though the peak
current is not seen in this phase the joint
found open, which clearly indicates the
crimping was not done properly during the
production stage.
Case Study 2: 4MVA, 33kV class transformer
cleared all the 9 shots during the dynamic
short circuit and cleared the entire routine
test post short circuit test. All the test results
found satisfactory. During the physical
inspection the neutral side busbar found bent
and one of the support nut found broken,
which is shown in Fig. 11.
(a)
(b)
Fig. 11 Effect of Improper support
The through verification from Fig 11
(b) it is clearly evident that the production
engineer is not arrested the busbar with nut
bolts in one place, instead tied with cable
ties. During the short circuit test, when the
current is seen by the transformer the cable
tie unable to attest the movement of busbar,
resulted in to the bus bar deformation and
breakage of the nearby nut. This clearly
indicates the complete negligence or a
human error during the production stage.
Case Study 3: During the short circuit testing
of 8MVA, 33kV transformer, the transformer
mineral oil found leaked through a welding
joint. The reactance variation is found
normal. This is most severe for the
transformer life. The condition of the leakage
is shown in Fig. 12.
It clearly indicates the human
mistakes during the welding of the
transformer tank and no proper quality
checks were done on this transformer during
the production stage.
Fig. 12 leakage through tank
Case Study 4: During the physical inspection
after dynamic short circuit testing of 3MVA,
33kV class transformer, the end blocks found
inside the transformer. During the test all the
parameter are found normal. The reason for
same is due to placing the lead near to the
end blocks.
As the lead is placed next to the end
blocks, during the short circuit test the lead of
the transformer came in contact with the end
block and dislocated the end block. This the
clearly observed in the Fig 13
Fig: 13 Improper position of terminal lead
IV. CONCLUSION
The effects of human error during
short circuit testing of the transformer are
clearly discussed in this paper. A very
thorough inspection and workmen skill are
required for the successful performance of
the transformers
Some very minor human errors effect
on the transformer short circuit test was
reported in the case study format. Short
circuit test on the transformer is a very
important and critical test even small mistake
in assembly will affect the results. So very at
most care should be taken in every stage of
the manufacturing.
REFERENCES
[1]. D K Chen: Field and Wave Electro-
magnetics, Addison-Wesley Publishing
Company Inc., USA, second edition,
November 1992.
[2]. G Bertagnolli: Short-circuit Duty of Power
Transformers, ABB Ltd., Zurich, third edition,
June 2006.
[3]. IEC standard 60076-5: 2006 Power
Transformers – Part 5: “Ability to withstand
short circuit”.
About The Authors:
Mr. Rajaramamohanarao chennu is
currently working at High Power lab of
Central Power Research Institute,
Bangalore. He is M.Tech from IIT Roorkee.
His areas of Interests are Power Quality,
Power electronics, power system stability
and High power testing of Electrical
Equipments.
Mr .S.Sudhakara Reddy is Joint Director
at High Power Laboratory, CPRI Bangalore.
He is M.Tech. in Heavy Electrical machine
design and PGDM in Energy Management.
He has 21 years of experience in testing and
development of Switchgear, research and
consultancy. Member of ET16-
Transformers and Member of institution of
Engineers
Mr. Anupam Awasthi is Addtional
Director and Head of the Division at High
Power Laboratory, CPRI Bangalore. He is
graduate in Electrical Engineering from S.
V. Regional College of Engineering and
Technology, Surat in the year 1982 and
completed Post Graduation in 1989 in
engineering materials from MACT, Bhopal.
He started his career with CPRI, Bhopal in
the year 1983 where as he was associated in
short circuit testing of LT Electrical
equipments as per national and international
[4].
standards.
Mr. T Gurudev is M Tech in electrical
Engineering with 17 years of experience in
Short circuit Testing of High power electrical
equipments such as power Transformers, HV
Circuit breakers, Generator Busducts,
Insulators, Line traps, Reactors, LT & HT
Panels, Current transformers & HV switches
Current transformers etc.,. Presented few
papers in International conferences organized
by IEEMA and National conferences organized
by CPRI. Third party Pre-dispatch Inspection
& witnessing of Routine tests, Special tests and
Type tests up to 500MVA,765KV
transformers and 200MVA, 400kV, 3-phase
power transformers .
Mr. Maroti is presently working as Test
Engineer at High Power Laboratory CPRI
Bangalore. He has professional experience
of 16 years in the field of short Circuit
Testing of Switchgear and Powergear
equipments . Participated as independent
Third Party Inspector for for electrical
power utilities in India & Abroad.
Graduated in Electronics & Communication
Engineering from the University of
Visvesvaraya College of Engineering,
Bangalore.
Mr. Arun Kumar obtained his M Tech
degree in Electrical and Electronics
Engineering from Anna University, Chennai.
He worked for CGL before joining CPRI, at
High Power Laboratory.

