IEEE Electrification Magazine - June 2016 - 41

installation and lower maintenance requirements for
commercial installations.
Even if cost targets for the components, installation, and
service in a dc microgrid are achieved, the project-specific
engineering cost has to be minimized. The design and construction of large centralized power plants require considerable engineering effort that ends up representing a small
cost on a per-watt basis. The microgrid market is foreseen
covering a wide range of powers but with a large section in
systems below 1 MW. The PV solar generation that mainly
operates exporting power to the ac grid has been successful
in achieving modularity and simplicity in commissioning
for solar modules and power electronics components.
Components are produced in series, tested at the factory,
and require minimum tuning or configuration on-site.
Communication is used for monitoring and is not critical
for power production. However, microgrids include functionality that is much more complex than exporting power
to a solid and stable source. A major difference between
classical components used in commercial or industrial
installations and those components that would be used in
microgrids lies in the fact that most
components in the microgrid require a
level of intelligence to allow them to
participate in the power and energy
management of the microgrid. Most
elements connected to the bus have to
operate in harmony, interacting with
each other and following specific algorithms to satisfy power management,
energy-demand response, power quality, and energy-management goals.
Without regard to using centralized or
distributed control philosophy, projectspecific configuration and communication have to be engineered and
commissioned, increasing the capital
cost of each installation. Even if the same control strategy
and communication platform are used, different size installations require redesign, tuning, and long commissioning
processes. If components are modified or suppliers
replaced, additional reprogramming would be needed. Such
a cost burden may limit the market of microgrids in lowerpower installations where most benefits can be achieved.
To get a wide deployment of microgrids, a different
model is needed in which components are tested by the
manufacturers, installed by electricians, and receive
simple and quick approval based on certifications similar to the installation of industrial equipment or to the
level being achieved in the residential and commercial
solar industry. Achieving standardization in the controls
and energy management of a microgrid is much more
difficult than reaching standardization in hardware and
electrical safety. In early concepts and demonstration
projects, different developers use different control principles, system management, and communication

platforms that, in many cases, include intellectual property and trade secrets. The variety of control methods
would need to converge into a few concepts, and those
approaching the plug-and-play concept successfully
used in the computer industry have a better chance
of success.

ARDA dc Microgrid Management Scheme
ARDA Power proposes a concept shown in Figure 5 in
which dc bus voltage is used as the communication mean
to provide information among the different elements connected to the microgrid. However, the dc bus signaling is
not used with the purpose of power management. Instead,
it is used for performance functions with slow dynamics
that may include battery management, energy management, and operating-cost optimization. This differs from
the classical droop control methods that fundamentally
achieve power balance among elements on the bus.
In ARDA's concept, the main control of the microgrid
or regulating function, which requires fast response and
quick decision making, resides in the energy-storage elements that execute power-management algorithms to maintain the dc
bus voltage at a set point. If a load or
source suddenly changes, creating a
power imbalance on the dc bus, only
the energy-storage elements respond
in a fast and coordinated manner following standard control algorithms
such as the one in Figure 6. The
power sources and power loads do
not respond to the fast transients in
the dc bus and are not involved in the
instantaneous power management of
the microgrid. Such a concept simplifies the system's stability and the
commissioning effort at the expense
of demanding a faster response and full power range for
the energy storage. Even if the energy-storage units are
operating at a high or low state of charge, they can be
used to maintain the dc link as long as the average power
in or out of the energy storage does not drive them to full
discharge or overcharge conditions. The proposal of using
the energy-storage elements to control the dc bus voltage
is supported by the fact that the energy storage is the
center of a microgrid capable of operating separately from
the ac grid.
In addition to executing the power management of
the dc bus, the energy-storage elements become the
energy manager for the microgrid. Simultaneously to
controlling the dc bus voltage, the energy manager
changes the dc bus voltage set point following an algorithm that involves one or several secondary sustaining
functions. The sustaining functions have the goal of
ensuring that the microgrid will continue operating
indefinitely by balancing the energy (or averaged power)

The next step is
the implementation
of dc microgrid
demonstration
projects to validate
the different
concepts in real
conditions.

IEEE Electrific ation Magazine / j une 2 0 1 6

Table of Contents for the Digital Edition of IEEE Electrification Magazine - June 2016