IEEE Power & Energy Magazine - November/December 2019 - 37

The islands of Hawai'i, Oahu, and Maui have among
the highest penetration levels of DERs in the United States,
in terms of installed capacity relative to system size.
of D-PVs is determined by the standard requirements at the
time of interconnection, with different possible aggregate
losses during these frequency and voltage excursions. The
largest contingency concern for Hawai'i island is the loss of
legacy DERs tripping at 60.5 Hz, as a large majority of the
existing D-PVs is subject to trip at or near that frequency.
The second-largest vulnerability is the loss of DERs during large-voltage excursions. On an island system like
Hawai'i's, 60.5 Hz is possible after delayed fault clearing
(due to transient swings following fault clearing) and major
loss-of-load events, which can occur with transmission outages (both N-1 and N-1-1). The potential aggregate loss
of D-PVs during high solar-production periods is roughly
double the largest single-generator contingency, resulting in
severe underfrequency conditions and the possible risk of
system failure.
The loss of D-PVs also exacerbates other loss-of-generation
events, with some additional D-PVs loss during the underfrequency situation. To address this concern, a battery energy
storage study was conducted to analyze the effect of legacy
DER tripping for low voltage during transmission faults,
low frequency following generating-unit trips, and high frequency following a transmission line fault. The study found
that retrofitting legacy DERs to full ride-through capability
was the most effective solution but that it was too costly and
impractical. The study also found the following:
✔ Storage can replace DER energy lost during transient
voltage and frequency conditions, preventing excessive underfrequency protection and reducing the risk
of system failure for disturbances during high solar
PV production.

UFLS STAGE DATA

✔ The size of the storage necessary to mitigate reli-

ability issues depends on the number of legacy
DERs installed.
✔ Increased numbers of spinning reserves could reduce
the size of the necessary storage but not eliminate the
need for it. Increasing reserves also exacerbates excess-energy concerns. D-PVs protection systems are
not consistently implemented, and performing sensitivity studies around the uncertainty of this DER behavior is critical.
Based on the studies, a BESS with 18 MW of capacity
and a 30-min duration was able to arrest frequency excursions and provide sufficient time to bring standby generation online. Increasing the duration to 1 h enabled the BESS
to respond to over-frequency conditions. Studies showed
that two BESSs geographically and electrically separated
from one another provided the best performance and
grid resilience.

Underfrequency Load Shedding
The reliable operation of the Hawai'i island systems heavily
relies on underfrequency load shedding (UFLS) that trips
distribution circuits at preestablished frequency thresholds.
However, as the penetration of DERs rapidly increases,
variability in net loading on any given feeder poses significant challenges to the conventional UFLS design, which
means that static UFLS arming is no longer effective. An
adaptive UFLS program has been implemented to address
this variability. Figure 2 is a dashboard of the adaptive program, including six instantaneous stages of UFLS arming
between 59.1 and 57.6 Hz and two kicker stages that operate

Change to Monitor

Mode: Active

Reset and Calculate

System Load: 151.629
Total Target:
120.559
Total Available: 127.133

Stage

Frequency

Percent

Target
Megawatts

Available
Megawatts

Tolerance %

Tolerance

Delta
Megawatts

Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Kicker 1
Kicker 2

59.1
58.8
58.5
58.2
57.9
57.6
59.3
59.5

5
10
10
15
10
20
5
5

7.53492
15.06987
15.06987
22.60476
15.06987
30.13968
7.53492
7.53492

7.68113
14.93933
15.45847
22.63691
14.63551
30.76797
7.90598
7.72132

5
5
5
8
8
25
5
5

0.377
0.753
0.753
1.808
1.206
7.535
0.377
0.377

-0.146
0.131
-0.1389
-0.032
0.434
-0.628
-0.371
-0.186

figure 2. The adaptive UFLS-arming dashboard.
november/december 2019

ieee power & energy magazine

IEEE Power & Energy Magazine - November/December 2019

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - November/December 2019