7PMVNF

t
/VNCFS

t
+BOVBSZ'FCSVBSZ


CYME power engineering

software for a smarter grid

Standing behind thousands of

T&D projects worldwide
Power Engineering Software and Solutions
Eaton's extensive line of the CYME power engineering
software features advanced analysis for transmission,
distribution and industrial electrical power systems. Our
services include engineering consulting, training, data
integration and customized IT developments.

CYME is the perfect solution for:

> Optimization of grid
efciency
> Distributed generation
impact
> Network-wide planning using
AMI/AMR data
> Key reliability assets
conguration improvement
> Downtown meshed grids
and secondary low-voltage
distribution network studies
> Ampacity calculations and
real-time thermal rating of
cables
> Protective device
coordination
> Design and optimization of
AC substation grids
> Data extraction from GIS and
quasi real-time simulations
> Customized plug-in and
applications

Contact us today to schedule a software demonstration or to

learn more about our solutions.

1-450-461-3655 1-800-361-3627 cymeinfo@eaton.com

www.cyme.com www.cooperpowereas.com

AD917001EN-2015
 magazine

7PMVNF

t
/VNCFS

t
+BOVBSZ'FCSVBSZ

www.ieee.org/power

on the 43
cover

ISTOCKPHOTO.COM/LEOWOLFERT
ISTOCKPHOTO.COM/AURIS

features

DPOUFOUT
16 Unleashing the Flexibility of Gas 43 Unlocking Flexibility
By Steve Heinen, Christian Hewicker, By Emiliano DallAnese, Pierluigi Mancarella,
Nick Jenkins, James McCalley, Mark OMalley, and Antonello Monti
Sauro Pasini, and Simone Simoncini
53 The Consumers Role in

Harnessing Flexibility from Flexible Energy Systems
Hot and Cold By Geertje Schuitema, Lisa Ryan,
By Juha Kiviluoma, Steve Heinen, and Claudia Aravena
Hassan Qazi, Henrik Madsen, Goran Strbac,
61 Flexibility Challenges for Energy Markets
Chongqing Kang, Ning Zhang,
By William Dhaeseleer, Laurens de Vries,
Dieter Patteeuw, and Tobias Naegler
Chongqing Kang, and Erik Delarue
34 The Triple Bottom Line for Efficiency
By Eoin Casey, Sara Beaini,
Sudeshna Pabi, Kent Zammit,
and Ammi Amarnath

columns &
72 departments
4 'SPN
UIF
&EJUPS
8 -FBEFST
$PSOFS


(VFTU
&EJUPSJBM


)JTUPSZ
85 4PDJFUZ
/FXT


$BMFOEBS


*O
.Z
7JFX
Digital Object Identifier 10.1109/MPE.2016.2620759

january/february 2017 IEEE power & energy magazine 1

 magazine

Editor-in-Chief IEEE Periodicals/Magazines Department

Michael I. Henderson 445 Hoes Lane, Piscataway, NJ 08854 USA
mih.psat@gmail.com www.ieee.org/magazines

Melvin I. Olken, Editor-in Chief Emeritus Geraldine Krolin-Taylor, Senior Managing Editor
Janet Dudar, Senior Art Director
Carl L. Sulzberger, Associate Editor Emeritus, Gail A. Schnitzer and Mark Morrissey, Associate Art Directors
History Theresa L. Smith, Production Coordinator
Peter M. Tuohy, Production Director
Editorial Board Felicia Spagnoli, Advertising Production Manager
J. Feltes, L. Goel, Dawn Melley, Editorial Director
N. Hatziargyriou, Fran Zappulla, Staff Director, IEEE Publishing Operations
Tao Hong, B. Johnson, IEEE prohibits discrimination, harassment, and bullying. For more information, Promoting Sustainable Forestry

Ben Kroposki, P. Kundur, N. Lu, visit http://www.ieee.org/web/aboutus/whatis/policies/p9-26.html. SFI-01681

A.P.S. Meliopoulos, M. Miller,

M. OMalley, N. Ochoa,
C.E. Root, H. Rudnick, IEEE Power & Energy Magazine
M. Shahidehpour, G.B. Shebl. C. Smith, IEEE Power & Energy Magazine (ISSN 1540-7977) (IPEMCF) is published bimonthly by the Institute of Electrical and
Electronics Engineers, Inc. Headquarters: 3 Park Avenue, 17th Floor, New York, NY 10016-5997 USA. Responsibility forthe
M. Thomas, E. Uzunovic, contents rests upon the authors and not upon the IEEE, the Society, or its members. IEEE Operations Center (for orders, sub-
S.S. Venkata, J. Wang scriptions, address changes): 445 Hoes Lane, Piscataway, NJ 08854 USA. Telephone: +1 732 981 0060, +1 800 678 4333.
S. Widergren Individual copies: IEEE members US$20.00 (first copy only), nonmembers US$82.00 per copy. Subscription Rates:
Society members included with membership dues. Subscription rates available upon request. Copyright and reprint permis-
Advertising sions: Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limits of U.S.
Erik Henson Copyright law for the private use of patrons 1) those post-1977 articles that carry a code at the bottom of the first page,
provided the per-copy fee indicated in the code is paid through the Copyright Clearance Center, 222 Rosewood Drive,
Naylor Association Solutions Danvers, MA 01923 USA; 2) pre-1978 articles without fee. For other copying, reprint, or republication permission, write
+1 352 333 3443, fax: +1 352 331 3525 Copyrights and Permissions Department, IEEE Operations Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. Copyright
ehenson@naylor.com 2017 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Periodicals postage paid at New
York, NY, and at additional mailing offices. Postmaster: Send address changes to IEEE Power & Energy Magazine, IEEE
Operations Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. Canadian GST #125634188
Printed in U.S.A.

IEEE POWER & ENERGY SOCIETY (PES)

The IEEE Power & Energy Society is an organization of IEEE members whose principal interest is the advancement of the science and practice of elec-
tric power generation, transmission, distribution, and utilization. All members of the IEEE are eligible for membership in the Society.
Mission Statement: To be the leading provider of scientific and engineering information on electric power and energy for the betterment of society, and
the preferred professional development source for our members.
Officers Standing Committee Chairs Technical Committee Chairs
D. Novosel, President H. Louie, Constitution & Bylaws A. Schneider, Analytical Methods for
S. Rahman, President-Elect C. Root, Finance & Audit Power Systems
F. Lambert, Vice President, Chapters M. Begovic, Nominations & Appointments K. Haran, Electric Machinery
M.P. Sanders, Vice President, Technical Activities S. Brahma, Power & Energy Education W.T. Jewell, Energy Development & Power
E. Uzunovic, Vice President, Education W.K. Reder & P.Sauer, Scholarship Plus Generation
M. Crow, Vice President, Publications C. Searles, Energy Storage & Stationary Battery
T. Mayne, Vice President, Meetings Chapter Representatives F. Frentzas, Insulated Conductors
J.C. Montero Q, Vice President, F. Al Dhaheri, W. Almuhtadi, Z. Bo, T. Koshy, Nuclear Power Engineering
Membership & Image R. Cespedes, S. Cundeva, C. Diamond, M. Dood, Power System Communications
S. Bahramirad, Vice President, New Initiatives/ & Cybersecurity
Outreach G. Gonzalez, J. Khan,
R. Nagaraja, N. Nair, M. Papic, I. Petruzela, C. Canizares, Power System Dynamic Performance
C. Root, Treasurer J. McBride, Power System Instrumentation
J. Bian, Secretary T. Salihy, G.N. Taranto, D. van Hertem,
M.C. Wong, Z. Zakaria & Measurements
M.M. Begovic, Past-President H. Chen, Power System Operation Planning &
IEEE Division VII Director Chapter Committee Chairs Economics
A. Rotz C. Diamond, Electronic Communications M. Pratap, Power System Relaying & Control
E. Carlsen, Awards & Resources S. Chandler, Smart Building, Load &
IEEE Division VII Director-Elect Y. Chen, Distinguished Lecturer Program Customer Systems
B. Meyer C. Diamond, Chapters Web site D. Watkins, Substations
R. Hotchkiss, Surge Protective Devices
Region Representatives Membership & Image P. Sullivan, Switchgear
D. Diaz, B. Gwyn, R. Krepps, Committee Chairs S. Antosz, Transformers
N. Logic, P. Pabst, C. Wong, V.Madani, Awards & Recognition D. Sabin, Transmission & Distribution
United States J.Hofman, Young Professionals
M. Armstrong, Canada Technical Council
Open, Ambassadors Coordinating Committees
J. Milanovic, Europe, Middle East, & Africa I. Riao S., Social Media
N. Segoshi, Latin America D. Houseman, Intelligent Grid & Emerging
D.Sharafi, Asia & Pacific J.C. Montero, Web Site Development Technology
S.Ninalowo, Women in Power D. Alexander, Marine Systems
Governing Board W. Bishop, Marketing D. Lew, Wind & Solar Power
Members-at-Large J.C. Montero, Membership Development
C.Y. Chung, B.Enayati, Technical Council
H. Koch, L. Ochoa Technical Council Standing Committees
M.P. Sanders, Chair J.H. Nelson, Awards
PES Executive Director F. Rahmatian, Vice Chair F. Rahmatian, Technical Sessions
Patrick Ryan, +1 732 465 6618, V. Vittal, Secretary V. Vittal, Organization & Procedures
fax +1 732 562 3881, p.ryan@ieee.org J.H. Nelson, Past-Chair T. Burse, Standards Coordination
D. Toland, Web Master
Digital Object Identifier 10.1109/MPE.2016.2620760

2 IEEE power & energy magazine january/february 2017

DIgSILENT announces

PowerFactory 2017
DIgSILENT has set standards and trends in power system modelling,
analysis and simulation for more than 25 years.
In the PowerFactory 2017 release a range of improvements have
been made to the handling and data management with a series
of additional convenience functions. The new version also comes
with a series of enhancements to the network diagrams and
graphic representation tools. A special focus in development has
been on the various analysis function capabilities of PowerFactory.
This includes a number of new power equipment models and ex-
tensions to existing ones. Most notably, further modelling flexi-
bility has been provided with the inclusion of user-defined models
for load flow and quasi-dynamic simulation.

Key Features
Enhanced Diagram Layout Tool for auto-drawing of feeders and branches,
protection device layout, as well as auto-layout of site and substation diagrams
Output window redesigned to be interactive, with flexible filter functionality
Add-on Modules: new framework for user-extendable function scope including
fully integrated result representation
New Project Combination and Connection Assistant
New and enhanced Power Equipment Models: harmonic filters, busbar trunking
systems, voltage regulator, 4-w transformer, power freq. control using merit order
New QDSL modelling language: User-definable load flow and quasi-dynamic
simulation models
IEC60909 Update - 2016 edition
New Protection Audit validation tool for protection settings and configurations
Connection Request Assessment: BDEW 2008 and VDE AR-N-4105 guidelines
New optimisation methods for Tie Open Point and Phase Balance Optimisation:
genetic algorithms and simulated annealing
Extended failure models and power restoration strategies for Reliability Analysis
New Outage Planning module
Extension of simulation scan by Fault Ride Through verification
IEC 61400-2-27 interface for external dynamic models
CGMES interface: functional extensions and performance improvements
New Integral export function

For more Information about DIgSILENT PowerFactory visit www.digsilent.com.

 from the editor
Michael Henderson

energy system flexibility

the importance of being nimble

O
OUR ENERGY SYSTEM FLEXIBIL-
ity issue discusses creative ways that the
power system can reliably accommodate
the large-scale development of variable
resources and load that fluctuates consid-
erably with weather, price, and behind-
the-meter generation. This issues guest
editors, Mark OMalley and Benjamin
Kroposki, collected six articles that dis-
IMAGE LICENSED BY INGRAM PUBLISHING
cuss the importance of flexibility and how
it can be achieved in ways not always
considered by electric power engineers.
The Guest Editorial summarizes the ever, remained unknown to the general Sulzberger as associate editor for the
flexibility theme and introduces articles public and most electrical engineers History column. Their extraordinary
that feature the following topics: until an article in this issues History service to this magazine and to IEEE
the role of the consumer in en- column on this topic. The use of re- is truly remarkable. Heartfelt thanks to
ergy systems actance-type dimmers both of you on behalf
the place of natural gas-fired gen- utilizing imaginary of the IEEE Power &
eration and its fuel supply power in the imaginary The Energy Society, its of-
the interaction between water world of the stage is ficers, and IEEE staff.
operations and electric power truly captivating. A technology of On a more personal
sectors special thanks to Tom stage lighting note, I am deeply in-
the integration and opt i m i za- Blalock, one of our most debted to you for your
t ion of multienergy system con- prolific history History remained guidance and tutelage.
trols authors, who presents And a special thanks
heating and cooling technologies the evolution of stage unknown to in advance for the con-
that provide storage lighting to us. The ar- the general tinued support of Mel
policies and markets that promote ticle is the last of Carl Olken, who always pro-
flexible electric system responses. Sulzbergers remarkable public and fessionally and patiently
It is my hope you find these articles in- contributions to IEEE answers my questions
teresting and informative. Power & Energy Maga- most electrical and provides true lead-
zine as the associate edi- engineers until ership as editor-in-
Reactors for the Roxy tor for History. chief emeritus.
At 6,000 seats, the Roxy Thea t r e i n an article in
New York City was widely known. Special Thanks Changing of the
The technology of stage lighting, how- IEEE Power & Energy this issues Editorial Guard
Magazine saw 84 is- History This issue represents a
Digital Object Identifier 10.1109/MPE.2016.2627098 sues with Mel Olken as changing of the guard
Date of publication: 2 February 2017 editor-in-chief and Carl column. for the magazines

4 IEEE power & energy magazine january/february 2017

Your NEPLAN experience

frick-partner.ch
is just one click away

Order now your NEPLAN 360 Cloud-Engineering Package with

one simple click. Visit www.neplan.ch/shop
Do you want to explore the future of cloud computing? Now we give
you the opportunity to work with the first fully browser-based
power system analysis tool!
We give you the chance to experience our powerful software with
the easy-start NEPLAN 360 Cloud-Engineering Packages, specially
designed for engineers and engineering companies with focus on
high cost efficiency and flexibility.

When it has to be fast, reliable and simple.

Visit NEPLAN360

NEPLAN AG Power Systems Engineering Oberwachtstrasse 2 CH 8700 Ksnacht Zurich

Phone +41 44 914 36 66 Fax +41 44 991 19 71 www.neplan.ch info@neplan.ch
 Editorial Board. A sincere thank you Mohammad Shahidehpour
for all the hard work by our outgoing Gerry Sheble
board members: Mani Venkata.
Massoud Amin And welcome to our new Editorial
Andrew Hanson Board members:
Stan Horowitz Jim Feltes
Ralph Masiello Tao Hong
Keene Matsuda Brian Johnson
Arun Phadke Ben Kroposki
Dick Piwko Ning Lu
Pete Sauer Mackay Miller
Ben Shperling. Nando Ochoa
My appreciation for continued ser- Charlie Smith
vice on the Editorial Board by Mini Thomas
Lalit Goel Edvina Uzunovic
Nickos Hatziargyriou Jianhui Wang
Prabbha Kundar Steve Widergren.
Sakis Meliopoulos Enrique Tejera will serve as our liaison
Mark OMalley with the newly formed Spanish version
Chris Root of IEEE Power & Energy Magazine.
p&e
Hugh Rudnick

6 IEEE power & energy magazine january/february 2017

 leaders corner
T
TWENTY-SEVEN YEARS AGO, I
joined the IEEE Power & Energy Soci-
ety (PES) as a student member. When
I attended my very first PES General
Meeting, I met famous professors who
I admired. Many of these professors
wrote the textbooks that I had used
since my junior year in college. I of-
ten asked, Im a big fan of yours and
Im curious: Who do you learn from?
I wanted to know who inspires them,
which books they read, and who their
mentors were. People who are the
best at what they domasters of their
craftknow that they wouldnt have
achieved the success they have if it
werent for their mentors along the way.
Throughout my 27 years with PES,
from connecting with professors, pub-
lishing my papers in IEEE publica-
tions, and searching for internships, to
growing in my career and profession,
I have not only used what PES offered
every step of the way but also became
a mentor and lifelong friend of many
PES colleagues.
In 2016, I started serving as the PES
secretary. Ive taken some time to re-
flect on the many interactions Ive had
with members around the world. Every
member has different needs. The ques-
tions I received the most, particularly
from student members, are Could you
be my mentor? and How can I find
Digital Object Identifier 10.1109/MPE.2016.2620619
Date of publication: 2 February 2017
8 IEEE power & energy magazine
more power to the future
Internships and more
an internship position? My answer to
the first question is absolutely yes. My
answer to the second
question is Get in-
volved with your local As the PES
PES Chapters and Stu-
dent Branch Chapters
secretary, I
when you first enroll want to build a
in college. Students
n o t o n ly f i n d m o r estronger support
mentors but also have
the opportunity to con-
network
tribute and network for student
in person with PES
professionals. members.
As the PES secretary,
I want to build a stronger support net-
work for student members. No one can
say they did it on their own. Weve
done so much because someone gave
us strength and helped us along the
way. PES Chapters and Student Branch
Chapters are local volunteer orga-
nizations that meet regularly, often
once a month. As of December 2015,
PES had 235 Chapters and 175 Student
Branch Chapters worldwide and is still
growing. This support network is the
PES identity!
Student members can learn more
about PES Chapters in their local areas
by going to www.ieee-pes.org/chapters.
Chapter meetings provide an excellent
way to meet other power and energy
professionals in local companies where
there may be internship openings. Many
PES Chapters are involved in outreach
Jessica Bain
activities and have programs for students
to grease the pipeline for future power
engineers. In addition to
socializing and network-
ing, Chapter meetings
can include technology
demonstrations, invited
lectures, plant trips,
and tutorials. Websites
and e-newsletters keep
Chapter members up to
date with local happen-
ings related to power
and energy. Awards
and recognitions are
available for Chapters
based on activities or students/engineers
within it.
As I met so many of you in 2016,
we talked about the problems and
challenges that the PES faces. I have
been reminded that there is a great
power in this community, the power
that we can use to solve problems and
meet the future challenges. My job is
to be your stepping stones, helping
every member succeed. We may play
different instruments, but we are in the
same orchestra.
My PES mentor and lifelong friend,
Don Ramey, filled me with dreams to
predict thunder and lightning 27 years
ago. My brilliant friend and supporter
has made me a better student, a better
engineer, and a better secretary.
p&e
january/february 2017
 3K\VLFDODQG2SHUDWLRQDO0DUJLQV
3206XLWH
3203K\VLFDODQG2SHUDWLRQDO
0DUJLQV
2302SWLPDO0LWLJDWLRQ0HDVXUHV
%25%RXQGDU\RI2SHUDWLQJ5HJLRQ
3&03RWHQWLDO&DVFDGLQJ0RGHV
))6 )DVW)DXOW6FUHHQLQJ
3$/03KDVH$QJOH/LPLW0RQLWRULQJ
526(5HJLRQRI6WDELOLW\([LVWHQFH
3206(3206WDWH(VWLPDWRU
320763207UDQVLHQW6WDELOLW\
32036 3203URMHFW6HOHFWLRQ
3206$ 6PDOO6LJQDO$QDO\VLV
%2576%257UDQVLHQW6WDELOLW\
320573205HDO7LPH
LQIR#YUHQHUJ\FRP

