IEEE Electrification Magazine - March 2020 - 33

The biggest
difference between a
seamless transition
and a black start is
the amount of
reactive power
needed to supply the
required transformer
inrush current.

60.3
60.2

a.
m
.

11
:1
9:
17

a.
m
.

11
:1
9:
16

11
:1
9:
15

11
:1
9:
14

a.
m
.

60.1
60

Grid Voltage

Lesson 7: Black Start versus Seamless Islanding
One important decision for the microgrid design engineer
is whether to implement seamless islanding or whether

81O = 60.5 Hz
81TD = 5 Cycles
31.25 Cyc

60.6
60.5
60.4

a.
m
.

Frequency (Hz)

to allow for microgrid blackout and
then use the black start to energize
the microgrid. Each approach has
its pros and cons. Seamless islanding enables customers within the
microgrid to experience no loss of power,
but achieving the secure and reliable
islanding-detection scheme can be
challenging and cost prohibitive. The
most challenging engineering part of
islanding detection is islanding nondetection zones, which typically occur
when the load is equal to the generation downstream from the protective
device that operated. In this case, the
DERs within the microgrid would continue to provide power, not only to the
microgrid customers but to all customers on the feeder
between the PCC and the upstream protective device that
was operated.
The microgrid islanding-detection scheme at Mount
Holly is based on local measurements and uses undervoltage (27), overvoltage (59), underfrequency (81U), overfrequency (81O), rate of change of frequency (81R), and 81-RF
elements and an 85-RIO scheme to detect islanding. This
scheme was tested numerous times, in excess of 50, when
the generation and load within microgrid were equal;
every time the upstream recloser (POI) tripped, the
scheme detected the island. If the total DER generation
within the microgrid is less than the load, then a loadshedding scheme must be implemented to prevent a
blackout. The scheme must operate within 20 ms or the
seamless islanding could be unsuccessful.
A black-start approach to energizing the microgrid
allows the feeder protection to operate first and then after
the feeder locks out, usually no more than 60 s after the
fault, the microgrid controller initiates the black start.
Depending on the type of DERs available within the
microgrid, a black start can be either more or possibly less

11
:1
9:
13

failed grid synchronization. Before
the microgrid controller issues a gridsynchronization command to the
BESS, it must torque control the voltage and frequency settings; otherwise, the microgrid will experience a
blackout. Once the BESS receives the
command to synchronize to the grid,
it changes its operating mode from
VSI-ISO to VSI-V/f with ±5% voltage
droop and ±1% frequency-droop settings. Since the control loops inside
the BESS PLC change during this process, the voltage and frequency transient response can result in having
these values outside the VSI-ISO setpoints. Figure 10 shows why is it
important to torque-control the settings during the gridsynchronization process, based on the frequency response
as recorded within the microgrid.
As seen from Figure 10, the transient frequency res--
ponse is such that the frequency goes above the 81O pickup
level (60.5 Hz) for the duration (31.25 cycles) longer than
the 81O time delay of five cycles. Therefore, during the
transition to island, which can take anywhere from 0.5 to 8 s,
frequency settings change and two new levels are implemented to enable a secure and reliable microgrid protection
scheme. First-level 81 protection was added below 58.5 Hz
and above 61.0 Hz with a time delay of 4,800 cycles (or 80 s).
Second-level protection was added as fast tripping for frequencies below 57.0 Hz and more than 62.0 Hz with a time
delay of five cycles. This protection is within IEEE Standard
1547 guidance and, based on the field measurement of
more than 10,000 transitions, most of the time, the moment
that BESS receives the command from the microgrid controller to start grid resynchronization, the frequency rises to
the 60.5-61.0-Hz range, which would cause the microgrid to
trip based on the steady-state settings.
Once the BESS completes the synchronization, it sends
the notification to the microgrid controller, which, in turn,
sends the command to PCC relay to close the PCC breaker.
At this time, the PCC relay checks its 25 settings to ensure
that all three analog points are within the desired limits. If
this condition is satisfied, the PCC closes and the grounding transformer is de-energized. This is important because
disconnecting the grounding transformer before the PCC
is closed will result in the PCC relay issuing a block PCC
close command for the reasons mentioned.
Field results show that, on average, it takes 32 ms to disconnect the grounding transformer, so this is a very fast
transition. Figure 11 presents the torque-controlled frequency settings during the grid-synchronization process.

Microgrid Voltage

Figure 10. The frequency response during grid synchronization.

IEEE Elec trific ation Magazine / MARCH 2 0 2 0

IEEE Electrification Magazine - March 2020

Table of Contents for the Digital Edition of IEEE Electrification Magazine - March 2020