IEEE Power & Energy Magazine - July/August 2021 - 60

can be deployed (started on and ramped up and down) to
deliver the required response. Therefore, starting from a certain
operating point, the flexibility NOEs for the faster raise
services are contained inside the envelopes for the slower
services, as shown in Figure 9.
The smaller operating range of reactive power for fast
FCASs (the blue line in Figure 9) reflects the fact that not all
units could start up within the response time of service. Furthermore,
the size of the NOE monotonically decreases with
the expected duration of the service as energy-constrained
resources cannot deliver a response for a longer time. For
instance, in Figure 9 this can be seen in the fact that the
" delayed lower " FCAS NOE is relatively smaller than the
" fast and slow lower " FCAS NOEs.
From Concept to Reality:
Challenges and Opportunities
Adopting Operating Envelopes
Traditionally, distribution networks are operated as passive
circuits such that minimal interactions are necessary from
the control rooms. Consequently, to make the transition
from fixed limits to operating envelopes a reality, there are
many challenges that distribution companies are expected
to face, from the data required by the algorithms to the
physical infrastructure into which these concepts will be
integrated. This section explores some of these challenges
(as well as potential avenues to address them) that are likely
to be encountered by distribution companies hoping to
make operating envelopes an integral part of the businessas-usual
operation.
✔ Adequate network models: The databases of distribution
companies often contain incomplete modeling
data for power flow analyses, particularly for LV
feeders. For instance, key information, such as the
phase connection of customers and the impedances of
LV conductors, are rarely fully recorded. Fortunately,
thanks to the increasing rollout of smart meters worldwide,
as well as ongoing research into data-driven
techniques, reconstructing these missing data is becoming
ever more plausible.
✔ Forecasting of passive customers at the head of a feeder:
The net demand of passive customers will affect
the available capacity in the corresponding LV feeder.
Similarly, as the head-of-feeder voltage sets a reference
point for the entire LV feeder, it will also play an important
role in determining the available capacity. Therefore,
the accuracy of these forecasts will directly affect
the validity of the calculated operating envelopes.
✔ Reactive power considerations: Undoubtedly, reactive
power from modern inverter-based DERs is a valuable
source of flexibility to provide services at the system
level. However, the incorporation of reactive power
into the operating envelopes algorithm is expected to
significantly increase the complexity. This is largely
60
ieee power & energy magazine
due to the combinatorial nature when considering
both active and reactive power as sources of flexibility
for the aggregator. For instance, during periods of
voltage-rise issues, absorbing reactive power (importing
vars) can help with voltage issues, whereas injecting
reactive power (exporting vars) will do the exact
opposite. Therefore, this introduces an inherent tradeoff
between allowing more flexibility for aggregators
and the complexity of the modeling behind the operating
envelopes algorithm for the distribution company.
✔ Speed and scalability: To ensure that operating envelopes
can be used in an operational setting in the
control room, the underlying algorithm needs to be
sufficiently fast and able to handle large-scale distribution
networks, i.e., thousands of LV feeders at
once. Consequently, ensuring the scalability of the
operating envelope calculation algorithm is one of
the key criteria to enable its successful adoption
by industry.
✔ Monitoring infrastructure at the head of a feeder: As
discussed earlier, head-of-feeder measurements are
essential to the algorithm that calculates operating
envelopes. However, the availability of this measurement
depends on the infrastructure available to distribution
companies. Therefore, this may entail additional
monitoring infrastructure to be installed by
distribution companies.
✔ Operationalizing SCADA measurements: Real-world
implementation requires operational data of the network
(i.e., the inputs, as shown previously). As discussed
earlier, depending on the frequency of updating
operating envelopes, this can have different
implications on the frequency with which SCADA
measurements of the networks are collected. For instance,
in Australia, smart-meter data are typically
collected in the control room twice a day. This means
that additional upgrades to the communication infrastructure
may be necessary if a higher data frequency
is required. Nonetheless, the implementation of more
frequent data collection provides a unique opportunity
that allows the distribution company to exploit other
ways of utilizing these data.
✔ Ease of implementation: A key factor to accelerate
the industry adoption of advanced solutions is its ease
of implementation. For instance, while specialized
third-party software exists to perform power flow
calculations, using such software requires developing
additional interfaces to the existing systems in
the control rooms. Alternatively, a potentially more
implementable approach is through rule-based algorithms
using equations that can be implemented in any
programming/scripting language already available in
the control room. This eliminates the need for developing
additional interfaces to third-party software,
thus reducing the integration risks.
july/august 2021
IEEE Power & Energy Magazine - July/August 2021

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - July/August 2021
