Available online at www.sciencedirect.

com

ScienceDirect
Available online at www.sciencedirect.com
Procedia Engineering 00 (2017) 000–000
www.elsevier.com/locate/procedia
ScienceDirect
Procedia Engineering 202 (2017) 20–28

4th International Colloquium "Transformer Research and Asset Management”

Transformer Fleet Monitoring

Tihomir Jaković*, Ivan Murat, Filip Klarić, Samir Keitoue
Tihomir Jaković, Fallerovo šetalište 22, 10002, Croatia
Ivan Murat, Fallerovo šetalište 22, 10002, Croatia

Abstract

Transformer monitoring systems are increasing in number and having a clear overview of the whole fleet of
monitored transformers is becoming crucial in an effort to provide the end user with an easy way to monitor its
transformer assets and react accordingly to the monitored situations. This paper will introduce a Fleet monitoring
solution that provides the architecture and tools to accomplish the set tasks of connecting and accessing information
of all the monitoring systems from their dedicated locations.

Standard transformer monitoring system configuration is made of a local unit on the transformer for collecting
and processing raw data from the sensors and transferring the data to another unit, usually a server located in the
substation which stores it in the database. The server unit is also responsible for the presentation of data to the end
user and communication with supervisory control systems. This configuration is well capable of achieving its
purpose of monitoring the transformer state and operation. The challenge is in accessing the monitored data.

Traditional approach for accessing the data was to access the servers locally in the substations. Responding to the
monitored conditions was done by connecting critical alarm signals from the monitoring system to the substation
control system and in case of an alarm indication the operator from the substation was required to access the
monitoring system for further details and inform personnel responsible for the maintenance of transformer units. The
transformer maintenance personnel were then required to check the monitored data either by connecting remotely to
the monitoring system, which in most cases is not possible due to security and technical restrictions, or to travel to
the remote location to investigate the issue. This approach was both very time consuming and costly. There is also a
consideration that the modern substations are becoming unmanned or have just a basic crew which may consist of
personnel not qualified or trained for the required tasks.

* Tihomir Jaković. Tel.: +385-99-313-8637; fax: +3851-366-7306.

E-mail address: tjakovic@koncar-institut.hr

1877-7058 © 2017 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the organizing committee of ICTRAM 2017.

1877-7058 © 2017 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the organizing committee of ICTRAM 2017.
10.1016/j.proeng.2017.09.691
2 TihomirTihomir
Jaković Jaković
et al. / Procedia Engineering
et al. / Procedia 00 (2017)
Engineering 202000–000
(2017) 20–28 21
2 Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000
The challenge was to enable a centralized access to the distributed monitoring systems across multiple
The challenge
substations. was to enable
An architecture, botha hardware
centralizedandaccess to the
software, wasdistributed
developed,monitoring systems across
that accomplishes multiple
this task. The
substations. An architecture, both hardware and software, was developed, that accomplishes this
architecture is highly modular so that the components of the system can be distributed and integrated into an overall task. The
architecture is highly modular so that the components of the system can be distributed
system which is fully customizable to the end user requirements. During development one of the mainand integrated into an overall
system which was
considerations is fully
to makecustomizable to of
the integration thealready
end user requirements.
existing During development
systems as seamless as possible. The oneresult
of the
is a main
Fleet
considerations was to make the integration of already existing systems as seamless as possible.
monitoring solution that centralizes access to the information from all connected monitoring systems while The result is aatFleet
the
monitoring solution
same time does that centralizes
not interfere with theaccess
work to
of the information
dedicated from systems
monitoring all connected monitoring
at remote systems
locations. while
This way theatdata
the
same time does not interfere with the work of dedicated monitoring systems at remote locations.
can be presented directly to the right personnel. This paper will also discuss further goals and directions of This way the data
can be presented
development directly tofleet
for transformer the monitoring.
right personnel. This paper will also discuss further goals and directions of
development for transformer fleet monitoring.
©
© 2017 The Authors.
2017 The Authors. Published
Published by
by Elsevier
Elsevier Ltd.
Ltd.
© 2017 The Authors.
Peer-review under Published by
responsibility of Elsevier
the Ltd. committee of ICTRAM 2017.
organizing
Peer-review under responsibility of the organizing committee of ICTRAM 2017.
Peer-review under responsibility of the organizing committee of ICTRAM 2017.
Keywords: transformer monitoring; fleet monitoring; asset management
Keywords: transformer monitoring; fleet monitoring; asset management