Mr. V.Sreeram obtained his BTech degree in

Electrical and Electronics Engineering from
NIT Calicut in the year 2010. He worked for
NTPC Ltd for two years. He obtained his ME
degree in Electrical Engineering from the
Indian Institute of Science in 2014. His
research interests include electromagnetics,
electrical machines and power systems. He is
currently working at High Power Laboratory at
CPRI.
 SFRA – Interpretation methods and Standards
Sreeram V, S Sudhakara Reddy, Maroti, Rajaramamohanarao Chennu, Arunkumar S, Rajkumar M, T Gurudev and A
Awasthi
High Power Laboratory
Central Power Research Institute, Bangalore.
Abstract—SFRA (Sweep Frequency Response Analysis) has
been used for condition monitoring of transformers for several
decades. From the time it was introduced, the measurement
practices and evaluation techniques have evolved to a point of
much sophistication. However SFRA has remained a tool for
indication in condition monitoring and has so far not been
successful as an objective evaluation criterion in any context.
This paper scans the evolution of SFRA and the associated
developments in interpretation methodologies.
Keywords — Frequency response, Transformers.
I. INTRODUCTION
Dick and Erven introduced the technique of sweep
frequency response analysis in the year 1978 [1]. The SFRA
method utilizes a variable frequency sinusoidal source of an
amplitude of 20 Vpp usually to extract the transfer function of
the transformer with different combinations of input and
output terminals. The magnitude plot reveals the ratio of the
magnitudes of output and input signals while the phase plot
reveals the phase difference between the output and input
signals. In addition to the transfer function approach, the ratio
of input voltage to input current is considered in the driving
point approach. However the present industry standard is to
use the transfer function approach. Similarly a low voltage
impulse response measurement had been in use to assess the
impulse response in time domain but now has yielded ground
to SFRA.
SFRA has been used in both offline and online modes,
with the online mode offering the most utility but presenting
huge signal processing and hardware design challenges. The
online mode usually prefers the low voltage impulse response
options.
II. THEORY OF TRANSFORMER RESPONSE
The transfer function of a physical system can be
represented as a rational function in the form of polynomials.
Vout
T.F. = H (ω) =
Vin
The transfer function is characterized by the internal
constituent elements of the system. The transformer can be
regarded as a complex combination of lumped circuit
elements in the low and mid frequency ranges. However in the
high frequency range, the response can be determined by a full
wave solution of the Maxwell’s equations. The experimentally
determined transfer function will be indicative of the internal
circuit elements. The transformer is a highly complex
electromagnetic system especially at high frequencies that its
behaviour is not entirely understood and an accurate analytical
formulation is almost impossible. A complete physical model
does not even exist for a single winding in air. In an actual
transformer, there will be inductive and capacitive coupling
with other windings and with the tank. The lumped parameters
are determined by the core magnetic characteristics, insulation
material, geometry of the constituent elements etc. Any
change in the geometry of the transformer will lead to changes
in the frequency response. While the changes in the values of
the parameters can be assessed by modelling studies, circuit
simulation will give the frequency response of the circuit at
hand. However analysis is complicated by the fact that the
changes in parameter values cannot be accurately predicted
and many unrelated changes can lead to same changes in the
frequency response. A model SFRA test circuit is shown in
Fig 1.
Fig 1. Model SFRA test circuit
The initial portion of the frequency response shows a
decreasing trend with frequency (below 2 kHz). This region is
dominated by the core inductance or leakage inductance
depending on the measurement configuration. The core
inductance is influenced by residual magnetism in the core
which is determined by previous tests as well as the phase of
voltage during interruption. There will be resonances in the
region caused by the magnetising inductance and the winding
to earth capacitances. The next region is dominated by the
interaction between windings (between 2 kHz and 20 kHz).
From 20 kHz to 1 MHz, the individual winding structure
determines the response. Interleaved disc-type winding
structure has an increasing response with little resonance
while continuous disc-type winding has steady response with
higher resonance. Above 1 MHz, the influence of radiated
noise becomes significant and hence the measurement set-up
becomes critical. The high frequency region is therefore
considered unreliable for diagnostics of transformer failure.
Different frequency regions are indicated in Fig 2.
III. METHODS OF INTERPRETATION
Generally there are four different methods for SFRA
interpretation as used in present day literature.
 (2) Correlation Coefficient
 XY
 X Y 2 2

(3) Standard Deviation

 Y  X 
2

N 1

Fig 2. Frequency Regions in SFRA

(4) Absolute Sum of Logarithmic Error
A. Methods based on circuit modelling
An RLC representation of the transformer is formed using
knowledge of the internal geometry or from terminal
| 20log Y  20log X |
measurements. A single circuit may not be sufficient to N
represent the transformer in all frequency ranges. It is
common to employ two or three different circuits to cover the
entire frequency range. The frequency response of the model (5) Sum of Squared Error

is obtained and matched with the measured SFRA. When the
second SFRA measurement is obtained, the changes are noted
and the possible changes in the circuit are predicted. From the
changes in the values of the parameters, the mechanical and
dielectric changes in the transformer are predicted. A variety
of methods belonging to this class has been described in the
literature.
B. Methods based on artificial intelligence
These methods generally utilize neural networks, genetic
algorithm and fuzzy logic for interpretation of SFRA traces.
These methods use AI techniques for two purposes – (i)
developing a model from the measurement data and (ii)
correlating faults with the observed variation in SFRA data.
While using neural networks, it is possible to use both
supervised and unsupervised learning. In supervised learning,
numerical indices of variation in SFRA data can be used for
training while in unsupervised learning the SFRA data is fed
as such for training
C. Methods based on exact calculation
These methods are based on calculation of various
numerical indices from the two SFRA traces to quantify the
difference between the traces. The numerical indices are
calculated not assuming any specialist knowledge regarding
the type of data. For example, resonant and anti-resonant
peaks are treated the same and all frequency regions are
treated the same. These techniques use statistical analysis of
the traces to detect possible faults. The threshold values for
the calculated indices have to be determined from experience
or expert knowledge. There are indices based on resonant and
anti-resonant frequencies as well. Recently non-parametric
statistical methods have been proposed as well. The common
indices which are used are defined below:
(1) Euclidean Distance
||X-Y||
 (Y  X )
2
N
Even though all these indices use the magnitude plot data,
recently phase plot data has also been added to increase the
efficiency in predicting faults.
D. Methods based on estimation
These methods are based on the estimation of a
rational function from the SFRA traces. The SFRA function
can be represented as
Y ( )
F ( ) 
X ( )
The important characteristics of a rational function are its
poles and zeros. Indices can be formed based on the shift in
the values of poles and zeros. The rational functions have to
be estimated using mathematical techniques and the common
method is known as “vector fitting”. Standardisation of the
estimation technique is crucial since the procedure has the
potential to cause massive differences in the estimated poles
and zeros.
IV. INTERNATIONAL STANDARDS
A. IEC
IEC 60076-18 has been the first standard to be published in
the area SFRA measurement and interpretation. IEC standard
was published in the year 2012. It borrowed heavily from
CIGRE technical brochure. The standard focuses on the
underlying theory and the measurement configurations rather
than on methods of interpretation. The standard gives the
requirements regarding the measuring instruments and the
precautions regarding measurement procedures. The standard
concludes by giving examples of variation in SFRA traces
due to common faults and false practices.
B. CIGRE
characteristics and leakage impedance will come into
CIGRE constituted study committee SC A2.26 to bring about picture.
the first attempt at standardization in SFRA measurement. A
technical brochure (reference number 342) was brought out
which serves a guide for mechanical condition assessment of
transformer windings. The TB also presents typical cases
where the SFRA traces differ due to common known faults.
The TB introduced four types of measurements

 End-to-end measurement: The signal is applied to

one end of a winding and the measurement is carried
out at the other end.

Fig 5. Capacitive inter-winding measurement

Fig 3. End-to-end measurement

 End-to-end short circuit measurement: The signal is

applied to one end of the winding and the
measurement is carried out at the other end with one
of the other windings short circuited.

Fig 6. Inductive inter-winding measurement

These four types have been adopted in the IEEE as well as

IEC standards. CIGRE has now tasked a new working group
WG A2.53 for developing an objective interpretation
criterion that commonly applied and accepted by all and can
be used as a pass or fail criterion in short circuit tests etc. WG
A2.53 is expected to submit its report in December 2019.

