IEEE Electrification Magazine - December 2015 - 19

calculate the GICs and the GICinduced reactive power losses in
transformers, followed by power flow
studies to analyze system voltages.
For GIC studies across a large footprint, a uniform electric field assumption is not realistic because of electric
local field variations caused by several
factors. One example of spatially varying nonuniform electric fields is the
peak benchmark electric field E peak (V/
km) for power system footprints in
North America, proposed in NERC TPL007-01 and given as
E peak = 8 ) a ) b,

(5)

Crust 0-100 km
Thick

Lithosphere
(Crust and Uppermost Solid Mantle)
Mantle

Asthenosphere

Mantle
2,900 km

Crust

Liquid
Outer Core
Inner Core

Core
5,100 km
Solid

Not to Scale

6,378 km
To Scale

where a = geomagnetic latitude Figure 8. The different conductivity layers of the Earth. [Image courtesy of USGS (http://pubs.usgs.
scalar, and b = Earth resistivity sca- gov/gip/dynamic/inside.html).]
lar. The scalar a accounts for the
xx
short-term or operational-this could include the
change in electric field strength with geomagnetic latiopening of certain lines or transformers or redispatchtude. This geomagnetic latitude differs from the geoing generation and reactive power resources to stabigraphic latitude, as it is with reference to the magnetic
lize system voltages
north and south poles of the Earth, which are not the
xx
long-term-GIC blocking devices such as the one
same as the geographic poles. The b factors model the
shown in Figure 10 could be installed in the neutrals
various conductivity zones of the United States and Canof key transformers, determined by GIC studies.
ada. Note that this resistivity factor accounts for the differIt could also mean installing additional reactive power
ent layers, as shown in Figure 8, since GICs can penetrate
support devices. The capital costs associated with these
far into the soil due to their low frequencies.
long-term measures necessitate sound GIC studies to jusThe a values range from 0.1 to 1, whereas the b values
range from 0.21 to 1.17 for the different resistivity zones.
tify their adoption.
This benchmark scenario can be studied for a given footOperational strategies can be studied by changing the
print by setting the electric field as 8 V/km in the electric
network topology, parameters, and putting elements
field dialog shown in Figure 3. The a and b values are also
in/out of service, using the models and tools described so
available in the software and are referenced by geocoordifar. GIC software tools also include an option to set a
nates. Hence, the appropriate values are used in a study
based on the geocoordinates of the substations in the
power system case. Figure 9 shows GICs calculated across
a model of the North American Eastern Interconnect,
using the benchmark electric field scenario.
This peak electric field may be applied in different
directions to determine the worst-case direction, for
example, the one that causes the highest reactive power
losses or largest dips in voltages. The Quebec blackout
during the March 1989 GMD event was mainly caused by
the GIC-induced harmonics tripping reactive power support devices such static var compensators (SVCs). GIC
power flow studies with the loss of such devices and other
system elements could be done. Harmonic analyses could
be performed to determine which SVCs or shunt capacitors are likely to trip during GMD events.

GIc mitigation
GIC studies can also be helpful in determining appropriate
strategies to mitigate the system impacts of GICs. Two key
types of strategies could be

Figure 9. The benchmark peak electric field applied to the Eastern
Interconnect model, with resulting GICs marked by yellow arrows.

IEEE Elec trific ation Magazine / d ec em be r 2 0 1 5

http://www.pubs.usgs

Table of Contents for the Digital Edition of IEEE Electrification Magazine - December 2015