See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325742525

Smart Grid The Future of the Electric Energy System

Article · June 2018

CITATIONS READS
5 21,534

3 authors, including:

Roger N. Anderson Hamid Gharavi

Columbia University National Institute of Science & Technology
1,561 PUBLICATIONS   9,589 CITATIONS    154 PUBLICATIONS   3,477 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Bell Geospace Gravity Gradiometry View project

Intelligent Building Management View project

All content following this page was uploaded by Roger N. Anderson on 16 July 2018.

The user has requested enhancement of the downloaded file.

 1

Smart Grid: The Future of the

Electric Energy System
Roger N. Anderson1, IEEE member, Reza Ghafurian2, IEEE Fellow, Hamid Gharavi3, IEEE Fellow

more proactive and are being empowered to engage

Abstract— This paper presents a discussion of the future of in the energy consumption decisions affecting their
the electric energy system, addressing the entire spectrum from
power generation, through substations, to distribution and the day-to-day lives. At the same time, they are
customer, and the feedback loops along the way necessary to expanding their energy needs. For example,
provide the computational intelligence necessary to make the consumer participation will ultimately include
“Smart Grid”. Both at the federal and state levels, governments
have recognized a need for modernizing the electric energy extensive use of electric vehicles (both cars and
system and establishing such Smart Grids around the world. We trucks), remote control of in-home appliances to
are at the point of a historic paradigm shift, with the opportunity promote energy conservation, ownership of
to implement new, more intelligent methods for producing,
distributing, delivering and using electricity in a much more distributed generation from ever more renewable
sustainable manner. Whereas the current electric system is energy sources, and management of electricity
based on a one-way flow of energy and information from the storage to locally match supply to demand. The
sources to the end users, the future Smart Grid will provide
multiple paths for the flow of electricity, and particularly availability of new technologies such as more
information about that flow, throughout the system. This paper abundant and aware SCADA sensors, secure 2-way
introduces this Special Issue by presenting a broad definition for communications, integrated data management, and
the Smart Grid. We discuss the necessary attributes for such a
system-of-systems, review the need for change, and identify the intelligent, autonomous controllers has opened up
technical challenges facing successful deployment and opportunities that did not exist even a decade ago.
implementation.
Index Terms—Smart Grid, Adaptive Stochastic Control,
Distributed Generation and Storage, Demand Response. The electric energy system of the future needs to
address all these needs and concerns by using
advanced technologies to create a smarter, more
I. INTRODUCTION efficient and sustainable grid. During recent years,
Three dominant factors are impacting the future there have been numerous articles and conferences
electric systems of the world; government policies, about the Smart Grid, but much confusion remains
efficiency needs of the consumer, and the among all constituencies about just what the term
introduction of new intelligent computer and entails. Although many different definitions have
hardware technologies. In addition, environmental been proposed for the Smart Grid, in most cases the
concerns have created governmental policies around users have chosen particularly focused definitions
the world, including at the federal and state levels, related to their specific applications and local needs.
which are driving the entire energy system to Below, we define the Smart Grid in its broadest
efficiency, conservation, and renewable sources of global terms. We begin with a description of the
electricity. These factors are the main drivers that make up of the present conventional electric energy
are expanding the use of all sorts of new renewable system, and we then identify the areas that must
energy and storage technologies on the one hand change in order to provide the intelligence and
and new energy efficiency and conservation control necessary to convert to the safe, secure, and
techniques on the other. Consumers are becoming efficient Smart Grid of the future. Papers that
follow in this Special Issue give a cross-section
through this vast new enterprise, and while not
Manuscript received by IEEE June 30, 2010. meant to be all-inclusive, are meant to be
1
R. N. Anderson is with the Center for Computational Learning Systems, illustrative of the changes coming to the Smart
Columbia University, NY, NY 10027
2
Grid.
R. Ghafurian is with the Consolidated Edison Company of New York,
NY, NY 10003
3
H. Gharavi is with the National Institutes of Standards and Technologies,
Washington, D.C.
 2