Fig 4. End-to-end short circuit measurement
 Capacitive inter-winding measurement: The signal
is applied to one terminal of a winding and
measurement is carried out at one terminal of a
different winding with the other ends of both the
windings kept open.
 Inductive inter-winding measurement: The signal is
applied to one terminal of a winding and
measurement is carried out at one terminal of a
different winding with the other ends of both the
windings kept shorted. This differs from the
capacitive measurements in the fact that the core
C. IEEE
IEEE standard C57.149 was also published in the year 2012.
The standard applies to frequency response interpretation in
oil immersed transformers. The method and type of
measurements are as per the CIGRE guidelines. Standard
measurement configurations are suggested. For common
categories of faults, example traces are given. Frequency
ranges have been identified where maximum variation is
expected in each category of fault. For example, in case of
axial deformation, maximum variation is expected in the
range 5 kHz to 100 kHz. In the case of radial deformation,
maximum variation is expected in the range 50 kHz to 1
MHz.
 V. COMMON SOURCES OF UNCERTAINTIES
One of the main reasons for the absence of objective
interpretation criterion in SFRA has been the presence of
varied sources of uncertainties in the measurement. The
contribution of sources cannot be easily separated out to
understand the real behaviour of the transformer. Most often
variations due to unrelated reasons are misunderstood as
being caused by faults within the transformer. Some of the
common causes of uncertainties in the measurement are:
 Variation in the connection of test leads
 Improper grounding of bushing flanges
 Improper core grounding connection
 Improper short circuiting jumpers
 Variation in the connections of windings which are
not under test
 Variation in temperature
VI. CONCLUSION
SFRA measurements have been used for condition
assessment for transformers for quite a long while. However
an objective interpretation criterion remains elusive. This
paper presents a brief summary of the theory and
measurement practices in SFRA. The paper also summarises
various methods for the interpretation of SFRA and the
common standards involved. The sources of uncertainty in
the measurements have also been presented.
REFERENCES
[1] E. P. Dick and C. C. Erven, "Transformer diagnostic testing by
frequency response analysis," IEEE Transactions on Power Apparatus
and Systems, vol. PAS-97, pp. 2144-2153, 1978.
[2] Patrick Picher, Stefan Tenbohlen, Mark Lachman, Alaor Scardazzi,
Poorvi Patel, “Current state of transformer FRA interpretation: On
behalf of CIGRE WG A2.53”, pp 3-12, Volume 202, Procedia
engineering,
[3] Mohammad Hamed Samimi and StefanTenbohlen, “FRA
interpretation using numerical indices: State-of-the-art,” pp 115-125,
Electrical Power and Energy Systems 89 (2017).
[4] "Measurement of frequency response," IEC 60076-18, 2012.
[5] "IEEE Guide for the Application and Interpretation of Frequency
Response Analysis for Oil-Immersed Transformers," IEEE Std
C57.149-2012, pp. 1-72, 2013.
[6] CIGRE Working Group A2.26, "Mechanical-Condition Assessment of
Transformer Windings Using Frequency Response Analysis (FRA),"
Brochure 342, 2008.
 A Case study of OLTC failures
Arunkumar S, Anupam Awasthi, Sudhakara Reddy, Gurudev T, Maroti, Rajaram, Rajkumar, Sreeram V
HPL, CPRI
Bangalore, India

