1

Case Study: Implementation of Dynamic Rating

for Transformers in a Distribution Network
Thahirah Syed Jalal, Member, IEEE

Abstract—In 2012, Unison Networks Limited implemented

Dynamic Power Transformer Rating (DPTR) on most of the
power transformers in its distribution networks. The
implementation involved 50 medium power transformers across
three different regions in New Zealand – Hawkes Bay, Taupo and
Rotorua. The DPTR scheme utilised an in-house developed
algorithm that has been designed based on the IEC60076-7
Loading Guide for Oil Immersed Power Transformer Standard.
Every three minutes, the algorithm samples real time data from
various sensors installed in the networks and calculates the
maximum loading that each individual transformer can safely
carry. This paper will discuss the outputs from the algorithm and
the benefits achieved from the implementation.

Index Terms—Power transformers, distribution networks,

smart grids

I. NOMENCLATURE
Advanced Distribution Management System (ADMS),
Dissolved Gas Analysis (DGA), Dynamic Power Transformer Fig 1. Unison’s networks in the North Island, New Zealand
Rating (DPTR), International Electrotechnical Commission
(IEC), megavolt-amperes (MVA), per unit (p.u.), Oil Natural As a part of the Smart Grid Initiative, in 2011, Unison
Air Natural (ONAN), Oil Natural Air Forced (ONAF), developed its own Dynamic Power Transformer Rating
Supervisory Control and Data Acquisition (SCADA), Unison (DPTR) to be implemented on its power transformers. The
Networks Limited (Unison). International Electrotechnical Commission (IEC) standard IEC
60076-7 [2] defined “medium power transformers” as
II. INTRODUCTION transformers with ratings between 2.5 to 100MVA for 3 phase

U NISON Networks Limited (Unison) is the electricity transformers. Unison’s 33/11kV power transformers fall under
distribution business that owns, maintains and operates this category since they range from 3 to 30MVA. Unison has
the electricity distribution assets that supply consumers in the fifty eight power transformers across Hawkes Bay, Taupo and
Hawkes Bay, Rotorua and Taupo regions of New Zealand’s Rotorua. In the 2012/13 financial year, the DPTR scheme had
North Island (see Fig. 1). It supplies over 100,000 consumers, been implemented on all its power transformers with ratings
has a regulatory asset base of NZ$400m and invests over above 5MVA, involving fifty power transformers in total.
NZ$30m per annum in electricity network infrastructure.
In June 2009, Smart Grid became a strategic direction for III. BACKGROUND AND OBJECTIVES
Unison. The following definition of Smart Grid has been The term “Dynamic Rating” implies that an asset’s rating
adopted by Unison: can be changed or made adaptive. Lachman et al. [3] defined
“The application of real time data, communications dynamic rating for a transformer as follows: “The maximum
technologies and emerging trends in electricity distribution to loading which the transformer may acceptably sustain under
improve capacity utilisation, enhance lifecycle asset time-varying load and/or environmental condition.” A
management and provide additional services on the modern transformer’s name plate rated power (typically in megavolt-
network, thereby optimising network investment to the benefit amperes, MVA) is the output from the secondary terminals,
of all stakeholders.” [1] which the unit can deliver on a continuous basis at:
• rated frequency,
• rated secondary voltage,
T. S. Jalal is with Unison Networks Limited, New Zealand (e-mail: • continuous ambient temperature of 30° Celsius (C)
[email protected]). and

978-1-4799-3653-3/14/$31.00 ©2014 IEEE

Authorized licensed use limited to: Batangas State University. Downloaded on February 08,2022 at 04:36:07 UTC from IEEE Xplore. Restrictions apply.
 2