II. THE CONVENTIONAL ELECTRIC ENERGY SYSTEM III. THE FUTURE SMART ELECTRIC ENERGY
A general description of today’s conventional SYSTEM
electric delivery system is represented in figure 1.
Traditionally the system is broken into mostly A general schematic of the future electric energy
isolated components (silos): generation, system, or Smart Grid, is presented in figure 2.
transmission, substation, distribution, and the The key requirements of this system will address
consumer. Key characteristics of this conventional the following transformational functionalities:
system that will be most strongly impacted by the
changes required to implement the Smart Grid are • Allow for the integration of renewable energy
the following attributes: resources to address global climate change,
1. centralized sources of power generation, • Allow for active customer participation to
2. uni-directional flow of energy; from the enable far better energy conservation,
source to the customers, • Allow for cyber-secure communications
3. passive participation by the customers; systems to address system safety,
consumer knowledge of electrical energy • Allow for better utilization of existing assets to
usage is limited to a monthly bill received, address long term sustainability,
after the fact, at the end of the month, • Allow for optimized energy flow to reduce
4. real-time monitoring and control is mainly losses and lower the cost of energy,
limited to generation and transmission, and • Allow for the integration of electric vehicles to
only at some utilities, does it extend to the reduce dependence on hydrocarbon fuels,
distribution system, • Allow for the management of distributed
5. the system is not flexible so that it is generation and energy storage to eliminate or
difficult to either inject electricity from defer system expansion to reduce the overall
alternative sources at any point along the cost of energy,
grid, or to efficiently and sustainably • Allow for the integration of communication and
manage new services desired by the users of control across the energy system to promote
electricity. interoperability and open systems and to
increase safety and operational flexibility.

Figure 1. Conventional Electric Grid Figure 2. The Smart Grid

It should be noted that the Smart Grid, as

These conventional attributes have adequately characterized above, does not replace the existing
served the needs of electric utilities and their electric system but rather builds on the available
customers in the past. However, the new needs of infrastructure to increase the utilization of existing
more energy knowledgable, computer savvy, and assets and to empower the implementation of the
environmentally conscious consumers, combined
new functionality. For example, centralized
with regulatory changes that promote sustainability
sources of generation will still play a major role in
and energy independence from foreign sources,
the Smart Grid, and large-scale wind and solar
availability of more intelligent technologies, and
ever greater demands for enough energy to drive the generation, wherever cost justified, will become
global economy, require an electric energy system major parts of the generation mix. Availability of a
of the future that is fundamentally different in all 5 2-way, cyber-secure, end-to-end communications
areas listed above. system will provide consumers with the knowledge
of their energy usage necessary to allow them to
 3

locally and/or remotely control their smart a system that is clean, safe, secure, reliable,
appliances and temperature settings. Monitoring resilient, efficient and sustainable. This definition
and control of the electric system components will covers the entire spectrum of the energy system
provide the utility with the real time status of the from the generation to the end points of
system. The use of this real time data, combined consumption of the electricity. The reader will note
with integrated system modeling and powerful new that many definitions proposed by other users are
diagnostic tools and techniques, will provide the subsets of this system-of-systems definition; as for
detection of incipient failures in order to drive example, if defined as smart metering, it addresses
preventive maintenance and dynamic work the consumption and to some extent the distribution
part of this definition but not the full spectrum of
management systems. Automatic reconfiguration of
integration required to implement the Smart Grid.
the system, powered by sophisticated, adaptive and
autonomous optimization controllers will maintain Achieving a smart grid will be a gradual and
the flow of energy without interruption when evolutionary process that will take many decades to
equipment failures do happen. Distributed be fully realized. To qualify as a Smart Grid, it is
generation and storage resources and remotely neither necessary nor feasible to incorporate all
controlled equipment will also play an important features at one time, but rather incorporation of
role in the management of the Smart Grid energy each new feature can be carried out independently.
system not only to address contingency needs but Each will require cost justification and reasonable
also to optimize power flow, eliminate load pockets, pay back on investments. However, interoperability
and minimize system losses. It should be noted that of open systems will allow each addition to “Plug-
building the Smart Grid, as envisioned here, will be and-Play” into the Smart Grid once the technologies
very costly and will require a sustained have been validated. Assuming fully realized, the
implementation process that evolves over decades. Smart Grid will have the following characteristics
that are not available in the conventional electric
IV. DEFINITION OF THE SMART GRID energy system:

The Smart Grid described in this Special Issue is • Secure communications (two-way) covering
not “pie-in-the-sky” but a true global transformation the system from end-to-end,
for which hundreds of billions of dollar-equivalents • All main components are sensed and
will be spent within the next decade on real variances detected: cables, joints,
technologies that will provide intelligent terminations, transformers, consumer usage,
management of the electric grid over the coming power quality, etc. will be monitored in real
decades. However, some aspects of the Smart Grid time.
system described herein may turn out not to be cost
effective, and they then must wait until cheaper The above characteristics will provide massive
technologies are developed or societal benefits amounts of incoming data that must be converted
justify the expenditures. That is, the ultimate Smart into situational awareness of the state of the grid.
Grid is a vision, keeping in mind that it requires Controller technologies will then have to automate
cost justification at every step before data and energy management so that information is
implementation, then testing and verification before streamlined, problems are diagnosed instantly,
extensive deployment. corrective actions are identified and executed
dynamically in the field, and feedback loops
Considering the above, the Smart Grid is defined as provide metrics that verify that the work done is
an electric system that uses information, two-way, producing the desired effects. Such Smart Grid
cyber-secure communication technologies, and controllers will have the following characteristics:
computational intelligence in an integrated fashion
across electricity generation, transmission, • Self healing: automatic repair or removal
substations, distribution and consumption to achieve of potentially faulty equipment from service
 4

before it fails, and reconfiguration of the management, diagnostic analysis, and work
system to reroute supplies of energy to management are required. The Smart Grid must
sustain power to all customers, operate as an integrated machine: a system-of-
• Flexible: the rapid and safe interconnection systems. As shown in figure 3, the Smart Grid will
of distributed generation and energy storage change the conventional concept of energy
at any point on the system at any time, management and operations since traditional
• Predictive: use of machine and “blind” demand will evolve to become controlled
reinforcement learning, weather impact “visible” demand (such as with curtailable load). In
projections, and stochastic analysis to some cases, demand will be convertable into supply
provide predictions of the next most likely (such as with electric vehicle and distributed battery
events so that appropriate actions are taken storage). Excellent industry summaries are
to reconfigure the system before next worst provided by Garrity (2008) and Some of these
events can happen, aspects of the Smart Grid are further described
• Interactive: appropriate information is below.
provided transparently regarding the status
of the system not only to the operators but Customer Demand, Demand Response and
also to the customers to allow all key Curtailable Loads
participants in the energy system to play an
active role in optimal management of
Many people, especially in the public sector,
contingencies.
consider the Smart Grid to be nothing if not
• Optimized: knowing the status of every Advanced Metering Infrastructure (AMI),
major component in real or near real time including– Automatic Meter Reading (AMR). More
and having control equipment to provide advanced features of an integrated AMI includes a
optional routing paths provides the Distribution Management System with full control,
capability for autonomous optimization of monitoring and Geographic Information System
the flow of electricity throughout the (GIS) interfaces. In addition, AMI Smart Grid
system. systems provide consumption control at the
• Secure: considering the two-way customer site, distributed load management, and 2-
communication capability of the Smart Grid way communications (c.f.,Mahmood, Aamir, and
covering the end-to-end system, the need Anis, 2008).
for physical- as well as cyber-security of all
critical assets is essential. AMI and Home Area Networks (HAN) provide
Figure 3. New equipment of the Smart Grid
added Demand Response functionality such as
automated control of refrigerators and/or air
conditioners by the utility and curtailable load based
on electronic communications only. Many conumer-
utility intermediary companies provide automated
curtailment programs through subscription services.
Certain “Self Healing” capabilities more common to
the Internet can be built into automated
reconfiguration regimes, as for example
V. MAJOR NEW COMPONENTS OF THE SMART GRID demonstrated in Tsoukalas, and Gao, 2008).
As indicated by the above characteristics, the Smart
Grid involves installation of much new, intelligent Photovoltaics
equipment at critical generation, transmission,
distribution, and consumption points. For this PhotoVoltaics (PV) provide mostly local power
equipment to become an effective part of the generation that can provide significant load relief
operations of an integrated Smart Grid, fundamental for the Smart Grid. However, the inherent
control technologies for communications, data
 5