film layer will not cause tap changer failures directly. The
I. ABSTRACT increased contact resistance due to the oil film layer will cause
On load tap changer [OLTC] is one of the important and coking (creation of hard and porous carbon material) at places
costliest components in the power transformer. OLTC’s are where the load current flows. The long term effect is
used to keep the secondary voltage at an acceptable level accelerated by high temperatures, high load current, infrequent
when the load changes or to adjust the transformer phase shift movement, low contact pressure and develops faster at copper
The OLTC’s failure can cause catastrophic failure of in- and brass contacts. Due to its infrequent movement, the
service power transformer. Most power transformer studies change-over selector is prone to the long term
indicated that the main cause of power transformers failures is effect.
the OLTC since it has mechanical parts that are in constant
motion. This paper discusses about the various failures of
OLTC and a case study during short circuit test conducted at
high power lab, CPRI.
II. INTRODUCTON
In most of OLTC’s failure cases, it has caused prolonged
power interruption, often due to severity of damage and timely
replacement, as a result of unavailability of spares and Figure 1 presents a formation of electrical treeing on the surface of the epoxy
incompatibility with the existing design. The root causes of cylinder
failures are often attributed to degradation and carbonization
of oil as well as contact problems
During normal tap change operation, switching arcs occur
in oil due to the making and breaking of currents. These arcs
cause degradation and carbonization that contaminate and
reduce the dielectric strength of the oil. The carbonization
over time causes accumulation of carbon deposits on the
surface of the fiberglass cylinder that result in the formation of
electrical treeing. Figure 2 presents a erosion and pitted marks on the roller contacts (left) and
Without proper maintenance, the electrical treeing develops carbon build up on the fixed contacts (Right)
into tracking and finally phase-to-phase arcing occurs in
OLTC. This leads to the permanent damage of the OLTC. Degradation of oil in the OLTC is not only influenced by the
Coking or low conductivity film buildup on fixed and roller switching arcs that occur due to making and braking of
contacts of OLTC. This film consists of layer of carbon currents during on-load tap change operation, but also affected
formation that bond to the oxide layer, which was formed as by the presence of moisture in the oil. Even with the use of
the result of surface oxidation of the contacts. Sometimes, vacuum switch type OLTC, moisture can still be presence due
coking has led to erosion or contact wear and pitted marks on to leakage or in most cases due to lack of maintenance of the
the contacts. This occurs as the result of continuous OLTC dehydrating breather
overheating, as coking worsen over time due to increased
contact resistance. Prolonged condition of such phenomenon
may result in excessive arcing that could trip the transformer III. CONDITION ASSESSMENT
protection. OLTC health condition is assessed through on site and off site
The long term effect on the OLTC is aging on the change over diagnostic technique and initiate maintenance tasks if
selector, which occurs when the tap changer is motionless. It abnormal results found from measured results. Following are
results in formation of thin layer on oil. This organic film is a the some of the diagnostic technique to check condition of the
less conductive layer build from polymerized oil organic OLTC.
components in the transformer oil bond to silver or copper
oxide and sulphide that is formed on the stator blocks. This oil
 I. DISSOLVED GAS ANALYSIS (DGA) TO DETECT
COKING OF CONTACTS
Condition assessment method using DGA was formulated
to interpret the health condition of the OLTC. The statistical
model used follows the method to determine two types of
condition indictor limits i.e. gas concentration and gas ratio
limits[1]. The gas concentration limits is defined in terms of a
statistical outlier limit identifying extreme values of gas
concentrationsuspected to be the results of faults or unusual
stresses. The gases that are used as indicators to discriminate
between normal and faulty conditions are C2H2, C2H4 and
CH4 together with the sum of CH4, C2H6 and C2H4 called
the total dissolved heating gases. The upper outlier limits U1, Table3. Interpretation of DGA results for OLTC
U2 and U3 were calculated for each set of gas concentration
data. U3 is introduced to represent the extreme values
requiring the highest attention (U3 =Q3 + 4.5IQR where Q3:
third quartile, IQR: interquartile range). Table 2 summarises
the gas concentration limits for OLTC.
The gas ratios that are used as indicators to
discriminate between normal and faulty conditions are
C2H4/C2H2 and TDHG/C2H2. The gas ratio values of non-
faulty OLTC operating under normal conditions are described
by the percentiles of the lognormal distribution representing Table4. Interpretation of DGA results for OLTC
the non-outlier gas ratios. The gas ratio limits calculated for
the study are based on 90th (C090), 95 th (C095), and 99th Table 4 shows the combustible gas concentrations of
(C099) percentiles of the lognormal distribution with burned contacts of OLTC. An internal inspection of the OLTC
significance level of 0.1, 0.05 and 0.01, respectively. Table 1 found a burned phase B contact. The damaged contact and the
summarises the gas ratio limits for OLTC. insulating fluid were replaced. The sharply reduced but still
considerable levels of combustible gas found in the samples
may represent residual gas, but further surveillance is required
for assurance that the fault is gone.
II. DYNAMIC RESISTACNE MEASUREMENT
Dynamic resistance measurement (DRM) is one of the
modern off-line diagnostic methods that is currently widely
accepted to assess the condition of OLTC during in-service.
This technique able to find defects of OLTC parts not
Table 1. OLTC gas ratio limits
accessible for inspection and of benefits to maintain the
reliability and availability of OLTC, thus power transformer as
Table 3 gives the overall interpretation to facilitate the well.
understanding on the condition of the arcing contacts based on Dynamic resistance measurement performed on power
the results of the gas concentrations and ratios. transformer can provide useful information particularly the
condition of load tap changer compared to static resistance
measurement, also known as DC winding resistance. Several
OLTC defects such as arcing contact wear, contact timing,
damaged of transition resistors and contact coking can be
detected by the method. A DC voltage is used to provide a test
Table2. OLTC gas concentration limits current on both techniques, except that OLTC is operated
throughout its tap position for dynamic measurement. So, any
excessive contact resistance and current interruption will be
detected when the current is circulated through the operation
of OLTC. The recorded current pattern of dynamic resistance
measurement is shown in Figure 3.
Figure 2 presents a current pattern of dynamic resistance measurement
A case study of infrequent operation of certain tap
position is presented. Infrequent operation can cause long term
ageing effect with the formation of thin layer of polymerized
oil at load tap changer contact. This will result in a non-
uniform of current density on the contact interface and thus
creates a local hotspot, which finally increase the contact
resistance and cause coking at area where the load current
flows.
Figure 3 demonstrates the result of field dynamic
resistance measurement on one of load tap changer contact
that having coking problem. It can be seen that the
deteriorated contact has caused collapse of current pattern.
Figure 3: Current pattern of coked contacts
The rapid changes in the operation of the load tap changer can
cause wear and tear of the arcing contact which finally can
changes its transition time. This situation is easily detected by
dynamic resistance measurement. Figure 4 shows the time of
the transition resistor which is affected by the worn contact.
The transition time of the load tap changer was
compared with the fingerprint results and its demonstrated that
arcing contact of tap position 8 was deviated significantly.
The transition time of the load tap changer was
compared with the fingerprint results and its demonstrated that
arcing contact of tap position 8 was deviated significantly.
Figure 4. Transition time of load tap changer having worn contact
at tap position 8.
III Mechanical problems
A case study of mechanical failure of OLTC is
presented. Loosening of bolt is seen in the figure 5. This lead
to arcing in the contacts and subsequently coking is formed on
the contacts. Most of the problems include failure of springs,
bearings, shafts and drive mechanisms. It can be attributed to
drive mechanism motor current, temperature difference
between OLT compartment and main tank due of coking of
contacts and it is due to thermal or dielectric in nature,
mechanism operating time and mechanical vibrations.
Figure 5: Loosening of bolt
One more reason for OLTC failure is improper alignment.
Improper alignment can cause high contact temperature and
ultimately result in failure of the power transformer.
A case study of OLTC failure during short circuit is
presented. Voltage class of transformer is 132kV. Transformer
withstood first short circuit. Second shot after 23ms tank and
ground current observed. Primary current increased and
continuous tank current observed. PRV operated and oil oozed
out from the transformer. Later study reveals problem in
OLTC and there is a dielectric failure happened between oil. Case study presented for OLTC failure during short circuit
OLTC and tank of transformer. condition.

Figure 6: OLTC failure

Another case study of OLTC presented. Voltage class of
transformer is 132kV class. First SC shot cleared by the
transformer. During first shot primary current distortion
observed on the transformer as shown in figure. This could be Figure 8: OLTC failure
due to the inititation of dielectric failure in OLTC.

Figure 9: Loosening of bolt

Figure 7: Primary current distortion

In second shot after 33ms dielectric failure occurred in
OLTC and primary current increased substantially.
Subsequently it leads to collapse of winding and it leads to
tank current and transformer tank bursted as shown in figure 9.
III. CONCLUSION
This paper presented the case study of OLTC failure
during dynamic resistance measurement on in service load tap
changer and also presented a study of dissolved gas analysis of
References
[1] MR OLTC instruction manual.
[2] IEC 66076-5
[3] Jur Erbrink: “OLTC Diagnosis on High Voltage Power
Transformers using Dynamic Resistance Measurements”,
Doctoral thesis, Delft University of Technology, 2011.
 On-line acoustic emission technique for identification of
Relative condition of Generator Transformers
in Hydro Power station- Case studies
Dr. T.Bhavani Shanker, V. Vaidhyanathan, G. Govinda Rao, A. Sheik Mohamed
Central Power Research Institute (CPRI),
P.O. Box. 8066, Prof. Sir. C.V.Raman Road,
Bangalore-560080, India
tbs@cpri.in, vaidhya@cpri.in