1. Introduction
1. Introduction
Transformer monitoring systems have seen an increase in number and in system complexity over the past 10
Transformer
years. Monitoringmonitoring systems
systems are have seen
generating moreaninformation
increase in that
number
mustand
beinpresented
system complexity
to the userover
in atheclear
pastand
10
years. Monitoring systems are generating more information that must be presented to the user in
organized way [1,2]. As the monitoring systems grow in numbers they are becoming an independent subsystem in a clear and
organized waythat
the substation [1,2].
hasAs the monitoring
a different purposesystems
that thegrow in numbers
traditional they
control are becoming
system an independent
in the substation. subsystem
The main in
purpose of
the substation that has a different purpose that the traditional control system
the monitoring system is object oriented monitoring and asset management [1]. in the substation. The main purpose of
the monitoring system is object oriented monitoring and asset management [1].
The paper will present Končar TMS Fleet monitoring, a system that rises to the challenge in accomplishing the
taskThe
of paper will present
networking Končarmonitoring
transformer TMS Fleetsystems
monitoring, a system
and giving thethat
endrises
usertothe
the best
challenge in accomplishing
possible the
insight into their
task of networking
equipment. transformer monitoring systems and giving the end user the best possible insight into their
equipment.
Nomenclature
Nomenclature
TMS Transformer Monitoring System
TMS
DGA Transformer
Dissolved Monitoring
Gasses System
Analysis
DGA
LCP Dissolved Gasses Analysis
Local Control Panel
LCP
NI Local Control
National Panel
Instruments
NI National Instruments

1.1. Traditional approach

1.1. Traditional approach
Standard transformer monitoring system configuration is made of a local unit on the transformer for collecting
andStandard transformer
processing raw data monitoring systemand
from the sensors configuration
transferringisthe
made
dataoftoaanother
local unit onusually
unit, the transformer for collecting
a server located in the
and
substation which stores it in the database [4]. The server unit is also responsible for the presentation of data in
processing raw data from the sensors and transferring the data to another unit, usually a server located the
to the
substation which stores it in the database [4]. The server unit is also responsible for the presentation
end user and communication with supervisory control systems. This configuration is well capable of achieving its of data to the
end userofand
purpose communication
monitoring with supervisory
the transformer state and control systems.
operation. This configuration
The challenge is well
is in accessing capable of data.
the monitored achieving its
purpose of monitoring the transformer state and operation. The challenge is in accessing the monitored data.
22 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28
Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000 3

Figure 1 Standard system configuration

Traditional approach for accessing the data is to access the servers locally in the substations (Figure 1). Usually
only a small subset of possible warning and alarm indications is connected from the monitoring system to the
substation control system. One of the biggest drawbacks in these kind of connections is that the alarm indications
are mostly grouped together with a generic message to the control system operators e.g. TMS Common alarm. This
confuses the operator and does not provide a clear information regarding the severity of the message.

Next step is for the operator from the substation to access the monitoring system for further details [2] and inform
personnel responsible for the maintenance of the transformer units about the situation. The transformer maintenance
personnel are then required to check the monitored data either by connecting remotely to the monitoring server or to
travel to the substation and investigate the issue. There are a few issues with this approach.

An approach with an alone remote connection to the monitoring server is usually accompanied with security and
technical restrictions [3]. This was one of the main reasons in promoting the idea of a separate subnetwork dedicated
to the communication of the monitoring systems.
Investigating the monitoring data locally by traveling to the substation is very time consuming and costly. If we
add into the consideration that the modern substations are becoming unmanned or with just the basic crew, which
mainly consist of personnel not qualified or trained for the operation of the monitoring systems, the need to establish
a new approach to accessing monitoring data becomes even more plausible.

The novel approach is to have a centralized access to the inherently distributed system with trained personnel in
charge of reviewing the monitoring system data. This way the end user has a clear and organized way of monitoring
the state of the transformers.
 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28 23
4 Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000

2. Fleet system overview and configuration

Basic overview of the fleet monitoring system is shown in Figure 2. The system is separated in 3 main tiers:
transformer tier, service tier and client tier.

Figure 2 Fleet system overview

Each tier has a dedicated set of tasks to execute.