unpredictability of the solar source caused by cloud grid where distributed generation sources such as
cover and weather make the dispatching of fixed PV and wind are linked to distributed storage, and
quantites of power impossible, unless distributed at least in concept, EV charging stations. They
storage is coupled with the PV systems. That said, provide local electric distribution for a
entire countries depend upon PV for hot water neighborhood, campus, manufacturing facility, etc.
subsystems for homes such as Cyprus. PV is that can be independent of the grid itself.
fundamentally a curtailment service, although small MicroGrids also include local load control, and
amounts of “Negawatts” are provided to the local often Heating Ventilation and Air Conditioning
grids surrounding PV arrays, especially in (HVAC) of large vertical buildings or groups of
California (c.f., Profiting from negawatts, J buildings. MicroGrids are designed to be able to
Steinberger, J Van Niel, D Bourg - Energy Policy, stand alone from the electric grid (islanding) in
2009). times of crisis so that the power in the local area can
be maintained via emergency generation.
MicroGrids can also be used for significant
EV,Charging Stations and Microgrids
curtailable load for utilities during peak load relief
periods (c.f. M. Dicorato, et. al., 2009).
Electric Vehicles and Plug-in Hybrid Electric
Vehicles (grouped as EVs here) present unique Actions between MicroGrids and the Smart Grid
problems for the electric grid because they are can then be coordinated to maintain optimal power
mobile sinks for power in the day and fixed sinks at flow, protection & switch coordination, while
night. The perception for most needed intelligent managing restoration plans and replacement
interaction via the Smart Grid comes during the day options, all the while responding to financial and
in large urban areas, when large populations of EV market variations. Pipattanasomporn and Rahman,
vehicles are predicted to plug into the grid for 2009) discuss how software agents can be deployed
recharging upon arrival at work, just as the by Smart Grid controllers to optimize just such a
electricity consumption of large urban cities is complex MicroGrid management scheme (also see
ramping up towards peak. A further homeland Liu and Su, 2008, for another good example).
security need is that each EV must receive at least a
25% recharge so all vehicles can make it out of the
city limits in case of an emergency. Thus, load Energy Storage
transfer to storage facilities linked to EV charging
stations is being considered in addition to grid A critical addition to the Smart Grid will be from
charging. In addition, so called “Green Garages” are the addition of significant energy storage capability.
beginning to appear, that certify that the power used Intermittent power sources like PV, Solar Thermal,
to charge the EV comes from renewable energy and Wind require someplace to store the electricity
sources, even as that fact is hard to verify. to fill needs during cloudy and/or windless times.
However, EVs could represent a significant mobile The Electricity Storage Organization tracks the cost
source of emergency power in case of crisis of both large and small scale energy storage
situations such as blackouts. Then, the EVs may systems, from Lithium-Ion, Nickle-Cadmium and
provide additional power to homes, assuming the Lead-Acid batteries, through fly wheels and super-
vehicles made it back home safely. There will be capacitors, to various large scale battery storage
much more on this topic as countries such as the devices, and finally to large scale cavern storage of
United States introduce new laws like “The Electric compressed air and hydroelectric that involves
Drive Vehicle Deployment Act” of 2010 that pumping water back upstream during nights (Figure
attempt to stimulate market penetration for EVs4. 4).

MicroGrids are small scale grids within the electric These technologies are all technologically viable, if
affordable: a barrier that has not yet been passed.
4
see for example: Until it is, large-scale deployment of alternative
http://markey.house.gov/index.php?option=com_content&task=view&id=400 energy sources will be limited. Other electricity
6&Itemid=141 ).
 6

storage devices that involve melting salt, heating many hours afterwards. This combination has
vegetable oils, freezing ice for HVAC chiller allowed the design of every large Solar Thermal
operations, and the use of fuel cells have attained power plants. It should be theoretically possible to
wider, though still limited, deployment (see below). build such plants that generate as much electricity
as the other two largest alternative energy sources to
burning hydrocarbons: nuclear and hydro electric.
Arizona has begun construction of the first 280 MW
of an intended 4300 MW Solar Thermal plant south
of Phoenix (Figure 5).

The outlook for wind combines strong but erratic

wind turbine generators such as those in West Texas
with Compressed Air Energy Storage (CAES) in
underground salt caverns or emptied natural gas
reservoirs. Swider (2007) has demonstrated the
economic market modeling needed to justify the
combined investment of wind generators with
CAES. Payback is minimized only if the laying of
regional transmission lines needed to get the power
to market is part of the up-front investment, a lesson
learned the hard way by Texas. There as much as
half of the 2000 MW+ of West Texas wind power is
dormant at any time because of transmission
Figure 4. The Relative Power Output, Discharge Time, and cost per KWH for limitations that are only now being alleviated, long
various Energy storage devices.
after construction of the wind generators (Anderson,
2004). A similar German study by Lerch (2008)
Distributed- and Co- Generation Demonstrated that CAES can be economic in the
case of large-scale wind power deployment in
European offshore waters.
The Smart Grid must be able to accommodate
small-scale generation owned by customers such as
combined heat and power co-generation facilities,
as well as from the previously mentioned PV, EV
and MicroGrid sources. Power management of
distributed generation involving everything from
Building Management Systems to solar generation
depends on accurate weather forecasting, which
adds to the uncertainties being optimized by smart
grid control systems. New methods linking these
erattic sources to storage are required to provide
dispatchable loads (c.f., Jiang, 2006, and
Chowdhury and Koval, 2005).

