IMPROVEMENT AND VALIDATION OF THE GEOMAGNETICALLY INDUCED CURRENT

Engineering Research for Excellence in Industrial

MODEL FOR THE 220KV-400KV NAMIBIAN POWER NETWORK
Hiskiel Stephanus & Gideon Gope
Department of Electrical and Computer Engineering, Polytechnic of Namibia, Windhoek, Namibia

Abstract Results 0.018

Single-Sided Amplitude Spectrum of the measured GIC signal

0.016

Geomagnetically Induced Currents (GICs) are Previous GIC models for the same High 0.014

Voltage(HV) network as presented in The simulation yielded the following GIC time 0.012

results of interactions between eruptive

variations at different nodes in the network. 0.01

Power
(Dax, 2014) utilised peak electric field
plasma from the sun and the earths 0.008
intensity as provided for in the Electric
protective magnetic field. The unavailability of GIC for 2013-07-24 0.006

Power Research Institute(EPRI) 1

Auas 0.004
GIC measurements for previous GIC studies OpenDSS. In this particular case, the
0

-1 0.002
has not allowed for the validation of previous instantaneous values of the electric field
0 500 1000 1500

GIC(Amperes)
0 1 2 3 4 5 6 7 8 9 10
GIC models. In this study, improvements were intensity were not applied. In this paper,
0.5

0
Obib
Frequency (Hz)

introduced in the modelling by applying the the use of the Natural Resources Canada -0.5

NRcan simulator to the magnetic field data (NRcan) simulator allows for the use of 0.5
0 500 1000 1500

Conclusion
Ruacana

measured in Namibia. Due to the absence of the instantaneous magnetic field data as 0

-0.5

surface impedance measurements in recorded from the observatory. 0 500 1000 1500

The study found a significant correlation

Namibia, the derived homogeneous 2

0
Kokerboom
between the polarised GIC measurements
Keetmanshoop surface impedance was Materials and Methods -2
0 500
Time in Minutes
1000 1500 and the modelled GIC values. Significant GIC
utilised. Validation was done by comparing magnitude values were observed at the Auas
model results to measured GICs from the The NRcan GIC simulator compiled up by The following directional sensitivities for substation with peaks close to 1A during the
Obib and Ruacana substations. A significant various network nodes were also yielded. K<4 storm and a peak of close to 8A during
Canadian researchers was used for GIC
correlation between measured and modelled the K=7 storm. GICs are essentially quasi in
modelling. The simulator has inputs such nature and are associated to frequencies
GICs when using polarised GIC values. Large as the magnetic field variations as well below 1Hz. A fast Fourier transform of both
GICs were modelled at the Auas substation network components. It then computes the the measured and modelled GIC signals has
during both the K<4 and K=7 storms. geo-electric field model using the Improved shown large magnitudes of low frequency
Introduction FFT method. The operation of the components. Phase spectrums of the two sets
simulator is summarised in the figure of signals indicated that the measured signal
below. is 180 out of phase from the modelled signal
During solar storms, development of due to the symmetric nature of the modelled
geomagnetic storms; as a result of erupting signal.
solar activity is observed as solar flares and References
Coronal Mass ejections. These two activities
interact with the earths magnetic field and
1. D. H. Boteler, L. Trichtchenko, R. Pirjola, J. Parmelee, S.
outer atmosphere in complex ways such that Souksaly, A. Foss and L. Marti, Real-time simulation of
The validation of the simulation model
they lead to the development of an electric geomagnetically induced currents, in VII International
produced the following correlation between Symposium EMC'2007, St Petersburg, 2007.
field on the earths atmosphere that thus GICs measured at the Obib substation and 2. J. Koen and C. Gaunt, Disturbances in the Southern
induces variable values of DC voltages, modelled GIC values. African power network due to geomagnetically induced
currents, in Cigr session, Paris, 2002.
causing Direct Currents known as 0
Polarised Modelled and Measured GIC variations
0.4 3. W. Chandrasena, P. G. McLaren, U. D. Annakkage and R. P.
Geomagnetically Induced Currents (GICs) to measured GIC Jayasinghe, An Improved Low-Frequency Transformer
GIC(Amperes)

GIC(Amperes)
-0.005 modelled GIC 0.2
Model for Use in GIC Studies, IEEE TRANSACTIONS ON
flow in transmission lines. GIC flow in the -0.01 0 POWER DELIVERY, vol. 19, no. 2, pp. 643-651, 2004.
network causes half cycle saturation of -0.015 -0.2
4. D. Boteler, KuanZheng, L.-g. Liu and R. Pirjola, Calculation
Analysis of Geomagnetically Induced Currents with Different
transformers which often creates transformer -0.02
0 200 400 600 800 1000 1200
-0.4
1400 Network Topologies, IEEEXPLORE, pp. 1-4, 2013.
hotspots leading to local or worse, Time in Minutes 5. K. Dax, Quasi-DC current model of the 220kV-400kV
Magnitudes of modelled and Measured GIC variations
0.02 0.4 Namibian Power Network BEng Research Project report,
widespread blackouts. GIC modelling is a tool The NRcan simulator then calculates the measured GIC Polytechnic of Namibia, Windhoek, 2014.
GIC(Amperes)

GIC(Amperes)
0.015 0.3
modelled GIC 6. J. Koen and C. Gaunt, Geomagnetically Induced Currents
used to investigate the susceptible prevalence DC source voltage. This is used as the in the Southern African Electricity Transmission Network, in
0.01 0.2
of GIC flow in a power network. A model is source for driving GIC in the transmission 0.005 0.1
IEEE Bologna, PowerTech, USA, 2003.
developed as an approximation of the network network. This is carried out using the 0 0
0 200 400 600 800 1000 1200 1400
behaviour and often includes discrepancies procedure below. Time in Minutes

introduced by the assumptions taken during

the modelling. No GIC model is ideally
perfect; however the extent to which the Single sided Frequency spectrums of the
Vdc E dl measured and modelled GIC signals were
model represents the behaviour and flow of R
also obtained through Fourier
GICs in the network can be validated through EL transformations.
physical measurements from the network. In

Research Day 2015

this study, GIC measurements from Namibia
were used for the validation. E X LX EY LY

17 July 2015
School of Engineering