Nordic Insulation Symposium - Nord-IS 13 - Trondheim, Norway, June 9 - 12, 2013

N. L. Fantana
ABB Corporate Research, Germany

considering both the changes in the condition of

equipment versus time, the blue condition curve, and the
In recent years there has been an increase in the related costs to restore the equipment condition, the red
availability of on-line condition monitoring devices and cost curve. The degradation starts at point S and is
systems and of monitoring installations in utility detectable at point D in time. The detection in D should
networks. Their installation is judged considering a lot allow reacting before failure occurs. The reaction time
of factors including: general utility strategy, expected available to act should be less than the time from
value from such an installation, asset and risk degradation detection to failure t_DF. The longer this
management, utility best practices, economic, time the better additional actions such as additional
environmental or regulatory pressures, and will vary diagnostics to get more details or better localize
from one utility to another. What is the present problem, perform maintenance or major repair can be
experience, the value from on-line condition monitoring planned and executed. The earlier the detection of
and what are the opportunities and challenges in future degradation is the lower the cost to restore equipment
grids and considering modern technologies? The paper condition and to limit the damage in the network. The
intends also to position on-line condition monitoring in degradation detection point is continuously evolving
life strategies and point to evolution and international with technology and knowledge.
activities related to the topic.

How to reduce equipment failure and minimize

consequential costs? This is a key issue in utilities and
for all power equipment users. Presently a set of factors
come together contributing in various ways to make the
situation in utilities especially challenging:
- Aging network equipment, average age is
continuously growing, 30, 35, 40..
- Economic constraints in utilities
- Required high availability of power supply.
- Increased loading, go-to-the-limit situations
- Changes in power flows, especially due to
distributed generation,
- The need to know - if possible instantly, in real
time - the situation in the network and status of
the equipment.
Increasing efforts have been done to respond to these Depending on the equipment or subsystem and the
challenges by advancing diagnosis and monitoring physical / chemical degradation mechanism, the
research, by using diagnosis and on-line condition materials used, construction details, etc., the time from
monitoring (OLCM), by having internal utility detection of degradation to failure occurrence, can be
strategies for asset management or for installing and very different. Cases exist where we talk about few
using OLCM systems. Installation of OLCM systems hours from detection to failure but also cases exist
has been done with new power equipment installations, where it can be considerably longer. Such rapid
but also for aged power equipment still in operation. evolution times can only be handled by on-line
It is important to position the OLCM in an integrated condition monitoring.
life concept, to understand the failure mechanisms and The degradation evolution shown in Fig. 1 highlights
evolution of failures and to highlight the value and the importance of detection of the condition degradation
importance of OLCM systems for an utility. as early and as precise as possible, such that a
meaningful action can be taken by operator or
maintenance team to minimize damage to the electrical
equipment and network. Early detection and precise
Degradation occurs over power equipment lifetime detection are both challenging and contradictory. Efforts
under normal operation and in case of special events. A are needed to avoid false alarms. A continuous
simplified, principle example is shown in Fig. 1. It is monitoring can be decided and put in operation after

63
 Nordic Insulation Symposium - Nord-IS 13 - Trondheim, Norway, June 9 - 12, 2013

considering a set of factors such as: actual network and taking into account specific degradation mechanisms
economic constraints, the equipment condition, the time and monitoring requirements.
constraints regarding replacement or spare provision,
the importance of the unit and the length of the time
from degradation inception to failure,
OLCM is a good mean supporting degradation detection
and failure avoidance, however at best it should be
integrated in a holistic context considering equipment
and network. Such a strategy for asset supervision and
management has to consider beside economic aspects
i.e. the importance of the equipment in the network, the
condition of the equipment. Such condition related
information can be derived based on few major
information pillars, such as: equipment history, e.g.
operation and maintenance, information from diagnosis
tests performed, information from OLCM systems,
expert knowledge and manufacturer’s expertise, Fig. 2.
Importance of OLCM is given by the value obtained by
applying it for the supervised equipment asset and for
the whole electrical network.
The aspect of “Obtaining value from OLCM” was
addressed in detail by the CIGRE working group B3-12
[4]. One important aspect which has been investigated
was to identify the main drivers to install condition
monitoring. The results found are shown in Fig. 3. Cost
saving as the major driver was expected, however
company image is noticeably the second highest
motivation, and it is linked to supply quality, reduced
outages and customer satisfaction.

