IEEE Power & Energy Magazine - May/June 2018 - 31

µPMU 1

Grid
Capacitor Bank
(No Sensor)
100

7.3

360
A

C
7.2
B
7.15

400

800 1,200 1,600 2,000
Time (ms)

Magnitude (A)

7.25

Magnitude (A)

A
A

80
B
60
C
40

400

B
120
C

800 1,200 1,600 2,000
Time (ms)

(a)

240

400

800 1,200 1,600 2,000
Time (ms)

(b)

(c)

figure 4. A switching-off event of a switched capacitor bank remotely observed by a µPMU. (a) Switching off the capacitor results in a permanent drop in voltage. (b) A severe current overshoot on phase A and a severe current undershoot on
phase C occur during the capacitor bank switch-off event. (c) The capacitor bank is apparently not initially de-energized
on phases A and C at the time of switching until several cycles later.

example 4 were identified by other existing sensors on the
utility grid. Both sets of anomalies were in the form of transient events that were visible only to the sensors with high
sampling rates at milliseconds, such as µPMus. although
the real-life µPMus that provided the measurements to perform the analyses in examples 3 and 4 were not initially
intended for the analyses in either example, they were able
to provide highly insightful measurements to detect incipient failures.
the application of incipient failure detection, coupled with
close to real-time communication, will allow early corrective

1.04
1.03
1.02
1.01
1
0.99
0.98
0.97
0.96

Primary Side
Secondary Side

0
20
0
40
0
60
0
80
0
1,
00
0
1,
20
0
1,
40
0
1,
60
0
1,
80
0
2,
00
0

Voltage (pu)

states), the effective turns ratio of the transformer is changed;
this is accomplished with a mechanical movement of a springloaded contact, often in an oil-filled environment. common
methods for sensing the failure of such a component include
dissolved gas analysis and oil-level measurements. these methods often require taking the device offline for complete testing.
In this real-world example, precise synchronized voltage
measurements from µPMus were also examined to determine any anomalous behavior related to the tap changes.
anomalies were, in fact, detected on the primary side of the
transformer in the form of synchronized voltage rise and
sag during a tap change, as seen in figure 5, despite the lack
of additional prior indications by other sensors on the field.
this could be related to the abnormal behavior of the resistor or reactor bridge within the oltc (used to limit the circulating current during the mechanical change) or to an oil
leak causing arcing within the device. Interestingly, neither
the oil sensor nor the smart meters had detected the presence of an anomaly. When the field crew visited the transformer location after the µPMu data analysis, it was determined that a repair was necessary. following the repair, the
anomaly disappeared. With continued oil leakage and in the
absence of µPMu data analysis, there would have been a
catastrophic transformer failure and customer interruption.

Time (ms)

A Note on Examples 3 and 4
It is worth emphasizing that neither the capacitor bank anomalies in example 3 nor the transformer anomalies in
may/june 2018

figure 5. The tap-change event of an OLTC observed by
two µPMUs.
ieee power & energy magazine

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