Massive Solar Thermal and Wind Farm Generation

Figure 5. An artist’s representation of the first 280 MW module of a Solar
Thermal Power Plant complex southwest of Phoenix, Arizona, that will be
Solar Thermal power generation, in particular, has selling power into the Western Interconnect and Southern California (image
been very successful in linking mirrors to a storage courtesy of Arizona State University).
medium, usually a salt that is melted or vegitable oil
that is heated. Both can be used to boil water for
 7

Nanotechnologies and Power Generation and • Nanotech lighting to replace incandescent and
Storage of the Future fluorescent lights
• Nano-Photovoltaics -- new paints for the
Above all, the Smart Grid must have the capacity to exterior of buildings that generate electricity
adapt to new technologies not yet invented or in
long term development such as fusion nuclear, or
more likely, nanotechnology breakthroughs.
Examples of future Nanotechnologies that might be
important energy sources or storage media within
the next 10 years (Smalley, 2007) include (Figure
6):
• Photovoltaic materials that may drop cost by
100 fold or more
• Photocatalysts that reduce CO2 emissions to
methanol
• Nano-materials that directly convert light and
water to produce hydrogen via
thermochemical catalysts that generate
Hydrogen from water and that work
efficiently at temperatures lower than 900 C
• Nano Fuel Cells that drop the cost by 10-100x Figure 6. Plenty of room at the bottom, according to Richard Feynman,
and provide low temp starting capacity and describing nanotechnology opportunities (c.f., Hey, J.G., Feynman and
are reversible computation: exploring the limits of computers, Westview Press, 2002).
Source: Baker Institute Study No. 30, Energy and Nanotechnology, Strategy
• Direct Hydrogen storage using lightweight for the Future, Rice University, April, 2005, accessable at:
http://www.rice.edu/energy/publications/energynanotechnology.html .
Nanl materials for pressure tanks and/or a
new lightweight, easily reversible hydrogen
chemisorption system (called material X) Taking the most likely of these to first appear on the
Smart Grid, Quantum wires (QW) have the
• Batteries, supercapacitors, and flywheels
electrical conductivity of copper at one-sixth the
improved by 10-100x for automotive and
weight and it will be stronger than Kevlar. They can
distributed generation applications
be spun into polypropylene-like “rope” and used for
• High current quantum wires (QW) that might
the transmission grid of the future. These Fullerene
rewire the transmission grid and enable
nanotude arrays form a “super-material” of extreme
continental, and even worldwide electrical
strength, lightness of weight, high temperature
energy transport; and also to replace
resistance, unidirectional thermal conductivity
aluminum and copper wires essentially
(electrons just fit into each tube, and so have only
everywhere -- particularly in the windings of
one place to go). The electrons quantum jump from
electric motors and generators
inside one QW tube to the next (c.f., Yakobson and
• Nanoelectronics to revolutionize computers, Smalley, (1997) and Anantram and Govindan,
sensors and devices. (,2000).
• Nano Robotics with Artificial Intelligence to
enable construction and maintenance of VI. ADAPTIVE STOCHASTIC CONTROL
solar structures in space and on the moon;
A key to the implementation of the Smart Grid is to
and to enable nuclear reactor maintenance
create the intelligent management of the margin
and fuel reprocessing on Earth
between the ever-expanding demand for electricity
• Super-strong, light weight materials to drop and its efficient, safe, secure, and sustainable supply
cost to launch solar arrays into space at all points along the distribution path. Electricity
• Thermochemical catalysts to generate H2 is no longer entering the grid exclusively at massive
from water that work efficiently at power plants on the transmission beginnings of the
temperatures lower than 900 C.
 8