Statistical information is also important to get feedback

from installed equipment and to set priorities of deeper
analysis or research and development of diagnosis and
OLCM methods, devices and sensors for degradation
detection and avoidance of failures. An excerpt from
failure statistics for power transformers is shown in
Table 1, [1],[2]. Failure statistics show themselves an
evolution over time and have intrinsic uncertainties
depending on the available data sources, still can give a
good picture. The results of the recent, transformer
failure survey from CIGRE, from about 1000 heavy
transformer failures from 1996-2010, see column
“CIGRE2012” in Table 1, show an increase of the
winding failures compared to previous statistics, an
approximately constant percentage of core failures from
total, some increase in tap changer failures, and a
varying evolution in bushing failure. They hint to the Maintenance cost reductions and reduced call out of
necessity to increase effort in diagnosis and monitoring staff are also important especially in cases where
to consider active part and insulation for diagnostics distances to installed equipment are long and terrain is
however continue efforts on monitoring and diagnosis difficult. The extension of life although on rank 4 is
of other major equipment parts like bushing or on-load often encountered in practice as a typical case of
tap changer are needed too. monitoring. A unit is often kept in operation, supervised
Similar statistical data exist for other substation by a monitoring system, e.g. until replacement unit or
equipment and recent publication of CIGRE study part becomes available. Finally, regulatory aspects can
committee A3 gives information for high voltage play an important role especially when they may require
equipment such as circuit breakers and GIS [3]. OLCM monitoring installations for important equipment for
is applied to such high voltage equipment too, however improved network operation.

64
 Nordic Insulation Symposium - Nord-IS 13 - Trondheim, Norway, June 9 - 12, 2013

Another important aspect regarding OLCM systems is historical data is leading to another problem related to
related to the question where the value/benefit of such a access do data during operation, on the necessary rights
monitoring installation comes from. How this aspect is and mechanisms needed in the organization and the
perceived by utility experts, according to the mentioned information technology (IT) infrastructure. The access
survey, is shown in Fig. 4. The highest value is data and the access rights have become critical especial
perceived from medium-term activities. One such after year 2000. These are linked to the cyber security
situation could be for example change of maintenance and are intensively discussed and considered in utilities
practices, e.g. going away from the time based to a and on country level. Access and use of data from
condition based or risk based maintenance. monitoring, together with data from operation is
technically meaningful, however not easy to achieve.
Important is especially to have secure network
operation, secure operation of the protection and control
systems, without being disturbed in any way, even
accidentally by the OLCM users. All these are guided
by utility IT security policies and restrictions.
The OLCM system is perceived as bringing value but
still as an valuable “add on” in most cases. However the
trend is toward considering OLCM system as a sort of
basis system a s sort of “insurance”, giving an early
warning and helping to avoid unplanned outages and
major failures.
It is difficult to estimate a priori the value of and OLCM
installation. In [4] a set of methods are indicated , which
can help estimate this for an OLCM project.

Second highest value from OLCM was perceived in the The OLCM systems are evolving and evolution is
long term aspects related to replacement of units, long driven mainly by
term strategies and planning, as well as in accumulating - The evolution of the grids and requirements
internal knowledge related to equipment long term - Evolution of user expectations and constraints
behavior and failure. Also, value from OLCM has been - The evolution of microelectronics, information,
found in support for operation, especially by supporting communication technology (ICT), sensing and
immediately operational decisions and using better other enabling technologies.
actual equipment capabilities. Some utilities are - Evolution of the equipment and degradation
considering OLCM similar to having supervision by a knowledge and data analysis.
(virtual) person on site 7 days a week x 24 hours a day, Evolution related to the future grids with their increased
and see the benefit of OLCM systems in the fact that complexity, changed operating conditions e.g. in terms
they can detect and inform timely, do not need humans of unpredictable and strongly fluctuating power flows,
on site and get info from distributed locations across the more use of power electronics, are expected to require
electrical network. In most cases the data from more real time supervision. Additional processing,
monitoring system is used as an early warning. The storage and communication functionality and also
warning can then triggering diagnosis or more detailed sensing capabilities are expected to be required e.g.in
assessment of the equipment condition based on all life terms of processing the collected information local on
data and knowledge and expertise available. The OLCM the monitoring device.
system is best positioned as an integrated part of the Expectations of the user are also continuously evolving.
utility strategy for life long asset management, In a recent CIGRE survey [3] user expect in few years a
maintenance and operation support. There are few significant move from the present situation where a
concerns related to OLCM too. One concern is related substation equipment expert is analyzing the data,
to the much shorter life of a monitoring system versus towards more automatically processing data and
the life of primary equipment such as transformer or involving a specialized condition monitoring expert
circuit breakers. This has to be considered in relation to more strongly in the process.
the life and maintenance of such a system and sensors to Evolution in OLCM will come also from technology
keep its required functionality for decades. advances like microelectronics and microsensing or
Secondly, there is the problem of the considerable communication and IT. Such advances will have a
amounts of data which can be accumulated over long considerable impact allowing new and more precise
periods of time. This problem is requiring ways of sensors, additional processing and storage capabilities,
systematical data storage, and methods for the analysis better communication and integration in the substation
of collected data of processing and extraction of environment, sharing data with the substation SCADA
relevant information. The access to on-line data or system or with the utility central system. The