ZZZYUHQHUJ\FRP
3206XLWHRQHVWRSVROXWLRQIRUDOO
\RXUSRZHUV\VWHPDQDO\VLVQHHGV
6PDUW7RROVIRUWKH6PDUW*ULG
8WLOL]LQJ308'DWDWRSUHGLFWV\VWHPLQVWDELOLW\DQGJHQHUDWHDODUPV
&RPSXWLQJSKDVHDQJOHOLPLWLQUHDOWLPHDQGROLQHHQYLURQPHQWV
,PSURYLQJWKHWUDQVPLVVLRQJULGUHOLDELOLW\WRIDFLOLWDWHLQWHJUDWLRQRI
WKHUHQHZDEOHHQHUJ\
([HFXWLRQWLPHIRUDEXVFDVH
320 $&VROXWLRQIRURQHFRQWLQJHQF\VHF
230 $OOHYLDWHVWKHUPDOYROWDJHDQGYROWDJHVWDELOLW\
YLRODWLRQVLQVHFIRURQHFRQWLQJHQF\
32076 $RQHVHFRQGWLPHGRPDLQVLPXODWLRQLVVHF
IRUG\QDPLFPRGHOV
$XWRPDWHVVWXGLHVWRFRPSO\ZLWKWKH1(5&5HOLDELOLW\
6WDQGDUG73/
$XWRPDWLFDOO\DVVHVVHV3ODQQLQJ(YHQWV
PDVVLYH111$&FRQWLQJHQF\DQDO\VLVDQG
YLVXDOL]DWLRQ
$XWRPDWLFDOO\DVVHVVHV([WUHPH(YHQWV
SUHGLFWLQJFDVFDGLQJRXWDJHVDQGYLVXDOL]DWLRQ
7UDQVLHQWVWDELOLW\EDVHGFDVFDGLQJDQDO\VLVZLWK
SURWHFWLRQV\VWHPPRGHOLQJ
2HUVIDVWDQGHDV\WRXVHWUDQVLHQWVWDELOLW\DQDO\VLV
3HUIRUPVIDVWIDXOWVFUHHQLQJWRGHWHUPLQHDQGUDQNPRVW
VHYHUHWKUHHSKDVHDQGXQEDODQFHGIDXOWV
2HUVDERXQGDU\EDVHGVROXWLRQIDVWDQGDFFXUDWH
DVVHVVPHQWRIRSHUDWLQJOLPLWV
'HWHUPLQHVUHPHGLDODFWLRQVWRVWD\ZLWKLQV\VWHP
RSHUDWLQJFULWHULDDQGLQFUHDVHVYROWDJHVWDELOLW\PDUJLQV
$XWRPDWLFDOO\FRPSXWHVFULWLFDOFOHDULQJWLPH
'HWHUPLQHVDQGYLVXDOL]HVWKHG\QDPLFVHFXULW\UHJLRQ
3HUIRUPVFRQYHQWLRQDOK\EULGDQGOLQHDUVWDWHHVWLPDWLRQ
:HRHU6SHFLDOL]HG&RQVXOWLQJ6HUYLFHVDQG
3RZHU6\VWHP6WXGLHVEDVHGRQWKHVWDWHRIWKHDUW
VXLWHRISRZHUV\VWHPWRROV
 guest editorial
E
ELECTRIC POWER SYSTEMS
around the world are experiencing
great changes, including the retirement
of coal and nuclear plants along with a
rapid increase in the use of natural gas
turbines and variable renewable tech-
nologies such as wind and solar. There
is also much more use of information
and communications technologies to
enhance the visibility and controllabil-
ity of the grid. Flexibility of operation,
the ability of a power system to respond
to change in demand and supply, is
critical to enable higher levels of vari-
able generation. One way to unlock this
potential flexibility is to tap into other
energy domains. This concept of inter-
connecting energy domains is called
energy systems integration (ESI).
ESI is the process of coordinating
the operation and planning of energy
systems across multiple pathways and/
or geographical scales to deliver reliable,
cost-effective energy services with mini-
mal impact on the environment. Integrat-
ing energy domains adds flexibility to
the electrical power system. ESI includes
interactions among energy vectors (e.g.,
electricity, thermal, and fuels) and with
other large-scale infrastructures includ-
ing water, transport, and data and com-
munications networks, which are an en-
abling technology for ESI.
The value of ESI is in coordinating
how energy systems produce and deliver
energy in all forms to reach reliable, eco-
nomic, and/or environmental goals at ap-
Digital Object Identifier 10.1109/MPE.2016.2629703
Date of publication: 2 February 2017
10 IEEE power & energy magazine
Mark OMalley and Benjamin Kroposki
unlocking flexibility
energy systems integration
propriate scales. The analysis of integrated
energy systems can inform policy makers
and industry on the best strategies to ac-
complish these goals. The benefits of ESI
include the integration of higher levels of
variable renewables, an increased reli-
ability and improved efficiency in power
systems, as well as significant savings
in, e.g., water, heating/cooling, and gas
system operations that can be achieved
by using the flexibility that emerges
from an integrated operation of mul-
tiple energy systems at multiple spatio-
temporal scales.
ESI, a multidisciplinary area ranging
from science, engineering, and technol-
ogy to policy, economics, regulation,
and human behavior, is most valuable
at the physical, institutional, and spatial
interfaces, where there are new chal-
lenges and opportunities for research,
demonstration, and deployment to reap
its commercial and societal benefits.
The simultaneous focus on multiple dis-
ciplines and stakeholders makes ESI a
challenging and exciting area.
This issue of IEEE Power & Energy
Magazine contains six articles that ex-
amine the topic of flexibility in energy
systems. The articles reach outside the
normal domain of electricity and con-
tinue to explore the concept of how ESI
can provide flexibility in future power
systems by tapping into the potential to
shift supply and demand across energy
vectors and networks.
The first article, Unleashing the
Flexibility of Gas by Heinen et al.,
discusses how the natural gas and elec-
trical infrastructures have continued to
become more tightly coupled than ever
before and addresses the capability of
the gas system to meet electricity sys-
tem flexibility requirements. Flexibility
of the system is analyzed in three parts:
from gas-powered generation, from gas
supply, and through multi-input, multi-
output plants and appliances.
Gas-powered generation plants are typi-
cally more flexible than many other forms
of generation, capable of starting quickly
and with significant ramping capability.
They are often an ideal complement to
variable renewable energy, although in-
creasing levels of variable renewable en-
ergy penetration may reduce the running
hours of these gas generators. The natural
gas supply system also has a large amount
of inherent energy storage in underground
storage caverns and the pipeline system.
This flexibility could be better exploited
through the use of excess wind or solar
energy that would otherwise be curtailed
to run an electrolyzer and produce hydro-
gen that could be put directly into the gas
system or converted to synthetic methane.
Other opportunities include multifuel
plants such as hybrid heaters that have
the ability to switch from gas to electric-
ity for generating heat at times of excess
renewable electricity on the power grid,
and, vice versa, at times of peak electric-
ity demand, they have the ability to switch
from electricity to gas.
The second article, Harnessing
Flexibility from Hot and Cold by Kivi-
luoma et al., covers a variety of ways that
thermal heating and cooling can provide
added flexibility to the electrical grid.
Moving thermal loads from electricity
january/february 2017
 PSCAD POWER SYSTEM SIMULATION

The Industry Standard for

Electromagnetic Transient Simulations
PSCAD / EMTDC has earned the trust of power system leaders around
the world for over 40 years. Our software provides unparalleled reliability,
speed, and accuracy for a range of applications, including:

4 Insulation Coordination: Lightning and Switching

4 Harmonics, Ferroresonance and Power Quality
4 HVDC and FACTS
4 Wind, Solar and Distributed Generation
4 Protection and Relays
4 Equipment Failure Analysis

Start your simulations today, our technical support

team is ready to help.

E: info@pscad.com
W: www.pscad.com

january/february 2017 IEEE power & energy magazine 11

 net-load peaks to net valleys can smooth
variations in load profiles. When the
heating devices have suitable controlla-
bility, they can also be valuable sources
of reserves that mitigate forecast errors
and faults. The article examines the pros
and cons of thermal-electricity integra-
tion and lays out generic principles and
characteristics related to thermal sector
flexibility as well as demonstrates its pos-
sibilities with specific examples.
A major potential for flexibility in the
heat sector is created by the low cost of
storing heat, which provides an opportu-
nity to shift electricity demand. Another
option is to utilize hybrid systems where
either electricity or fuel can be used to
produce heat depending on the price
variations of either option. This category
includes many different options, starting
from dual heaters in buildings all the way
to large district heating systems with CHP
plants, fuel boilers, and electric heaters. In
addition, the article examines flexibility

^/
 W'K


 K^
,QQRYDWLRQSUHYDLOV
created by thermal loads in several coun-
tries around the world.
In The Triple Bottom Line for Effi-
ciency, authors Casey et al., examine the
energy-water nexus and address efficien-
cy financial, environmental, and social
impacts. The article looks at examples
that demonstrate the emerging integration
of the water network and energy grids,
particularly where water is used in the
power production, wastewater treatment,
and the end-use part of the electric grid.
Water presents a variety of challenges for
the electric sector including the reduction
of water use in power plants, and using
efficient electric technologies for water
treatment, transport, desalination, indus-
trial processes, and end uses can conserve
electricity, hence reducing water use in
power plants. Water and wastewater treat-
ment facilities are good candidates for
demand response because they are energy
intensive and have significant flexibility
through variable speed pumps and inher-
K

2QOLQH

2QOLQH9ROWDJH
ent storage in water tanks. This opera-
tional flexibility can make these facilities
ideal partners for electric utilities seeking
to manage electric load through demand-
response programs.
Unlocking Flexibility by DallAnese
et al., examines integrated operational so-
lutions that strike a balance among cost,
reliability, and environment while ac-
counting for the affordability of energy,
sustainability, and social acceptability.
This article gives an overview of possible
joint control and optimization approaches
for multienergy systems. It also dis-
cusses the core challenges related to the
development of distributed control and
optimization algorithms that allow dif-
ferent parties to retain controllability of
their own energy assets while maintain-
ing their individual performance and reli-
ability objectives. The physical couplings
among systems is represented in mathemati-
cal terms through the so-called energy hub
approach that offers the flexibility to
t>



2QOLQH5HDFWLYH5H
^/ 6\VWHP2SHUDWLQJ 7UDQVLHQW6WDELOLW\ VHUYH9$50DUJLQ
 /LPLW62/(QJLQH 3-07(3&2 0RQLWRULQJ6\VWHP
&$,62  79$ 

  6XSHUDQG*OREDO
/RRN$KHDGDQG 1H[W*HQHUDWLRQ 2SWLPDO3RZHU)ORZ
2QOLQH9ROW9$5 6WDWH XSWRFRQ
&RQWURO (VWLPDWLRQIRU WUROYDULDEOHV
8.31 7 '


%LJZRRG6\VWHPV2QOLQH9ROWDJH6WD
ELOLW\SURGXFWLVVLPSO\WKHEHVWLQWKH ([WHQVLYHVRIWZDUHSRUWIROLRIRUERWKWUDQVPLVVLRQDQGGLVWULEXWLRQV\VWHPV3OHDVH
ILHOG YLVLWRXUZHEVLWHWROHDUQPRUHDERXWVXFFHVVIXOVWRULHVSURGXFWVDQGXWLOLW\FRQVXOWLQJ
&$,623URMHFW/HDGHU ZZZELJZRRGV\VWHPVFRP

%6,2QOLQH7UDQVLHQW6WDELOLW\$VVHVV ^/
PHQWLVDVXSHUESURGXFW
7(3&20DQDJHU /^KW:D/>
 dWdWKds
%6,9$50DQDJHPHQW6\VWHPVLVD h<WEEz/^K
EHVWSUDFWLFHWRROIRUDOOXWLOLWLHV
6(5&$XGLW6WDII ^/
 

12 IEEE power & energy magazine january/february 2017

capture energy-conversion factors at vari-
ous spatial scales. The resultant energy
hub model involves an inputoutput ma-
trix formulates the solution to multisys-
tem level optimization problems. From an
operational perspective, the coordinated
control of various energy infrastructures
represents a significant change that fa-
vors a local view and makes city quarters,
residential neighborhoods, and industrial
areas the fundamental building blocks of
the integrated energy system.
The fifth article, The Consumers Role
in Flexible Energy Systems by Schuitema
et al., examines the important role that con-
sumers play on different levels in achiev-
ing flexibility in energy systems. They can
help make energy systems more flexible
by changing their energy-consumption
patterns, resulting in load shifting and re-
duction in energy demand. To do this,
customers must accept smart technolo-
gies and energy-efficiency measures. This
is something that electrical power system
operators dont necessarily control or un-
derstand. The area of consumer acceptance
and energy use is a multifactor challenge
that typically needs the disciplines of so-
cial sciences, economics, and humanities
to understand. There are also factors such
as policy interventions and market design.
The article notes that a more integrated ap-
proach is needed within social sciences and
humanities as well as with energy research-
ers in other disciplines, such as engineering,
environmental science, and computer sci-
ence, to design the best energy system of
the future.
The final article, Flexibility Challeng-
es for Energy Markets by Dhaeseleer et
al., discusses the influence of policy and
regulation on the efficient behavior of
energy markets and warns of inadequate
performance of the overall energy sys-
tem if the wrong policy choices and their
practical implementation are made. The
article gives several examples of how the
speed of investments in certain areas and
ignoring system interactions can lead to
targets that may even be opposed to each
other, thereby compromising the strategic
objectives. The authors give some pos-
sible alternative routes for increasing sys-
tem flexibility and promotes better energy
systems integration.
The issue concludes with the In My
View column, in which Mattias An-
derson makes the case for the need of an
interdisciplinary approach to achieve ap-
propriate levels of flexibility in our energy
system. He points out the interdisciplinary
opportunities within the developing area
of ESI. Andersons background in intel-
lectual history and political science and
his many years of working with research
engineers gives him some interesting per-
spectives, in particular, on the transition
toward a renewable-based energy system
with its large share of intermittent energy
sources such as wind and solar.
We hope you enjoy this selection of ar-
ticles that examine the concept of ESI and
how it can be used to unlock flexibility in
the electric power system.
p&e
 T
THE ENERGY SYSTEM IS A COMPLEX NETWORK
of physical infrastructure and markets interacting closely
with one another. Within this network, the gas and electric-
ity systems have become the backbone of modern energy
production. Both systems are closely interconnected due
to the vast deployment of efficient combined-cycle gas tur-
bines (CCGTs) over the first decade of the 2000s, mainly in
Organization for Economic Cooperation and Development
countries. This increased interdependence and rapid pen-
etration of variable renewable energy sources (varRE) make
the gaselectricity nexus a primary concern and opportunity
for energy system flexibility.
The significant discrepancies in gas prices across the
world (Figure 1) bear witness to the fact that gas markets
remain largely regional. The role of gas in electricity sys-
tems and the interaction between gas and electricity differ,
however, across the globe.
In the United States, natural gas has become a major
energy source (primarily at the expense of coal invest-
ment), due to the shale gas revolution and to natural gass
potential in implementing the CO2 reduction policies
introduced by U.S. Environmental Protection Agency.
Consequently, the gas network has been stressed during
times when demand from both the electricity system and
direct gas consumers is high. To ensure reliable opera-
tion, several coordinating initiatives between the gas and
electricity sectors have emerged across the country
(for example, in California, Texas, New England, and
the Midwest).
ISTOCKPHOTO.COM/DIMA_ORIS

Unleashing the
Flexibility of Gas
By Steve Heinen, Christian Hewicker, Nick Jenkins,
James McCalley, Mark OMalley, Sauro Pasini,
and Simone Simoncini
Digital Object Identifier 10.1109/MPE.2016.2621838
Date of publication: 2 February 2017

16 IEEE power & energy magazine 1540-7977/172017IEEE january/february 2017

 Latin America is one of the worlds most hydro-energy dependent regions. However,
recent prolonged droughts have sparked numerous operational and system planning
issues, thus renewing interest in conventional thermal investments for dispatchable
generation, with demand for both fossil fuel and electric power increasing rapidly
across the region. The gas market continues to grow, owing in part to ample new in-
vestment in liquefied natural gas (LNG) infrastructure planned by Chile, Colombia,
and Uruguay.
In Eastern Asia, natural gas tends to be a scarce resource and today is mainly imported through
LNG terminals. Gas-fuelled electricity generation is used, at best, for mid-load generation
except in Japan, where, after the Fukushima nuclear plant disaster, CCGT plants have operated
as base-load plants. Although coal remains king, the role of gas could expand, given local con-
cerns about air pollution and the increasing availability of pipeline gas from Russia and central
Asia. Technically, China has the worlds largest recoverable shale gas resources; however, this
potential is constrained by geological complexity, shortages of water, and land access diffi-
culties, as well as by limited industrial experienceall of which led the country to lower its
production targets. Chinas gas consumption has been increasing faster than its production over
the past five years. This trend is likely to continue because the 13th Five-Year Plan (20162020)
stipulates that coal in nonpower sectors be replaced with either natural gas or electricity.
In Europe, gas power is considered an important technical resource for renewables inte-
gration, but it is currently struggling to be economically competitive: several gas power
stations have been mothballed, and utilities are calling for payment mechanisms to keep
plants online. The situation has been aggravated by flat or, in some countries, declining
electricity demand, low coal prices, and weak carbon markets. In parallel, efforts to decar-
bonize the gas network and reduce import dependence are increasing: biogas production
is growing, although from a small base, and several power-to-gas demonstration projects
have been commissioned.
In all these regions, the inevitable penetra-
tion of variable generation and electrification of
heat and transport will lead to increasingly vari-
Innovating Gas Systems to able operation of thermal dispatchable genera-
tors (this is already being observed in Europe).
Meet the Electricity Systems The growing net-load variability affects not only
power stations but also networks and gas supply
Flexibility Requirements systems (e.g., gas storage and LNG tanks).
In this article, we discuss the gas systems
ability to meet the electricity systems flexibility
requirements and also explore some of the tech-
nical, economic, and policy measures required
if gas is to become a flexibility resource. We analyze flexibility in three parts of the system:
from gas power generationtechnology and electricity market design
in gas supplygas storage and gas/electricity market coordination
through multiple-input, multiple-output plants and appliances.

Flexibility from Gas Power Generation:

Technology and Electricity Markets

Impacts of Flexible Operation and Technology Development

From a technical perspective, gas turbine-based plants are typically more flexible than
many other forms of generation, able to start quickly and having significant ramping
capability (Figure 2). In many cases they are an ideal complement to variable renew-
able energy. For example, Ireland has, simultaneously, a very large penetration of wind
and gas-fired electricity generation. Modern gas turbine plants excel, with startup times
of lower than 1 h and ramp rates above 50 MW/min. Older coal plants, heavy oil, and
nuclear plants often require four to eight hours for start up and have lower ramp rates
(few MW/min).

january/february 2017 IEEE power & energy magazine 17

 Gas turbine research and development (R&D) focuses on
18 improving the technical flexibility as well as economic prof-
16
14 itability in electricity markets by minimizing startup times,
USD/MBtu

12 enhancing ramping capabilities, and reducing gas power sta-

10
8 tions minimum stable output. The R&D priorities for gas
6 power plants are as follows.
4
2 Using advanced materials to minimize cycling im-
0 pact and cost. For CCGTs in particular, this involves
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
improving the balance of plant (or BOP) and the
heat-recovery steam generator (HRSG), which often
Japan LNG Europe (NBP)
limits startup times because the material and equip-
United States (Henry Hub)
ment cannot sustain higher temperature gradients.
Improved maintenance procedures in combination
figure 1. Gas pricing at different global trading hubs, with new control instruments (potentially in real time)
20032016. The United Kingdom is representative of can also optimize the startup procedure and reduce
European gas prices. NBP: national balancing point. startup times.
(Source: International Energy Agency, Tracking Clean
Increasing the use of monitoring and automation for
Energy Progress 2016, Paris: OECD/IEA, 2016.)
reliable startup sequencing. For CCGTs, additional
monitoring systems help identify stress and residual
Open-cycle gas turbines (OCGTs) and aeroderivative gas life on steam turbines and HRSGs.
turbines (ADGTs) were always designed to provide flexibil- Maximizing load gradients during load changes. This
ity, but CCGT power plants were initially designed to operate can be accomplished by using advanced materials and
mid-load to base load. Over the last decade, European CCGTs real-time monitoring systems to minimize wear and
have already evolved to a point where it is common that they tear of the material.
have a more flexible operating schedule. Since the mid-2000s, Improving combustion stability in gas turbines during
CCGT dispatch in some European regions has moved from load change.
base load to mid-load to several startup/down cycles per day Reducing turn-down ratio and maximizing part-load
(Figure 3). Consequently, operating times decreased to as low efficiency, especially of the gas turbine. This can be
as 1,300 h per year, while startup rates have increased from achieved by improving the combustion process and
25 starts per terawatt-hour (TWh) to more than 80 starts per burner materials.
TWh produced (Figure 4). The increased cyclic operation Turbine-based plants today completely dominate the gas
exposes gas plants to more wear and tear and, consequently, power sector, but in some cases they could potentially face
increases cycling costs. competition from reciprocating engines. In the past, recipro-
cating engines have mainly been
used only for small, decentralized
applications because turbine ef-
Ramping Minimum Turndown Hot Startup ficiencies are considerably lower
OCGT for these applications (<10 MW)
CCGT and reciprocating engines can burn
ICE a broader range of fuel composi-
ICE CC tions (pipeline quality gas and,
Hard Coal e.g., synthetic natural gas, landfill
Lignite gas, and biogas). However, recip-
0 10 20 30 0 20 40 60 0 2 4 6 rocating engines are now avail-
FL/min (%) FL (%) Hours able in sizes of up to 20 MW and
can be organized as banks of en-
figure 2. The flexibility characteristics of thermal electricity generation plants. gines to form a large power plant
Typical plant size (MW) is as follows: OCGT: 50200; CCGT: 300500; internal (>200 MW). Today, in fact, many
combustion engine (ICE)/reciprocating engine: 20200; ICE combined cycle (CC):
gas turbine manufacturers also
250450; black (hard) coal: 5001,000; lignite: 5001,000. Note that nuclear plants
own reciprocating engine compa-
are excluded because they perform worse. FL: full load. (Sources: International
Energy Agency, Energy Technology Perspectives 2014, Paris: OECD/IEA, 2014; nies (Table 1). These plants pro-
German Institute for Economic Research, Current and Prospective Costs of Electricity vide cost-effective N-1 reliability
Generation, Berlin, 2013; Verband der Elektrotechnik Elektronik Informationstechnik, for islanded power systems due
Erneuerbare Energie braucht flexible KraftwerkeSzenarien 2020, Frankfurt am to the scalability of the cascad-
Main, 2012.) ing plants, which require only one

18 IEEE power & energy magazine january/february 2017

additional reciprocating engine to meet the reliability stan-
dard. With regard to flexibility characteristics, reciprocating 400
engines could provide 300
higher efficiency than OCGT and ADGT (up to 48%)

higher part-load plant efficiency and very low mini-
mum output, given that the plant (20200 MW) is
based on small units (<20 MW)
very quick startup time (a few minutes) and good ramping.
Electricity Market Design to Reward Flexibility
From an economic perspective, increasing levels of vari-
able renewable energy penetration may reduce the run-
ning hours for these gas generators. Modeling based on the
United Kingdoms system indicates that the capacity factor of
CCGTs could drop from around 45% today to as low as 10%
in 2050. Additionally, the more frequent startups and higher
ramps result in higher cycling costs, which potentially raises
economic concerns for such gas plants as the result will be a
reduction in revenue and an increase in cycling costs. This
challenge has contributed to an extensive global debate about
designing electricity markets that reward flexibility and main-
tain adequacy.
For example, in Europe many state-of-the-art gas plants
have been mothballed as they are no longer in the merit order
(a problem caused not only by increased variable renewable
generation but also by the relative market prices for coal and
gas), and many others are struggling to continue to operate.
This has led to the development of capacity mechanisms
and other market measures for some of these gas plants. In
Ireland, new ancillary service products are being defined to
reward flexibility, and gas plants are potentially providers of
(MW)
200
100
0
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
2005 2012
figure 3. A comparison of the operating schedule of the
same Italian CCGT owned by Enel in 2005 and 2012.
(Source: Enel.)
these services. New market products are also being devel-
oped in some parts of the United States, notably in Califor-
nia and by the MISO system.
Determining the best type of gas plant to have in scenar-
ios with a high penetration of varRE depends largely on local
circumstances. OCGTs are more flexible than CCGTs but
less efficient; therefore, there is a market tradeoff between
energy and flexibility. Other designs, such as ADGTs and
reciprocating engines, combine flexibility and efficiency,
but they do so at the expense of additional capital expenditure.
Economic profitability, based mainly on market revenues and
technology cost, will also define how gas power plants per-
form against other flexibility sources in the electricity system
(e.g., interconnection, demand-side control, consumer interac-
tion) and in the wider energy system (heat, water). The most
economic form of flexibility will be system specific. However,

6,000 90

80 78
5,100
5,000 71
Startup Rates (Starts per TWh)

70
4,000
Full Load Hours (h)

4,000 60 55

50
2,800
4,000
37
40
2,200
2,000 28
2,000 30

20
1,000
10

0 0
2007 2011 2012 2013 2014 2007 2011 2012 2013 2014
(a) (b)

figure 4. (a) The equivalent operating hours and (b) startup rates (starts per TWh) for an Italian CCGT plant owned by
Enel. (Source: Enel.)

january/february 2017 IEEE power & energy magazine 19


table 1. Gas turbine manufacturers and
their related reciprocating engine companies.
Gas Turbine ICEs
GE turbines Jenbacher, Waukesha, Dresser
Rolls Royce Bergen
Solar turbines CAT Power
MAN Turbo MAN
Mitsubishi Heavy Industries Mitsubishi Heavy Industries
when combined with capacity requirements, gas plants would
be particularly suitable. Market design and rewards for flex-
ibility will be essential to encouraging investment in flexibil-
ity and ensure reliability.
Flexibility in Gas Supply: Gas Storage
and Gas/Electricity Market Coordination
Gas Storage
In the past, gas network flexibility appeared abundant com-
pared to electricity systems and was largely ignored in elec-
tricity reliability assessments. The variability of varRE in the
electricity system will lead to the more flexible operation of
gas power stations, ultimately translating to diurnal variability
in gas supply, and may require that gas be stored in prepara-
tion for a ramping event in the electricity system. Compared
to electricity systems, gas systems typically offer significant
flexibility due to different storage options: line pack, under-
ground storage, and LNG tanks.
Line pack is the volume of gas stored in pipelines and can
be used to meet abrupt diurnal changes in gas demand. It is
proportional to average system pressure. During a period of
low renewable energy output (for example, wind), gas gen-
erators may be called upon, which would lead to a large and
rapid decrease of the gas line pack. If this happens when
peak gas and electricity demand coincide, the resulting pres-
sure drop in the gas network could limit its ability to meet
rapid changes in gas demand (including gas for power gen-
eration) and cause interruption of gas supplies to CCGTs.
Other flexibility sources (e.g., electricity imports through
interconnectors or demand-side response) would be required
to ensure reliable system operation. To mitigate such line-
pack shortages, coordination between gas network and elec-
tricity system operators will be increasingly important.
Modeling results for the United Kingdoms 2020 system
suggest the increasing need for coordination between gas
and electricity system operation. In a low wind scenario,
the line pack decreases strongly at times when high demand
coincides and limits the gas supply to CCGTs. As a result,
the power output from CCGTs during peak hours may drop
by almost 3 GW (Figure 5). This reduction in CCGT power
output was compensated for by the import of more expensive
electricity through the U.K./France interconnector.
Underground gas storage facilities include depleted
gas/oil fields, aquifers, and salt caverns. Depleted fields
and aquifers are typically used as seasonal storage facili-
ties. Natural gas is injected into storage during the sum-
mer (low-demand season) and withdrawn during the winter
(high-demand season). The withdrawal rate and capacity
are often very large, but the cycling capability is limited.
Salt caverns are commonly used as fast-cycle storage due
to their ability to support several cycles of gas injection into
and gas withdrawal from storage within a year. This type
of storage is better suited to providing gas supply flexibil-
ity to electricity systems with high penetrations of varRE.
Despite the receding demand for gas in Europe, the number
of European storage facilities is increasing due to growing
flexibility requirements as well as concerns over the secu-
rity of supply. The completion of currently planned projects,
mostly salt cavern facilities, will increase storage capacity
by 20% in 2020 compared to current levels.
LNG storage tanks and gasification stations are used as
peak shaving facilities that can respond rapidly to sudden

30
340 Base Case: 2009 6h Low Wind Case: 2020
25 3.1 GW
10 mcm

338
Line Pack (mcm)

Power (GW)

20
336
15
334
332 10
High Wind: 2020
330 5
Low Wind: 2020
Base Case: 2009 High Wind Case: 2020
328 0
0 4 8 12 16 20 24 28 32 36 40 44 48 0 4 8 12 16 20 24 28 32 36 40 44 48
Time (h) Time (h)
(a) (b)

figure 5. (a) The aggregate gas network line pack and (b) power generation by CCGT. (Source Qadrdan et al., 2010.)