grid, but it will also be generated from distributed programming algorithms. These systems are very
resources at customer sites throughout the good at identifying the “next worst” condition that
distribution grid, and even from energy storage at can happen to the electric grid at any given time,
consumer sites and substations. As indicated in but not so good at determining actions to prevent
Figure 7, intelligent controllers must receive, digest, the “next most likely” condition to occur on the
and interpret all manner of new data coming from electric grid.
SCADA sources and send commands to manage
contingencies, optimize power flows, initiate Controlling the new complexities of the Smart Grid
preventive maintenance, control switching and load, is a multi-stage, time-variable, stochastic
minimize capital investment, deal with erratic solar optimization problem to the Operation Research
and wind generation, and optimize distributed engineer and operator. The Adaptive Stochastic
storage, all the while dealing with potential and real Controller (ASC) for the Smart Grid requires the
equipment failures as well as weather and price import of Approximate Dynamic Programming
variations. (ADP) and Mixed-Integer, Nonlinear Programming
solvers that are more familiar to the petrochemical
and transportation industries. If the electricity
industry successfully imports these intelligent
controllers, economic benefits will be on the order
of savings we have measured after transitions to
such autonomous, adaptive controllers in these
other industries (c.f. Anderson et al, 2008).

ADP Adaptive Stochastic Control optimizes by

solving the Hamilton-Jacobi-Bellman equation
using ADP interacting with the electric system
model used by engineering to plan improvements in
the grid today. Feedback loops and a critic function
are critical to establish cause-and-effect, similar to
Figure 7. The Smart Grid must optimally interpret incoming data from many the ways that models are used in Model Predictive
new sources (green) with new controls such as Contingency Analysis
Programs (CAP in the upper right) and Capital Asset Prioritization Tools
Control (MPC) in other industries. The ADP
(CAPT in lower right). Adaptive Stochastic Controller delivers these real-
time feedback loops to Machine Learning Systems
This Smart Grid data and energy management that enable simultaneous Load & Source Control
system must be, by definition, adaptive and and Dynamic Treatment Optimization to drive the
stochastic, meaning that it is prepared to respond to Smart Grid to fewer and fewer emergency failures
varying weather conditions, crew status, and over time.
equipment performance changes while optimizing
supply to meet demand within economic constraints Momoh (2006) offers an excellent summary of the
that simultaneously minimize costs for consumers, present state of the art in ADP Adaptive Stochastic
regulators and industry stockholders. Control and the technologies that it is replacing in
Control Centers across the world (Table 1 is
Some utilities now use complex, computational replicated below).
command and control systems similar to those used
in petrochemical and nuclear plant management,
such as decision support and portfolio management
tools, activity based accounting, and preventive
maintenance programs. However, the computational
systems utilized in these controller calculations are
generally policy-and-rules-based decision systems.
Risk and variance are considered using linear
 9

Current business plans for utilities are also based on

risk/reward optimization algorithms, although these
are usually Net Present Value (NPV) computations
used in a portfolio management context. The future
Smart Grid will be too dynamic and too complex
for such linear methodologies. Real options
optimization and autonomous control will be
required throughout the grid, from consumer to
utility to generation, as described by Ernst, et al,
(2008).

Load & Source Control:

The control of distributed energy storage for the real

time Load nad Source Control has been mostly
Werbos (2009) further describes the intelligence limited to date to pumped hydro. However, recent
that must be mathematically managed using state- demonstration projects are providing new
of-the-art computational control theory. Chuang and opportunities to show the value of energy storage in
McGranaghan (2008), further develop the
the control of grid stabilization, operations support,
requirements for such intelligent controllers for power quality management, and load shifting
Smart Grids to include distributed generation and applications. Candidate high value applications are:
storage devices and the interfaces needed connect to
• “Instantaneous” versus Ramp Rate Limited
electricity market participation. Thusly, the ADP
Generation Based Spinning Reserve as
Adaptive Stochastic Controller coordinates
Bridge Power to Standby Generation in the
generation, utility and customer response to actual
Event of Loss of Generation or
system and market conditions in real-time.
Transmission
Within the Adaptive Stochastic Controller, ADP is • Cycling Power Supply and Load Arbitrage
used as a mathematical decomposition strategy that • Regulation Control Support
breaks the problem of continuous grid management, • Reserve Power for System Power
with its long time horizons, into a series of shorter Reliability, Security and Quality
problems that the Mixed-Integer Nonlinear • Utility Load Shifting for Supply
Programming solver can handle, thereby solving the Infrastructure Asset Optimization and
‘curse-of-dimensionality’ caused by the exponential Emergency Response
increase in complexity as dimensionality of sources • End-user Energy Management and Power
and loads increase as the Smart Grid is built-out. Quality/Reliability Requirements
The ADP Adaptive Stochastic Controller • Intermittent Renewable Power (e.g., Wind)
framework also provides a way of treating Stability and Optimization
uncertainty from both operational and financial
standpoints, simultaneously. To achieve this, ADP Adaptive Stochastic Control for Load and
Anderson and Boulanger (2009) describe the Source management must exercise real options
mathematical combination of Real Options decisions in real time based on price and market
evaluation of the cost/benefit with operational condition information fed to the controller. The real
policy and action optimization using the ADP time options include:
algorithms within the Adaptive Stochastic • option value of arbitrage,
Controller. The result is maximization of real option • option value of peak shaving,
value as a control objective along with efficient and • option value of greater network reliability,
safe operations. • option value of environmental benefits
• Dynamic Treatment
 10