Abstract –Generator transformers in hydro power
stations are capital intensive and form a vital link
between generation and transmission of power to the
grid stations of ac power system. Identification of
condition of these transformers relative to the working
of other transformers is very essential and important
in order to avoid unforeseen breakdown. These
unforeseen breakdowns result in interruption of power
supply, thereby causing inconvenience to the end user
and financial loss. Periodic assessment of the
condition of the transformers and also the relative
condition is ascertained by carrying out various tests
by the utilities. There are several tests like off-line and
on-line tests to assess the condition of these
transformers. Except the tests on oil, all other tests are
off-line and require outage of the transformers,
thereby causing disruption of the power supply. Of
several on-line techniques, acoustic emission
technique is gaining focus of most of the researchers
for monitoring and identification of the condition for
defects like partial discharges and hotspots within the
transformers, and also to ascertain the relative
condition by comparing the results with other identical
transformers working in the power station. The
principle of acoustic emission technique along with in-
situ findings of the condition of two groups of single
phase generator transformers of 400kV class (Three
transformers in each group) for defects like partial
discharges and hotspots, in a hydro power station are
discussed in this paper as case studies. Also the
results of condition of these transformers over a period
of 5 years were compared by carrying out the test by
acoustic emission technique during year 2012 and
year 2017 are also presented in this paper.
Index Terms – Condition monitoring, hydro generator
transformers, partial discharge (PD), insulation condition,
acoustic emission (AE) technique.
1. INTRODUCTION
There are various ways of generation of power, be
it thermal (steam turbine), hydro (hydro turbine), gas
(gas turbine) etc., which form prime mover for the
generators (alternators). Hydro generator
transformers are the important and vital components
which are used to transform the generating voltage to
transmission levels of 400kV. Unwarranted and
unforeseen interruptions of these transformers would
lead to major financial losses in terms of generation.
Hence, condition assessment of these transformers at
regular intervals is essential in order to maintain
uninterrupted power to the end user. On-line
technique based on acoustic emission (AE) detection
is being employed as one of the test for condition
assessment of generator transformers. AE technique
has also been utilised by some of the power utilities for
assessing the relative condition of similar equipment
operating under identical conditions
The case studies of on-line testing based on AE
detection for partial discharges and hotspots in six
numbers(Three transformers in each group) of single
phase 400kV class hydro generator transformers
(GT’s) in operation at hydro power station of India, are
presented in this paper . Also the results of condition
of these transformers over a period of 5 years were
compared by carrying out the test by acoustic
emission technique during year 2012 and year 2017,
are also presented.
2. CONDITION MONITORING TECHNIQUES FOR
TRANSFORMERS
Condition of transformers is generally
monitored by various off-line and on-line techniques.
The tests were discussed extensively in IEEE-62-1995
RA 2005 Part 1 [2]. This specification also includes
some of the on-line diagnostic tests such as tests on
insulating oils and Thermal Imaging by Infra-Red
method. The off-line techniques require disconnection
of transformers from the circuit, whereas an on-line
technique does not. Off-line diagnostic tests are
presently being used for insulation condition
monitoring which include Insulation resistance (IR),
Polarization Index (PI), Recovery Voltage
Measurement (RVM) measurement of Capacitance &
tan delta of the insulation [1]. Dissolved gas analysis
(DGA) is one of the one-line tests which are in use
since several years.
In addition to DGA various other on-line diagnostic
techniques are being developed for assessment of
overall condition of transformers by way of periodic
monitoring. Acoustic emission (AE) technique and
Ultra High Frequency (UHF) techniques are some of
them. These techniques are gaining importance and
are being considered as promising area of research.
On-line technique overcomes some of the practical
difficulties for testing of transformers at site as they do
not call for the shutdown /interruption of power supply
and are non-interfering/non-destructive.
Utilities/Industry normally follows certain
maintenance schedules to check the health of the
transformers by taking them out of service apart for
monitoring the condition of the oil at more frequent
intervals. Tests on transformer oil along with DGA
form a primary indicator of condition of the bulk
insulation of transformer. However, tests on oil do not
indicate the sources of the degradation which may
lead to failure of power transformers. Therefore
detection and localisation of sources of degradation in
early stages by way of periodic condition monitoring
would help in avoiding failure of the transformer. On-
line condition monitoring based on acoustic emission
technique would help in monitoring these insulation
degradation parameters by carrying out the acoustic
emission test periodically. Acoustic emission
technique is non-invasive, non-destructive tool for
detecting and locating electrical discharges and other
incipient faults in transformers. With the availability of
wide variety of acoustic emission sensors and
advanced instrumentation, it is possible to assess the
relative condition of the transformers by acoustic
emission test and comparing with that of the similar
transformers working under similar conditions and in
same power station. Hence, Acoustic emission test
would help the utility to assess the relative condition of
transformers without forced outage thereby avoiding
inconvenience to the end users and also avoiding
revenue loss.
3. ACOUSTIC EMISSION TECHNIQUE
Acoustic emissions (AE) are transient elastic
waves generated by the rapid release of energy from
localized sources within the material. Acoustic
emissions can occur inside the transformer as a result
of small-scale breakdown (PD) and other events such
as arcing, tracking and hot spots. When this happens,
acoustic emissions are produced which can propagate
through the oil and other insulating media. High
frequency piezoelectric sensors are mounted
strategically on the wall of the transformer (tank
surface) to detect these acoustic emissions.
Measurements like amplitude (dBae) of the acoustic
emissions and signal features will help in identifying
the amplitude and source location of acoustic
emissions inside the transformer. With the latest
advancements in electronics and computer
technology including data acquisition and analysis has
resulted in the application of acoustic emission (AE)
as one of the important and useful on-line tool for trend
monitoring of condition assessment of insulation for
transformers.
CPRI has employed a PC based 16-channel
AE system consisting of AE sensors made out of
piezoelectric material with integrated pre amplifiers for
condition monitoring of transformer for partial
discharges [3.4,5,6]. AE Sensors are mounted on the
walls of the transformer tank using magnetic holders
at locations of importance, after cleaning the surface
and applying suitable couplant grease. Couplant
grease ensures proper sensor contact to the surface
of the transformer tank. Identification numbers are
assigned to the sensors and their coordinates
(x,y,z) are noted with reference to lower left corner of
the high voltage side as origin (0,0,0). The sensor
mounting layout depends on the size, design, rating of
the transformer. Condition monitoring of similar units
is done by mounting the sensors at the same positions
for relative performance of similar units operating in
the same power station. Necessary safety precautions
were followed while mounting the sensors on high
voltage transformers.
4. CASE STUDIES
The hydro power station under the study has 4
