IEEE Power & Energy Magazine - September/October 2021 - 41

Western Australia Rate of Change of Frequency Control Service
In August 2019, a Western Australia energy transformation
task force found that a real-time co-optimization of all frequency
control services, including inertia, was most appropriate
for the future WEM, driven by a mixture of physical,
operational, and market considerations. Historically, the
WEM relied on an empirically derived rule of thumb: 70%
of the largest generation contingency (in megawatts) was
allocated as headroom across a set of designated facilities.
An analysis and comparison of this approach identified that
the combination of isolation and relatively small size resulted
in the WEM being run close to its technical limits.
The transition to the greater penetration of renewables
has necessitated a more sophisticated market design and
led to a preference for a real-time spot market to optimize
system inertia and primary response speed. In this context,
initial design options focused on the correct balance of service
definition " segmentation, " for example, adding 1-, 2-,
and 3-s services to complement the co-optimized 6- and
60-s markets, as done in the NEM. With system requirements
abstracted to fundamental quantities (i.e., generic
megawatt specifications), the optimal delivery of these services
would be by market dynamics, irrespective of the underlying
technology.
Unfortunately, investigations and analysis revealed issues
with the multisegment approach from the following
physical and market perspectives:
* Physical
˚ Each segment adds complexity and increases the
degree of " fantasy " space in which the commercial
abstraction diverges from physical reality. In
practice,
there is no clean,
megawatts into convenient buckets.
˚ Inertia is only superficially the same as primary
response. True rotating machinery has a fundamentally
instantaneous reaction, while power
electronics suffer from an electronic detection
delay on the same order (<1 s) of the critical
ROCOF period.
* Market
˚ Each segment adds complexity, resulting in additional
infrastructure/systems overhead plus an opportunity
to game/manipulate market systems.
˚ Especially in a relatively " shallow " market (pool of
suppliers), more complexity increases the chances
of a participant effectively exercising power over
a market.
The task force decided that a single segment was most aplinear
separation of
propriate. While multiple segments facilitate more service differentiation,
in practice, such gains were marginal, while the
downsides were guaranteed. The implementation of this direction
required a fundamental change in the perspective of
service definitions. Rather than split physical responses across
multiple segments, the entire response profile is characterized
in reference to a perfect exponential response (see Figure S3)
chosen to approximate the output of a physical turbine. The
response factor is then converted into a multiplier that incentivizes
speed. Inertia is split from the primary response in recognition
of the underlying physical differences, while inverterbased
generation is credited through very high-performance
multipliers. The task force, however, noted the ongoing research
and development of inverter-based technology, and
named the inertial service ROCOF Control in recognition that
future developments may open this segment to power electronic
devices.
60
50
40
30
20
10
Measured
Measured (Clean)
Inertial (Shifted)
t0 (-1s)
Primary Frequency Response
τ (s)
50
0.2
0.5
1
2
3
4
5
6
7
8
9
10
-5 05 10
Time (s)
figure S3. The physical response of a gas turbine is measured and compared against an array of hypothetical " perfect exponential "
responses of different speeds.
september/october 2021
ieee power & energy magazine
41
15
20
25
49.8
49.6
49.4
49.2
49
30
Facility Output (MW)
Frequency (Hz)

IEEE Power & Energy Magazine - September/October 2021

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - September/October 2021