ADP Adaptive Stochastic Control also provides years that will be required for a full conversion to a
Dynamic Treatment optimization for maintenance comprehensive Smart Grid in any country.
operations. Machine Learning and statistical Documenting these improvements requires the
models for failure that use causal inference, establishment of an initial baseline for all major
propensity and survival analysis developed for the components of the existing grid, and then
medical industry have been shown effective to continuous measurement of the impact of new
arrive at treatment actions to prevent electric grid construction and implementation against that
failures (Rudin, et al, 2011). The dynamic treatment baseline. A benefit from this documentation will be
output of Adaptive Stochastic Control is a that Adaptive Stochastic Controllers of the Smart
prioritization of work needed and control actions to Grid will have been validated to redirect load
be taken, either discretely or continuously, to keep around congestion, manage peak demand, weather
grid devices such as distributed generation and and equipment problems that will eliminate the
storage, sectionalizing switches, and load pockets need for expensive new power plants and
within optimal performance bounds. substations. Internationally, computers operating
these Adaptive Stochastic Controllers managing
every level of the new Smart Grid could eventually
Adaptive Stochastic Control in Transmission
save the need to build Terra-watts of new
generation worldwide.
The Adaptive Stochastic Control functions for the
Smart Grid also include static and dynamic security It is expected that over time the Smart Grid will
assessment capabilities along with self-checking of improve the capacity factor of the electric system
relay settings on critical transmission facilities. through more optimal supply and demand
With the deployment of phasor measurement units management. It allows for the re-use of existing
to monitor grid performance across heavily loaded hardware infrastructure in a more efficient manner
regional transmission interconnections, there will be by adding modern controller intelligence to the
advances in state estimators that are capable of real- existing system. Understanding the risks and
time simulations for large networks that will need to consumer impact of using the available resources
be incorporated into the overall management of the optimally should allow Smart Grid utilities to lower
Smart Grid. For example, mobilizing grid capacity peak demand and reduce Capital and O&M costs by
reserves through active management to avoid mitigating emergencies of all kinds during peak
overloads will enable operators to relieve load periods. It is our joint task as an industry to
bottlenecks and redeploy necessary generation and maintain the tracking metrics worldwide to
transmission assets from both the transmission and document that these predicted benefits are actually
distribution grids to eliminate congestion points and realized by the Smart Grid implementation we are
load pockets in the integrated transmission and all beginning. Best practices should be shared easily
distribution grids. and efficiently.

Challenges to the future success of the smart grid

VII. CHALLENGES TO ACHIEVING A
come from many fronts, such as consumer buy-in:
COMPREHENSIVE SMART GRID consumers have to see real savings and efficiency
improvements; better regulation: governmental
A primary objective of the Smart Grid is to improve control must stay up to date technologically and in
our capacity to use more, but cheaper, electricity to touch with consumers; cost justification; Smart Grid
power the improvements in the standard-of-living of components must be individually as well as
all people on Planet Earth. However, the transition systemically cost effective; education: utilities,
must be cost effective, or we will never get there service companies and universities must produce
from here. The tracking of key performance educated consumers as well as a new generation of
metrics that continuously and automatically score electrical engineer savvy in computer sciences and
improvements generated by the Smart Grid will be systems engineering; and new inventions and
required if the effort is sustainable over the 20 to 30 technologies must be easily adopted and adapted
 11

into the Smart Grid since it will evolve over the [20] Anderson, R., and A. Boulanger, Innervated Stochastic Controller For
Real Time Business Decision-Making Support, United States Letters Patent,
next 20 to 30 years. Number 7,395,252, (2009).

