IEEE Power & Energy Magazine - July/August 2020 - 60

With the changing energy landscape and the trend for digitalization
and decentralization, there is the opportunity as well as the
necessity to augment the traditional reliability approaches.

schematically represents a range of possible outcomes for
the overall resilience of power systems and illustrates how
physical hardware redundancy can increasingly be replaced
by smart flexibility. However, smart flexibility exposes the
system to malicious cyberattacks and malfunctions. Therefore, plausibility checks, validation, and fallback settings
decrease the potential damage from loss (or even adversary
takeover) of communication and control systems.

Distribution System Islands
to Enhance Resilience
In case of a large-scale blackout, i.e., when either the transmission system or the connection to it is unavailable, the
formation of distribution system islands can help mitigate
the adverse impact. Even if the extent of local generation
capacity is limited, there is often the potential to supply at
least part of the normal load. If part of the power system
can continue operating as an island without interruption, the
damage to the customers served can be completely avoided.
However, even if customers are disconnected by the initial
disturbance, there is significant benefit in reducing the local
outage duration.
The increasing penetration of DERs potentially improves
the resiliency of distribution islands. However, the distributed
nature of many renewable generators presents operational
challenges. Furthermore, the stochastic, weather-dependent

availability of generation from sources such as solar and
wind power increases the number of possible different scenarios a distribution system operator (DSO) may face, which
greatly complicates matters.
Here, we present a case study about the possibility of
local load coverage while respecting limits for active power
exchange with the surrounding power system. Figure 7 displays the amount of load that can be restored in a representative section of a distribution system with a given penetration of DERs for different starting times over a year,
depending on the required duration. The study considers nameplate values of 133 MW for PV plants, 108 MW
for wind farms, and 12 MW for biomass plants; the average load for normal operation is 22.4 MW. The analysis
accounts for the availability of remote controllable
switches and the actual feeder configuration, as well as
time series data for load and generation of the year 2014,
with the renewable generation scaled up according to a
midterm future scenario.
In Figure 7, the y-axis shows the amount of load that
can be covered under the given constraints. The x-axis represents starting times during the year, sorted by coverable
load. The individual lines represent different constraints
concerning the required minimum duration of supply. The
(small) area below the line "18 h" represents what can be
done (and how often) if only the connection of feeders is

Covered Load (MW)

25
Desired Time Horizon for Load Coverage
20
3h

10
9h
5
0

12 h
18 h
0

30
40
50
60
70
Eligible Starting Time Points per Year (%)

100

figure 7. The suppliable load over the year.
60

ieee power & energy magazine

july/august 2020

IEEE Power & Energy Magazine - July/August 2020

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - July/August 2020