IEEE Power & Energy Magazine - November/December 2019 - 43

november/december 2019

Frequency (Hz)

1:10:00 p.m.
1:11:00 p.m.
1:12:00 p.m.
1:13:00 p.m.
1:14:00 p.m.
1:15:00 p.m.
1:16:00 p.m.
1:17:00 p.m.
1:18:00 p.m.
1:19:00 p.m.
1:20:00 p.m.
1:21:00 p.m.
1:22:00 p.m.
1:23:00 p.m.
1:24:00 p.m.
1:25:00 p.m.
1:26:00 p.m.
1:27:00 p.m.
1:28:00 p.m.
1:29:00 p.m.
1:30:00 p.m.
1:31:00 p.m.
1:32:00 p.m.
1:33:00 p.m.
1:34:00 p.m.

Normalized Power
(MW/MW During Interval Ending 1:10:55 p.m.)

greater than 30°, ROCOF levels exceeding 0.4 Hz/s, with managing frequency. However, more than 15% of the
and an inability to meet the ride-through requirements D-PVs in Queensland and more than 30% in South Australia did not behave according to the standard, suggesting
specified in standard AS/NZ4777.2-2015.
✔ Laboratory testing also showed sluggish responses for issues with compliance.
Those findings demonstrate a range of areas in which the
control modes (frequency-watt and volt-watt) that are
compliant with the standard but provide little support DER standards in Australia need to be reviewed to ensure
the reliability of the BPS. The review and implementation
to the grid.
Those findings highlight a number of potential security of advanced standards should be pursued years in advance
risks that the AEMO is trying to address. They also sug- of significant DER penetration due to the extensive stakegest a need to review Australian standards to more explicitly holder consultation that is required and the need for transition periods for manufacturers to implement new capabilispecify required responses.
Collecting suitable data related to D-PVs has been one of ties. Further, dynamic models that represent DER behavior
the main challenges that AEMO has faced in this program. during grid disturbances need to adapt quickly to those
Feeder-level monitoring captures aggregate DERs and load changes since these studies are of paramount importance
behavior but does not separate the two. Furthermore, distur- for grid reliability.
bances during high D-PVs output periods are rare (for example, voltage disturbances are often associated with lighting The Duke Energy Progress
strikes where cloud cover impacts D-PVs output). AEMO Experience With DERs
has been working with Solar Analytics, Redfern, Australia, Solar PV has grown significantly in North Carolina during
and the University of New South Wales, Sydney, Australia, to the past five years due to government policies, the avoided
develop new techniques to analyze DER responses.
cost of energy, available land, and falling prices for solar
Solar Analytics provided AEMO with 1-min-resolution data technology. Approximately 3,100 MW of solar PV capacfor 5,000 D-PVs systems during a disturbance on 25 August ity were connected in North Carolina, as of January 2019,
2018. A fault occurred at the Queensland-New South Wales across the Duke Energy North Carolina balancing areas.
interconnector, causing a voltage dip across the northern part Roughly 2,300 MW are located in the Duke Energy Progress
of New South Wales. Figure 8 shows the D-PVs responses for (DEP) region, with a winter peak demand of approximately
systems installed prior to October 2015 (AS/NZ4777.3-2005 15,000 MW. About two-thirds of the solar PV capacity is in
compliant), after October 2016 (AS/NZ4777.2-2015 compli- utility-scale DERs, with the most common DER capacity
ant), and during the transition period (which could meet either being 5 MW, due to past incentives. The rest is transmission
standard). Nearly 45% of the pre2015 D-PVs in zone 1 (a 540-kmdiameter area) disconnected. Com1.1
51
pliance newer standard appeared to
1.05
have improved voltage disturbance
ride-through capabilities, especially
1
50.5
in zone 2 where the voltage dip was
0.95
less severe. Future versions of DER
50
0.9
standards can build on this improve0.85
ment in capabilities, minimizing
risks to power system security.
49.5
0.8
D-PVs can also provide valuable
0.75
grid support. AS/NZ4777.2-2015
49
0.7
specifies that D-PVs should pro0.65
vide over-frequency droop response
48.5
when frequency exceeds 50.25 Hz,
0.6
with a linear ramp to zero generation by 52 Hz. This was observed
during the August 2018 event, as
illustrated in Figure 9. The fault
led to the separation of Queensland
AS4777.2-2015:Specified Response
AS4777.2-2015:Specified Response
and South Australia from the rest of
Observed Power, <30 kW, 878 Sites
(1-min Average)
QLD Frequency
Observed Power, 30-100 kW, 92 Sites
the NEM, causing over-frequency
50.15 Hz
50.25 Hz
in those regions. Many D-PVs systems behaved according to the standard and reduced output to assist figure 9. The behavior of post-2016 D-PVs inverters in Queensland on 25 August 2018.
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