20 IEEE power & energy magazine january/february 2017

 The inevitable penetration of variable generation
and electrification of heat and transport will lead to increasingly
variable operation of thermal dispatchable generators.
gas demand changes. Therefore, they not only contribute
to energy security by diversifying supply but also provide
operational flexibility.
Gas/Electricity Market Coordination
Gas and electricity markets interact via gas power plant
operators buying fuel on gas markets to generate power,
which is sold to the electricity market. Plant operators may
do this by trading in a variety of markets, i.e., from long-
term contracts and forward markets until shortly before real
time. While longer-term transactions are mainly important
with a view to the need for sufficient gas network capacity,
flexibility needs are primarily driven by trading in the day-
ahead and intraday/within-day markets as well as the need
to provide ancillary services and balance energy to power
system operators in real time.
Electricity and gas markets are often operated in isola-
tion on different time frames throughout the day and have
often failed to create a homogenous structure. Among oth-
ers, some of the key challenges include the following:
different time scales, such as the difference between
the gas market day (6 a.m.6 a.m.) and the calen-
dar day or the use of subhourly settlement intervals in
electricity systems
a system of fixed gates (day-ahead and/or during the
day) at which electric power and/or network capacity
is traded in the electricity markets, as opposed to con-
tinuous trading in the gas market
different product definitions and mechanisms for al-
location of network capacities
widespread use of interruptible network capacities in
the gas market.
As a result, gas plant operators may be required to commit to a
certain gas volume before knowing if their electricity market
bids have been accepted, or vice versa. As gas plant opera-
tors need to account for such risks in their bidding behavior,
this may result in a suboptimal market outcome and increased
costs to consumers. Similarly, gas network operators are often
unable to predict the variability in gas off-take induced by the
electricity market, making it difficult to manage diurnal flex-
ibility (such as line pack) in an optimal way.
As the deployment of varRE progresses, limited mar-
ket coordination may lead to serious risks for flexibility,
such as the need for quickly ramping up generation by gas-
fired power plants. In recent years, the lack of coordination
between gas and electricity has already threatened reliabil-
ity. For instance, insufficient stocks of natural gas in local
january/february 2017
storage contributed to the need for rolling blackouts in Texas
in February 2011. Similarly, in February 2012, parts of the
power system in southern Germany were close to break-
down because the interdependence of the gas/electricity sys-
tem had not been considered. A cold spell drove electricity
demand to record highs, while direct gas demand for heating
was also high. As only interruptible gas pipeline capacity
had been contracted for, some gas power plants could not be
dispatched as required, and a rolling blackout could only be
avoided by actively reducing demand.
In response to these concerns, regulators and governments
are increasingly encouraging coordination between both
markets. In the United States, gas and electricity coordina-
tion activities and interdependence assessment are ongoing in
various regions, including California, Texas, New England,
and the Midwest. Likewise, this topic has been picked up
both at national and continent-wide levels in Europe. Besides
improved coordination and information exchange between
system and network operators for gas and electricity, some
of these initiatives have also suggested changes of market
design and network access arrangements. For instance, in
France, the gas transmission system operator introduced a
set of specific operating restrictions for a growing fleet of
gas-fired plants but in combination with a new commercial
product that allows such plants to purchase additional diurnal
flexibility on a daily basis.
To a certain extent, the coordination challenges are
linked to different time constants in electricity and gas bal-
ancing (see Table 2). While an efficient integration of varRE
requires shorter gate-closure times and settlement intervals,
physical gas pipeline flows can be changed only with a sig-
nificant delay. This creates a dilemma for gas/electricity
market coordination as well as natural barriers for aligning
market opening/closing times. Regulators and system opera-
tors thus have to make a choice between either 1) exposing
gas power plant operators to the risk of diurnal restrictions
and different time scales for the gas and electricity market
or 2) allocating the risk of variations in the final two or three
hours to gas network operators.
Flexibility Through Hybrid Energy
Conservation Systems
Integrated energy conversion systems can provide high lev-
els of flexibility when they are able to switch between input
(energy resources) and output (production service) as well as
to store the input resource and/or some intermediate or final
form of the converted resource. Such a system is commonly
IEEE power & energy magazine 21
 table 2. Different time constants in electricity and gas balancing.

Issue Electricity System Gas System

Balancing requirement
Balancing process
Balancing time frame
Adapted from DNV GL.
Need to maintain system frequency within
strict limits in real time
Close to real time (<1 s) power balance
Focus on immediate action in the last
minutes/hour before real time
Maintenance of operating pressures within a certain
range due to line-pack capability
Cumulative energy deviation over balancing time
frame or day
Focus on delayed actions (ex-post), typically
2 h

referred to as a multiple-input poly-generation conversion HES. However, a large variety of alternative HES designs are
system; a multiple-input,multiple-output conversion system; conceivable through the combining of different inputs (elec-
energy hubs; or a hybrid energy-conversion system (HES). tricity, heat, fuels, and/or biomass) and outputs (electricity,
The flexibility can be utilized for the electricity system, the heating and cooling services, water, hydrogen, transportation
natural gas system, or both. Two kinds of flexibility can be fuels, and/or commodity chemicals). The flexibility benefits
distinguished for HESs. of HES deployment are exemplified here based on three dif-
Operational flexibility enables meeting highly vari- ferent HES designs: 1) an advanced HES based on anaerobic
able net loads or maximizing operation at steady digestion (AD), 2) hybrid residential heaters, and 3) wind-
state of certain HES appliances to minimize wear electrolyzer systems.
and tear. An advanced HES can be conceived around AD. As
Economic flexibility enables arbitrage between input illustrated in Figure 6, such an HES has three energy in-
resources and output services, i.e., utilizing least-price puts (natural gas, biomass, and electricity), three energy
input resources while providing highest-price out- output services (biomethane, electricity, and cooling/heat-
put services, subject to contractual and physical con- ing), and three storage devices (biomass, heat, and biogas).
straints. A traditional single-input, single-output power Additionally, it contains two heat sources: the combined
plant may provide significant operational flexibility, heat and power (CHP) unit and the low-temperature geo-
but it would not have economic flexibility. thermal system with heat pump. The CHP and heat pump
serve both the anaerobic digester and the district energy
Examples of Advanced system. AD utilizes low-grade heat to support the diges-
HES Designs tion of organic materials (e.g., wasted food, plant clippings,
A combined heat-and-power, or cogeneration, plant fueled by animal manure, sewage) to produce biogas. The biogas can be
natural gas and biogas is a familiar design that exemplifies an used directly to fuel the CHP, it can be stored, or it can be
cleaned and upgraded before its
injection into a natural gas pipe-
line system. The district energy
Biomethane system distributes heat obtained
Biogas Cleaning/ from the CHP and heat pump
Injected
Storage Upgrading
into Pipeline systems to the demand; it could
Natural Gas also provide cooling, if an ab-
Combined sorption chiller is included.
Biogas
Biomass Heat and Electricity The heat storage (or accumu-
Anaerobic
Mixing Power
Digester lator) facility and the heat pump
Heat provide that the HES meets heat-
Heat
ing and cooling demand while the
Heat Heat CHP meets the flexibility require-
Biomass Storage Heat
Storage ments from the electricity system;
Low-Temp Cooling alternatively (or additionally), the
District Energy Services
Geothermal w/ heating demand may be controlled,
Electricity System Heating
Heat Pump
Services reducing the need for the heat pump
or the heat storage.
figure 6. An advanced HES design based on an anaerobic digester (three inputs and The integration of AD into this
three outputs). HES is motivated in four ways.

22 IEEE power & energy magazine january/february 2017

 First, although AD requires heat, the necessary heat is
low grade, ranging from 30 to 38 C, and can be ex-
tracted from a wide range of processes including heat
pumps and CHPs. This sets AD apart compared to pro-
cessing biomass via gasification or pyrolysis because
these latter two methods require high-quality heat.
Second, many attractive regions for wind-energy
development (for example, in the United States, the
Midwestern northsouth belt from about Wyoming
on the western side to Illinois on the eastern side) are
also highly agricultural with a diversity of biomass
feedstock including animal waste, grass and maize
silage, and grains (e.g., wheat and triticale). Thus, as
wind and solar penetrations grow, so will these re-
gions need for flexibility, a need that could be met
by HESs through a biomass resource indigenous to
the region. Although there are over 11,000 AD fa-
cilities in Europe and 2,100 in the United States, the
potential for agricultural biomass digestion remains
underutilized. Most U.S. facilities (about 1,880) are
associated with wastewater treatment plants or land-
fill gas projects; only 247 are on farms and thus make
use of agricultural biomass.
Third, if the input feedstock decomposed naturally, un-
dergoing the same biological process as in AD, then it
would emit methane directly into the atmosphere. Con-
sidering that the global warming potential of methane
is at least 21 times higher than the CO2 released if AD
is used, then AD operation can represent a significant
reduction of greenhouse gas emissions.
Finally, investment in AD provides an effective hedge
against long-term natural gas price volatility, a mani-
festation of the economic flexibility of the HES.
Another example of an HES is a hybrid of heat pumps/gas
boilers that uses both gas and electricity to supply residen-
tial heat. The smart integration of such heaters enables shifts
in real time between the different fuels to respond to system
conditions. For example, the hybrid heaters can switch from
gas to electricity for generating heat during times of excess
renewable electricity on the power grid; conversely, at times
of peak electricity demand, they can switch from electricity
to gas. Wide-scale deployment minimizes electricity capacity
expansion compared to single-fuel heat pump deployment and
reduces up-front costs for consumers because the expensive
heat pump can be downscaled. Such deployment also supports
decarbonization and reduces natural gas dependency com-
pared to single-fuel gas boilers. Technical flexibility is limited
by consumer comfort, which depends on personal preferences
as well as on building properties. An investment study of this
technology for Ireland indicates that its deployment is cost-
effective and enables system-wide cost reductions compared
to boiler- or heat pump-only deployment.
A third HES example is based on electrolyzers, which
are fueled by electricity and produce synthetic natural gas
and hydrogen. These outputs can be stored locally, fed into
january/february 2017
the gas grid, or used as fuel or feedstock. Excess wind
energy that would otherwise be curtailed can be used to
run an electrolyzer and produce hydrogen. The resulting
hydrogen can be used as transportation fuel or industry
feedstock. Alternatively, the hydrogen can be fed directly
into the gas system or processed to synthetic natural gas.
The admissible hydrogen concentration for direct injec-
tion into the gas grid is limited mainly by gas combustion
equipment because the different combustion properties of
hydrogen lead to flame speeds and reactivity, while hydro-
gen embrittles the pipeline. This HES enables storage of
excess renewable electricity in a gaseous fuel, thus provid-
ing access to the vast storage capabilities of the gas infra-
structure. The gas network offers storage capabilities over
all time frames, from daily cycling as line pack to intersea-
sonal storage underground, and is thus much more flexible
than other storage technologies. Therefore, this HES pro-
vides considerable operational flexibility, but its economic
potential depends largely on the price spread between wind
energy and hydrogen, as well as between hydrogen and nat-
ural gas. Also, the electrolyzers cost itself makes deploy-
ment today prohibitive. A few pilot plants have been built
in Germany since 2013 with support from both industry
and government. The Danish system operator expects to
rely on electrolyzer systems by 2030 to provide flexibility.
Benefits of Wide-Scale HES Deployment
The economic flexibility resulting from wide-scale HES
deployment is manifest by an increase in resilience. An
indication of resilience of a certain system is typically the
total price increase of electric and natural gas services
nation- or continent-wide after a disturbance or disruption.
A lower price increase indicates a higher level of resilience.
As shown in Figure 7, HES deployment increases the link
density of the overall networked system, better delivering
Output Side
Output Side
Input Side
Input Side
Conventional: HES:
Single-Input, Multiple-Input,
Single-Output Multiple-Output
Energy Converters Energy Converters
figure 7. The impact of HESs on network link density.
IEEE power & energy magazine 23
 energy from raw resources to service demands relative to
single-input, single-output power plant designs.
HES provides increased path redundancy between energy
resources and energy services, offering alternative means to
satisfy energy service demands. After the 2005 hurricanes
Katrina and Rita in the Gulf of Mexico, a large percentage
of U.S. natural gas supply was shut down for many weeks.
Although electric and natural gas demand was interrupted
for only a short time in a localized region, electric and natu-
ral gas prices rose steeply throughout the nation, and they
did not return to their prehurricane levels for months. An
increased link density due to HES deployment would have
enabled shifting between supplies and products after the
disaster occurred, thus minimizing the price spike.
HES plants are scalable from conventional utility-sized gen-
eration to distributed resources (DRs) to meet flexibility needs
at different scales. DRs are connected at the electric distribution
level (consistent with IEEE standard 1547), which means that
for some small-scale applications, the capacity is constrained
by distribution circuits. Distributed HES plants with capaci-
ties of 10100 MW, however, could be connected directly to
the distribution substation rather than the distribution circuit,
which means that their DR potential is less limited. Distribu-
tion-substation connections of resources at this capacity main-
tain the partial benefit of proximity to loads while still retaining
the ability to utilize the transmission system without reversing
flows on distribution circuits. Therefore, wide deployment
throughout a region of many HESs of this scale would satisfy
the operational flexibility necessary for high wind and solar
penetration, enable economic flexibility, and balance the ben-
efits of load proximity with transmission accessibility.
HES deployment can potentially provide both opera-
tional and economic flexibility by combining different energy
resources (inputs) and energy services (outputs). However, the
industry is often disaggregated, and many plant users are only
active in one specific market. Therefore, plant owners and com-
panies may not see opportunity arising from a by-product, lack
the skills to expand to new markets, or shy away from the risk of
expanding into unknown markets. Collaboration and research
are essential to develop skills and confidence.
Conclusions
Today, gas turbines are the main flexibility source (next to
interconnectors) for balancing demand and variable supply and
achieving stable grid operation. Gas-fueled power plants typi-
cally start up quickly and provide excellent ramping capabili-
ties that cannot be understated. The profitability of gas power
plants is decreasing, however, because operational hours are
dropping and material wear and tear is increasing. Gas turbine
R&D efforts are focused on reducing cycling costs and maxi-
mizing flexibility capabilities, but market design that adequately
rewards flexibility is essential to ensure system reliability.
The cyclical operation of gas power plants increases
gas supply variability and requires the increased use of
short-term storage and intraday market trading. Increased
24 IEEE power & energy magazine
coordination between gas and electricity infrastructure is
critical due to the different time constants for real-time oper-
ation of gas and electricity networks.
Further integrating the energy resources and energy ser-
vices through HESs can increase both the operational and eco-
nomic flexibility of an energy system. Various HES designs
are possible that enable the use of existing infrastructures and
meeting local demands. Additionally, the deployment of HES
plants improves system reliability and resilience by increas-
ing link density and enabling the switch between different
supply sources and products. However, collaboration among
sectors and industries is essential to realize this potential.
The gas infrastructure is a major flexibility resource for
the electricity system. A holistic perspective including both
systems captures couplings and interactions, and, if those
are significant, then it reveals integration challenges and
opportunities to further increase the flexibility options
Acknowledgment
Steve Heinen is supported by the Fonds National de la Recher-
che, Luxembourg (project reference 6018454), and the CIT-
IES project, Denmark (project reference 1305-00027B/DSF).
For Further Reading
COWI and DNV KEMA, Study on synergies between
electricity and gas balancing markets (EGEBS), European
Commission for Energy, Brussels, Belgium, Rep. TRENTA-
LOT3-015, Oct. 2012.
Report on outages and curtailments during the South-
west cold weather event of February 15, 2011Causes and
recommendations, Federal Energy Regulatory Commission
and North American Electric Reliability Corp., Aug. 2011.
Accommodating an increased dependence on natural
gas for electric power, NERC, Atlanta, GA, 2013.
P. Ostergaard, Comparing electricity, heat and biogas
storages impacts on renewable energy integration, Energy,
vol. 37, pp. 255262, 2012.
M. Qadrdan, M. Chaudry, J. Wu, N. Jenkins, and J. B.
Ekanayake, Impact of a large penetration of wind genera-
tion on the GB gas network, Energy Policy, vol. 38, no. 1,
pp. 56845695, Jan. 2010.
N. Szarka, F. Scholwin, M. Trommler, H. Jacobi, M.
Eichhorn, A. Ortwein, and D. Thran, A novel role for bio-
energy: A flexible, demand-oriented power supply, Energy,
vol. 61, no. 10, pp. 1826, Oct. 2013.
Biographies
Steve Heinen is with University College Dublin, Ireland.
Christian Hewicker is with DNV Energy, Germany.
Nick Jenkins is with Cardiff University, United Kingdom.
James McCalley is with Iowa State University, United States.
Mark OMalley is with University College Dublin, Ireland.
Sauro Pasini is with Enel Thermal Generation, Italy.
Simone Simoncini is with Enel Thermal Generation, Italy.
p&e
january/february 2017
 Harnessing
Flexibility from
Hot and Cold

A
AS HAS BEEN OFTEN REPORTED, ELECTRICITY SYSTEMS
with high levels of variable wind and solar power generation would ben-
efit from demand flexibility. What is not as often mentioned is that elec-
trification of the transport and heat sectors could exacerbate the need
for flexibility, if they are implemented as inflexible loads. This demand
could also be made more flexible, but it comes with a cost. The main
issue is to identify the cases in which the benefits will outweigh those
costs, a matter that will naturally depend on the evolution of specific
energy systems. In this article, we
lay out some generic principles
and characteristics related to heat-
sector flexibility and demonstrate
Heat Storage and
its possibilities using specific
examples. While we generally
Hybrid Systems Can
use the word heat here, most of
the discussions also apply to cool,
Play a Major Role
which, after all, is just another
form of temperature difference.
A major potential for flexibil-
ISTOCKPHOTO.COM/MIMACZ, ROPEIMAGE LICENSED BY INGRAM PUBLISHING
ity in the heat sector results from the low cost of storing heat, which
allows opportunities to shift electricity demand. Another possibility is
to utilize hybrid systems in which either electricity or fuel can be used
to produce heat depending on price variations between the two options.