[21] Chuang, J., and M. McGranaghan, Functions of a Local Controller to

REFERENCES Coordinate Distributed Resources in a Smart Grid Angela Chuang, IEEE,
2008.
[1] Mahmood, A., M. Aamir, and M.Anis, Design and Implementation of
AMR Smart Grid System, 2008 IEEE Electrical Power & Energy Conference, [22] Ernst, D., M Gravic, F. Capitanescu, and L. Wehenkel, Reinforcement
2008. Learning Versus Model Predictive Control: A Comparison on a Power System
Problem, IEEE Transactions On Systems, Man, And Cybernetics—Part B:
[2] Garrity, T., Innovation and Trends forFuture Electric Power Systems, Cybernetics, Vol. 39, No. 2, April 2008.
IEEE Power and Energy, 38-45, March-April, 2008.
[23] Rudin, C., et al, Machine Learning for the New York City Power1 Grid,
IEEE Proceedings, Special Issue on the Smart Grid, 2011.
[3] Katz, J., Educating the Smart Grid, IEEE Energy 2030, 2008.
[4] Tsoukalas, L., and R. Gao, From Smart Grids to an Energy Internet [24] McDonald, J., Leader or Follower: Developing the Smart Grid Business
Assumptions, Architectures and Requirements, IEEE DRPT Conference, Case, IEEE Power &Energy, p. 18-24, Nov-Dec, 2008.
April, 2008.
[25] Schulz, A. et al, Agile engineering versus Agile Systems Engineering,
[5] Steinberger, J., J. Van Niel, D. Bourg, Profiting from Negawatts: Systems Engineering, V. 3, Issue 4, p. 180-211, 1999.
Reducing absolute consumption and emissions through a performance-based
energy economy, in Elsivier, Energy Policy, 2009. [26] Hamlyn, A., H. Cheung, T. Mander, L. Wang, C. Yang, R. Cheung,
Computer Network Security Management and Authentication of Smart Grids
[6] Pipattanasomporn A., and A. Rahman, Multi-Agent Systems in a Operations, IEEE, 2008.
Distributed Smart Grid: Design and Implementation, Proc. IEEE PES 2009
Power Systems Conference and Exposition, Mar. 2009. [27] Hauser, C., D. Bakken, a. Bose, A Failure to Communicate, IEEE Power
and Energy, 2005.

[7] Liu, X., and Su, B., Microgrids - An Integration of Renewable Energy
Technologies, in Protection, Control, Communication and Automation of
Distribution Networks, S3-25,CT 1800, CICED, 2008.

[8] Dicorato, M., G. Forte, and M Trovato, A procedure for evaluating

technical and economic feasibility issues of MicroGrids, IEEE Bucharest
Power Tech Conference, 2009.

[9] Jiang, Z., Power Management of Hybrid Photovoltaic - Fuel Cell Power
Systems, IEEE paper 1-4244-0493-2, 2006.

[10] Chowdhury, A., and D. Koval, Impact of PV Power Sources on a Power

System’s Capacity reliability Levels, IEEE I&CPS-05-4, 2005.

[11] Swider, D., Compressed Air Energy Storage in an Electricity System,

with Significant Wind Power Generation, IEEE Trans of Energy Conversion,
v. 22, no. 1, 95-102, 2007.

[12] Anderson, R., Texas Wind Energy Plan, Railroad Commission of Texas
Report to the Texas Energy Planning Council, July 24, 2004.

[13] Lerch, E., Storage of Fluctuating Wind Storage: Cse for Compressed Air
Energy Storage in Germany, IEEE, 2008.

[14] Smalley, R., Our Energy Challenge, at

http://video.google.com/videoplay?docid=-4626573768558163231#

[15] Yakobson, A., and R. Smalley, Fullerene Nanotubes: C1,000,000 and

Beyond, American Scientist, 85-4, p.324-337, 1997,

[16] Anantram, M., and T. Govindan, Transmission through carbon

nanotubes with olyhedral caps. M. P. Phys. Rev. B, 61(7) p. 5020, 2000).

[17] Anderson, R., et al, Computer-Aided Lean Management in the Energy

Industry, PennWell Press, 2008.

[18] Momoh, J., Optimal Methods for Power System Operation and
Management, PSCE, p. 179-186, 2006.

[19] Werbos, P., Putting More Brain-Like Intelligence into the Electric Power
Grid: What We Need and How to Do It, Proceedings of the 2009 international
joint conference on Neural Networks, IEEE Computational Intelligence
Society, 2009.

View publication stats