This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
A. Nesbitt, B.G. Stewart, S.G. McMeekin, S. Conner, J.C. GamioSchool of Engineering and Computing, Glasgow Caledonian University, Glasgow, UK.
L. Moodley, K. Govender
Doble Engineering Africa, Durban, South Africa
K. Liebech-Lien, H.O. Kristiansen
Doble TransiNor AS, Trondheim, Norway
ABSTRACT
A substation asset management strategy is key to ensuring that critical substation apparatus operate reliably, meetavailability requirements and provide adequate capacity for the future. This depends heavily on predictive
maintenance assessments conducted periodically on the apparatus for signs of deterioration. Radio frequency
interference (RFI) measurement offers a front line non-invasive technique to assess the condition of individual high
voltage (HV) apparatus as part of a substation surveillance program.
This paper presents a case study that demonstrates the precautions and procedures to record and map the RFI
emissions observed within a strategically important 400kV substation switchyard and assess for RFI disturbances
that are characteristic of potential insulation degradation. The results were recorded to allow future trending and
therefore facilitate an assessment of individual HV apparatus insulation over time. In the course of conducting the
survey, three sites of potential insulation degradation were discovered, the more serious of which was localised,
characterised and identified as a fault condition by analysing and correlating the measured frequency spectrum with
the observed pulsed behaviour of the RFI emissions.
INTRODUCTION
High-voltage (HV) substation apparatus is subjected to electrical, mechanical, and thermal stresses as well as
environmental conditions. These stresses act to accelerate the deterioration of the insulation and hence the electrical
integrity of the HV equipment eventually leading to failure. A substation asset management strategy is key to
ensuring that critical substation apparatus operate reliably, meet availability requirements and provide adequate
capacity for the future.
In order to use the existing equipment most efficiently, system operators need better operation and diagnostic
support tools to identify developing or incipient problems, and for longer-term asset management. An asset
management strategy depends heavily on predictive maintenance assessments conducted periodically on the
equipment. This approach adds value to the maintenance work that is actually conducted on each asset by ensuringthat the asset actually needs the service. This approach has the effect of limiting the overall work scope
accomplished on each asset and ensures that only necessary work is accomplished.
The deployment of radio frequency interference (RFI) measurement has gained increasing acceptance as a front line,
non-invasive technique allowing engineers to assess the condition of individual HV electrical apparatus as part of a
substation surveillance program [1]. It has long been observed that corona discharge pulses and micro-gap
discharges occurring on high-voltage power transmission and distribution lines could disrupt radio and television
reception [2, 3,4]. This noise was frequently the result of a defect in the system, which, unchecked, sooner or later
could result in an outage. RFI scanning can alert the engineer to the presence of possible sites of insulation
signals. The discrimination of these signals from the narrowband sources is crucial to the process of recognising the
presence of a discharging source. Average detection in an EMI instrument is intended to recover and measure the
amplitudes of narrowband signals in the presence of impulsive broadband noise.
The provision of dual peak and average detectors in an RFI instrument can provide the additional information at the
point of measurement to allow practical discrimination between the narrowband characteristics of broadcast RF
emissions and the broadband impulsive RF emissions from insulation defects. If there is little or no difference in the
observed amplitudes of the peak and average measurements at specific frequencies then the signal can be consideredto be narrowband. If there is a substantial difference in the amplitudes where the average value is considerably
lower, then the signal can be considered to be broadband in nature. Figure 2 illustrates how ambient RFI and
impulsive discharge signals are discriminated by observing and comparing the output from the peak and averages
detectors.
Discriminating Between Ambient RFI and Impulsive Discharge Signals.
Figure 2
Measurement of Low Repetition Rate RFI
For low repetition rate phenomena, the challenge of conducting an RFI measurement is to optimise resolution
bandwidth, frequency span, sweep time and measurement duration, and the selection of detector type to ensure a
high probability of signal detection and accurate measurement of their amplitude and frequency [5].
Emissions from partial discharge events cover a frequency spectrum wider than the receiver IF bandwidth.
Therefore the frequency spectrum is resolved by conducting a frequency sweep and constructing the pulse spectrum.