By Juha Kiviluoma, Steve Heinen,

Hassan Qazi, Henrik Madsen, Goran Strbac,
Chongqing Kang, Ning Zhang,
Dieter Patteeuw, and Tobias Naegler
Digital Object Identifier 10.1109/MPE.2016.2626618
Date of publication: 2 February 2017

january/february 2017 1540-7977/172017IEEE IEEE power & energy magazine 25

 Hybrid systems may take many different forms, from dual 30% of primary energy is used to produce heat, 30% is
heaters in buildings all the way to large district heating sys- used in the transport sector, and 40% is used for electricity,
tems with combined heat-and-power (CHP) plants, fuel boil- including electricity to heat (see Figure 1).
ers, and electric heaters. The share of electricity in primary energy consumption
is higher than in final energy consumption because a large
The Flexibility Potential proportion of energy is wasted in electricity generation pro-
of Heating and Cooling cesses using the Carnot cycle (Figure 1). The opposite is true
Together, heating and cooling represent a huge part of for heat, because nearly 100% of the energy in fuel is con-
energy consumption. However, the heat system is often not verted to useful heat. The large share of heat in the final
considered as a single system, but ratherdue to the historic energy use translates to a large potential for power system
emphasis on energy supplyas subsystems of different sup- flexibility. For example, the value of surplus wind or solar
ply sources (e.g., gas, coal, and electricity). Therefore, size power is zero, but if that electricity can be used to replace
and flexibility potential are often overlooked. According to heating fuels, the value rises to the price of the fuel. The
2014 Eurostat figures, in the European Union (EU) around value is affected by the conversion efficiencythe value
gets higher if heat pumps are used instead of less-efficient
direct-resistance heaters. However, there are investment
costs that need to be factored in as well.
Primary Energy Use Final Energy Use
Heat is consumed in most end-use sectors, except transport,
but its use is very diverse. In residential and commercial build-
ings, heat is used for space and water heating. In the indus-
22%
30% trial sector, heat additionally provides process heat. In terms
40%
45% of flexibility, some demands are more amenable to being con-
trolled than others. Figure 2 divides final energy use (heat, in
33% particular) among the EUs 28 member states (EU-28) into
29%
several categories. Most heat in the EU is generated from natu-
ral gas, while coal, oil and biomass make up much of the rest.
In Europe, cooling demand is considerably lower
Heat Transport Electricity than heating demand. However, it is growing quickly with
increasing space-cooling requirements and also due to the
figure 1. The primary and final energy use in the EU for heavy urban development in warmer climates and new uses
2014 (based on Eurostat figures) divided into three main needed in emerging services and industries, such as cooling
categories of energy end use. large data centers.
The heating/cooling usage types and demand profiles
will largely impact the flexibility potential, We use the fol-
lowing categories to describe the major parts of the heat/
400 cooling system:
Energy Use (Mtoe)

350
300
Residential local heating and cooling of buildings
250 district heating and cooling
200 heat for industry.
150
100 Local Heating and Cooling
50
0
Final energy use in residential buildings is dominated by
space and water heating demandroughly 80% in Europe
gh s
an y

Sp /For s
ry

g
er ting

g
t

e
ce
r

or

tin

in

in
st

st

nc

and 60% in the United States. Worldwide, most residential

ok

tn
sp

i
du

ea
W Hea
Ag erv

ia
tri Co
In

pl
S

Li

homes and commercial establishments produce heat locally

Tr

Ap
e
ro
ac

at

within the building using a variety of different heating/cool-

ec
El

ing technologies and are not connected to district heating

Wood Heat Gas Coal
figure 2. The final EU-28 energy use for 2014 (based on
Eurostat figures). The residential sector is split into compo-
nents using 2008 data (ODYSSEE). Losses in transformation
and transfer are not included. Heat as energy use refers
to the heat being a conversion by-product from, e.g., a
CHP plant or an industrial process.
Oil Electricity networks. Heating technologies can be mainly differentiated
by their fuel (wood, natural gas, biogas, and electricity) and
conversion process (combustion in a boiler or burner, liq-
uid evaporation for heat pumps, and Joules law in resistance
heaters). Cooling is generally based on electricity.
The heating/cooling appliance couples the building to
the energy supply system. Because solid and gaseous fuels
can be stored easily over weeks, if not seasons, these supply

26 IEEE power & energy magazine january/february 2017

 Electric storage heaters make use of the solid materials
around the resistance heater as a heat store and may utilize a fan
to release the heat in a more controlled manner.
systems inherently have considerable flexibility, and plan-
ning boils down to managing supply and storage infra-
structure, as with most commodities. On the other hand,
electricity-fueled heaters/coolers require the power system
to balance supply and demand instantaneously, changing the
dispatch in the short run and impacting the generation port-
folio in the long run.
Electric heating, if deployed in an uncoordinated man-
ner, requires additional power-system flexibility to meet
daily, seasonal, and annual variations. For example, in
France most residential heating is based on electric heating
and causes considerable temperature sensitivity in the power
demand (2,300 MW/C); this represents a major driver for
extreme peak loads and security of supply.
However, a controllable electric heating/cooling can
draw on the potential flexibility of heating (thermal iner-
tia) to facilitate renewables integration and manage peak
loads. For example, some demand-side management pro-
grams are being carried out in France as ad hoc measures
to improve flexibility. In another example from Germany,
it was realized that electric overnight storage heaters can
be a valuable source of flexibility, so an earlier decision to
remove them was reversed in 2013. The use of information
and communications technologies in electric heaters could,
therefore, provide the option to shift demand loads accord-
ing to power system conditions, while also meeting the
building occupants comfort requirements. These exam-
ples illustrate the strong interaction between residential
electric heating and the power sector, but they also raise the
question on how such integration of electric heating should
be managed to provide flexibility in the most cost-effective
and nonintrusive manner.
Electric heating systems are mostly based on resistance
heaters (including electric storage heaters) and on the more
efficient heat pumps. Heat pumps make use of the natural tem-
perature difference between the outdoors and indoors during a
condensation/evaporation cycle. The heat cycle only requires
electricity to run the compressor and other auxiliary equip-
ment, thereby producing two to four units of heat for each unit
of electricity consumed in air-source heat pumps (although
the gain tapers off in colder temperatures). The efficiency can
be even higher for ground-source heat pumps, reaching per-
formance coefficients of four to five. This minimizes genera-
tion requirements and peak load but will not allow as much
flexibility to utilize excess renewable electricity.
Local cooling is usually provided by heat pumps (either
with air conditioning systems that only cool or by reversible
january/february 2017
heat pumps that can also heat). When the heat pump is used
in cooling mode, it is called a chiller. The energy efficiency
ratio is typically lower, ranging 1.52.5 units of cooling for
one unit of electricity. In regions with high cooling loads,
the coordinated cooling of buildings results in the dominant
peak electricity demand.
In many cases where a heat pump is installed, it is
complemented by use of an electrical resistance heater or
a gas boiler. The combination with a gas boiler (the so-
called hybrid heat pump) shows vast flexibility poten-
tial. Its smart integration into the electric system could
enable the power system to access the flexibility of the gas
system by switching from the heat pump to the gas boiler
whenever the electricity system is under stress (this could
represent an extended period of several days). Hybrid fuel
boiler/resistance heater systems could provide the option
of using excess renewables by switching from fuels (often
gas) to electricity.
Thermal storage in buildings can enable the optimization
of electricity consumption and charging based on electricity
market prices while still providing thermal comfort to the
user. If the resistance heater is integrated with high-thermal-
capacity materials, then the heaters are referred to as stor-
age heaters. Electric storage heaters make use of the solid
materials around the resistance heater as a heat store and
may utilize a fan to release the heat in a more controlled
manner. Using resistance coils or hydronic systems in under-
floor heating enables some energy to be stored in the thermal
capacity of the floor as well. Other technologies for thermal
storage include water storage tanks and solid materials. In
particular, in hydronic systems, a water tank can be added
relatively easily, although there is a cost related to the space
use in addition to the cost of the storage device.
Energy can also be stored with a cold storage. Tempera-
ture differences are smaller than in heating though; conse-
quently, cold storages would need more volume for the same
energy content. However, it is possible to take advantage
of the latent heat of freezing/melting, which corresponds
to approximately 80 C of the temperature difference in
water. Available commercial chillers use ice storage; these
can achieve more operating hours, and, consequently, the
chiller can be downsized while also decreasing electricity
use during daily peaks compared to traditional air condi-
tioning. Conceivably, these chillers could also offer flex-
ibility for higher shares of wind and solar power, although
there would probably be a different optimum in the sizing
of the components.
IEEE power & energy magazine 27

Space Heating and
14%
Hot Water <100 C
34% 12% Process Heat <100 C
Process Heat 100500 C
21% Process Heat 5001,000 C
20% Process Heat >1,000 C
figure 3. The use of different grades of heat in EU-28
industries (based on Naegler et al., 2015).
District Heating and Cooling
District heating pipes carry hot water from centralized heat
plants to buildings with heat exchangers. After heat has
been transferred to the buildings heating system, cooled
water flows back to the plants through an adjacent pipe.
District heating is mainly used in more densely populated
areas in northern latitudes (although it is not widespread in
North America).
In addition to economizing with large fuel boilers, dis-
trict heating offers the possibility to use CHP plants. In
some countries (e.g., Germany and Denmark), even small
district heating systems often have CHP units, while in oth-
ers (e.g., Russia and Finland) CHP units are found mainly
in bigger systems that can accommodate larger, more
economic plants. When used alongside CHP plants, fuel
boilers cover heating peaks and back up the CHP units.
Combining CHP plants and fuel boilers enables sensitivity
to power prices. Some CHP units can also change the ratio
between heat and electricity production, which increases
their flexibility.
The flexibility of a district heating system can be fur-
ther increased with heat storages (accumulators) that offer
a very low-cost form of energy storage at district heating
scale (thousands of cubic meters in insulated steel tanks or
caverns). When power prices are sporadically very low [e.g.,
high levels of wind or solar photovoltaic (PV)] and there
are no regulatory hurdles, it can become feasible to install
heat pumps and electric resistance heaters in district heating
systems. Electric heaters offer a low-cost solution to utilize
cheap power, while heat pumps give considerably more heat
per unit of electricity for a higher investment cost.
It is costly and inconvenient to install district heating
pipelines into existing cities. However, new neighborhoods
are a potential target for small-scale networks. In compari-
son to building-level heating, they decrease the relative cost
of heat generation units with a limited investment in heat
pipelines. But, more importantly, from a flexibility view-
point, they offer considerable economies of scale for heat
storage, the specific cost of which decreases nearly logarith-
mically with increasing size.
District cooling is much less common than district heat-
ing. The challenge has been that economies of scale are more
28 IEEE power & energy magazine
difficult to achieve in cooling units than in heating units.
However, because most people live in climates where cool-
ing is an aspiration, district cooling may have a more impor-
tant role in the future. District cooling can provide better
access to more efficient ambient heat sinks (e.g., sea water)
than heat pumps located in buildings. This would also help
to keep the cities themselves cooler because waste heat is
transferred away from the city. Cooled fluid, typically water,
could also be stored in accumulators to gain more flexibility
toward the power system.
Heat Use in Industries
Figure 3 shows six grades of industrial heat use, dominated
by process heat, which makes up roughly 85% of the total
energy demand for industrial heat in Europe. The remain-
ing 15% is due to space heating. Heat pumps can serve low-
temperature loads, while CHP units can serve somewhat
higher-temperature levels and still be able to produce
electricity. A large fraction of industrial heat loads, cur-
rently dominated by natural gas burners, requires higher
temperatures. However, electric heating technologies such
as resistance heating, electric arc heating, induction heat-
ing, and dielectric (radio-frequency) heating can provide
temperature levels above 500 C and so can replace, e.g.,
natural gas burners. These alternatives can achieve a high
range of temperatures and offer accurate temperature con-
trol. They could provide system flexibility if combined
with a heat storage or used in a hybrid configuration with
fuel burners. The costs of energy, equipment, and grid con-
nection have so far limited the use of electric heating as
compared to combustion.
Also, the type of electric heating capable of replacing or
supplementing an industrial gas burner strongly depends on
the process. Quite a few industrial processes also use the
fuel as a raw material. For example, steel production in blast
furnaces requires coke not only as an energy carrier but also
as a reducing agent that takes part in the chemical reaction in
the blast furnace. Thus, the electrification potential of indus-
trial process heat has to be analyzed carefully for each type
of process and will strongly differ across countries.
Characteristics of Heat Demands
and Thermal Storages
Heat Consumption Profiles
Heat demand profiles are determined by the weather,
building characteristics related to thermal losses, occupant
behavior, and occupant expectations about indoor tempera-
ture. Consequently, typical daily demand profiles can be
quite diverse across countries (Figure 4).
In Finland, buildings are typically well insulated, and
thermal losses are relatively small even though outside tem-
peratures can get very cold. In district heating systems, heat
is stored in the pipelines and also in the building envelopes,
resulting in a daily profile with little variationmainly
january/february 2017
driven by longer-term ambient temperature variations. Inside
temperatures are kept nearly constant, even when occupants

Demand of Yearly Demand (%)

0.045
are not present. By contrast, houses in Ireland leak more, 0.040
and the small share of buildings that rely on electric radia- 0.035
tors use them mainly when occupants need the extra heat 0.030
for example, in the morning and evening during weekdays. 0.025
When occupants are not present or they are sleeping, inside 0.020
0.015
temperatures are often allowed to drop. Despite the weather
0.010
being more moderate in Ireland than in Finland, the average 0.005
Irish living room is probably colder than its Finish counter- 0.000
part due to different occupant expectations.

0:00
2:00
4:00
6:00
8:00
10:00
12:00
14:00
16:00
18:00
20:00
22:00
Annual profiles can also be quite different, although they
follow more closely the inverse of the ambient temperature. Large District Heating Large District Heating
Figure 5 shows that systems where the heat source also pro- System in Finland System in China
vides hot domestic water have some load during the summer. Domestic Gas Demand Electric Space Heating
In China, district heating systems can be shut down outside in Belgium in Ireland
the heating period.
While not shown in Figure 5, cooling could complement figure 4. Hourly heat profiles from a winter weekday.
the annual space heating profiles. In some climates where
heating and cooling needs are comparable, similar flexibil-
ity from cooling could complement flexibility from heating.
Wherever there are interconnected power grids spanning 1.00
across warm and cold climate zones, part of the variations,
0.90
Demand of Yearly Demand (%)

depending on the relative strength of the interconnections,

0.80
can be smoothed at this continental scale. In either case,
heating and cooling could provide a rather stable source of 0.70
potential flexibility for the power sector. Furthermore, in 0.60
hot and sunny countries, cooling loads and PV generation 0.50
may complement one another well. 0.40
Industrial heat demand at the country level does not 0.30
exhibit strong seasonality and could, therefore, provide
0.20
year-round flexibility (e.g., the industrial heat demand from
0.10
Finland shown in Figure 5). Also, daily profiles, especially
in heavy industries, are typically relatively flat. In individual 0.00
1 51 101 151 201 251 301 351
industrial sites, the profiles can have more variationfor
example, lower demand for products can cut work shifts. Large District Heating Large District Heating
System in Finland System in China

Time Constants and Thermal Storages
It is technically possible to store heat from one season to
another, but this has proved economically challenging.
Storing heat becomes more viable when considering time
spans of several days (or shorter). The storage time constants
depend on the storage size or on end-user comfort or needs,
which might be affected by the operation of the heating
device. Here is an approximate list of time constants for dif-
ferent heat uses:
domestic refrigerator/freezer: 15 min1 hour
supermarket refrigeration systems: 15 min3 hours
thermal mass of buildings: 212 hours
buildings with local hot water storage: 224 hours
district heating pipelines: 15 hours
district heating storages: hours to several days.
For economic reasons, water is commonly used as a
medium, even though other viable heat storage materials
exist. A cubic meter of water changing 5595 C offers
Domestic Gas Demand Electric Space Heating
in Belgium Demand in Ireland
Industrial Heat Demand Heat Demand in
in Finland the United Kingdom
figure 5. The average daily heat demands over a year.
about 58 kWh of thermal energy storage. In a not-so-well
insulated house on a cold day, this would last about half a
day. It can become quite impractical to install much larger
hot-water tanks inside residential buildings because they
require considerable space and would likely not fit through
door frames. Most existing water tanks are much smaller.
Consequently, for the most part, hot-water tanks offer flex-
ibility constrained by a limited time constantalthough the
flexibility can still be valuable for the power system when
aggregated over millions of houses.

january/february 2017 IEEE power & energy magazine 29

 The cost of storing thermal energy in water tanks
decreases rapidly with larger tank size (Figure 6). Longer
time constants and, consequently, more flexibility could be
achieved if the hot-water tank were oversized. Sharing the
tank between several buildings would, in turn, decrease the
60
50
US$/kWh ( 50 K)
specific costs. District heating systems already often utilize
large tanks to make their operation more flexible.
Thermal electricity end uses such as water heating,
space heating/cooling, and refrigerator/freezers are a
suitable source for flexibility due to their discretionary
nature, inherent thermal inertia, and large volumes. Large
thermal inertia means that these loads can be switched
off for a while without affecting consumer comfort. Fur-
thermore, because the flexible loads consist of a number
of appliances dispersed across the system, reliability can

be statistically greater compared to an individual conven-

40 tional power plant.
30 The building envelope itself also provides thermal iner-
20
tia depending on the insulation level and thermal mass of
the building. In a well-insulated house, electric load can be
10
shifted (212 hours depending on building), while consumer
0 comfort is still met. Preheating or precooling increases flex-
0 5 10 15 20 ibility but typically also increases energy usage (depending
m3 on the insulation level). Thermal storage and building pre-
(a)
heating enable considerable demand-shifting potential at a
10 comparably low cost.
US$/kWh ( 50 K)

8
6 Thermal-load forecasting is often used in district heating sys-
4
2 heating or cooling. Forecasting failures can lead to unwar-
0
0 5,000
m3
(b)
figure 6. The cost of hot water tanks per unit of storable
heat in relation to the storage tank size. (Small tank sizes
are based on market data; large tank sizes are from the
European Commission Joint Research Centres 2012 report.)
16 January 2015 08:00
Normalized Load ()
Forecasting
tems and for estimating electric heating loads. The uncertainty
of heat-load forecasts is important when trying to optimize
ranted costs or uncomfortable inside temperatures. For exam-
10,000 ple, when a heat-load forecast error persists in one direction
15,000
for muliple hours, heat storage may become emptied or filled,
after which it is not useful any more. When uncertainty is not
considered, optimizing the use of heat storage is too easy, and
model results or control strategies are too optimistic. Figure 7
demonstrates the quality of heat forecasts 096 hours ahead
for the Snderborg district heating system in Denmark.
If heating and cooling loads are to be controlled in a
manner that provides flexibility to the power system, accu-
rate forecasts will be important. Good forecasts will need
to include climatic variables, building characteristics, and
often predictions about user behavior. Building characteris-

2.2 tics can be considered either through direct modeling or past

2.0 behavior. Occupancy behavior varies from family to family:
1.8 some families have a very systematic and easily predicted
1.6 load pattern, whereas others seem to have a very random and
1.4
less predictive load profile.
1.2
1.0
Similar to electric loads, it has been shown that the
aggregation of individual loads can decrease relative fore-
0 24 48 72 96
cast errors and smooth out rapid variations. However,
Horizon (Hours)
when heating/cooling takes place in individual buildings,
the control algorithm cannot aggregate. On the other hand,
figure 7. The quality of heat forecasts 096 hours ahead for
the Snderborg district heating system in Denmark (the blue when several buildings are connected to a district heating
line). The 5% and 95% percentiles are shown in red, and the system, the forecasts for control can be aggregated and
actual load as it was observed later is shown in black. The relative forecast errors decrease. This is different from just
plot is generated using the adaptive heat-load forecasting forecasting loads without control, where aggregation can
system PRESS (www.enfor.dk/products/press.aspx). take place in all cases.

30 IEEE power & energy magazine january/february 2017

 In district heating systems, heat is stored in the pipelines
and also in the building envelopes, resulting in a daily profile with
little variationmainly driven by ambient temperature variations.
Harnessing Heat Flexibility
Flexibility from heating and cooling can be used in several
ways in a power system. Moving loads from electricity net
load peaks to net valleys can smooth variations. When the
heating devices have suitable controllability, they can also
be valuable sources of reserves that mitigate forecast errors
and faults. Because heat is such a diverse and heterogeneous
load, a selection of system studies and applications is pre-
sented in the following to highlight how heat flexibility can
potentially be harnessed.
Integrated Energy System Studies
Electricity System Benefits of
Heat-Pump Deployment in Belgium
Belgium is a densely populated country with significant
plans for variable-power generation. Cost-effective integra-
tion will be a challenge; consequently, there is strong inter-
est in finding workable sources of flexibility. In one study,
the deployment of 1-GW electricity from flexible heat pumps
managed to reduce curtailment of variable generation and
avoided 100 GWh of gas-fired generation. However, it was
identified that performing demand response with the heat
pumps increased the buildings heat demand by 110%. This
increase in electricity use poses a challenge in terms of com-
pensating consumers for participating in demand response,
especially since residential consumers are typically exposed
to prices higher than wholesale market prices. Another case
study shows that the contribution to the peak demand in
winter due to electrical space heating can be significantly
reduced, by 2 GW on a total of 16.5 GW.
Integration of Heat and Electricity Sectors
in the United Kingdom
In the United Kingdom, over 80% of households use nat-
ural gas for space and water heating, and this consumes
more than 1.5 times more energy than U.K. electricity con-
sumption. Peak heating requirements in winter are more
than five to six times higher than electricity peaks. While
the electrification of heat could, in principle, provide flex-
ibility to the power sector, the danger is that if electric
heating is uncontrollable, it will magnify power system
variability and peak demand. Additionally, the parallel
deployment of wind and solar power, in combination with
relatively inflexible nuclear generation, could exacerbate
flexibility demand. Studies suggest that, in an inflexible
U.K. electricity system having 30 GW of variable renew-
january/february 2017
ables in combination with inflexible nuclear generation,
more than 25% of wind energy may need to be curtailed if
no additional measures are taken.
In this context, a well-designed integration of heat and
electricity systems can lead to a more cost-effective transi-
tion toward a low-carbon energy system (see Figure 8). When
heat is supplied by heat pumps via district heating systems
or through a controllable electric heating device at customer
premises, analysis demonstrates significant benefits accrued
from three sources:
1) There will be less need for heat production capacity
when heat storages that use electricity cut peaks.
2) Curtailment of renewable generation is reduced when
heat storages can utilize excess generation.
3) With reduced curtailments, less renewable generation
capacity is needed to meet emission reduction targets.
Heat Versus Other Flexibility Sources
in the North European Context
The value of flexibility from the heat sector is influenced
by the cost-effectiveness and availability of flexibility from
other sources. In a Northern European study, a combined
generation-planning/operational model was used to evaluate
the benefits of adding heat pumps, electric heat boilers, and
heat storages to a district heating system in a future where
there would be much greater variable generation in the sys-
tem. The results demonstrated that, while new transmission
lines probably have the best cost-benefit ratio, heat-sector
flexibility comes as a close second, far ahead of electricity
14
Backup
Integration Cost (/MWh)
12
Balancing (OPEX)
10 Balancing (CAPEX)
Saving in RES
8
6
4
2
0
2030 2050
figure 8. The reduction in integration costs of renewable
generation (electricity sector only) enabled by the integrat-
ed operation of heat and electricity sectors.
IEEE power & energy magazine 31
 The parallel deployment of wind and solar power,
in combination with relatively inflexible nuclear generation,
could exacerbate flexibility demand.
storages or peak shaving-demand response. Buildings with
local heating and the transport sector were not included in
the study.
Heat Loads in Primary-Frequency Reserves
When a power system is running mainly with nonsynchro-
nous generation, one option in providing upward primary-
frequency reserve is to curtail wind or solar power plants to
get the necessary headroom for reserve operation. However,
flexible heating and cooling loads can also provide a fast
response, typically in the order of 100400 ms, when using
local frequency detection. Consequently, some generation
curtailments could be avoided, and system-wide fuel use
could be decreased.
In one case study, system-wide operating cost savings
ranged 8.5500 per heating appliance, depending on vari-
ous factors (fuel costs, wind penetration, etc.), for Great
Britains power system. The cost savings from the activation
of flexible heating loads improve with increasing shares of
wind power and with increases in reserve requirements. In
addition to direct economic benefits, using flexible loads in
primary reserve reduces the number of conventional plant
startups, enables higher levels of wind penetration, and
improves frequency stability.
When implementing heating- or cooling-based reserves,
one needs to consider the load pickup that takes place after the
load has been curtailed for reserve provision. The magnitude
of the pickup varies with the thermal inertia characteristics
of the heating/cooling load, the duration of the response, and
the control mechanism. For example, activation of 60 MW
of primary reserve from a domestic cold load (refrigerator/
freezer) for 90 s (5-min recovery duration) requires the addi-
tion of 20 MW (35% of activated flexible load) to subsequent
reserve categories to allow for the load pickup.
Residential HeatElectricity Integration
Using Hybrid Heaters in Ireland
Hybrid heating systems, such as a combination of a heat pump
and a gas boiler, enable shifting between the two different
sources of heat. If equipped with smart controls, it is possible
to shift in real time, depending on electricity market conditions.
An investment study of the Irish 2030 system, with 40%
electricity from wind power, found that the large-scale
deployment of such systems can provide electricity system
benefits. An optimization model was used to find the least-
cost heater capacities and operation schedule. If a gas boiler
is combined with a resistance heater, those hybrids will
32 IEEE power & energy magazine
operate primarily on gas but will shift to electricity when-
ever low-price electricity is available. When compared to a
gas boiler alone, the results showed annual system-wide sav-
ings of 1865 per household, depending on the gas price.
If a gas boiler is combined with a heat pump, they will oper-
ate mainly on electricity and shift to gas during periods of
low wind-power supply or high demand. Then, the annual
savings were 46159 per household. The flexibility from
hybrid heaters enabled the lowest-cost energy system.
Benefits of Electric Boilers in Reducing Wind-Power
Curtailment in Northern China
In the northern provinces of China, 20~40% of wind energy
was curtailed in 2015 due to inflexible operation of coal-fired
CHP plants. In winter, these plants must operate at nearly full
capacity to meet the demand for building heat (delivered as
hot water through district heating systems) and must produce
electricity at the same time. Combined with a high output from
wind power plants, this often causes an oversupply of electric-
ity, and wind power plants need to be curtailed. A series of
numerical studies tested the use of thermal storage and/or heat
pumps to increase the flexibility of the system. The results
demonstrated a significant reduction in wind-power curtail-
ments. On the other hand, air-source heat pumps suffer from
low efficiency in the cold winter conditions of Northern China
and may not be an economic choice.
Some Real-World Experience
and Applications
Denmark is one of the leading countries in the integration of
large amounts of wind power. In 2015, 42% of its electricity
was generated by wind turbines. Apart from its large inter-
connectors to neighboring countries, the integration of wind
power was enabled by its district heating networks. These
networks can store excess wind power generation through
a combination of electric heaters and heat storages. Mean-
while CHP plants can be operated when there is not enough
low-price electricity available.
In residential buildings, smart thermostats can give function-
ality beyond temperature and time-of-use control. Communica-
tion with the Internet or an aggregator enables the utilization
of power prices and weather forecasts. Meanwhile, occupant-
modeling intelligence can consider the actual needs of the occu-
pants in the control scheme. For example, the model predictive
control (MPC) algorithm can make use of the additional infor-
mation to better utilize lower power prices and improve energy
efficiency. From a power system perspective, this appears as
january/february 2017
 versity acknowledges support from the National Natural Science
Foundation of China (51620105007). S. Heinen acknowledges
Fraction of Total Load