• continuous winding hottest spot rise over ambient

temperature of 80°C
resulting in a continuous hottest spot temperature of 110°C
[4].
The name plate definition implies that theoretically, when
ambient temperatures are lower than 30°C, it is possible to
sustainably load a transformer above its name plate rating
without causing the transformer to overheat above 110°C.
Under this scenario, the dynamic rating calculated for the
transformer will be above the name plate rating. Conversely,
when ambient temperatures are higher than 30°C, the Fig. 2. Inputs and outputs for the DPTR algorithm
calculated dynamic rating for a transformer will be lower than
its name plate rating. The ability to utilize these properties of Fig. 3 shows the flowchart of the algorithm. The first step
dynamic rating is distinctly beneficial, as discussed in Section involves getting real time data from the various sensors in the
VII. network as inputs to the algorithm. The algorithm stores all the
The objective of developing the DPTR scheme was to parameters of each transformer that are required for
provide Unison the ability to load its power transformers using calculation. The detailed explanations of these parameters and
dynamic rating in a safe and reliable manner. This is associated equations are provided in Jalal et. al [1]. Some of
particularly useful in meeting winter peak demands in New the parameters are unique for each transformer since they
Zealand which tend to be higher than in other seasons due to depend on the transformer’s design. Some parameters change
space heating. The DPTR scheme gathers real time data with cooling operation and tap position whereas some remain
through sensors and online monitoring and using intelligent fixed. Based on each transformer’s real time tap position and
algorithms, translates the data into meaningful information on cooling operations, the corresponding transformer parameters
a power transformer’s conditions and loading ability, therefore are selected and used in the calculations. The calculated
minimizing the risks of transformer failure and allowing the outputs are then made available to be viewed by various users.
potential benefits of dynamic rating to be achieved safely and These steps are repeated every three minutes to provide real
successfully. time data. It is also possible to use forecasted ambient
temperature data to calculate the forecasted dynamic rating for
the transformers.
IV. DPTR ALGORITHM DEVELOPMENT
Worldwide, DPTR is a relatively new application and not 1. Perform data sampling from the relevant sensors in
many electricity distribution networks have ventured into it. A the network
possible reason for this scenario is due to the fact that
implementing DPTR using existing commercial products is
2. Select and calculate the transformer parameters
not cost effective for power transformers with small ratings.
based on transformer status (e.g. cooling operation, tap
By developing the DPTR algorithms in-house and utilizing position)
existing infrastructure, Unison is able implement the scheme
at much lower costs.
The DPTR algorithm developed by Unison has been based 3. Calculate the
on the IEC60076-7 [2] standard since the transformers in its i. dynamic rating
ii. hot spot temperature
networks have been designed to the standard or its
iii. insulation loss of life
predecessors. The American equivalent of this standard is the
IEEE Std C57.91: Guide for Loading Mineral-Oil-Immersed
Transformers [4-6]. Fig. 2 shows the inputs and outputs of the 4. Display outputs to various users
DPTR algorithm. From the network sensors, the following real Fig. 3. DPTR algorithm flowchart
time operation data: a transformer’s load, top oil temperature,
tap position, cooling operation (natural, fan or oil pump
assisted) and its surrounding ambient temperature are sampled
through the Supervisory Control and Data Acquisition
(SCADA) system. Using these data as inputs, the DPTR
algorithm calculates in real time the dynamic rating for the
transformer. The transformer’s hot spot temperature and its
insulation paper’s loss of life are also calculated to ensure that
the transformer does not breach the limits permitted by the
standard even if its name plate rating has been exceeded.

Authorized licensed use limited to: Batangas State University. Downloaded on February 08,2022 at 04:36:07 UTC from IEEE Xplore. Restrictions apply.
 3

V. DPTR IMPLEMENTATION
The DPTR algorithm has been rigorously developed and
tested using Matlab software and then prototyped for
production in PI software by OSIsoft. The PI system collects
the ambient temperature data from the weather stations
through an external service provider. The data from power
transformers are collected via the remote terminal units at the
substations and remitted to PI system by SCADA. PI ACE
(the calculation program of PI) performs the DPTR
calculations and sends the outputs to the relevant PI tags. The
PI tags are then linked to SCADA tiles to allow the outputs to
be accessible by the control room operators. The
implementation of DPTR using PI is illustrated in Fig.4. The
Realflex SCADA system in Unison is currently being replaced
by an Advanced Distribution Management System (ADMS)
developed by Telvent and Schneider Electric. The ADMS is
expected to function well with PI in delivering the DPTR
outputs.

Fig 5. The PI webpart for DPTR displaying collected data from sensors and
calculated dynamic rating for one of Unison’s 15/20MVA transformer

The IEC60076-7 [2] dictates that the operational limits for

medium power transformers under a normal cyclic loading
scenario are as shown in Table II.

TABLE II
IEC60076-7 CURRENT AND TEMPERATURE LIMITS APPLICABLE TO LOADING A
MEDIUM POWER TRANSFORMER BEYOND ITS NAME PLATE RATING
Fig 4. The DPTR scheme implementation
Types of Current Winding hot Other Top oil
loading (p.u.) spot metallic temperature
The outputs of the PI ACE calculations stored by the PI temperature hot-spot (°C)
tags are accessible via PI Processbook, PI System Explorer or (°C) temperature
PI Excel Datalink. Customized PI webparts have also been (°C)
created to display the calculated outputs for each transformer Normal 1.5 120 140 105
using appropriate plots such as radial gauges and trend graphs. cyclic
Some of the webpart components are shown in Fig.5. loading