units of 120MW, 13.8kV, 3 Phase Generators, Each
generator is connected to 3 numbers of Single Phase
13.8kV/(400/√3)kV,50MVA Generator transformers
(GT’s). On-line condition assessment of generator
transformers using acoustic emission technique has
been carried out on 6 numbers of single phase
13.8kV/(400/√3)kV generator transformers. Case
study-1, comprising of 3 generator transformers (GT’s)
(Group-1) of identical ratings connected to Generator
Unit #1 and Case study-2, study comprising of 3
generator transformers (GT’s) (Group-2) of identical
ratings connected to Generator Unit #2. Relative
condition of these generator transformers are
compared with respect to partial discharges and
hotspots over two different years with a span of 5
years(Measurement in the year 2012 and in the year
2017) between the measurements are presented in
the paper.
4.1 Case Study – 1: Assessment of condition of
Generator transformers during the year 2012
Three generator transformers of rating 50MVA,
13.8kV/(400√ 3)kV, single phase and Oil Forced
Water Forced (OFWF) type of cooling system
(Fig.1(a)) are in operation for more than 15 years, at
one of the hydro power station in India. At the request
of the power station authorities, measurements for
partial discharges and hotspots, were carried out
employing on-line acoustic emission technique. As all
the transformers operating under identical conditions,
the request was received for comparing the relative
condition of the 2 groups of three generator
transformers (designated as Group -1 and Group-2).
Fig.1(a) : A view of the generator Transformer
A 16 channel acoustic emission work station
was employed for executing the study. The layout of
16 numbers of sensors (channel #) with reference to
the sides of the transformer is shown Fig.1 (b). The
threshold for the measurement was selected as
30dBae ie., AE signals of magnitude greater than
30dBae will be captured and recorded by the AE
system. Some of the sensors (channels) will not
receive acoustic emission signals as the magnitudes
of the AE signals are lower than the set threshold
value of 30dBae.
Fig.1 (b): Sensor (channel) layout for Generator
Transformers
AE signals from the Generator Transformers (GT’s)
were acquired and analysed. The maximum magnitude
(dBae) of AE signals captured during the year 2012 for three
generator transformers(Group-1) were depicted in Fig. 2(a)
for GT-1R, Fig. 2(b) for GT-1Y, & Fig. 2(c) for GT-1B and
for another three generator transformers(Group-2) were
depicted in Fig. 3(a) for GT-2R, Fig. 3(b) for GT-2Y, & Fig.
3(c) for GT-2B. Although the transformers are of identical
rating and operating under identical conditions, acoustic
emission activity due to partial discharge and hotspots was
observed to be different for each of the transformer.
Acoustic emission technique is found to be suitable for
identifying the relative condition of the generator
transformers with regard to partial discharges and hotspots.
The magnitude of AE signals captured the test is depicted
in Table-1 for the two groups of generator transformers
during the year 2012.
TABLE -1
M AXIMUM LEVEL OF ACOUSTIC EMISSION SIGNALS CAPTURED
FROM Single Phase 13.8kV/(400/√3)kV GENERATOR
TRANSFORMERS
As can be seen from Table-1, the
measurement carried out in the year 2012, the
maximum magnitude of AE signals were found to be
within 40dBae for GT-1R and GT-1Y, whereas for GT-
1B, the magnitude of AE signal was found to be 45
dBae, which is marginally higher compared to the
other transformers. On analysis of the acoustic
emission waveforms captured during the
measurement, it was informed to the power station
operating personnel that the acoustic emission activity
due partial discharges and hotspots was marginally
higher for GT-1B as compared with the other GT’s in
Group-1. It was informed that, based on the CPRI
experience, the condition of the transformers in
Group-1 were found to satisfactory. It was informed
that GT-1B requires to be conditioned by filtration of oil
during the annual maintenance schedule and also to
monitor Dissolve Gas Analysis (DGA) at regular
intervals in order to ascertain the condition of the
transformer (GT-1B).
The maximum magnitude of AE signals were
found to be within 38 dBae for GT-2R and GT-2Y,
whereas for GT-2B, the magnitude of AE signal was
found to be 57 dBae, which is higher compared to the
other transformers. On analysis of the acoustic
emission waveforms captured during
measurement, it was informed to the power station
operating personnel that the acoustic emission activity
due partial discharges and hotspots activity was
higher for GT-2B as compared with the other GT’s in
Group-2. It was informed that, based on the CPRI
experience, the condition of the transformers in
Group-2 were found to satisfactory for GT-2R and GT-
2Y, but the magnitude of AE signal for GT-2B was
found to be much higher compared to GT-2R and GT-
2B. After analysis of acoustic emission waveforms,
power station personnel were informed that GT-2B
requires to be inspected internally during the annual
maintenance schedule and also to monitor Dissolve
Gas Analysis (DGA) at regular intervals of 6 months in
order to ascertain the condition of the transformer (GT-
2B).
Based on the experience of CPRI, it was
recommended to repeat the on-line acoustic emission
measurement to be carried out for both the group of
transformers after 3 years of operation.
Fig. 2(a) Amplitude versus channels for GT- 1R
(Measured in Year 2012)
Fig. 2(b) Amplitude versus channels for GT- 1Y
(Measured in Year 2012)
the
Fig. 2(c) Amplitude versus channels for GT- 1B
(Measured in Year 2012)
Fig. 3(a) Amplitude versus channels for GT- 2R
(Measured in Year 2012)
4.2 Case Study – 2: Assessment of condition
of transformers during the year 2017
Based on the recommendation of CPRI and at
the request of the customer and as a part of periodic
condition assessment of the transformers, the same
groups of transformers (Group-1 and Group-2) were
monitored for the assessing the condition of the
transformers with regard to partial discharges and
hotspots, using on-line acoustic emission technique
in the year 2017.
As all the transformers operating under identical
conditions, request was received for comparing the
relative condition of the six generator transformers
with reference to the measurement carried out in the
year 2012
A 16 channel acoustic emission work station was
employed for executing the study. The layout of 16
numbers of sensors (channel #) with reference to the
sides of the transformer is shown Fig.1 (b). The
sensors were mounted at the same locations (Fig.1
(b)) as was done in the year 2012. The threshold for
the measurement was selected as 30dBae ie., AE
signals of magnitude greater than 30dBae will be
captured and recorded by the AE system.
 understood that some maintenance activities like oil
filtration and inspection was carried out on GT-1Y
during annual maintenance activities, which was the
cause for reduction in acoustic emission activity. On-
line acoustic emission technique was found to be
useful for the power station personnel for identifying
the condition of transformers It was informed that,
based on the CPRI experience, the performance of
the transformers in Group-1 were found to
satisfactory as the magnitude of AE signals
measured were within 50dBae for all the
transformers of Group-1.
As suggested by CPRI, maintenance activities
Fig. 3(b) Amplitude versus channels for GT- 2Y were carried out by the power station personnel, on
(Measured in Year 2012) Group-2 transformers in the year 2016. On-line
acoustic measurement was carried out in the year
2017, as can be seen from Table-2, the maximum
magnitude of the AE signals for Group-2, were found
to be less than that of the magnitudes captured during
2012, indicating that the condition of the these
transformers with reference to partial discharges and
hotspots is satisfactory and have shown a reduced
magnitude compared with that of the measurement in
the year 2012. Condition of Group-2 transformers
were found have improved after carrying out
maintenance activities based on the suggestions by
CPRI.
TABLE -2
MAXIMUM LEVEL OF ACOUSTIC EMISSION SIGNALS CAPTURED FROM
Single Phase 13.8kV/(400/√3)kV GENERATOR TRANSFORMERS
Fig. 3(b) Amplitude versus channels for GT- 3Y
(Measured in Year 2012)