Precooling Energy
Rush Hour support from the Fonds National de la Recherche, Luxembourg
1.0
0.8 (project reference 6018454) and CITIES (DSF 1305-00027B/
0.6 DSF). J. Kiviluoma acknowledges support from the FLEX-e
0.4 program funded by TEKES.
0.2
0.0
For Further Reading
01:30 p.m.
02:00 p.m.
02:30 p.m.
03:00 p.m.
03:30 p.m.
04:00 p.m.

04:30 p.m.
05:00 p.m.
05:30 p.m.
06:00 p.m.
06:30 p.m.
D. Patteeuw, K. Bruninx, A. Arteconi, E. Delarue, W.
Dhaeseleer, and L Helsen, Integrated modeling of active
Time demand response with electric heating systems coupled to
thermal energy storage systems, Appl. Energy, vol. 151,
Actual ac Use pp. 306319, Aug. 2015.
Projected ac Use Without Rush-Hour Rewards
S. Heinen, D. Burke, and M. OMalley, Electricity, gas,
heat integration via residential hybrid heating technologies:
figure 9. Employing heat flexibility in residential buildings An investment model assessment, Energy, vol. 109, pp.
to avoid electricity demand when the electricity grid faces 906919, Aug. 2016.
high loads. (Source: nest.com.) K. MacLean R. Sansom, T. Watson, and R. Gross, Com-
paring the impacts and costs of transitions in heat infrastruc-
ture, Centre for Energy Policy and Technology, Imperial
increased flexibility. MPC is applied by companies such as College, London, U.K., Apr. 2016.
BuildingIQ or QCoefficient when they exploit chiller efficiency J. Kiviluoma, Managing wind power variability and un-
variations due to ambient temperature to achieve energy sav- certainty through increased power system flexibility, Ph.D.
ings without overly affecting occupant comfort. An example of dissertation, Aalto Univ., Dept. Applied Physics, VTT Sci 35,
smart-thermostat-based control is shown in Figure 9, where the Espoo, Finland, 2013, p. 77.
smart thermostat Nest performed large-scale peak shaving by T. Naegler, S. Simon, M. Klein, and H. C. Gils, Quantification
precooling American residential buildings. of the European industrial heat demand by branch and tempera-
In 2011, China initiated a series of pilot projects that ture level, Int. J. Energy Res., vol. 39, pp. 20192030, Oct. 2015.
substitute electric boilers to for coal-fuelled CHP boilers N. Zhang, X. Lu, C. P. Nielsen, M. B. McElroy, X. Chen,
in the Jilin province. The electric boilers often use surplus Y. Deng, and C. Kang. Reducing curtailment of wind electric-
wind generation as their energy source. From 2015, the ity in China by employing electric boilers for heat and pumped
project has expanded to all the northern provinces such as hydro for energy storage, Appl. Energy, vol. 184, pp. 987994,
Hebei, Liaoning, Inner Mongolia, and Xinjiang. The elec- Dec. 2016.
tric boilers are equipped with water tanks capable of pro- P. Bacher, H. Madsen, H. A. Nielsen, and B. Perers, Short-
viding 10~15 hours of storage. term heat load forecasting for single family houses, Energy
Buildings, vol. 65, pp. 101112, Oct. 2013.
Conclusions
Heating and cooling offer considerable flexibility potential Biographies
for power systems. Much of this could become cost-effective Juha Kiviluoma is with the VTT Technical Research Cen-
as the share of variable and uncertain generation increases. tre, Espoo, Finland.
Simultaneously, electrification of heating offers a possibility Steve Heinen is with University College Dublin, United
for heat sector decarbonization. However, the picture is not Kingdom.
entirely rosy. Seasonality in space-heating needs makes it a Hassan Qazi is with University College Dublin, United
less attractive source of flexibility. On the other hand, flex- Kingdom.
ibility from heating could be partially complemented by flex- Henrik Madsen is with the Danish Technical University,
ibility from cooling or from more stable loads in the industrial Lyngby, Denmark.
sector. At the same time, the industrial sector is very diverse Goran Strbac is with Imperial College London, United
and will require elaborate research to understand the true Kingdom.
flexibility potential in heat-consuming industrial processes. Chongqing Kang is with Tsinghua University, China.
Ning Zhang is with Tsinghua University, China.
Acknowledgments Dieter Patteeuw is with Katholieke Universiteit Leuven,
H. Madsen acknowledges support by CITIES (Center for IT- Belgium.
Intelligent Energy Systems, DSF 1305-00027B). G. Strbac Tobias Naegler is with the German Aerospace Center,
acknowledges support from the U.K. Research Council-funded Stuttgart.
p&e
project Energy Storage for Low-Carbon Grids. Tsinghua Uni-

january/february 2017 IEEE power & energy magazine 33

 By Eoin Casey,
Sara Beaini,
Sudeshna Pabi,
Kent Zammit,
and Ammi Amarnath

E
ENERGY AND WATER HAVE BEEN INTEGRATED
throughout most of modern history, and that linkage
will continue into the future, not only in the physical
infrastructure but also through digital infrastructure
(e.g., the Internet of Things). The term energywater
nexus is quickly expanding to refer to more than simply
water used for energy production and energy used for
water treatment and transport. Just as the energy grid
is changingbecoming more flexible and resilient and
providing energy-efficiency gainsthe water network
is also changing. The integration of these two systems
can provide optimization and opportunities that would
not otherwise be possible. This integration of electrons
and molecules is being enabled by advances in Inter-
net connectivity and wireless communications, so that
energy in all its forms can be employed most effectively
by end users to optimize efficiency, reliability, security,
economics, and environmental performance.
ISTOCKPHOTO.COM/TCMAKE_PHOTO

Benefits of System Integration:

A Triple Bottom Line
Addressing the energywater nexus invites a system-
integration approach (Figure 1). Todays electric grid has
become the integrated energy network, which requires
versatility across power paths and adapting to the new

The Triple Bottom

Line for Efficiency
Digital Object Identifier 10.1109/MPE.2016.2629741
Date of publication: 2 February 2017

34 IEEE power & energy magazine 1540-7977/172017IEEE january/february 2017

 Integrating Systems
Within Water and
Energy Networks

commercial buildings, with benefits for both cus-

tomers and energy suppliers
carbon reductions achieved through the electrification
of water systems, replacing fossil fuel point sources and
increasing efficiency for the pumping, heating, treat-
ment, and transport of water and wastewater
off-peak water treatment, water heating, and water
pumping to reduce both energy costs and generat-
ing-unit downturn
distributed water treatment/WWT and local reuse,
providing greater system security and resiliency
expanded availability of Energy Star and Water-
Sense appliances for water and energy savings, en-
abling easy ways to achieve customer savings and
satisfaction as well as to implement energy and wa-
ter conservation methods.
In particular, water presents strategic challenges and
opportunities for the electric sector.
Using less water for power production conserves a
scarce resource for other necessary uses.
Minimizing the environmental impacts of water use
for power production preserves environmental re-
sources and protects human health.
Using efficient electric technologies for water treat-
ment, transport, desalination, industrial processes,
and other end uses can conserve electricity, thereby
reducing any imbedded water demand.

energy end user, called a prosumer, who may also produce

energy. Similarly, the water network has taken on an in-
creasingly dynamic role as more decentralized, small-scale
systems are implemented. Think of the interaction between
the water network and the smart electric grid as the integrat-
ed energy network (or integrated network of resources),
where further efficiencies may be gained through this inte-
gration of systems. With this integration, we drive resource
use toward a more sustainable and efficient utilization and
management of resources (Figure 2)
Many potential benefits can be realized by integrating
the electricity and water sectors. Some examples include
load leveling, e.g., pricing signals used to control figure 1. The energywater nexus. Integrating the
demand for systems such as water pumping to head energy grid and water network to create an integrated
tanks, desalination, wastewater treatment (WWT), ir- energy network can drive more sustainable and efficient
rigation pumps, and water heaters in residential and resource utilization.

january/february 2017 IEEE power & energy magazine 35

 Todays electric grid has become the integrated energy network,
which requires versatility across power paths and adapting to the
new energy end user, called a prosumer.
Smart grids are characterized by connectivity, flexibil-
ity, and resiliency, all of which effectively optimize their
efficiency, reliability, security, economics, and environmen-
tal performance. Numerous system integration technolo-
gies are emerging today that promote these smart features.
Energy systems integration (ESI)a multidisciplinary
area ranging from science, engineering, and technology to
policy-making, economics, regulation, and human behav-
ioris coming to the fore in the planning, design, and oper-
ation of the global energy system. ESI seeks to optimize
the energy system and other large-scale infrastructuresin
particular, waterby leveraging synergies across all scales
and pathways.
Evaluating an integrated energy network allows us
to address efficiency with a triple bottom line: financial,
environmental, and social impacts. In other words, the sys-
tems optimization will be driven by a primary target with
consideration given to secondary impacts. The primary or
secondary targets might include greenhouse gas emissions,
Ecosystem Electric Power
Demand
Agricultural Industrial
Municipal
figure 2. Resource management has become more
intricately integrated, specifically for water and energy
resources. Although not as obvious as energy consump-
tion, water usage covers a variety of applications, from
power generation and industrial and agricultural applica-
tions to municipal needs with water treatment/WWT, and
also includes the goal of sustaining of natural ecosystems.
(Image courtesy of the Electric Power Research Institute.)
36 IEEE power & energy magazine
power-load profiles, water conservation metrics, and the
population density served, among others. Deciding which
are the primary or secondary targets depends on the driv-
ing or pressing factors. Utilities, municipalities, and building
and facility operators may reap the benefits of the integration
of water and energy networks with the triple bottom line.
The examples that follow here demonstrate the emerging
integration of the water network and energy grids, particu-
larly in the WWT part of the water sector and the end-use
part of the electric grid. Specifically, for both the water and
energy sectors, these examples highlight demand response
(DR) opportunities, energy-efficient technologies, and the
important role that reconceptualizing WWT plants can play
as part of future energy systems in terms of virtual storage
and as resource factories.
Demand Response
DR, as defined by the U.S. Federal Energy Regulatory Com-
mission, refers to changes in electric usage by end-use cus-
tomers from their normal consumption patterns in response
to changes in the price of electricity over time, or to incentive
payments designed to induce lower electricity use at times of
high wholesale market prices or when system reliability is
jeopardized. This implies the shedding of some loads when
the electric system reaches critical peaks and loads cannot
be served by existing generation plants (Figure 3). Because
water treatment and conveyance require large amounts of
energy, the potential exists for large DR capabilities in these
sectors. Examples of DR implementation include
agricultural irrigation DR signaling so that pumps
operate when electricity demand and associated rates
are low
pumping to head tanks at off-peak hours
off-peak water treatment
flexible load management and energy storage in
water heaters.
Water treatment and WWT facilities are good candidates
for DR because they are energy intensive. In some cases, a
water storage capability could offer some flexibility in the
operation of certain processes, including pumps and centri-
fuges. This operational flexibility, in turn, can be leveraged
for DR if properly coordinated, making these facilities ideal
partners for electric utilities seeking to manage electric load
through DR programs. Furthermore, water storage can be
used in conjunction with on-site power generation to provide
greater demand reduction. For example, water storage can
be used to shave power requirements at high electricity load
january/february 2017
 Evaluating an integrated energy network allows us to address
efficiency with a triple bottom line: financial, environmental,
and social impacts.

points so that on-site generation can be sold into the grid at
the higher rates. The stored water can then be treated later,
with power purchased at lower rates.
Water distribution systems contain potentially large
amounts of storage, which provide system pressure and
backup. When properly managed, water utilities can reduce
distribution system pumping and allow the water supply
system to coast during peak electrical periods. Wastewa-
ter systems, on the other hand, may divert a portion of the
incoming sewage into holding cells or reduce aeration dur-
ing peak electrical periods (Figure 4). Under the right cir-
cumstances, DR from water and wastewater facilities can
be significant, benefiting both electric utilities (by reducing
the need for peak generation) and water treatment/WWT
facilities (through DR incentives; see Figures 5 and 6). For
example, a third-party DR aggregator has enrolled in excess
of 100 MW of curtailable loads and on-site generator capac-
ity at about 700 U.S. water treatment and WWT facilities.
WWT systems struggle with significant flow changes,
particularly during rainfall events. Infiltration into the collec-
tion system represents a challenging problem for many sys-
tems, so many plants have storage available at the front of
the treatment plant to capture excess flow for treatment at a
later period. Basin management decisions are usually based
on keeping storage available for the next storm event, but,
depending on the plant, this storage could be used to man-
age peak electric demand by diverting wastewater to these
basins during peak electric periods. Fortunately, peak electric

Dams
Produce
Energy Is Used for Electricity
Pumping Water

Power Plant Water Used

Cooling Uses for Mining Fuels
Water

Water Supply
Uses Energy

Energy Uses in
Water/Wastewater
Treatment

Water Flows Water and Energy Use

in the Home Are Related
Energy Flows

figure 3. Water system quality can be maintained while pumps, heaters, and other equipment can provide DR,
which shifts electrical load away from peak periods to off-peak periods. (Image courtesy of the U.S. National
Renewable Laboratory.)

january/february 2017 IEEE power & energy magazine 37

 Mid-Atlantic WWTP1, Processes 681,390 m3/day
1) 2) 3) 4)

Screens Primary Activated Sludge Secondary Sand Chlorination Discharge

Clarifier Reactors Clarifier Filter

Mid-Atlantic WWTP2, Processes 7,570 m3/day

1) 2) 3) 4)

Screens Primary Primary Secondary Secondary Multimedia Chlorination Discharge

Clarifier Aeration Tank Aeration Tank Clarifier Filter

Midwest WWTP1, Processes 1,363 m3/day

1) 2) 3) 4)

Screens Activated Sludge Clarifiers Seasonal Discharge

Lagoons Chlorinationa

Midwest WWTP2, Processes 1,439 m3/day

1) 2) 3)

Screens Sequencing Batch Lagoon Lagoon Lagoon Lagoon Lagoon Discharge

Reactor Cell A Cell B Cell C Cell D Cell E

figure 4. WWT plants provide opportunities for electric system flexibility.

demand periods often coincide with hot and dry weather, giv-
ing plant managers some flexibility in keeping storage avail-
able while reducing electrical demand.
WWT plants comprise a number of unit processes that
typically operate in continuous mode. Energy demand
is closely correlated with liquid flow rates into the plant.
Because flow rates are variable (diurnal cycles) or weather-
related (rainfall), energy demand is not constant. Waste-
water utilities typically pay for their electricity according
to a fixed rate, although dynamic pricing structures do
exist. When the electricity grid experiences peak demand,
the opportunity exists for a plant to voluntarily curtail its
electricity usage by turning down or shutting off equip-
ment in return for rebates from the electrical utility. Load
shifting strategies in WWT facilities include pre-aeration,
utilization of storage capacity, and the scheduling of some
operations (e.g., dewatering and filter back-washing) during
off-peak periods.
However, without proper design, the duration of this cur-
tailment has potential implications for WWT operations in
terms of water quality. There are some documented exam-
ples of turning off aeration blowers for several hours as a DR
measure. However, this can have a negative impact on the
water quality (for example, turbidity) in the treated waste-
water. Another interesting approach is the concept of over-
oxygenating of wastewater by over-aerating wastewater prior
to a DR event. This load-shifting strategy shows promise for
significant DR reductions. Other constraints include meeting
discharge consents, dealing with major fluctuation in storm
water flows, and public health risks.

38 IEEE power & energy magazine january/february 2017

 With proper management, water utilities can reduce distribution
system pumping and allow the water supply system to coast
during peak electrical periods.

An integrated energy system approach allows opportu-
nity for decision support on process operation strategies that
increase/decrease loading to follow electricity tariffs and/or
short-term loading response to provide power system flex-
ibility. Thus, there is a need for real-time data analysis and
forecasting systems that will inform process-control strate-
gies. The benefits of such an approach include cost savings,
improved control and decision-support systems for planning
plant upgrades as part of long-term wastewater throughput,
and tightening of nutrient discharge limits.
Water storage tanks in municipal water treatment facili-
ties and water heaters in residential and commercial buildings
represent an energy storage opportunity for the electricity grid.
For example, there are roughly 53 million homes in the United
States with electric water heaters; a direct load-control program
could result in an estimated 0.40 kW in savings per home,
thereby providing peak demand reduction of 5,300 MW,
assuming a 25% participation rate.
To realize this waterenergy load management and storage
potential, there is a need to develop standardized communi-
cation protocols and ubiquitous communication networks for
the secure messaging of energy price and event information
related to distributed energy resources, such as storage tanks
and water heaters. In addition to these enabling standards
and communication technologies, more research is needed to
determine how best to integrate and aggregate large numbers
of small resources, such as electric water heaters, into the
overall energy management system.
Efficiency Approaches
The sourcing, treatment, and distribution of water require
significant energy. Approximately 3% of the electricity in
the United States is used to move and treat water and waste-
water. There are opportunities to optimize energy and water
consumption in the water sector by deploying technologies
such as advanced supervisory-control and data-acquisition

Irrigation 2010
Washin
g ton e
pshir
Ham t ine
Montana Lake Superior New o n Ma
North Dakota Minnesota Verm
La
igan

Oregon
ke
H

Wisconsin Lake Ontario Massachusetts

Mich

ur

Idaho South Dakota

on

n New York
Wyoming iga e E rie Rhode Island
Lake

ch Connecticut
Iowa L a k Mi
Pennsylvania
Nebraska New Jersey
Nevada
Ohio District of Columbia
Utah Illinois Indiana West Delaware
Colorado Virginia Maryland
Califo
rnia Kansas Virginia
Missouri Kentucky
North
Tennessee Carolina
Arizona Oklahoma Arkansas South Explanation
New Mexico
Water Withdrawals,
Mississippi

Carolina
in Million Gallons
Georgia
Alabama Per Day
Texas 0200
Hawaii 2011,000
Louisiana
Flo

1,0015,000
rid

5,00115,000
a

Alaska West-East Puerto U.S. 15,00123,100

Division for Virgin
Rico Islands
This Report

figure 5. The operation of irrigation systems can provide DR services to the electrical system, especially in the western
United States. (Source: U.S. Geological Survey.)

january/february 2017 IEEE power & energy magazine 39

 Methane Generation Potential from Wastewater Treatment

Tonnes/Year
>5,000
2,5005,000
1,0002,500
5001,000
<500

This analysis estimates the methane generation potential of wastewater treatment plants using methodology from the EPAs
inventory of U.S. Greenhouse Gas Emissions and Sinks: 19902011 and data from the EPA Clean Watersheds Needs
Survey (2008). The results were further aggregated to country level.

figure 6. Wastewater plants provide opportunities for electric power generation fueled by biomass. EPA: U.S. Environ-
mental Protection Agency. (Image courtesy of the National Renewable Energy Laboratory.)

