IEEE Power & Energy Magazine - January/February 2014 - 72

power for almost two weeks, and cost towns millions of
dollars. the widespread power outages in the wake of hurricane sandy cast light on the weakness of a centralized
electric power system and highlighted the benefits of distributed control of distribution systems.
it is well understood that the destructive power of storms
can spread through the utility grid much quicker, farther,
and wider than single faults. the transmission grid in the
united states has been continually struck by major outages
caused by cascading failures, such as the 2003 blackout that
knocked out power to 60 million people in the northeast and
most recently the september 2011 blackout in the territory
of san diego Gas & Electric, which shut down all of southern california around san diego. the transmission grid is
designed to prevent these cascading failures, but they seem
to happen with a regularity that disproves the expressed
assurances about the efficacy of the technological safeguards already in place.
as the utility grid remains quite vulnerable and exposed
to natural disasters, it is understood that there is no way
to completely protect citizens and towns against extreme
conditions such as flooding, wind gusts, heavy storms and
rains, downed trees, and flying debris. rather than protecting the system from storms like sandy, the power industry has therefore focused on ways of restoring the system
quickly after disasters. however, the repair of damaged
equipment and a complete system restoration after a major
storm can take months, and expert information regarding
the operation and maintenance of a power system that spans
many square miles could play a major role in identifying
trouble spots, repairing critical equipment, and restoring
power after such disasters.

Conventional Service
Restoration Strategies
the limited operating margins embedded in centralized
power systems have increased the risk of system collapse
and power blackouts resulting from natural disasters. after
blackouts, restoration processes are initiated to maximize
load restoration within a minimum elapsed time, taking
into account operating conditions and the need to manage
system security. in such cases, the individual characteristics
of regional power systems determine the specific restoration procedures, which are often developed using heuristic
methods that reflect human experiences along with numerical simulations. the procedures usually include step-by-step
guidelines for system operators.
in every major disaster, healthy sections of power distribution networks may also be deenergized, which leads
to additional customer interruptions. to reduce the chance
of deenergizing healthy sections, power utilities resort to
switching sequences, alternative resources and substations, and additional pathways to supply electricity to the
deenergized healthy sections without violating operating constraints. however, tighter operating constraints in
72

ieee power & energy magazine

centralized distribution systems may limit the feasibility of
restoration procedures at individual load points.
traditionally, most utilities initiate fault detection and
system restoration after receiving calls from customers
who experience power outages. the distribution utility
sends out a field crew to investigate fault locations and find
the most feasible switching scheme to isolate the faults
and restore services to affected customers as quickly as
possible while repairing faulty feeders. But automation
can speed up restoration, especially with large distribution systems. By employing smart grid technologies,
some utilities have made the process of fault detection
and healing 50% faster. smart meters can send reports
automatically to the outage management system, which is
quicker than having a person call in and overcomes any
possible communication barriers. checking for restored
service from central offices via smart meters also creates
time and personnel savings. in addition, feeder-switching
devices equipped with intelligent electronic devices for
protection and control applications have been developed
to provide feeder automation. the automated capabilities
of these devices include measurement, monitoring, control, and communication, which all facilitate automated
fault identification, isolation, and power restoration. for
example the intelliteam sG automation restoration system
can search for possible power sources so as to maximize
the number of restored customers after an outage in the
shortest possible time, with due consideration given to the
limitations of distributed energy resources (dErs) connected to the distribution systems.
feeder automation alters the topological structure of feeders by changing the status of switches under contingencies in
distribution networks. When the system suffers a sustained
fault and the switch located upstream of the fault opens, the
feeder automation selectively isolates the fault and restores
power to as many loads as possible in order to decrease the
interruption indices of the distribution system and customers
and reduce the duration and magnitude of power outages during the fault troubleshooting. With an adequate network communication structure, automated fault detection and isolation
are relatively unchallenging. But automatic power restoration is still a challenging task because of practical issues in
identifying a feasible and optimal restoration plan, including load balancing and operating constraints. in a nutshell,
smart grids are about reacting to storms and outages and
limiting the amount and severity of power outages. another
tool-microgrids-is about stopping power outages from
the get-go.

The Emergence of Microgrids
the advent of modern distribution networks with dErs has
changed the service restoration paradigm. such distribution networks are equipped with local generation sources
that behave like virtual power plants and active loads that
provide ancillary services such as frequency and voltage
january/february 2014

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - January/February 2014