Frequency resolution or step size is defined by the resolution bandwidth RBW filter. Once the receiver is tuned to a
different frequency at each instant of time, there is a finite probability the signal will physically be present and
therefore detected and measured. With low pulse repetition rates the pulse spectrum can be resolved as long as the
total measurement time is sufficient to ensure there is convergence of the pulse spectrum. Otherwise, signals will be
undetected at specific frequencies because they were not present during the measurement. To increase the likelihoodof the presence of the signal at a spec ific tuned frequency, the receiver must „dwell‟ at each frequency for a time and
apply „MAX HOLD‟ over a number of scans. In substation measurements the „dwell time‟ or „gate time‟ has a
default setting of 40ms, however 80ms is used if repetition rate is low.
EMI standards reference CISPR 16-1:1999 defines the specifications for EMI receivers. The standard specifies the
design parameters for the peak, quasi-peak, rms and average detectors [16]. The quasi-peak detector is mandatory in
the determination of compliance with legal national and international EMC standards [14,15]. The quasi-peak
detector provides a response that simulates the human perception of radio frequency disturbances on broadcast radio
receivers resulting in the weighting of broadband impulsive signals as a function of their repetition rate. However, it
Site Mechanical Layout Plan (with survey points marked)
Figure 5
RFI Disturbance Mapping – North Eastern Section of 400kV Switchyard
Planning and conducting a RFI disturbance map of a substation must consider the main overhead busbar and
electrical bay configuration. Therefore organizing a site mechanical layout plan before visiting the site is an
advantage. Points of measurement are located along overhead busbar routes and at intersections of the electrical
circuit bays.
An overlay plot of the RFI measurements taken along and underneath „Busbar 1‟ in the north eastern section of the
substation is provided in Figure 6, the survey locations marked on the site mechanical layout plan, Figure 5, refer to
the relevant stored trace filenames. This represents the observed RFI emissions in this section of the switchyard.
In general the plotted RFI emissions show a 10dB uplift at frequencies up to 200MHz in comparison to that
observed from the baseline measurement taken at survey point TRAC0298. This increase may be attributable to the
loud and audible levels of corona that are pervasive across the site and to the level of pollution on the apparatusbushing insulators. As the measurement point is moved towards electrical circuit bay „BUS COUPLER B‟, a
broadband uplift in RFI amplitude can be observed at frequencies above 200MHz. More notability at frequencies
beyond 335MHz and at 580 – 750MHz where the amplitude peaks at -53dBm. The broadband nature of the uplift
and the range of frequencies over which this is observed, highlights a potential site of insulation degradation that is
not attributable to the audible level of corona that is pervasive across the site. The upper range of frequencies this
uplift is observed appears to be 750MHz which at present suggests the presence of a site of insulation degradation.
Note, extensive localization and characterisation of the site of degradation did not place at this time.
An overlay plot of the RFI measurements taken along circuit bay „BUS COUPLER A‟ is provided in Figure 9. The
comparison of RFI traces and the relative uplift in measured amplitude at the higher frequency ranges suggest thelocation to be close to the intersection of „BUSBAR 2‟ and „BUS COUPLER A‟, in the vicinity of measurement
point TRAC0317. It can be observed from Figure 9 that the broadband nature of the RFI emissions extends up to
1000MHz. A peak amplitude of -55dBm is recorded at 1000MHz.
Exploiting the high levels of attenuation observed at these frequencies, provides a very effective means of localising
the source of the RFI and hence the site of degradation. Using a spot frequency of approximately 900MHz, the
source of RFI emissions was located and identified as Bus Coupler 2A isolator – phase Blue. PHOTO 2 and 3
provide visual identification of the suspect apparatus.
The Doble PDS100 unit was deployed to conduct an RFI emissions survey and mapping exercise. The results are
recorded to allow future trending and therefore facilitate an assessment of individual HV apparatus insulation over
time. In the course of conducting this survey, three sites of potential insulation degradation were discovered, the
more serious of which has been localised, characterized and identified as a fault condition on „BUS COUPLER 2A‟
Isolator – blue phase. The other potential sites were not investigated at this stage.
It was recommended that the identified fault condition on „BUS COUPLER 2A‟ Isolator be monitored further to
establish the seriousness of the fault condition and whether it poses a significant risk. A recommended first stage
action was to deploy an infrared camera to assess the thermal signature of the apparatus.
Further recommendations were to allocate resources to locate and characterise the nature and severity of the other
two sites of degradation, one in the vicinity of „CAPACITOR 11‟, the other in the general vicinity of electrical bays
„BUS COUPLER B‟ and „STATIC COMPENSATOR 1‟.