(SCADA) systems, forward osmosis, bubbleless aeration,
membrane distillation, capacitive de-ionization, and emerging
biological treatment processes. Deploying these energy-effi-
cient technologies will require a comprehensive assessment,
from lab-scale to field demonstration, to characterize their
performance in real-world applications.
In addition, the adoption of end-use energy and water con-
servation methods by the business and residential sectors would
reduce the imbedded energy and water required to meet those
needs. Current research into new technologies to conserve water
and energy at end use will have ripple effects on water and
energy conservation upstream. In addition to the potential for
energy- and/or water-use savings, these technologies could also
provide opportunities to more effectively treat water in munici-
pal and industrial applications that currently use chemical treat-
ment. End-use conservation would directly reduce water and
energy demand. Some examples of efficiency opportunities in
water treatment/WWT are discussed in the following.
Efficiency via Data Monitoring
and Process Control
As with any complex industrial process, the potential for
efficiency gains via computer-based monitoring and con-
trol in water and wastewater systems is significant. Moni-
toring and control technologies vary from simple devices
to advanced SCADA systems. SCADA systems are used
for precise control of key equipment and processes, in-
cluding raw-water wells, water treatment, and distribution
pumping. Typically, these SCADA systems pull data from
field devices, such as programmable logic controllers, re-
mote terminal units, and electric meters, and analyze/for-
mat the data to be viewed by operations staff or used for
process control. SCADA systems are being implemented
in the water and wastewater industries, but traditionally
there has been a greater focus on improving process qual-
ity and reliability than on controlling systems to optimize
energy efficiency.

40 IEEE power & energy magazine january/february 2017

 The adoption of end-use energy and water conservation methods
by the business and residential sectors would reduce imbedded
energy and water requirements.
It has been estimated that an electric energy savings
potential of 510% across the U.S. public water supply can
be achieved with advances in pumping and water treatment
process control. Assuming the public water supply currently
uses about 39 billion kWh per year, the potential electric
energy savings associated with advanced SCADA systems
ranges from 2.0 to 3.9 TWh per year. This translates into
electricity savings ranging 5.410.9 million kWh per day
across the United States. One such energy-saving technique
is to use a SCADA system for automatically selecting the
best pump combination, reducing system pressure when pos-
sible, checking the system efficiency in real time, and then
notifying the operator when changes are required.
The most sophisticated control systems learn the char-
acteristics of the distribution system, relying on predictive
modules to assist in scheduling pumping. This option is
extremely valuable in systems where the pump station takes
advantage of time-of-day electric rate schedules.
Efficiency via Water Conservation
Water conservation is an overlooked challenge as an energy-
efficiency measure in both water treatment and WWT.
Lowering water demand reduces the volume of water drawn
from public water supplies; this, in turn, reduces the energy
required to pump and treat the water supplied to end users.
A lower demand for fresh water also translates directly into a
reduced demand for wastewater transport and treatment and
a corresponding reduction in energy used.
There are two main challenges for water conservation
in water supply and wastewater disposal. On the water sup-
ply side, the opportunity lies in detecting and eliminating
leaks in the supply system. On the wastewater side, inflow
and infiltration lead to significant increases in flow to the
treatment facility, particularly during rain events. The addi-
tional volume of inflow water combines with wastewater
effluent and increases the amount of wastewater that must be
pumped and treated.
Reducing Demand for Water in End Uses
Considerable opportunities exist for reducing fresh water
demand for landscape irrigation. Based on U.S. Environ-
mental Protection Agency and U.S. Bureau of Reclamation
data, the potential savings from advanced irrigation con-
trols in residential and commercial applications is estimated
to be 1.53% of total electricity use in the public water sup-
ply. At a current electricity use rate of 39,000 million kWh
per year, this equates to potential savings of 0.51.2 TWh
january/february 2017
per year in the public water supply. While this is not a small
number, the nature of the savings through numerous, small
actions makes the impact of this measure extremely chal-
lenging to measure.
Providing timely information on usage patterns has
proven to be an effective way to increase awareness and
transform consumer behavior in both the energy and
water industries. There is a substantial opportunity to
modify consumer behavior and detect leaks by providing
a greater degree of visibility into use patterns. An exam-
ple is energy savings due to reducing hot-water demands
with low-flow devices.
Energy Recovery and Generation
A new and growing trend in the water and wastewater indus-
try is the emphasis on recovering energy whenever possible.
In water treatment, the focus is on recovering some of the
pumping energy through the use of energy-recovery devices
in the distribution system. In WWT, the emphasis is on bio-
logical treatments combined with opportunities in capturing
energy in the wastewater itself. These include cogeneration
using digester biogas and the recovery of excess line pressure
to produce electricity (microhydro).
Advanced Technologies in Water
Treatment for Energy Efficiency
There are significant growth opportunities in advanced
technologies in water treatment and WWT spurred mainly
by drivers associated with water scarcity and the need to
meet stricter discharge limits. However, many of these pro-
cessesincluding, for example, reverse osmosis for desali-
nation, advanced ionization for micro-pollutant removal,
and membrane bioreactorsare expected to continue to
be highly energy intensive. Some emerging developments
to address this problem include forward osmosis or mem-
brane distillation using low-grade waste heat. Another
significant opportunity is to couple desalination with
renewable-energy systems. Energy efficiency can also be
improved through the integration of space-conditioning
and water-heating systems. For residential and commer-
cial building applications, newer systems are under devel-
opment that use waste heat from outdoor air-conditioning
compressor units to heat water. Research is underway to
determine the overall efficiency of such systems. This
technology is fairly mature in the industrial sector, where
heat pumps are used recover heat from industrial processes
to heat-process water.
IEEE power & energy magazine 41
 In WWT processes, energy-efficiency measures include
retrofitting plants with high-efficiency pumps, variable-
speed drives, and advanced monitoring and process con-
trol for the biological reactor. Traditionally, WWT plants
are fitted with oversized pumps to cope with hydraulic load
fluctuations; as such, these pumps are inherently inefficient.
Thus, there is a need for real-time data analysis and fore-
casting systems that will inform process-control strategies.
The benefits of such approaches include cost savings, flex-
ibility, and improved control and decision support for plan-
ning plant upgrades.
Reconceptualizing the WWT Plant as an
Energy- and Resource-Generation Facility
Future WWT plants will no longer be just pure waste man-
agement facilities but rather recovery systems for clean
water, energy, and minerals. The chemical energy in waste-
water is in the form of biodegradable and inert chemical
oxygen demand and reduced nitrogen (NH4+ or organic N).
Large-scale WWT plants recover 515% of this energy
through anaerobic digestion. However, because other impor-
tant parts of the treatment plants are energy-intensive (aero-
bic biological oxidation requires 0.400.65 kWh/m3), it
follows that most treatment plants are net users of energy.
With the advent of new and improved treatment technolo-
gies, a net-energy-positive WWT plant is now considered
achievable. Emerging process technologies, including anam-
mox, bubbleless aeration, and aerobic granular sludge, can
deliver significant savings in energy demand.
Biogas is central to energy neutrality in WWT. Biogas
can be produced continuously, and there are several possible
uses: on-site electricity generation, combustion for ther-
mal processes (e.g., Cambi), or direct injection into the gas
grid; however, there are upsides and downsides with each
(for example, for injection into the grid, the biogas needs
to be pretreated). Currently, large-scale WWT plants typi-
cally recover only 520% of the chemical energy through
anaerobic digestion (the production of biogas and cogenera-
tion). Codigestion of excess sludge with external solid/liquid
organics is a potential approach toward carbon neutrality.
Additionally, the transformation of WWT plants can be
seen with the emergence of decentralized, small-scale set-
ups. By reducing the scale of and decentralizing some water
and wastewater operations, it is possible to lower costs and
improve efficiency.
Decentralized WWT lowers the cost for pumping
wastewater to central stations for treatment, then back
to communities for reuse.
Not all water uses require potable water quality: resi-
dential wastewater can be treated to use for local ir-
rigation, flushing toilets, etc.
Decentralization can also have benefits in minimizing
the need to increase infrastructure for high-density in-
fill projects. Building codes could minimize runoff;
storm water reuse reduces drain flows.
42 IEEE power & energy magazine
Forward osmosis and other emerging low-energy
technologies are proving beneficial as ways to treat
water for potable use locally.
Conclusions
Transitioning toward an integrated energy network requires
many drivers. Important players for implementing and
accelerating this change in resource management include
new regulations from resources agencies, economic incen-
tives for end users, and commercialization of new technolo-
gies with improved performance and lower costs. When
rallying the diverse stakeholders toward the triple bottom
line of efficiency, the dialog will revolve around the advan-
tages of the system integration approach for the energy
water nexus. The emerging system-integration case stud-
ies presented here in terms of DR, efficiency approaches,
and reconceptualizing the WWT plant as an energy- and
resource-generation facility all work to enable the goals
of efficiency, reliability, security, flexibility, collaborative
excellence, and technology leadership for both the smart
energy grid and the water network.
Acknowledgment
This publication has emanated from research conducted
with the financial support of the Science Foundation Ireland
(SFI) under the SFI Strategic Partnership Programme Grant
Number SFI/15/SPP/E3125.
For Further Reading
R. Hamilton, B. Braun, R. Dare, B. Koopman, and S. A. Svo-
ronos, Control issues and challenges in wastewater treat-
ment plants, IEEE Control Syst. Mag., vol. 26, no. 4, pp.
6369, 2006.
B. Sparn and R. Hunsberger, Opportunities and chal-
lenges for water and wastewater industries to provide ex-
changeable services, National Renewable Energy Labo-
ratory, Golden, CO, Tech. Rep. NREL/TP-5500-63931,
2015.
S. Pabi, L. Reekie, A. Amarnath, and R. Goldstein.
Electricity use and management in the municipal water
supply and wastewater industries, Electric Power Research
Institute. Palo Alto, CA, EPRI Rep. 3002001433, 2013.
Biographies
Eoin Casey is with University College Dublin, Ireland.
Sara Beaini is with the Electric Power Research Insti-
tute, Palo Alto, California.
Sudeshna Pabi is with the Electric Power Research Insti-
tute, Palo Alto, California.
Kent Zammit is with the Electric Power Research Insti-
tute, Palo Alto, California.
Ammi Amarnath is with the Electric Power Research In-
stitute, Palo Alto, California.
p&e
january/february 2017
By Emiliano DallAnese,
Pierluigi Mancarella,
and Antonello Monti

Unlocking
Flexibility
ISTOCKPHOTO.COM/LEOWOLFERT

E
ELECTRICITY, NATURAL GAS, WATER, AND DISTRICT HEATING/
cooling systems are predominantly planned and operated independently.
However, it is increasingly recognized that integrated optimization and con-
trol of such systems at multiple spatiotemporal scales can bring significant
socioeconomic, operational efficiency, and environmental benefits. Accord-
ingly, the concept of the multi-energy system is gaining considerable atten-
tion, with the overarching objectives of 1) uncovering fundamental gains (and
Integrated potential drawbacks) that emerge from the integrated operation of multiple
systems and 2) developing holistic yet computationally affordable optimi-
Optimization zation and control methods that maximize operational benefits, while 3)
acknowledging intrinsic interdependencies and quality-of-service require-
and Control ments for each provider.
On a much broader scale, the main drivers for the integrated operation
of Multienergy of multiple infrastructures include the impetus toward a decarbonization of
various energy and transportation sectors and the potential for resolving the
Systems so-called energy quadrilemma by putting forward integrated operational
Digital Object Identifier 10.1109/MPE.2016.2625218
Date of publication: 2 February 2017

january/february 2017 1540-7977/172017IEEE IEEE power & energy magazine 43

 solutions that strike a balance among cost, reliability, and
the environment, while accounting for energy affordability,
overall stainability (which pertains not only to carbon emis-
sions but also involves the use of natural resources such as
water), and social acceptability. The latter includes visual
impacts, safety and privacy concerns, and thermal (dis)com-
fort, to name just a few.
Shifting Supply and Demand
Across Spatiotemporal Scales
From an operational perspective, the coordinated and seam-
less control of various energy infrastructures represents a
significant change, which favors a local view that renders
city quarters, residential neighborhoods, and industrial areas
the fundamental building blocks of the integrated energy
system. Along with the growing role of distributed energy
resources (DERs), the envisioned control architectures is
based on a multi-area view, whereby the local neighborhoods
and districts are driving factors in a bottom-up approach.
Benefits such as the integration of higher levels of renew-
able energy resources, increased reliability and improved
efficiency in power systems, and significant savings in, e.g.,
water, heating/cooling, and gas system operations can be
achieved by uncovering (and capitalizing on) the intrinsic
flexibility that emerges from an integrated operation of mul-
tiple energy systems at multiple spatiotemporal scales. Flex-
ibility can generally be seen as a systems ability to provide
secure and economical supply-demand balance across spa-
tial and temporal scales by leveraging and seamlessly coor-
dinating various controllable assets. In the context of future
low-carbon power systems, major flexibility challenges are
associated with
Adjacent
Systems Solar
Industrial
Area
H1
H2
Commercial H3 H4
Area
Electricity Heating Natural Gas
figure 1. A complex system of systems providing basic electricity, heat/cooling, gas,
and water services to end customers. Dark areas designated Hi (energy hubs) represent
locations and facilities where the various infrastructures are coupled.
44 IEEE power & energy magazine
the integration of variable and uncertain renewable
energy sources
low-inertia operational settings, with the consequent
impacts on the frequency-response task
the high cost of utility-level and community-level en-
ergy storage systems.
In this respect, multi-energy systems could represent a key
option to provide flexibility in future power systems owing
to their untapped potential to shift supply and demand across
energy vectors and networksand, in doing so, across spa-
tiotemporal scalesby exploiting energy storage in the form
of energy, heat, or gas storage.
For example, coupling combined heat-and-power (CHP)
plants with electric heat pumps (EHPs) can introduce oppor-
tunities for an energy-shifting arbitrage between gas and
electricity to supply electricity, heat, and cooling to end
users, something that is particularly useful in the presence of
variable renewable energy. The presence of thermal energy
storage (TES)which may come at a much lower cost than
electricity energy storage but can provide similar services
in the context of multi-energy systems (a process that over-
all might be considered a form of virtual storage)could
provide further possibilities for energy-shifting flexibility.
Similarly, considering the natural interaction between gas
and electricity networks, there are major flexibility oppor-
tunities that can be exploited via coordinated control, as
explained later in the article.
The Concept of Energy Hubs
Figure 1 exemplifies a complex system of interconnected
infrastructures providing basic electricity, heat, gas,
and water services to end customers. The physical couplings
among systems could, for instance,
be represented in mathemati-
cal terms through the so-called
energy hub formalisma ge-
ner ic, sca lable, a nd modula r
modeling approach that offers the
flexibility to capture energy-con-
version factors at various spatial
Residential
scales. In particular, the math-
Area
Pumped Hydro ematical model associated with
(Storage) the energy hub shown in Figure 2
could offer representations of the
interdependencies among energy
carriers at the household, com-
mercial, and energy-district lev-
els and facilitate the formulation
Wind
of (and solutions for) control/op-
timization schemes to control
Water
co-generation and tri-generation
plants in distribution settings, as
well as gas-to-power and power-
to-gas and hydro facilities at the
transmission level.
january/february 2017
 The generic schematics provided in Figure 2 include a
number of input energy vectors (electricity, water, natural gas,
hydrogen, and district heating); typical conversion elements
(such as CHP units, electric and absorption chillers, and water
pumps); and relevant outputs representing local generation and/
or electricity, heat and cooling, gas, and water demands. The
energy hub depicted in Figure 2 also contains a heat exchanger
for district-heating-network connection, energy-storage sys-
tems, and TES in the form of, e.g., hot water storage and ice-
thermal storage. Additional elements not explicitly represented
in Figure 2 could include reversible EHPs, absorption chill-
ers, power-electronics-interfaced renewable energy resources,
electrolyzers (with a hydrogen output), and different types of
engines or turbines.
The conversion stages can be captured through coupling
factors and/or conversion efficiencies; this black-box approach
reduces the level of complexity and the number of parameters
necessary to describe the energy hubs operation and, hence,
facilitates the development of computationally affordable con-
trol/optimization schemes, while maintaining adequate accu-
racy in the representation of the underlying system physics. A
generic energy hub model of Figure 2 can be tailored to spe-
cific configurations (e.g., the unit, facility, plant, or geographi-
cal area to be modeled) by retaining relevant components and
conversion stages and incorporating different levels of sim-
plifications/reductions. Overall, the resulting model involves
an inputoutput efficiency matrix that can serve as a building
block for the formulation (and solution) of multisystem-level
optimization problems. Relevant optimization problems can be
used to compute the optimal energy mix for the hub to mini-
mize operational costs or to optimize the operation of an inter-
connected system of energy hubs.
This article provides an overview of possible joint con-
trol and optimization approaches for multi-energy systems;
it also elaborates on core challenges related to the develop-
ment of distributed control and optimization algorithms that
allow different parties to retain the ability to control their
own energy assets and pursue their individual performance
and reliability objectives, while acknowledging interdepen-
dencies among energy subsystems.
Example of Transmission-Level
Modeling and Applications
Recent efforts have looked at the transmission-level infra-
structure from a multi-energy system viewpoint, with the
objective of assessing the potentials for flexible and reli-
able operation and developing innovative control approaches
that tap into the identified opportunities for flexibility. This
refers primarily to interaction between electricity and gas
networks, as well as their interaction with other energy sec-
tors and vectors, such as heating and potentially hydrogen.
Conceivably, the gas network could represent a very large
storage facility that can assist in managing any excess
renewable-based power generation. Using renewable elec-
tricity to produce hydrogen (via electrolizers, through the
january/february 2017
so-called power-to-gas process) that could be blended
with natural gas or processed into methane, this otherwise
curtailed clean energy could then be transported and stored
in the gas network for successive use (even across seasons).
This process could also contribute to the decarbonization of
the hard-to-decarbonize gas sector.
A multi-energy setting could also highlight hidden opera-
tional cross-system flexibility constraints that may arise due
to, for example, limits in the local amount of gas that can be
provided to gas turbines to follow the net load, e.g., due to a
sudden, unexpected decrease in wind power. In fact, recent
events in various countries have pointed out the importance
of considering limitations of the gas infrastructure for simul-
taneous energy supply to the electricity and heating sectors,
especially under very cold conditions. From the standpoint
of future low-carbon settings, with more and more variable
renewable energy sources as well as greater options for deliv-
ery of heat (e.g., through EHPs and/or CHP), the evolving
electricity and heat sectors could bring about further new flex-
ibility requirements in the gas network.
For instance, it might happen that there is not enough flex-
ibility in the gas infrastructure (in the form of line pack, i.e.,
gas stored in the pipelines, especially in the presence of local
bottlenecks) to deliver fuel to gas-fired power plants to pro-
vide balancing and reserve power in the case of uncertain and
fast-changing renewable production. In cases like this, there
may be a need to constrain gas plants ability to follow varia-
tions in the net electrical load they see and, instead, provide
reserves by introducing intertemporal (across time scales that
consider the line-pack storage and its limitations) and inter-
sector (across gas and electricity) ramp-like constraints in the
(electrical) optimal-power-flow problem. Delivery of heat can
exacerbate this issue, because a more or less electrified heat-
ing sector may change the magnitude of the ramps in covering
the net load variations and the volume of gas (and, thus, the
line pack) in the gas network.
A challenge in this direction pertains to how to capture
this feature in mathematical terms and accommodate it into
+/
Electricity Electricity
Water Water
Natural Gas
Cooling
Hydrogen
Heating
Distributed
Heat
figure 2. A generic energy hub model capturing a variety
of inputs/outputs and conversion stages, along with storage
of different energy types.
IEEE power & energy magazine 45
 Relevant optimization problems can be used to compute the
optimal energy mix for the hub to minimize operational costs or to
optimize the operation of an interconnected system of energy hubs.
underlying optimization and control problems. Hence, the
introduction of joint optimization and integrated control for
electricity and gas networks could bring substantial benefits
by preventing the rise of such situations (see Figure 3).
On the other hand, joint optimization of electricity and
gas, also through the coordination of market activities,
could improve overall system efficiency and the utiliza-
tion of resources; this is particularly relevant in terms of
the scheduling of reserves by gas-fired generating units and
access to all the substantial flexibility available in the gas
network. In addition, there are new forms of multi-energy
flexibility that the gas network can enable through emerg-
ing technologies, including the power-to-gas processes, that
could be accounted for in the development of integrated
optimization and control strategies considering not only
potential flexibility benefits but also relevant flexibility con-
straints. These constraints may relate, in particular, to regu-
latory limits imposed on the volume of hydrogen that can be
blended in natural gas (with an upper physical bound on the
order of 20% in volume to prevent leakages, malfunction-
ing of devices, etc.) and to the fact that more hydrogen in
the network can reduce the flow capacity and lead to greater
line-pack swing.
Examples of Distribution-Level
Modeling and Applications
Departing from conventional operational settings, the
main goal at the distribution level is to leverage the fast-
acting capabilities of power-electronics-interfaced DERs,
as well as the flexibility offered by a variety of other con-
trollable assets, to enable sustainable capacity expansion,
respond to service requests precipitated by distribution
and transmission-systems operators, and achieve network-
level coordinationand so ensure reliable operation of the
whole distribution infrastructure. Forward-looking control
strategies for renewable-based DERs are complemented by
load-side optimization mechanisms, with the intention of
providing the necessary flexibility to cope with the volatil-
ity of renewable generation and provide services at the bulk
level. Additionally, (micro) CHP units as well as TES can
provide flexibility at the generation side. A coordinated
operation of various controllable energy assets can locally
supply electricity and heat (as well as cooling, if required),
while substantially reducing operational costs and environ-
mental impacts. And such coordination can offer increased
flexibility in providing electricity grid services in the form
of demand response in real or close-to-real time.
46 IEEE power & energy magazine
Flexibility in distribution systems is related to three main
factors (all of which affect the development of distributed
optimization and control strategies at various time scales):
the availability of energy storage, more typically in
forms of energy other than electricity (e.g., TES and
thermal inertia)
a significant difference in the value of time constants
between electricity and other processes such as heat-
ing or cooling
effective user flexibility in the net consumption/gen-
eration profile.
While the domestic and industrial sectors contribute equally
to the first two factors, the last option is more pronounced in the
industrial setting, where, for example, production processes can
be rearranged to achieve a given demand curve. In contrast, pro-
viding such flexibility at the residential level without impacting
customer comfort may be more challenging, and only a few and
limited options have been proposed, e.g., scheduling of appliances
or heating, ventilation, and air-conditioning (HVAC) systems.
The following sections explain some example applica-
tions of multi-energy models to suggest opportunities for
increasing flexibility at the district, neighborhood, and resi-
dential levels.
Flexibility from Distributed Multigeneration
Focusing on multi-energy districts (e.g., downtowns, industrial
areas, and neighborhoods), recent studies have demonstrated
that distributed multigeneration (DMG) plants can offer key
advantages by integrating complementary technologies such
as CHP units, EHPs, and TES. In fact, they can locally supply
electricity and heat (as well as cooling, if required), while sub-
stantially reducing operational costs, thus offering enhanced
flexibility in provisioning the electricity grid.
An example of a general electricity-and-heat DMG struc-
ture is illustrated in Figure 4, where seven operational con-
figurations are possible:
1) auxiliary boiler (AB), which serves as a reference case
2) EHP
3) EHP + TES
4) CHP
5) CHP + TES
6) CHP + EHP
7) CHP + EHP + TES.
Numerical tests performed with this DMG setting indicate that
the operational cost is expected to significantly decrease when
TES is added to both CHP-only and EHP-only settings, as well as
by operating CHP and EHP synergistically. Case 7) dramatically
january/february 2017
 Actual Wind Output
Domestic Heat Demand Industrial Heat Demand
Curtailed Wind Generation
Commercial Heat Demand Domestic Hot Water
Wind Generation Capacity
350
300 Electrical Network Model 35
250 30
200
25

january/february 2017
150

System Heat
100 20

Demand (GW)
50 15
0 10
0600 1200 1800 2400
Time 5

Wind Generation (GW)

Regional Electricity and Heat 0
08 Mar. 15 Mar. 22 Mar. 29 Mar.
Demand Model +
Heating Technology Scenario Eimp Renewables Curtailment
EDS
EEHP Integrated
ED
Electricity-H2-Gas
TES HO Model
EHP
Hs
HEHP

HD,ED
Consumers
Hydrogen and
Synthetic Natural
Gas from Excess
Renewable
Integrated Electricity-
Electricity
Heat OPF
(Power to Gas)
15
Solar Generation Capability
10
5 Use Gas Network as
a Means to Store
0 and Transport
Wind Generation Capability
40 H2-Gas Blend

20
Eexp Eimp

Generation Capability (GW)

0 Use Clean Gas
EDS May Aug. Nov. Feb.
(1-)Echp Echp EEHP for Heating or
ED Date Electrical
Fchp H
CHP EHP EHP Renewable and Conventional Generation Model Generation
h e, h t H HD
chp Gas Network Model
Faux
Boiler Haux TES

HD,ED
Hs

Consumers
haux

IEEE power & energy magazine

figure 3. An illustration of the methodology for assessing integrated electricity-heat-gas-hydrogen multi-energy systems, with specific application to Great Britains
electricity-gas transmission networks. HD: heat demand; ED: electric demand; EDS: electric distribution system.