65
 Nordic Insulation Symposium - Nord-IS 13 - Trondheim, Norway, June 9 - 12, 2013

monitoring systems will allow continuous access to - CIGRE WG B3.12 “Obtaining value from
information from power devices and inform the network substation on-line condition monitoring”, finalized,
operator, maintenance team or other utility group [4], TB 462, 2011;
specified, as well as feeding central utility data stores - CIGRE WG B3.26, Guidelines for the Design and
with desired data. How a group of experts see the role Construction of AC offshore substations for Wind
and position of OLCM in future network is shown in Power Plants, finalized, TB 483, 2011, considers
Fig. 4, based on a survey [5]. monitoring aspect for this special situation;
- CIGRE WG A2.44, Transformer intelligent
condition monitoring, work ongoing, expected
technical brochure 2014;
- IEEE PC57.143, IEEE standard, Application of
Monitoring of Liquid-Immersed Transformers and
Components, finalized, published 2013
- CIGRE WG B2.36, Guide for Application of Direct
Real-Time Monitoring Systems, finalized, TB 498,
2012, considers real time monitoring of overhead
lines, mainly the “ampacity” aspects.
- CIGRE WG B5-05, Modern Techniques for
Protecting, controlling and monitoring power
transformers, finalized, TB 463, 2011
Interestingly the number of CIGRE working groups
dealing with monitoring grew from 2, before 2000, to
11 after year 2000, showing a considerable increase in
the interest towards monitoring of power equipment.
An essential evolution factor in OLCM will come from
the increasing knowledge on degradation phenomena
and materials. Also advanced data analysis algorithms
will contribute to make good use of collected data.
Other possible changes could stem from the integration On-line condition monitoring systems are an important
of some basic monitoring functionality in the protection element in present networks and will play an important
devices or on substation and network level. Such a so- role in future grids in avoiding failures, reducing costs
called network monitoring has been described in [4] and allowing a reliable operation. Monitoring and
allowing an integrated look at a network wide diagnosis are both essential parts utility asset
phenomena by using all available installed network management and life strategies. The value of OLCM is
supervision or recording devices, however it requires a in most cases seen as a support for decision making. In
considerable effort in collecting, storing and processing case of medium and long term strategies decision
of data from the whole network devices. support comes from combining for decision support
In the view of the WG B3-12 [4] OLCM is expected to information on the network and from operation and
play an important and central role in future grids. events, use OLCM information, performed diagnoses
However there is also a downside of growing condition and maintenance over time. The more experience will
monitoring and supervision, and there are some be accumulated by the world-wide engineering
challenges and aspects to be considered: community the and the more major failures will be
- lack of skilled personnel to deal immediately with a avoided, the more the question of importance and value
large number of such systems from on-line condition monitoring will get an positive
- organizations difficulties to deal with the huge answer.
amount of data which can easily be collected,
- security regarding accessing and using the
monitoring data, cyber-security aspects,
- dealing with lifetime aspects and integration. [1] S.Tenbohlen et.al. Trends of the Diagnosis of High
The evolution and status and best practices regarding Voltage Equipment, ETG 2012, Fulda, 15-16
on-line condition monitoring systems are subject of 11.2012, ISBN 978-3-8007-3465-8
investigations and work in various international [2] A. Wilson, personal communication
organizations. Such activities in committees or working [3] C. Sölver, M. Runde et.al, “Final report on 2004-
groups are fully dedicated to on-line condition 2007 International Enquiry of High Voltage
monitoring or contain parts related to OLCM. Some Equipment”, CIGRE TB 509, 2012
activities in CIGRE and IEEE dealing with on-line [4] A.Mackrell, N.Fantana, et.al. , “Obtaining value
condition monitoring are mentioned below including from on-line substation condition monitoring”,
their status, as examples of activities related to OLCM CIGRE TB 462, 2011
and without claiming completeness: [5] N.Fantana, et.al. , Expert round table IEEE-CIGRE
on OLCM, IEEE PES Substation meeting,
Raleigh, 2012

66