Transformer Tier:
• Collecting the raw data from the sensors on the transformer and performing estimations and analysis on the
collected data
• Triggering warning/alarm signals
• Communication to the service tier

Service Tier:
• Communication with the transformer tier and client tier
• Database manipulation – storing and retrieving monitoring data
• Client servicing – providing data to the client

Client Tier:
• Data visualization
• Communication with the service tier
• Client validation

With usual software approach service tier is realized with monolith architecture and meant to be run as a single
process accompanied by relational database. Running single process applications simplifies development and
maintainability of software during operations as there is only single point of functionality and failure and by that, a
single point of setup and recovery. While functional and adequate for single station environments, monolith
24 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28
Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000 5

architecture is not easily expandable on horizontal scale required for fleet monitoring systems. For that purpose,
TMS system delivers a distributed architecture of processes each responsible for distinct functions of service tier. As
distributed processes are not necessarily located on same hardware inter-process communication is configurable to
standard network protocols. Such system allows us to adjust point of data converging and distribution in overall
system and to accommodate various hardware layouts.

Every tier can have one or more system components in it. The components can be setup in a way to enable the
fleet monitoring system to have a high degree of flexibility and modularity. System component can be separated into
software and hardware components.

Physical location of the system tiers and components can vary and depends on the end users’ specifications and
requirements.

2.1. Hardware

2.1.1. Končar TMS Controller

The TMS controller is the basic building block of the monitoring system. The component can be located either in
the TMS cubicle on the transformer or in LCP of the transformer. The recently developed Končar cVISK controller,
is the product of Končar own development, both the hardware and the software. Also in use, there are NI controllers
on legacy systems.

The controllers are highly modular in configuration allowing the monitored quantities to be setup according to
the exact clients’ specifications.

2.1.2. Networking equipment

The networking equipment used to connect the different tiers consist of routers, switches, media converters,
firewalls, GSM-GPRS modems and other equipment normally found in networking solutions. The base for the
network is the TCP/IP protocol.

In recent years, there has been a trend to network all the monitoring systems into one subnetwork. As indicated
earlier this was the threshold in allowing the monitoring system to have their own dedicated network to exchange
data. The practice in most cases is to use an isolated subnetwork of the already existing networking infrastructure.

2.1.3. Workstation computers

Workstation computers are used for running various software components. The system utilizes standard industrial
PC or server configurations. The software distribution allows for these components to have different configuration
i.e. a more powerful server configuration for running multiple software components, or a few lesser server
configurations running a separate software component on their own.

2.2. Software

The software suite consists of components that run on different system tiers with different tasks. Software
component can communicate with each other using standardized industrial protocols (IEC 61850, IEC 60870-5-104,
DNP3, Modbus, http) and proprietary protocols. One of the development tasks was to keep the fleet monitoring
system capable to seamlessly integrate legacy systems that are based on our proprietary protocol and to enable the
legacy systems to communicate using modern protocols. The result is that the legacy system migration to the new
platform requires only software update of the controller and/or the workstation.
 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28 25
6 Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000

2.2.1. e-Trafo RT application

Application is located at the TMS controllers on the transformers. The application monitors all the data from the
transformer and communicates with the service tier. Application has data redundancy and can store data during
communication outages on its internal memory. The period for system redundancy depends on the number of
monitored quantities and their rates of change since the memory size has a fix value. The application operates in the
transformer tier.

2.2.2. e-Trafo NET service

This is the service tier application. Responsibilities of the application are communication with the TMS
controllers, servicing the client applications with data, archiving the data to the database and communication to the
supervisory control systems in the network e.g. SCADA. The application can also communicate to other TMS
service applications.

2.2.3. Database
Database application is used for long term data storage. Currently the system uses Microsoft SQL Server, but
other database systems can be supported. Database is accessed through e-Trafo NET service. Database is a service
tier application.

2.2.4. e-Trafo Fleet Web

The e-Trafo Fleet Web is the client application that enables users to get access to the monitoring systems. The
application runs in the client tier. Application can connect to multiple e-Trafo NET services at once. Figure 3 shows
an example of a main window of the application when the user logs in.