AE signals from the GT’s were acquired and

analysed. The maximum magnitude (dBae) of AE
signals captured during the year 2017 for three
generator transformers(Group-1) were depicted in Fig.
4(a) for GT-1R, Fig. 4(b) for GT-1Y, & Fig. 4(c) for GT-
1B and for three generator transformers(Group-2)
were depicted in Fig. 5(a) for GT-2R, Fig. 5(b) for GT-
2Y, & Fig. 5(c) for GT-2B. The maximum magnitude of
AE signals captured due to partial discharges and
hotspots are depicted in Table-2 for the two groups of
generator transformers during the year 2017. Some of
the sensors (channels) have not received acoustic
emission signals as the magnitudes of the signals are
lower than the set threshold value of 30dBae.
As can be seen from Table-2, the measurement
carried out in the year 2017, the maximum magnitude
of AE signals were found to be almost the
same(within ±2dBae) for GT-1R and GT1B, the
maximum magnitude of AE signals for GT-1Y was
found to be less than that was in the year 2012,after Fig. 4(a) Amplitude versus channels for GT- 1R
discussion with power station personnel, it was (Measured in Year 2017)
Fig. 4(b) Amplitude versus channels for GT- 1Y
(Measured in Year 2017)
Fig. 4(c) Amplitude versus channels for GT- 1B
(Measured in Year 2017)
Fig. 5(a) Amplitude versus channels for GT- 2R
(Measured in Year 2017)
Fig. 5(b) Amplitude versus channels for GT- 2Y
(Measured in Year 2017)
Fig. 5(c) Amplitude versus channels for GT- 2B
(Measured in Year 2017)
5.0 SUMMARY
On-line acoustic emission technique was
adopted to assess the relative condition of hydro
generator transformers at one of hydro power
stations in India. In-situ, measurements have
revealed the effectiveness of on-line acoustic
emission technique for assessing the relative
condition for partial discharges and hotspots in
400kV class generator transformers. The study
carried out has helped to localise the source of
defects like partial discharges and hotspots and
helped in taking necessary steps to improve the
condition of the transformers. Based on the
recommendations of CPRI, repeat acoustic
emission measurements were carried out for the
assessment of the condition of the transformers. The
repeat measurements indicated improvement in the
condition of the transformers. The measurements
using on-line acoustic emission technique was found
to be effective and useful for the power station
operating personnel in assessing the condition of the
generator transformers and also to take necessary
remedial measures to keep the transformers
healthy, thereby making the transformers available
in service continuously.

ACKNOWLEDGMENT

The authors acknowledge CPRI Bangalore India for

the support extended in carrying out this work.

REFERENCES
[1] M.Wang, A.J.Vandermaar, K.D.Srivastava,
“Review of Condition assessment of Power
Transformers in service” IEEE Electrical
Insulation Magazine, Feb. 2002, pp 12-25.
[2] IEEE-62-1995 RA 2005“Diagnostic Field Testing
of Electrical Apparatus-Part 1: Oil Filled
transformers, Regulators and Reactors”
[3] IEEE Std C57.127-2007 – “IEEE guide for the
detection of acoustic emissions from partial
discharges in oil immersed power transformers
and reactors.
[4] Adrian A Pollock, “Acoustic Emission Inspection”,
Physical Acoustics Corporation, TR 03-96-
12/89.
[5] H.N.Nagamani, T.BhavaniShanker,
V.Vaidhyanathan and S.Neelakantan “Acoustic
emission technique for detection and location of
simulated defects in power transformers”
Presented at IEEE Powertech 2005, at St.
Petersburg, Russia, during 27-30 June 2005.
[6] T.S.Ramu, H.N.Nagamani, “ Partial Discharges
based condition monitoring of High voltage
equipment” text book, January 2010 , New Age
International publishers, New Delhi, India .
 Improvement in stray loss calculation of distribution transformer
Rajkumar.M, Sudhakara Reddy S, Gurudev T, Maroti S, Rajaramamohanarao Ch, Arunkumar S, Sreeram V

Abstract leakage field. Designer used to keep extra

Stray losses occur due to the leakage field of
winding in a transformer. The stray loss in
windings can be considerably high in large
transformers. If conductor dimensions and
transposition methods are not chosen properly
winding stray losses results in significantly
higher even in distribution transformers, winding
stray losses due axial leakage field is more
predominant than radial leakage field. To
calculate transformer stray losses accurately it is
required to estimate stray losses by considering
both axial and radial leakage field.
Objective of this paper is to estimate stray
losses for different ratings of distribution
transformers by considering axial and radial
leakage filed and compare with tested results.
margin between guaranteed losses and
calculated loss to accommodate stray losses
due to radial leakage field, but with today’s
competitive market we cannot keep much
margin between calculated losses and
guaranteed losses, to operate with less safety
margins it is required to calculate losses
accurately. Unable to meet guaranteed losses
during testing results leads to penalty or risk of
rejection of the supply of transformer order.
Hence it is essential to calculate losses
accurately. To calculate losses accurately it is
required estimate stray losses by considering
both axial and radial leakage field.
Typical transformer leakage field shown in
figure. 1

Introduction

Transformer is highly efficient equipment in

power system; even though quantum of energy
loss due to transformer losses is significant.
Load loss is the major contribution in total losses
of the transformer.

The load loss of a transformer consists of the

2
loss due to the ohmic resistance of windings (I R
losses) and some additional losses. These
additional losses are generally known as stray
losses which occur due to the leakage field of
windings. The stray losses in the windings are
further classified as eddy and circulating current
losses. The remaining stray losses occur in
structural steel parts. The stray loss in windings
can be considerably high in large transformers.
If conductor dimensions and transposition Figure 1 Transformer leakage field
methods are not chosen properly winding stray
losses results in significantly high even in From above figure it is evident that most of
distribution transformers. leakage field is axial, but at the ends of the
winding radial leakage filed density is more.
Stray losses occur due to the leakage field of
windings in a transformer. Leakage field is Mean eddy loss per unit volume of the winding
classified as axial and radial field. Winding stray due to the axial leakage field is expressed as
losses due axial leakage is more predominant
than radial leakage field. Previously for small
W e mean = 5555555.. (1)
distribution transformers, the error introduced by ∗
neglecting the radial field may not be
considerable; it was common practice to Where:
calculate winding stray losses due to only axial Bp peak value of the flux density in the HV-LV
Page 1 of 5
gap, and is given by Case study: 1
√ Rating : 1000 kVA
Bp = 5555555.. (2) Voltage ratio : 33/0.420V
Percentage impedance
‘t’ thickness of winding conductor i.e. radial
at rated tap : 6.0%
dimension of conductor
Standard : IEC 60076
‘ω’ Angular displacement
Energy efficiency level : -
‘ρ’ resistivity of winding conductor
Calculated stray losses only
‘Hw’ is winding height
with axial leakage field : 815 W
‘N’ Number of turns
Calculated stray losses
‘I’ current in ampere
with both axial and
radial leakage field : 1001 W
The leakage field incident on a winding
Measured stray losses : 977 W
conductor (see Figure 2) is resolved into two
components, viz. By and Bx, and losses due to Case study: 2
these two components are calculated separately Rating : 1600 kVA
by equations (3) and (4) and then added Voltage ratio : 11/0.415V
together to get total winding stray losses. Percentage impedance
at rated tap : 6.25%
Standard : IS 1180-1
Energy efficiency level : 1
Calculated stray losses only
with axial leakage field : 1115 W
Calculated stray losses
with both axial and
radial leakage field : 1381 W
Measured stray losses : 1360 W