The temperature and current limits shown in Table II are not

intended to be valid simultaneously. The current may be
limited to a lower value than that shown in the table to meet
the temperature limitation requirement. Conversely, the
temperature may be limited to a lower value than that shown
in Table II in order to meet the current limitation requirement
[2]. The limits imply that on a normal daily basis, a medium
power transformer can potentially be loaded to 150% of its
name plate rating for certain durations as long as its thermal
limits are not breached.
Observations on the DPTR scheme outputs implemented in
Unison networks confirmed that during periods when the air
temperature is low, the dynamic rating calculated for the
transformers are higher than the name plate rating. However,
the DPTR algorithm limits the calculated dynamic rating to
150% of the name plate ratings to provide a safety margin for

Authorized licensed use limited to: Batangas State University. Downloaded on February 08,2022 at 04:36:07 UTC from IEEE Xplore. Restrictions apply.
 4

any input error. The dynamic rating calculation also defaults In principle, the calculated dynamic rating for a transformer
to the name plate rating when it does not receive any data is high when its surrounding air temperature is low and its
from the sensors. The transformer temperature protection is current loading is not high. For Flx T2, the maximum
still maintained to trigger an alarm at a top oil temperature of calculated dynamic rating is limited 15MVA, which is 150%
80°C and trip the transformer at a top oil temperature of 95°C of its ONAF name plate rating of 10MVA. Fig. 6a shows that
as an additional safety feature. The control room operators and the calculated dynamic rating only reaches 15MVA on some
asset managers are also automatically notified by the PI days in January 2013 when the measured air temperatures are
system if any abnormal reading is observed. less than 30°C.
The DPTR scheme has now been implemented on the On the 18th of January, Flx T2 carried a higher loading, up
transformers that have been installed with thermal sensors and to 67% of the ONAF name plate rating, due to an outage,
ambient temperature sensors. This involves 50 transformers resulting in a calculated dynamic rating of lower than
across 26 substations, including substations that are more than 15MVA. The air temperature on that day was the lowest in the
250 kilometers of distance from the control room. The author month, resulting the calculated dynamic rating of 14.5MVA,
shared the challenges of implementing the scheme in another still close to 15MVA. On the 26th of January, there was a brief
publication [1]. communication outage, causing the DPTR to default to the
ONAF name plate rating of 10MVA. The average dynamic
VI. CASE STUDY rating for January 2013 is 14.18MVA.
This section presents a case study on one of our 33/11kV Fig. 7 shows the PI trend plots for the winter month of July
power transformers. It is a 7.5/10MVA transformer, known as 2013. Fig. 7a shows the measured load and calculated
“Flx T2”, with an ONAN/ONAF cooling mode. Fig. 6 shows dynamic rating (in MVA) for ONAF mode whereas Fig. 7b
the PI trend plots for the summer month of January 2013. Fig. shows the corresponding measured air and top oil
6a shows the measured load and calculated dynamic rating (in temperatures and the calculated hot spot temperatures.
MVA) for ONAF mode whereas Fig. 6b shows the
corresponding measured air and top oil temperatures and the
calculated hot spot temperatures.

Fig 6a. The PI trend plot for the Flx T2 DPTR displaying measured load and
calculated dynamic rating for ONAF mode for January 2013 (summer)
Fig 7a. The PI trend plot for the Flx T2 DPTR displaying measured load and
calculated dynamic rating for ONAF mode for July 2013 (winter)

Fig 6b. The PI trend plot for the Flx T2 DPTR displaying air, top oil and Fig 7b. The PI trend plot for the Flx T2 DPTR displaying air, top oil and
calculated hot spot temperatures for January 2013 (summer) calculated hot spot temperatures for July 2013 (winter)

Authorized licensed use limited to: Batangas State University. Downloaded on February 08,2022 at 04:36:07 UTC from IEEE Xplore. Restrictions apply.
 5