47

Thermal
Heat Demand
Gas
Electricity
Electricity Market
EHP
Electricity (Net)
Demand
CHP
figure 4. An example of a general electricity-and-heat
DMG structure in an integrated electricity-heat-gas
market setup.
decreases the operational costs when the DMG responds to
time-varying market prices and/or regulating signals.
Further, when grid services are requested, the DMG unit in
Figure 4 can adjust the power provided to or withdrawn from
the grid around determined set points, e.g., during the hour- or
day-ahead planning phase, by ramping down/switching off the
EHP and/or ramping up/switching on the CHP plant, among
other possibilities. Of course, different DMG components can be
controlled in real time, based on the underlying time scales of the
regulating commands. The so-called profitability map provides a
way to assess the tradeoff between the operational loss incurred
by deviating the operating points from the optimal ones and the
economic benefits achieved by providing services to the grid.
A qualitative example is provided in Figure 5. The cost to
provide flexibility to the grid generally increases monotonically
Electricity
8 Shifting
7 Energy Cost Variation Potential
Costs and Benefits ( )
with the respect to the amount of flexibility provided (indicated
in the figure as reduced electricity input from the grid). It can
be seen that a nonprofitable region exists, where demand
Energy
Flow
response incentives are not sufficient to make up for the extra
costs in moving from optimal set points to provide flexibility.
TES Flexibility from Residential Neighborhoods
and Commercial Buildings
Buildings and homes are basic, spatially defined examples
AB
of multi-energy hubs, given the heterogeneous setting that
supplies essential electricity, thermal, and manufacturing
needs. At the commercial level, HVAC units are prospec-
tive candidates for providing services to the power grid at
various time scales, especially because of the favorable flex-
ibility offered by thermal inertia in buildings. In particular,
control strategies for fans in air-handing units of commercial
buildings can be designed to provide fast time-scale regula-
tion services to the distribution grid, while simultaneously
minimizing the thermal discomfort of building occupants.
At a slower time scale, and taking advantage of the build-
ings thermal inertia, optimization strategies for retail offices
can, for instance, precool during low-power-demand periods
to contribute to power peak-shaving efforts in the summer.
Load-control mechanisms in residential neighborhoods can
enable end users to provide services to the power grid by
offering more favorable tariffs as well as economic incentives
to respond to system-driven signals. As previously mentioned,
load-control mechanisms can be integrated with distributed
and decentralized control strategies for renewable sources of
energy so that the aggregate net power consumption of a num-
ber of end customers can follow regulating commands dis-
patched by electrical distribution systems operators.
Flexibility from Data Centers
A compelling example of flexible multi-energy systems in a
smart city context is suggested by the growing presence of
differently sized data centers. Data centers can be described
0.14 as energy hubs that are characterized by
DR Incentive ( /kWhel)

6
Demand Response (DR) 0.12 local generation (more and more data centers are
Benefits equipped with their own local generation unit)
5 0.10
electrical energy storage for reliability purposes,
4 0.08
which can also offer flexibility products
3 0.06
thermal cooling processes
2 0.04
a flexible load, provided by the possibility of schedul-
1 Non-Profitable 0.02
Region ing computational power.
0 For this reason, recent research has been moving toward the
0 10 20 30 40 50
Reduced Electricity Input from Grid (kWhel)
creation of efficient but also flexible data centers that will
play a critical role in providing energy services in a general
smart multi-energy system context.
figure 5. An example of profitability map (in monetary
units ) of the flexibility provided by a DMG plant. The thick
piecewise linear curve represents the energy cost increase Flexibility from Joint Water-Power
when moving away from the optimal set points for a given Optimization and Control
load at a given time to provide flexibility to the grid; the Operators of municipal water systems (MWSs) and waste-
dashed lines represents the potential benefits, parameterized water systems (WWSs) have the core objective of providing
with respect to different demandresponse incentives. clean water and treated wastewater, according to well-defined

48 IEEE power & energy magazine january/february 2017

water-quality requirements. MWSs
and WWSs are operated in a price-
taking setting, with tasks such as
pump scheduling aimed at minimiz-
ing operation costs based on the price Min f (Water) Min g (Power) Min h (Gas)
s.t Water s.t Power s.t Gas
of electricity. On the other hand, the Constraints Constraints Constraints
power consumption of MWSs and
WWSs is used as an input to the opti-
mal power-flow problems utilized by
Adjacent
power-systems operators. Systems Solar
However, it is increasingly rec- Industrial
Area
ognized that a synergistic control
of power and water systems may H1 Residential
H2
bring significant benefits from Area
operational, reliability, and societal Pumped Hydro
standpoints. In particular, control- (Storage)
lable assets in MWSs and WWSs
can provide valuable services to the
power grid at multiple spatial scales
to enhance reliability and efficiency,
as well as to cope with the volatility Commercial H3 H4 Wind
Area
of distributed renewable-based gen-
eration; services include frequency Electricity Heating Natural Gas Water
regulation, regulating reserves, or
even contingency reserves. figure 6. Traditional operations, where optimization and control tasks are typically
On the other hand, the incen- used to operate power, natural gas, water, and district heating systems in an
tives for provisioning grid services independent and decoupled way. One systems demand is used as a fixed input for
to electric utilities could be used by relevant optimization and control problems of the other systems. s.t: subject to.
water-system operators for capital
improvements and capacity expan-
sion, while operational savings emerging from joint optimiza- power networks to compute the optimal steady-state set
tion and control would lower costs for water-utility customers points of (renewable) generation units, controllable loads, and
while meeting stringent water-quality standards. Pumping in an storage devices; optimal pump-scheduling and water-flow
MWS accounts for the majority of the power consumption; a problems; and gas load-flow problems.
joint water-pump scheduling and power-flow task could, there- However, these optimization and control strategies
fore, be used to provide optimal regulating and contingency are typically used to operate power, natural gas, water, and
reserves to the power grid while maximizing the economic district heating systems in an independent and decoupled
benefits to MWS operators. way (see Figure 6). Grounded on the understanding that joint
Overall, the envisioned control architecture would enable optimization and control of multi-energy systems enables sig-
a seamless system-level coordination of controllable assets at nificant benefits from socioeconomic, flexibility, operational
multiple temporal scales to enable flexible and efficient oper- efficiency, and environmental perspectives (see the examples
ation of the multi-energy infrastructure, while systematically given earlier), the objective of recent research is to formulate
addressing customer needs and well-defined performance (and solve) global optimization problems where a variety of
objectives as set forth by system operators. Core challenges performance objectives and (economic) indicators that pertain
in this direction are outlined in the following section. to single-energy and multi-energy providers as well as end
customers are optimized, while intrinsic interdependencies
Technical Challenges in the among systems and operational constraints are acknowledged
Optimization and Control (see Figure 7).
of Integrated Multi-Energy Systems Inheriting the characteristics of, e.g., ac optimal power
Core optimization tasks in the domains of power, water, ther- flow, water flow, and gas load-flow settings, the resulting
mal, and gas system operations and control enable operators multi-energy optimization problems are, unfortunately, hard
to compute the set points of controllable assets that are op- to solve for global or local optimization in a computation-
timal in a well-defined sense, while concurrently satisfy- ally efficient manner. Problem complexity is primarily due
ing operational, quality-of-service, and security constraints. to the nonlinear equations that govern the underlying phys-
Examples include optimal power flow-type problems for ics of power, water, heat, and gas networks and the curse of

january/february 2017 IEEE power & energy magazine 49

 dimensionality associated with joint decision-making across
different systems at multiple spatiotemporal scales.
Nonlinearity leads to nonconvex problems (i.e., with multi-
ple locally optimal solutions and a globally optimal solution
difficult to find and assess simultaneously) and, oftentimes,
nondeterministic polynomial-time-hard (NP-hard) pro-
grams that may be computationally prohibitive to solve even
for local optimization. Nonconvexity stems from the under-
lying physics governing water, power, and gas flows as well
as heat-transfer models describing the systems couplings
and energy conversion stages. Further nonconvexities may
arise from the optimization approach or structure, such as
in the original energy-hub formulation that, by aggregating
elements that contain decision variables in a synthetic input-
output representation, generates a nonlinear optimization
problem structure, where decision variables may serve to
multiply one another.
along the pipes, as well as the hydraulic characteristics of
variable-speed pumps. Computationally intensive problems
in gas (transmission) networks pertain to finding solutions
for highly nonlinear gas-flow equations in both steady-state
and transient (where intertemporal couplings add to the non-
linearity complexity) forms.
Discrete Control Decisions
The binary and discrete decision variables required to con-
sider on/off decisions and discrete power consumption levels
(relevant to appliances, HVAC units, water heaters, EHPs,
and water pumps, to mention just a few) exacerbate the prob-
lem complexity. The dimension of the search space increases
exponentially with the number of binary variables (especially
if binary decisions are required over a prolonged optimiza-
tion horizon), thus rendering solutions for the underlying
optimization problems computationally prohibitive.

Nonconvexity from Flow Physics Nonconvexity Due to Coupling Factors

The power-balance constraints in power systems are non- For modeling simplicity, the coupling factors among energy
convex due to nonlinear ac power-flow equations; addi- carriers are often assumed to be constant. This is the case for
tional sources of nonconvexity include constraints on the the fuel-to-power and fuel-to-heat conversion efficiencies of
minimum voltage service levels and on branch thermal CHP units, for example. However, a number of coupling fac-
limits applied to power flows. In pump-scheduling prob- tors are, in fact, nonlinear; examples include the efficiency
lems and water-flow problems, nonconvex constraints natu- and power consumed by a variable-speed water pump, which
rally emerge from the relevant flow mathematical models are nonlinear functions of the pumps frequency. Further, the
to capture the head losses or pressure losses due to friction operational region of some types of co-generation units or
absorption and compression chill-
ers may be nonconvex.
Min f (Water) + g (Power) + h (Gas)
s.t Water Constraints Tradeoff Between
Power Constraints Complexity and Flexibility
Gas Constraints
Energy hubs and multigeneration
+ Coupling Constraints
models based on so-called dispatch
Setpoints factors, which are used to specify
Adjacent how energy is split among conver-
Systems Solar sion units and outputs, may intro-
Industrial
Area duce other sources of nonconvexity
via bilinear or trilinear terms ap-
H1 Residential
H2 Area
pearing in equality constraints.
Nonlinearity could be, in this case,
Pumped Hydro bypassed either by introducing aux-
(Storage)
iliary variables or by simplify the
hub model. It is, thus, apparent that
tradeoffs between the complexity
of any modeling approach to the
Commercial H3 H4 optimization problems and achiev-
Wind
Area able flexibility must be taken into
account in the systems design and
Electricity Heating Natural Gas Water
operational processes.
figure 7. A global control problem optimizes a variety of performance objectives
and economic indicators, while acknowledging intrinsic interdependencies among Opportunities
systems and operational constraints. Flexibility is naturally enabled by such a prob- Off-the-shelf solvers for mixed-
lem formulation, since the coupling constraints capture shifts in supply and demand integer nonlinear programs could,
across energy vectors and networks. in principle, be used to find solutions

50 IEEE power & energy magazine january/february 2017

for optimization problems where the operation of power, to problem formulations where multi-energy systems are jointly
water, thermal, and gas systems is jointly optimized. Given optimized is largely unexplored. Ample research opportunities ex-
the underlying nonconvexity and NP-hardness, suboptimal ist for how to leverage advances in convex relaxation and approxi-
solutions are largely accepted (e.g., multistart techniques can be mation to develop computationally efficient solution methods for
used to possibly improve the quality of the solution and estimate multi-energy system optimization problems.
its attraction basin). It is, however, worth pointing out that the Key research issues involve the development of mecha-
computational complexity becomes rapidly prohibitive with nisms that guarantee that the relaxed or approximated convex
increasing of the system size and the optimization horizon, problem yield feasible operational set points and the deriva-
especially in the presence of various forms of energy storage; tions of conditions under which the solution of the convex
consequently, underlying optimization tasks for integrated multi- surrogate retains locally or globally optimized properties.
energy systems may not offer decision-making capabilities at Successfully proven convex relaxation and approximation
appropriate time scales and may not be adequate for operational techniques for nonlinear models can then be leveraged to
landscapes where the rate of set-point updates has to match fast derive mixed-integer convex surrogates for challenging non-
dynamics of system and ambient conditions. In the context of convex mixed-integer nonlinear programs.
future deregulated multi-energy markets, nonconvexity inher-
ently challenges the development of provably convergent dis- Distributed Control
tributed solvers utilized to strategically decompose the decision- Different energy actors may own and operate different
making process across energy providers, users, and devices. energy assets. In lieu of centralized problem formulations
Accordingly, recent research has focused on developing that require one single authority to supervise and control the
computationally affordable solution methods by leverag- overall multi-energy infrastructure, powerful decomposition
ing advances in convex relaxations and approximation methodologies can be leveraged to strategically decouple
of nonlinear (nonconvex) flow constraints. In the power- the solution of multi-energy system optimization problems
systems context, convex relaxation techniques enable a solu- across actors. Distributed strategies allow different parties
tion of the ac optimal power-flow
task with reduced computational
Communication Links
burden at both transmission and
distribution scales, while possibly
identifying globally optimal pow-
er-flow solutions. Sufficient condi-
Min f (water) Min g (Power) Min h (gas)
tions for tightness of semidefinite s.t Water Constraints s.t Power Constraints s.t Gas Constraints
programming and second-order + Consensus + Consensus + Consensus
Constraints Constraints Constraints
code programming relaxations
are available for some classes of
system topologies and problem
Adjacent
setups, while the efficacy of these Systems Solar
methods for general topologies Industrial
Area
can be demonstrated by numerical
evidence. Further, powerful linear Residential
H1 H2
approximation methods enable Area
one to develop linear and quadrat- Pumped Hydro
ic programming surrogates of the (Storage)
ac optimal power-low problem,
the computational complexity of
which scales more favorably with
system size. Considering other sec-
tors, relaxations have, for instance, Commercial H3 H4
Wind
been recently proposed for water- Area
flow problems based on, e.g., a
Electricity Heating Natural Gas Water
second-order cone relaxation of the
relevant flow equations.
figure 8. Distributed optimization and control methods are leveraged to strategi-
Although the virtues of convex cally decompose the solution of the global optimization problem across energy hubs
optimization tools have been dem- (H ), networks, and operators. Each energy operator retains the ability to control
i
onstrated for specific problems its system, while global coordination is achieved to increase operational flexibility.
within the power-, water-, and gas-en- Online optimization tools can also be leveraged to offer decision-making capabilities
gineering domains, their application at appropriate time scales.

january/february 2017 IEEE power & energy magazine 51

 to retain the ability to control their own energy assets and
pursue their business, performance, and reliability objec-
tives, while acknowledging interdependencies among energy
subsystems. Although decomposition methods can conceiv-
ably be applied to nonconvex programs, convex surrogates
of the power-water-thermal-gas flow equations and convex
relaxations of binary constraints typically facilitate the
development of distributed schemes with improved conver-
gence properties.
Identifying constraints that couple energy assets and sub-
networks managed by different entities is key to achiev-
ing these goals. For example, water-power coupling con-
straints pertain to the power/speed of pumps; coupling among
energy hubs may be via power lines, district heating/cool-
ing pipes, and gas pipes connecting the hubs. After formulating
partial Lagrangian functions based on the coupling constraints,
a variety of techniques (including primal-dual-gradient-type
methods and the alternating direction method of multipli-
ers) can be leveraged to derive a relevant distributed optimiza-
tion procedure.
In particular, these methods allow one to develop algo-
rithms for cases where a) different systems are managed by
different operators, as illustrated in Figure 8; b) energy hubs
are owned and controlled by different multi-energy actors; c)
commercial and residential customers retain control of their
own assets and participate in the provision of different power
system services (optimization task); and d) a combination of
a)c). In all these configurations, the entities partaking of the
relevant optimization task retain control of their assets and
pursue their specific operational objectives. However, by
exchanging relevant optimization variables, they will achieve
the solution of the global optimization problem illustrated in
Figure 7, which naturally encapsulates multiflexibility options.
It is worth pointing out that distributed solutions involve
iterative schemes where the set points of the energy assets
are dispatched by each energy actor only upon convergence
of the algorithms. For fast time-varying operational land-
scapes, it is more desirable to resort to online optimization
schemes, where the set points of the devices are dispatched
as and when available, without necessarily waiting for the
distributed algorithm to converge. A challenge in this direc-
tion is to ensure that the set points produced by the online
algorithm do not induce violations of operational and secu-
rity limits. A cross-fertilization of control and online opti-
mization tools is, therefore, the key to enable the synthesis
of distributed feedback control schemes that ensure satis-
faction of physical and security limits while tracking solu-
tions of underlying multi-energy optimization problems.
Conclusions
Coordinated control of multi-energy systems at multiple
spatiotemporal scales promises significant benefits from
socioeconomic, operational efficiency, and environmental per-
spectives by leveraging the flexibility of various controllable
52 IEEE power & energy magazine
assets to ensure a secure and economical supplydemand bal-
ance and provide reserve services. To enable such a level of
coordination at appropriate time scales, one possible approach
consists in formulating global optimization problems where
various performance objectives and (economic) indicators that
pertain to single-energy and multi-energy providers as well as
end customers are maximized and, subsequently, in leveraging
relevant optimization and control tools to develop computa-
tionally affordable distributed (and online) algorithms.
At a slower time scale, distributed algorithms enable
energy hubs, network operators, and (possibly) customers
to coordinate to achieve solutions of system-level dispatch
problems; at a faster time scale, control algorithms strategi-
cally decompose the decision-making process across actors,
while offering fast time-scale flexibility options and steer-
ing the operating points of multi-energy systems toward the
solution of global optimization problems. To this end, it is
critical to represent multi-energy system and network mod-
els as relevant optimization and control tasks and uncover
intrinsic convexity structures that lead to computationally
efficient distributed solutions.
For Further Reading
P. Mancarella, G. Andersson, J. A. Peas-Lopes, and K. R.
W. Bell, Modeling of integrated multi-energy systems:
Drivers, requirements, and opportunities, in Proc. 19th
Power Systems Computation Conf., Genova, Italy, 2016, pp.
122.
M. Geidl, G. Koeppel, P. Favre-Perrod, B. Klockl, G. An-
dersson, and K. Frohlich, Energy hubs for the future, IEEE
Power Energy Mag., vol. 5, no. 1, pp. 2430, Jan./Feb. 2007.
S. Clegg and P. Mancarella, Integrated electrical and
gas network flexibility assessment in low-carbon multi-en-
ergy systems, IEEE Trans. Sustain. Energy, vol. 7, no. 2, pp.
718731, Apr. 2016.
P. Mancarella and G. Chicco, Real-time demand response
from energy shifting in distributed multi-generation, IEEE
Trans. Smart Grid, vol. 4, no. 4, pp. 19281938, Dec. 2013.
D. Mller, A. Monti, S. Stinner, T. Schlsser, T. Schtz,
P. Matthes, H. Wolisz, C. Molitor, H. Harb, and R. Streblow,
Demand side management for city districts, Build. Envi-
ron., vol. 91, pp. 283293, Sept. 2015.
E. DallAnese and A. Simonetto, Optimal power flow
pursuit, IEEE Trans. Smart Grid, May 2016.
Biographies
Emiliano DallAnese is with the National Renewable En-
ergy Laboratory, Golden, Colorado, United States.
Pierluigi Mancarella is with the University of Mel-
bourne, Australia, and the University of Manchester, United
Kingdom.
Antonello Monti is with RWTH Aachen University,
Germany.
p&e
january/february 2017
 The Consumers
Role in Flexible
Energy Systems
An Interdisciplinary
Approach to Changing
Consumers Behavior

By Geertje Schuitema,
Lisa Ryan, and
Claudia Aravena

A
A TRANSITION TO FOSSIL-FREE ENERGY
systems is necessary to secure a safe, reliable, and
sustainable future. This implies increasing shares
of renewable energy sources, such as solar and
wind, and introduces new challenges in terms of
flexibility, storage, and energy transmission. Con-
sumers play a crucial role in achieving this energy
transition, as consumer flexibility is required to
accommodate variable generation and peak loads.
This implies that consumers become more flexible
in their energy use and adopt technologies that facil-
itate greater reliance on renewable energy sources.
Increasing consumers flexibility is complex,
and the solution lies in a combination of many disci-
plines, such as psychology, marketing, economics,
anthropology, computer science, and engineering.