Figure 3 Fleet Web main screen example

User has a geographical map that marks the location of the substations with a basic information regarding the
current warning/alarm statuses at the substation. This main view can be configured to show more information if
necessary. If there are any indications of a fault the user can then focus on the issues by clicking on the map or by
26 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28
Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000 7

navigating the tree-menu on the left side of the main screen. Other tools such as database access and analysis tools
are available from the upper screen icons.

The web application is independent of the service nodes it connects to. The application can run on a different
workstation than the services and if the service is unavailable it will not cause the web application to malfunction.

3. Putting it all together

In the previous chapters a brief overview of the system and its components was presented. These components
combined make the Končar TMS Fleet monitoring solution. In this chapter, two examples of the fleet monitoring
system will be presented. One will be a straight forward 3 tier solution that was implemented at two utilities in
Croatia. The other will be a solution that was prepared for the integration of all the transformer monitoring systems
in a prominent foreign country into one fleet monitoring system. Networking in all solutions was done through
already infrastructure with having a dedicated subnetwork for the monitoring systems.

3.1. Small scale fleet monitoring

Two small scale fleet monitoring system were delivered in Croatia for 2 utilities. The system architectures are
described in Figure 4 and Figure 5.

Figure 4 Utility 1 system Figure 5 Utility 2 system

Both system have one workstation that hosts application from 2 tiers: service and client. From that workstation
the e-Trafo NET service connects to the distributed substations in the area. The first one has 4 transformers at 4
different substations and the second has 5 transformers across 4 substations and 1 more transformer in preparation to
be connected to the system.

The main difference between the solutions is the way that the data is transferred from the substations. The first
solution has the newest Končar cVISK controllers that are directly connected to the service application at a remote
location. The second solution has a mix of directly connected controllers (legacy NI and new cVISK) and other
workstations located at the dedicated substations that transfer data and give access to local databases i.e. transformer
and service tier in the substation.
 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28 27
8 Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000

3.2. Large scale fleet monitoring

The fleet monitoring requirements for this project presented bigger challenges. The request was to provide the
solution to integrate all the transformer monitoring systems in the state and to keep the fleet monitoring system
expandable to integrate additional systems at request. One more interesting detail was the request to integrate data
from different transformer monitoring vendors. This would be integrated by connecting to the system via industrial
protocol interfaces. Figure 6 shows the proposed architecture of the system.

Figure 6 Large scale fleet monitoring

When faced with this quantity of transformers to monitor additional system components have to be added in the
tiers to split the work load and provide system expandability. As can be seen in the configuration the service tier has
more levels of workstations. The workstations are split between service and database workstations, and the service
workstations are clustered to split load of data transfer and provide more system independence (in case of
component failure) and expandability. The client tier has also got more workstations for the web application for the
same reasons as with the service workstations. The main client workstation provides the user interface and connects
to chosen substation in lower tiers.

In this solution, the client has a special dedicated team of trained personnel for transformer monitoring that are
meant to use this system.
28 Tihomir Jaković et al. / Procedia Engineering 202 (2017) 20–28
Tihomir Jaković et al. / Procedia Engineering 00 (2017) 000–000 9

4. Conclusion

The examples in the paper show possible implementations of a fleet monitoring system. There is a greater degree
of system flexibility and modularity, since adding or removing monitoring systems is made relatively easy, and the
system can be configured to the exact customer specifications. The data is immediately accessible to end user to be
used for monitoring, maintenance and asset management.

Such architecture is scalable and its design is robust so that replacing or updating components of the system will
not affect the integrity of the fleet. The future installations of the different generations of the monitoring systems
will be seamlessly integrated into already existing network, which will add value to the end user.

As mentioned in previous chapter this architecture also allows for the integration of monitoring systems and
devices from other vendors into its fleet, allowing the users to have one unified platform to monitor the equipment
and the gathered data.

References

[1] S. Keitoue, R. Gardijan, Transformer Monitoring System – Important Component of Smart Grid, ITCE, Teheran, 2014.
[2] M. Banović, V. Cindrić, S. Keitoue, A. Keller, I. Tomić, Monitoring system for the most important transformers, HRO CIGRE, Cavtat, 2007.
[3] M. Banović, V. Cindrić, S. Keitoue, Mobile remote access to the monitoring system, HRO CIGRE, Cavtat, 2006.
[4] M. Banović, R. Gardijan, Hardware and measurements in the transformer monitoring system Končar TMS, HRO CIGRE, Cavtat, 2005.