Subsequent feedback from the Client reported that a Corona camera detected problems with this isolator in that there
were problems with the shed extenders and silicon coating that was used to increase the creepage distance from 21
mm/kV to 31 mm/kV. A thermal camera was deployed but no hot spots were discovered. It was also reported that
the isolator support insulator did flashover in the past. The isolator in question has been placed on a plant
replacement program as a result of the RFI survey that has taken place with further evidence that it has caused
problems in the past.
Subsequent communications from the Client also indicated that a surge arrester had failed in the vicinity of „BUS
COUPLER B‟ and „STATIC COMPENSATOR 1‟ electrical bay before additional resources were allocated to
localise and characterise the suspect apparatus. The RFI survey had provided warning that a site of insulation
degradation was present in that vicinity.
REFERENCES
[1] P.J. Moore, I.E. Portugues and I.A. Glover, “Radiometric location of partial discharge sources on energized
high-voltage plant”, IEEE Trans. on Power Delivery, Vol. 20, No. 3, July 2005.
[2] E.N. Limberger, “Locating radio frequency interference on power transmission lines”, 36 th
Annual Conf. on
Rural Electric Power , Vol. C, May 1992, pp 1-5.
[3] K. Arai, W. Janischewskyi and N. Miguchi, “Micro-gap discharge phenomena and television interference”,
IEEE Trans. Power Apparatus and Systems, Vol. 104, No. 1, Jan 1985.
[4] M.O. Loftness, “Power line interference – sounds, patterns and myths”, IEEE Trans. Power Delivery, Vol. 12,
No. 2, April 1997.
[5] A. Nesbitt et al, “A novel approach to high voltage substation surveillance using radio frequency interference
measurement”, IEEE Electrical Insulation Conf. Proc, Montreal, Canada, May 2009, pp. 159 – 163.
[6] E. Gulski, “Discharge pattern recognition in high voltage equipment”, IEE Proc. Sci. Meas. Technol., Vol. 142,
No. 1, January 1995.
[7] G.P. Gleary, M.D. Judd and O. Farish, “Investigation of partial discharge current pulses in transformer insulating oil”, Proc. Int. Conf. on Advances in Processing, Testing and Application of Dielectric Materials,
Wroclaw, Poland, Sept 2001, pp. 50 - 53.
[8] T. Babnik, R.K. Aggarwal, P.J. Moore and Z.D. Wang, “Radio frequency measurement of different
discharges”, IEEE PowerTech Conf. Proc., Vol.3, June 2003
[9] M. Hikita et al., “Electromagnetic noise spectrum caused by partial discharge in air at high voltage
substations”, IEEE Trans. on Power Delivery, Vol. 13, No. 2, April 1998.
Alan Nesbitt, Glasgow Caledonian UniversityAlan Nesbitt graduated from Queens University Belfast with a BSc in Electrical and Electronic Engineering (1
st
Class Honours) and then with an MSc in Digital Techniques from Heriot-Watt University, Edinburgh. He has over
24 years experience of high performance embedded systems design, firstly in the avionics industry and latterly
leading University collaboration with industry. Latterly over the past 10 years his attention is to the fundamental
process of partial discharge (PD) and the characterization of failure modes through academic research and
consultancy for clients. He was lead investigator for Glasgow Caledonian University on a number of industriallyfunded projects on the investigation of novel methods for detecting and measurement of deterioration in insulation
materials in power plant. He is currently engaged in transferring academic knowledge to industrial through the
development of the next generation of condition monitoring instruments.
Brian Stewart, Glasgow Caledonian UniversityBrian G Stewart graduated with a BSc (1st Class Honours) in 1981 and PhD in 1985 both from the University of
Glasgow and is currently a Professor in the School of Engineering and Computing at Glasgow Caledonian
University. Brian has been involved in the condition monitoring of high voltage plant for the past 10 years and has
over 20 years of research and design experience of the electronic systems. His research interests are in the areas of
partial discharge, high voltage condition monitoring and communication systems. He has carried out numerous
industrial consultancies and reports on HV and partial discharge measurement.
