IEEE Power & Energy Magazine - May/June 2018 - 29

inserted into the loop, sometimes conducting the diagnosis
with a power flow model. In this section, we discuss two
such use cases utilizing the µPMu data collected from Riverside, california.

Example 2
figure 2 depicts the synchronized voltage transients during an
animal-caused short-circuit fault event. the fault occurred on
a lateral not too far from the substation, affecting mostly one
phase. only the affected phase is shown in this figure.
voltage phasors are measured by three µPMus at three
locations. µPMu 1 is installed toward the end of a lateral
(not the faulted lateral) on the faulted feeder. the shortcircuit fault momentarily brings the voltage down to zero
at the location of µPMu 1, causing a brief interruption, as
shown in figure 2(a). µPMu 2 is the point of common coupling of the faulted feeder at the substation. the short-circuit
fault creates a severe voltage sag at this location, as shown
in figure 2(b). finally, µPMu 3, installed on another feeder
several miles away from the faulted feeder, records a considerable voltage sag, as shown in figure 2(c), albeit much less
severe than what µPMu 2 has captured.
this type of synchronized distribution grid voltage data
can be used to calibrate the coordination of the distribution
system protection devices. But, more importantly, these
data could enable a quick deployment of field crews to the

particular fault location, with an indication of what may have
been damaged based on the observed fault characteristics.

Example 3
the high reporting frequency of µPMus allows them to capture hundreds of voltage events every day. some of these
events have root causes at the transmission level, which may
not be of much interest to distribution operators. therefore,
it is necessary to distinguish transmission-induced events
from distribution-induced ones. the latter can then be used
in various distribution-level diagnostics. We can make such
a distinction by comparing synchronized voltage measurements at neighboring distribution feeders.
an example is shown in figure 3, where four voltage
events are marked. event 1 appears in feeder 2 but not in
feeder 1, so it is induced by the loads and equipment on
feeder 2. events 2 and 4 appear in feeder 1 but not in
feeder 2, so they are induced by the loads and equipment on
feeder 1. event 3, which is the most severe voltage transient
event in this example, appears in both feeders. therefore, it
is most likely caused by issues at a higher voltage level, such
as the transmission or subtransmission systems.
data from µPMus have also been used recently for other
descriptive analytics, such as evaluating photovoltaic site
performance; determining voltage controller behavior, topology, as well as for phase detection; and performing resource

Feeder 1

Grid
Circuit
Breaker

Fuse

µPMU 2

µPMU 1

Animal
Fault

µPMU 3

Feeder 2

240

Voltage (V)

300

180
120
60
0

µPMU 1
0

200

400 600
Time (ms)
(a)

800 1,000

9,000

8,000

8,000
Voltage (V)

360

7,000
6,000
5,000
4,000

µPMU 2
0

200

400 600 800 1,000
Time (ms)
(b)

7,000
6,000
5,000
4,000

µPMU 3
0

200

400 600 800 1,000
Time (ms)
(c)

figure 2. The impact of an animal-caused short-circuit fault recorded at three geographical locations. (a) The short-circuit
fault momentarily brings the voltage down to zero at the location of µPMU 1, causing a brief interruption. (b) The shortcircuit fault also creates a severe voltage sag at µPMU 2. (c) µPMU 3 records a considerable voltage sag on another feeder,
although much less severe than what µPMU 2 captured.
may/june 2018

ieee power & energy magazine

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2018