In the winter month of July 2013, the Flx T2 average load without any external expert consultation.
is higher than in January due to space heating. However, the
air temperatures are considerably lower than the temperatures
in January 2013, resulting in the calculated dynamic rating to VIII. SUMMARY AND CONCLUSION
be 15MVA most of the time. The average dynamic rating for A case study of a DPTR implementation in a distribution
July 2013 is 14.98MVA. Even during a continuous 2 days network is presented. Real time data - loads, top oil
outage of Flx T1 in the first week of July where Flx T2 was temperatures, ambient temperatures, tap positions and cooling
carrying a load of 73% of its name plate rating, the calculated operations - from each medium power transformers in the
dynamic rating is still 150% of its ONAF name plate rating Unison networks are used to determine their individual
due to the low air temperatures. dynamic ratings. The DPTR algorithm is able to accommodate
the various parameters unique to each transformer. The hot
spot temperature and insulation loss of life for each
VII. BENEFITS transformer are also calculated and monitored.
The case study presented in Section VI demonstrated that The DPTR scheme optimizes a transformer’s operation and
in most situations in Unison networks, the calculated dynamic enhances the network’s post contingency capability. In the
rating for a transformer is around 150% of its name plate long term, it also defers investments in transformer upgrades.
rating. Hence, operating a transformer to its dynamic rating The visualized data collected from the sensors allow potential
optimizes its operation to its actual designed thermal potential. transformer problems to be quickly identified.
The DPTR scheme also allows Unison to have better security
and contingency whereby Unison can still continue to deliver
electricity supply if some of its transformers failed during IX. REFERENCES
emergencies and natural disasters, since the remaining 1. Jalal, T.S., N. Rashid, and B. van Vliet. Implementation of
transformers can be safely loaded above their name plate Dynamic Transformer Rating in a distribution network. in IEEE
International Conference on Power System Technology
ratings. Constraining the dynamic rating to 150% of name
(POWERCON). 2012.
plate rating and including the temperature trip mechanism at 2. International Electrotechnical Commission, International Standard
95°C jointly ensure that the extended capacity afforded by Power transformers, in Part 7: Loading guide for oil-immersed
dynamic rating is attained without any operational risk. power transformers. 2011.
3. Lachman, M.F., et al., Real-time dynamic loading and thermal
Operating a transformer at its full name plate rating diagnostic of power transformers. Power Delivery, IEEE
continuously in 30°C results in a typical insulation life of Transactions on, 2003. 18(1): p. 142-148.
20.55 years [4]. If a transformer’s load is less than its name 4. Institute of Electrical and Electronics Engineers (IEEE), IEEE
Guide for Loading Mineral-Oil-Immersed Transformers, in IEEE
plate rating and its ambient temperatures are lower than 30°C, Std C57.91-1995. 1996. p. i.
it will have a lower rate of insulation deterioration and hence a 5. Institute of Electrical and Electronics Engineers (IEEE), IEEE
longer insulation life. Conversely, in temperatures higher than Guide for Loading Mineral-Oil-Immersed Transformers -
Corrigendum 1, in IEEE Std C57.91-1995/Cor 1-2002. 2003. p. 1-
30°C, faster deterioration occurs when the transformer’s name 16.
plate rating is exceeded. However, since ambient temperatures 6. Institute of Electrical and Electronics Engineers (IEEE), IEEE
often vary and thermal ageing of the insulation is a cumulative Guide for Loading Mineral-Oil-Immersed Transformers and Step-
Voltage Regulators - Redline, in IEEE Std C57.91-2011 (Revision
process, transformers may be operated above their name plate of IEEE Std C57.91-1995) - Redline. 2012. p. 1-172.
ratings with an acceptable loss of life, provided they are
operated for sufficient periods below their ratings. X. BIOGRAPHIES
A power transformer is generally the most costly asset in a
distribution network. Having the ability to operate Thahirah Syed Jalal completed an MEng. in
transformers beyond their name plate ratings sustainably for Engineering Science from Oxford University, United
Kingdom in June 2002, under the scholarship from
certain durations will enable monetary savings. In the long the main power utility company in Malaysia, Tenaga
term, dynamic rating utilization defers investing on larger Nasional Berhad (TNB). She worked in TNB’s
transformers to meet growing load demands. wholly owned university, Universiti Tenaga Nasional
(UNITEN) from July 2002 as a lecturer. Her research
The online monitoring of transformers parameters and and consultancy work there were on electromagnetic
archiving their data also informs the assessment of a fields, energy efficiency and renewable energy.
transformer’s condition. An example of this was observed
earlier in 2013 when on 18 April, a lab report on Dissolved In 2008, Thahirah pursued her PhD at University of Canterbury, New
Gas Analysis (DGA) of one of Unison’s transformer indicated Zealand. Her thesis was on the development of a System Dynamics (SD)
model for generation expansion in New Zealand. In 2011, she joins Unison
that there was an increased level of ethylene in the transformer Networks Limited, New Zealand as the Network Investment Specialist. In
oil, representing an early warning of a possible impending Unison, she works on developing intelligent algorithms for smart grid data
fault. A Duval triangle analysis on the combustible gases involving transformers. She is a senior member of the team that is responsible
for developing, maintaining and managing decision support tools in Unison.
confirmed that there was a thermal fault of more than 700°C
present in the transformer. The archived top oil temperature
data helped in identifying the possible source of the
overheating and allowed Unison to address the problem

Authorized licensed use limited to: Batangas State University. Downloaded on February 08,2022 at 04:36:07 UTC from IEEE Xplore. Restrictions apply.