Scott McMeekin, Glasgow Caledonian UniversityScott McMeekin has worked on the design, fabrication and characterisation of photonic devices for the past 20
years. He was appointed as a Lecturer in the School of Engineering Science and Design at Glasgow Caledonian
University in 2004 and promoted to Reader in 2007. His present research activities include next generation digital
communication systems, condition based monitoring systems and optical systems. Prior to joining Glasgow
Caledonian University he was the Process Development manager at Alcatel Optronics Ltd (formerly Kymata Ltd)
where he was responsible for the development and qualification of novel optical components for advanced optical
telecommunication systems. Previous positions have included being a Lecturer in the Cardiff School of Engineering
at the University of Wales, Cardiff from 1994 to 2000 where his research activities included the fabrication of sub-
micron electronic and optoelectronic devices, visible lasers, and BioMEMs structures.
Carlos Gamio, Glasgow Caledonian UniversityJ. C. Gamio received the B.S. degree in electronic and communication engineering from the Mexico National
Polytechnic Institute, Mexico City, in 1989 and the M.Sc. degree in instrument design and application and the Ph.D.
degree in electronic engineering from the University of Manchester Institute of Science and Technology (UMIST),
Manchester, U.K., in 1993 and 1998, respectively. He is a member of the Institution of Engineering and Technology
(IET) and the Institute of Physics (IoP). From 1984 to 2005 he was with the Mexican Petroleum Institute, Mexico
City. During that period, he designed a sonar-based tool used to measure man-made underground oil-storage caverns
and undertook research on multiphase flow imaging using capacitance tomography. In 2005 he joined Glasgow
Caledonian University where he designs novel instrumentation systems for condition monitoring of high-voltage
plant, including an optically-isolated partial discharge sensor and a radio-frequency based portable partial discharge
detector.
Steven Conner, Glasgow Caledonian UniversitySteve Conner received his PhD from the University of Strathclyde in 2003. From 2003-2007 he worked at Optosci
Ltd., where he designed a tunable diode laser spectroscope for gas leak detection. In 2007 he moved to Glasgow
Caledonian University to develop a portable RF instrument for partial discharge detection. He has also worked as a
part-time consultant on many projects, including precision laser diode drivers for the telecoms industry, audio
mixers for Numark, LLC., and a computer-controlled Tesla coil display at Danfoss Universe, and is named as aninventor on several patent applications.
Luwendran Moodley, Doble Engineering AfricaLuwendran Moodley graduated from the University of Natal with a BSc in Electrical Engineering. Upon his
graduation he joined Durban Electricity as an Engineer in training. He held the positions of Technical Support
Engineer, Manager: Technical Support and finally the Manager: Transmission Substations within Durban
Electricity. In 2007 he joined Doble Engineering and manages the Doble Africa offices based in South Africa. He is
also currently involved in transformer condition assessment.
Kamendran Govender, Doble Engineering AfricaKamendren Govender graduated from the University of Natal in 2006 with BSc in Electrical Engineering. Upon
graduating he worked as a Candidate Engineer in Durban Electricity. In 2008 he joined Doble Africa as an Engineer.
He is predominantly involved in transformer condition assessment work for the utilities in Africa.
Kjetil Liebech-Lien, Doble TransiNor ASKjetil Liebech-Lien has been with Doble TransiNor since May 2006. His current position is Product Support
Engineer and his responsibilities are within product support, product training, field & lab testing and product
development & modification. Kjetil graduated from Norwegian University of Science and Technology with M.Sc. in
Electronics in 2005. Since joining Doble, Kjetil has gained extensive world wide field experience and product
knowledge for in-service testing of surge arresters, cable terminations and SF6- insulated systems, and he has been
co-author and presenter of papers at conferences like SIPDA, ICOLIM and GCC Cigré.
Hans Ove Kristiansen, Doble TransiNor ASHans Ove Kristiansen has been employed with Doble since January 1st, 2008. As Product Manager he is responsible
for the acoustic online insulator analyzer, surge arrester leakage current tester and insulation pollution monitor. Hans
holds a degree from Trondheim Technical College, Electronics and Automation 1986 and Norwegian Institute of Technology, Electrical Engineering in 1991. He started his own company in 1992 called Stretek AS where he
worked as Engineering Manager, developing electrical drives for HVAC. From 1995 he was employed as sales
manager for a company in Trondheim selling pumps and electrical drives until he started as Service Manager at
Metron AS in 2003. This company was acquired by Fluke Corp in 2005. Hans left the position as Global Technical
Support Manager in 2007 to join Doble TransiNor AS.