Figure 2 winding conductor in a leakage field. Case study: 3

Rating : 1600 kVA
Eddy losses per unit volume due to the axial (By) Voltage ratio : 11/0.433V
and radial (Bx) components of the leakage field Percentage impedance
are given in the below expressions at rated tap : 6.25%
Standard : IS 1180-1
Energy efficiency level : 1
W e axial = 5555555.. (3) Calculated stray losses only
with axial leakage field : 1505 W
Calculated stray losses
W e radial = 5555555.. (4) with both axial and
radial leakage field : 1751 W
Where: Measured stray losses : 1790 W
w is width of the conductor i.e. axial dimensions
of winding conductor

Equations (3) and (4) are generally used with

some empirical correction factor applied to the
total calculated stray loss value.

The winding stray losses estimated with

a) Considering axial leakage field only

b) Considering both axial and radial
leakage field.

These results are compared with measured

values for different ratings of distribution
transformer. Figure 3: 1600 kVA Transformer under study
Page 2 of 5
Case study: 4
Rating : 1600 kVA
Voltage ratio : 11/0.433V
Percentage impedance
at rated tap : 6.25%
Standard : IS 1180-1
Energy efficiency level : 1
Calculated stray losses only
with axial leakage field : 1611 W
Calculated stray losses
with both axial and
radial leakage field : 1941 W
Measured stray losses : 1907.5 W

Figure 5: 2000 kVA Transformer under study

Case study: 7
Rating : 2000 kVA
Voltage ratio : 3.3/433V
Percentage impedance
at rated tap : 10%
Standard : IEC 600076
Energy efficiency level : -
Calculated stray losses only
Figure 4: 1600 kVA Transformer under study with axial leakage field : 2607 W
Calculated stray losses
Case study: 5 with both axial and
Rating : 2000 kVA radial leakage field : 3426 W
Voltage ratio : 6.6/0.433V Measured stray losses : 3386.3 W
Percentage impedance
at rated tap : 6.25%
Standard : IS 1180-1
Energy efficiency level : 1
Calculated stray losses only
with axial leakage field : 1714 W
Calculated stray losses
with both axial and
radial leakage field : 2151 W
Measured stray losses : 2208 W

Case study: 6
Rating : 2000 kVA
Voltage ratio : 11/433V
Percentage impedance Figure 6: 2000 kVA Transformer under study
at rated tap : 6.25%
Standard : IS 1180-1 Case study: 8
Energy efficiency level : 1 Rating : 2200 kVA
Calculated stray losses only Voltage ratio : 11/2x0.380V
with axial leakage field : 2105 W Percentage impedance
Calculated stray losses at rated tap : 6.0%
with both axial and Standard : IS 2026
radial leakage field : 2614 W Energy efficiency level : -
Measured stray losses : 2670 W Calculated stray losses only
with axial leakage field : 1991 W
Page 3 of 5
Calculated stray losses Standard : IEC: 60076
with both axial and Energy efficiency level : -
radial leakage field : 2596 W Calculated stray losses only
Measured stray losses : 2590 W with axial leakage field : 3777 W
Calculated stray losses
with both axial and
radial leakage field : 4911 W
Measured stray losses : 4819.3 W

Figure 7: 2200 kVA Transformer under study.

Case study: 9
Rating : 2500 kVA
Voltage ratio : 11/433V
Percentage impedance
at rated tap : 6.25% Figure 9: 2500 kVA Transformer under study
Standard : IS 1180-1
Energy efficiency level : 3 All the cese studies are tabulated in table 1.0
Calculated stray losses only Calculated total stray loss Measu
only with axial with both axial red
with axial leakage field : 3678 W leakage field and radial total
Calculated stray losses leakage field stray
Case study No.

with both axial and % % loss in

Rating in kVA

radial leakage field : 4851 W change change watts

w.r.t. w.r.t.
Measured stray losses : 4728 W
in watts

in watts
measur measur
ed ed
values values
1 1000 815 83.4% 1001 102.5% 977.0
2 1600 1115 82.0% 1381 101.5% 1360.0
3 1600 1505 84.1% 1751 97.8% 1790.0
4 1600 1611 84.5% 1941 101.8% 1907.5
5 2000 1714 77.6% 2151 97.4% 2208.0
6 2000 2105 78.8% 2614 97.9% 2670.0
7 2000 2607 77.0% 3426 101.2% 3386.3
8 2200 1991 76.9% 2596 100.2% 2590.0
9 2500 3678 77.8% 4851 102.6% 4728.0
1
2500 3777 78.4% 4911 101.9% 4819.3
0

Figure 8: 2500 kVA Transformer under study Table 1.0: comparison of calculated stray loss
with measured stray loss.
Case study: 10
Rating : 2500 kVA
Voltage ratio : 11/433V
Percentage impedance
at rated tap : 11.0%
Page 4 of 5
 leakage field in transformer stray loss
calculations.

Acknowledgement:
The authors would like to express my gratitude
to management of CPRI for support and
encouragement in preparation of this paper.

References:

[1] Transformer Engineering Design, Technology

and Diagnostics, S.V. Kulkarni, S.A. Khaparde
Second Edition.

[2] IS 1180-1 : 2014 , Outdoor Type Oil

Immersed Distribution Transformers Up to and
including 2 500 kVA, 33kV — Specification
Figure 10: Graphical representation of
comparison of calculated stray loss with
measured stray loss.

Ten different ratings of distribution transformers

are considered for estimation; firstly losses are
calculated only with axial leakage field and then
considering both axial and radial leakage field.
Transformer measured stray losses are obtained
from load loss measurement. Results are
analyzed with help of tabular form and graphical
representation.

From Table 1.0 and Figure 10 it evident that

stray loss calculated with consideration of both
axial and radial stray field are close to measured
stray loss values than stray loss calculated only
with consideration of axial leakage field.

It was the common observation/understanding

that conductor thickness (radial dimension) is
significant in stray loss control, but as long as
loss due to only consideration with axial leakage
field conductor thickness is significant, whenever
radial leakage field is considered conductor
width(axial dimension) also significant.

Conclusion:

Stray losses are estimated for different ratings of

distribution transformers by considering axial
and radial leakage field and compared with
tested results.

It is evident that, to calculate losses accurately it

is required to consider both axial and radial

Page 5 of 5