Digital Object Identifier 10.1109/MPE.2016.2620658

Date of publication: 2 February 2017
ISTOCKPHOTO.COM/MACROVECTOR

january/february 2017 1540-7977/172017IEEE IEEE power & energy magazine 53

 Despite the importance of consumers in a more flexible energy
system, only a small part of current energy research has so far
focused on this issue; moreover, most of this research is con-
ducted in monodisciplinary settings. Yet consumers flexibility
involves many different elements that are best represented and
investigated using an interdisciplinary approach.
This article focuses on the role of consumers in a flexible
energy system from an interdisciplinary point of view.
What is the consumers role in a flexible energy system?
How can consumers behavior be changed?
What barriers are there for change?
What influences consumers investment in flexibility-
enhancing technologies?
How can policies and market design be used to stimu-
late flexible behavior?
In addition, we briefly discuss the public acceptance of
energy infrastructure, an important precondition for any
flexible energy system. We conclude with some remarks on
how research on consumer behavior could be better inte-
grated into energy research more generally.
The Consumers Role in
a Flexible Energy System
An increase in variable renewables (e.g., wind and solar
photovoltaic) in the energy system means that energy sup-
ply fluctuates more; without flexible demand, supply will
most likely require additional storage and capacity in the
system. However, consumers can improve the flexibility of
the energy system by playing a more active role in both the
demand for and supply of energy.
When talking about consumers energy-using behavior,
researchers distinguish between curtailment behavior and
efficiency behavior; both facilitate the flexibility of the energy
system. Curtailment behavior refers to energy-using activities
that typically occur on a frequent basissuch as showering,
changing thermostat settings, and switching on/off lights or
appliancesand is therefore much more linked to habitual
and occupancy behaviors. To enhance the flexibility of the
energy system, changing consumers curtailment behavior is
mainly linked to load shifting and energy-demand reduction.
Efficiency behavior refers to investment in energy-efficient
solutions, such as insulation, heat pumps, and electric vehi-
cles, and it includes the adoption of technologies that facili-
tate curtailment behavior change, such as smart meters and
smart appliances. Efficiency behavior involves the adoption of
new technologies and investments leading to energy-efficient
actions and measures that will have a longer-lasting impact.
Consumers also increasingly play a role in improving the
flexibility of the energy supply. This is why they are some-
times referred to as prosumers, which means that they are both
consumers and producers, or active agents on the supply
side. This new role involves new types of relevant consumer
behaviors: generation of ones own energy, for example. Pro-
sumers can provide flexibility to centralized energy generation
through their decisions on whether to sell their energy back to
54 IEEE power & energy magazine
the grid, store it, or consume it themselves (self-consumption).
Flexibility is also provided when consumers allow automated
control over their appliances, implying that third parties align
their energy use with the supply and demand on the grid.
The increasing share of variable renewables, such as solar
and wind energy, associated with the requirement for flexi-
bility also introduces new challenges in terms of major infra-
structure changes such as the building of wind farms, pylons,
and transmission lines. A major factor, and often a barrier, in
these transitions is public acceptance. Hence, a discussion
of the contribution of consumers to flexibility would not be
complete without reference to the issue of public acceptance,
so this topic is included here as well.
Changing Consumers Energy Behavior:
Load Shift and Demand Reduction
Demand Response
Consumer flexibility is needed in terms of both shifting
energy demand to times of the day when (renewable) energy
is available and also reducing energy demand when the sup-
ply of energy is limited. Demand-side management (DSM)
is the term used to describe a range of measures for improv-
ing the efficiency and flexibility of energy demand from the
consumer side. Demand-response (DR) measures are a part
of DSM programs; they are designed to encourage consum-
ers to change their energy consumption. Some of these DR
programs rely on changes in curtailment behaviorthat is,
consumers need to change their behavior based on alerts
or real-time information. This is problematic because it
requires considerable effort on the part of consumers: they
need to pay attention to the information, decide on appropri-
ate actions, and ultimately undertake those actions. A basic
assumption here is that consumers have the knowledge to
take appropriate actions, which often may not be the case.
In fact, there have been reports of tension in households as a
result of members having different views about how best to
change their energy consumption. As a result, the effects of
DR programs on consumers load shifting and energy con-
sumption fluctuate and may be fairly limited.
To overcome these problems, future DR programs are likely
to be designed to include a significant share of direct load
control (DLC) by energy providers. DLC allows appliances
whose time of use is not critical within reasonably narrow time
periods (such as refrigerators, air conditioners, water heaters,
and pumps) to be switched off or have their energy reduced
remotely by suppliers at times of peak demand. Although DLC
is likely to increase load shifting and reduce energy demand,
consumers are often reluctant to accept automated control
over their energy use. Surveys show that fewer than a third of
consumers use time-of-use (ToU) controlling functions or are
willing to accept contractual arrangements that allow utilities
to directly control appliances to deliver load shedding. This
reluctance is very much based on the fear of losing control
over ones own energy consumption, use of appliances, and
january/february 2017
basic comfort. Figure 1 shows how consumers acceptance of
remote control of their appliances energy use increases if they Your Electricity Provider Can Remotely Limit the Use of
have the ability to override this outside control and thus feel Certain Major Home Appliances and You Cannot Choose
to Reverse This Course of Action
personal control over the decision making.
Overall, consumers perceive the risks related to new tech- Your Electricity Provider Can Remotely Limit the Use of
Certain Major Home Appliances and You Cannot Choose
nologies and DLC to be much higher than any benefits they
to Reverse This Course of Action
may provide. And, while a positive experience with one smart
100
technology may increase a consumers acceptance levels of 90

(Certainty + Probably)
smart technologies generally, this is unlikely to occur if the 80
consumers experience is frustrating due to unrealistic expecta- 70

% of Sign Up
60
tions or a lack of policy and market support.

29%
50
The characteristics of energy use partly depend on 40

19%
household structures and socio-demographics. For exam- 30
20
ple, energy consumption increases with the number of 10
inhabitants in the house, the age of the head of the house-

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
hold, and the number of unemployed household members. In
% of Savings in Energy Bill
addition, homeowners usually have higher energy consump-
tion as compared to renters. The extent to which interven-
figure 1. The influence of consumers sense of personal
tions are effective to change consumers behavior depends, control on their likelihood of accepting programs allowing
to some extent, on such household characteristics. utilities to control their appliances energy use. (Data derived
from Accenture, Understanding Consumer Preferences in
Time-of-Use Tariffs Energy Efficiency, 2010.)
ToU tariffs are price instruments
that support consumer flexibility
by realigning price signals in favor
of more flexible energy use. ToU
tariffs are aimed at encouraging
energy use during off-peak hours
Midnight
11 p.

(when energy supply is high and

nt te

m.
h
/kW
Ce Ra

10

1 a.

.
demand is low) by setting lower
.m
m.
14 Day

p.m

9
2a

Nig ent/k
.
24
energy prices compared to high- p. m
m a.
.

C
ht
. 3
peak hours. The overall objective is
Ra h
8 p. .
m. .m
te
to shift energy demand from high- 4a W
to low-peak hours, rather than to 7 p.m m.
. 5 a.
20 Cent/kWh

reduce overall energy use. ToU

Peak Rate*

tariffs are designed to better reflect 6 p.m. 6 a.m.

the true energy generation costs
at a certain time of the day. How- . 7 a.
5 p.m m.
ever, this also implies that energy
8a
prices will vary more, which may . m. .m
4p .
cause uncertainty among consum-
.

9
m

a.
p.

ers about how much energy will m

10
3

.
.

11 a.m.
.m

cost. Generally, consumers prefer

Noon

a.m
1 p.m
2p

certain outcomes over uncertain

.

outcomes; hence, they may ini- Day

tially object to ToU tariffs. How-
ever, in the longer term, if average
energy costs go down, resistance
may decrease.
ToU tariffs have already been-
put in place in various countries,
such as France, Italy and Sweden.
Figure 2 represents an example
of ToU tariffs designs for a smart
meter trial in Ireland. Studies
14 C Rate
ent/
kWh
* Peak Rates Applied Moday to Friday Only, Excluding Public Holidays
figure 2. An example of ToU tariffs designs for a smart meter trial in Ireland that ran
1 January31 December 2010. Note that all prices excluded VAT. (Data derived from:
CER, Electricity Smart Metering Customer Behaviour Trials (CBT) Findings Report.
Dublin: Commission for Energy Regulation, 2011.)

january/february 2017 IEEE power & energy magazine 55

 showed that the introduction of ToU tariffs could shift the displays (IHDs) but also more static measures such as labels
energy consumption at peak times to off-peak hours by 8%. relating to the energy performance of buildings and a host
However, average consumption was found to reduce little of energy-using appliances. With or without IHDs, smart
or remain the same. If energy efficiency is also a goal, ToU meters facilitate large-scale flexibility by allowing control of
tariffs should be combined with the provision of informa- electricity use remotely.
tion on energy consumption and costs (e.g., feedback infor- Smart meters provide consumers with real-time informa-
mation through paper bills and technology displays or tion on their energy consumption via IHDs, websites, or
smart meters). mobile apps. This is an advantage over traditional commu-
nication methods, where more generic information was pro-
Information and Feedback vided via billing or other (paper-based) information. With
Another category of instruments or interventions relates smart meters, utilities can tailor the information provided to
to the provision of feedback and information. This can a particular costumer in real time, which is considered a great
have several objectives. First, information policies may be advantage for both the consumer and the utility. This infor-
required to increase knowledge and raise awareness among mation could be related to individual energy consumption,
consumers as to the importance of their energy decisions. prices, forecasts of future energy use, generation sources,
Many governments have introduced mandates to provide and payments. Targeted information to the individual user
information to consumers on their energy use. This includes has been shown to be more effective in changing behavior
feedback technologies such as smart meters with in-house than generic information. Providing tailored information in
real-time is important to avoid over-
load or redundancy, which would be
counterproductive in increasing con-
What Organizations Do You Trust to Inform You About Actions You Can Take sumers knowledge and awareness of
to Optimize Your Electricity Consumption? their energy consumption.
Do Not Trust Neither Trust Nor Distrust Trust Although the provision of infor-
mation can be effective in making
13 13
20 28 consumers energy behavior more
29
49 flexible, the effects are fairly small
51 53
(in the range of 515%) and tend to
59 60 be short term. It has been suggested
66 48 46 that to create a long-term learning
effect, consumers need help in inter-
43 42 36 preting their energy consumption
and frequent reminders, which can
28 27 26 25
14 11 be accomplished by adding infor-
8 7
mation on the consequences of their
Home Service Provides**

Retainers/Equipment Manufactures

Online Service Providers**

Utilities/ Electricity Provides

Consumer Associations

Academics/Schools/Scientific Associations

Environmental Associations
Government/Governmental Organizations

consumption (e.g., financial conse-

quences or carbon footprint). Also,
combining energy consumption with
feedback about ones own historical
energy use, others energy use (e.g.,
neighbors or similar consumers)
can be effective. The electric utility
OPOWER in the United States, for
example, has had success using this
technique with their consumers.
An important note: it is crucial
that consumers trust both the infor-
mation and the provider of it if this
*e.g., Google, Microsoft technique is to have any effect. Fig-
**e.g., Cable Television Providers, Telecommunication Providers, Home ure 3 shows that the same informa-
Security Companies
tion may be trusted when it comes
from one source but not when it
figure 3. Consumers level of trust in sources that provide information on actions comes from another. That is, infor-
they can take to optimize their electricity consumption. (Data derived from Accen- mation on how to optimize ones
ture, Understanding Consumer Preferences in Energy Efficiency, 2010.) electricity consumption is deemed

56 IEEE power & energy magazine january/february 2017

more trustworthy when it comes from sources that are seen
as objective in the protection of consumers interests (e.g.,
environmental organizations, scientists, or consumer orga-
nizations) than when it comes from sources that are not
typically seen to have consumers interest as a main prior-
ity (e.g., industry or government).
It is well known that, when the provision of information
enhances transparency on energy use and utility billing, con-
sumers are more likely to trust the information and act upon
it. However, consumers trust is not guaranteed just because
information is provided, and good intentions may backfire.
Fear of infringement on their freedom or privacy can lead to
strong objections by consumers to engaging in any activity
that may lead to a more flexible energy system. It is impor-
tant to acknowledge and address such responses, as they
can seriously undermine the effect of any program aimed at
changing consumers energy behavior toward enhancing a
more flexible energy system.
Other Drivers of Behavioral Change
Economists see economic and information policy instru-
ments as means to overcome market failures such as the pres-
ence of externalities, split incentives, and transaction costs
associated with investment in and use of smart appliances.
However, from a psychological point of view, some barriers
to flexibility arise because consumers do not always behave
rationally; therefore, additional measures may be required.
For example, DR and DLC may lead to lower energy bills
for consumers, which is a rational reason for them to change
their behavior. However, consumers may feel that they need
to invest considerable effort and give up control for this flex-
ibility, and, as a result, financial compensation may simply
not be sufficient to motivate this behavioral change. Stronger
price signals may overcome this reluctance to some extent
but not fully, as consumers tend to feel uncertain about
whether they will benefit and for how long.
Consumers respond not only to economic benefits; there
are also other motivations underlying their behavior. For
example, in addition to their own benefits, consumers may
also consider the collective consequences, such as the impor-
tance of having a secure, sustainable, and affordable energy
system for all. If consumers understand how their behavior
can contribute to this, they are more likely to engage, despite
little financial gain, because they want to contribute to society.
Also, it is important not to overestimate consumers
interest level in the topic of energy. For most, the services
provided by energy including watching television, doing laun-
dry, or being warm are more important than energy and
energy demand in themselves. Nudging is another strategy
to change consumers behavior toward becoming more flex-
ible. This strategy relies on using consumers automatic gut
feeling, rather than on their consciously considered behavior.
Nudging is based on the idea that the context is changed so
that, when people follow their gut feeling, they automatically
choose (and thus are nudged toward) the desired option. This
january/february 2017
100
% People Saying yes to Smart
Meter with Automated Control
90 Opt-In Opt-Out
80
70
60
50
40
30
20
10
0
Denmark Norway Switzerland
Note: In the original study, a third group was included.
See Broman-Toft et al. for more information.
figure 4. The effect of different default options on con-
sumers adoption of smart meters with remote control.
(Source: M. Broman Toft, G. Schuitema, and J.Thgersen,
The importance of framing for consumer acceptance of
the Smart Grid: A comparative study of Denmark, Norway,
and Switzerland, Energy Research & Social Science, no. 3,
2014, pp. 113123.)
has been accomplished, for example, by suggesting a differ-
ent default option to consumers when choosing between gray
and green energy or about whether or not to have a smart
meter with automated control (Figure 4).
In the second example, after receiving general information
about the smart meter, one group (opt-in) of participants could
check a box indicating Yes, I would like to have a smart meter
with remote control installed in my house; it was made clear
that if they did not check the box, they would automatically not
get the smart meter. The other group (opt-out) could only check
a box No, I would not like to have a smart meter with remote
control installed in my house, and it was implied that if they
did not check the box, they would get the smart meter. This
study was carried out in three different European countries, and
the results show a similar pattern: in all cases, consumers in
the opt-out group were more likely to accept smart meters with
automated control compared to those presented with the opt-in
conditions. Such studies show that people tend to go along with
the choice that is presented as the default option. This implies
that when the desired option is presented as the standard
choice (e.g., green energy, smart meters with automated con-
trol), consumer uptake often increases.
Technology Adoption
and Investment Behavior
Consumers can facilitate a flexible energy system by adopt-
ing new technologies and investing in them. Such invest-
ments are usually fairly large, which is one of the barriers
to consumer investment. A range of economic instruments
exists that can support consumer investment in smart and
efficient technology that supports flexibility.
IEEE power & energy magazine 57
 when hardly anybody else has; they
are usually called innovators or early
Deployment of Smart Electricity Meters in EU Members States by 2020
adopters and are characterized by a
strong sense of innovation. For other
Yes
laggards, a large number of others
Yes, Official Decision Pending
need to have adopted a technol-
No Based On Countrys Assessment
ogy before they will consider do-
No Decision Yet
Selective ing so too.
FI
That social influence is impor-
SE
tant for the uptake of new technolo-
EE gies is observed, for example, in the
LV
installation of solar photovoltaic
DK
IE LT panels. Solar contagion is a term
UK that is used for the phenomenon
NL PL that, in neighborhoods where so-
BE DE
LU
lar panels are visible, more solar
CZ panels appear. This is probably
FR SK
AT because the visibility of solar pan-
HU
SI RO els triggers others to adopt them,
HR
which is likely to be connected to
PT ES IT the symbolic functions of technol-
BG
ogy and the desire to express ones
GR identity, for example as an inno-
CY vator or a green consumer.
In the deployment of new tech-
MT
nologies, grants and subsidies are
often proposed as a way to encour-
age or kick start the market by
figure 5. The EU smart meter rollout in 2020. (Source: Briefing European Parliament,
Smart Electricity Grids and Meters in the EU Member States, Brussels: European Parlia- spreading the adoption of technolo-
ment, Sept. 2015; image reproduced with permission from the European Parliament.) gies through social networks. Once
there is more uptake in the market,
familiarity with the technologies,
One possibility is to increase energy prices overall. If and product visibility, these subsidies should be slowly phased
energy use becomes more expensive in general, it will be out. Split incentives can be addressed by ensuring that there
more attractive for consumers to invest in energy-efficient are benefits to the agent investing in energy efficiency or that
technologies due to a higher rate of return or shorter pay- the investment costs can be passed on to the tenant or benefi-
back time. There may be distribution issues, however, where ciary of the reduced energy bills.
low-income groups are the most adversely affected. As an Another way to kick start the market is to encourage
alternative to increasing energy prices, incentives or subsi- grass-roots innovations and community initiatives to advance
dies can be provided to encourage investment in technolo- adoption of technologies collectively. Such a bottom-up
gies that enhance the flexibility of energy systems. approach is more likely to engage and involve consumers,
There are also other barriers to investment in technolo- because the initiative is their own rather than imposed on
gies. For example, the perceived effort to install and use new them via top-down procedures. Moreover, a sense of owner-
technologies influences consumers investment behavior as ship and contribution is likely to keep consumers interested
well. In addition, consumers tend to be uncertain whether and committed to the project.
their investments will pay off and whether the adoption of Finally, the benefits of community projects may be
flexibility-enhancing technologies (such as smart meters) larger because, in contrast to individual consumers, the total
will really reduce their energy bills. energy demand of communities is larger, which makes it
The theory of the diffusion of innovations suggests that more likely that they can negotiate a contract for DR and
social influence is important for the adoption of technol- ToU. This may increase the benefits for the community in
ogies: the more people who have adopted a certain technol- general, as well as for each individual member. Although
ogy, the more likely it is that others will do so as well due to a bottom-up approaches are promising, a well-documented
neighboring effect. The threshold for how large the number of challenge is their upscaling. As in most organizations, they
others must be depends on an individuals personal char- will be more efficient and powerful if they grow. However,
acteristics. Some consumers will adopt new technologies community-based and grass-roots initiatives are always

58 IEEE power & energy magazine january/february 2017

small scale and deal with local community issues. Upscal-
ing very often leads to problems in terms of the management
and generalizability of the project.
Policies to Enhance Flexibility
Government or regulatory intervention has a role to play
in supporting consumer flexibility. The elements neces-
sary to achieve more flexible energy demand from con-
sumers, such as installing smart appliances and smart meters
and negotiating contracts that include provisions for auto-
mated DR, are likely to require policy support to facilitate
their introduction.
Many governments have introduced mandates to ensure
that steps are being taken to stimulate flexibility behavior.
These can involve mandating the rollout of a technology
that facilitates consumer flexibility such as smart appli-
ances or smart meters (see Figure 5) or providing subsidies
on investments in solar panels or heat pumps. The 2009
Electricity Directive recommended the rollout of smart
meters among European Union (EU) member states but
left deployment decisions up to the individual states. When
individual member states positively assessed the costben-
efit analysis of the rollout of smart meters, at least 80%
of consumers were expected to be equipped with smart
meters by 2020. However, the progress on the rollout of
smart meters across the EU shows a mixed picture: Finland,
Italy, and Sweden are advanced in their rollouts (install-
ing 45 million meters), and another 13 states declared
their intention to proceed with rollouts, although they are
in different stages. In seven member states (Belgium, the
Czech Republic, Germany, Latvia, Lithuania, Portugal,
and Slovakia) the costbenefit analysis was negative or
inconclusive, implying that the rollout will be delayed or
not started. Four states (Bulgaria, Cyprus, Hungary, and
Slovenia) have not conducted a costbenefit analysis or
instituted any rollout plans as yet.
Alternatively, command-and-control policies may be
used to ban or discourage the use of less flexible technolo-
gies (for example, the interdiction of natural gas boilers
in buildings in Denmark since 2013 or the prohibition of
incandescent light bulbs in the EU since 2009), which may
indirectly encourage the introduction of more flexible tech-
nologies such as heat pumps (in the former case) and light-
emitting diodes (in the latter case). Finally, some mandates
have been introduced to provide information to consumers
on their energy use and bills (e.g., via smart meters with
IHDs). But more static measures have also been taken, such
as the introduction of energy labels relating to the perfor-
mance of buildings and a host of energy-using appliances.
DSM programs have been introduced in many countries
to facilitate more flexible consumer energy-using behav-
ior. While the majority of existing DSM programs are util-
ity-driven, policy interventions are required as well to 1)
encourage utilities to offer such programs and 2) facilitate
the appropriate price signals and information that make it
january/february 2017
worthwhile for a large percentage of consumers to sign up
for them, especially in jurisdictions where energy prices
are regulated.
Ultimately, the measures described here should be inte-
grated, as far as possible, in the design of energy markets to
enhance consumers flexibility. In other words, if consumer
flexibility is valued and incentivized in the market, it will
be more likely to occurfor example, through utility DR
and DSM programs or other intermediaries such as aggre-
gators. There is also increasing recognition that DR involv-
ing customers changing their operating patterns can assist in
balancing electricity systems and gives transmission system
operators more flexible options.
Consumers Acceptance of
Renewables and Energy Infrastructure
A topic that is often not considered when discussing the con-
sumers role in flexible energy systems is the public accep-
tance of renewables and energy infrastructure. However, we
believe this to be an important topic in this context. Major
infrastructure projects are often hindered or stopped by
public resistance and lack of consumer acceptance. Conse-
quently, understanding consumer acceptance of renewable
energy sources and energy infrastructure is an important
precondition for the realization of a flexible energy system.
It is often assumed that lack of support for energy infra-
structure is the result of a general resistance to change or
not-in-my-back-yard reaction. However, this view has been
criticized as simplistic and erroneous and is often used as an
excuse to ignore valid public objections. Instead, it is argued
that people do not resist change per se, but they resist change
for various, often valid, reasons such as increased costs and
exposure to risks and seeing limited or reduced benefits. The
(perceived) costs, risks, and benefits are steered by underly-
ing psychological factors such as values and lifestyles as well
as attachment to place, personal identities, symbolic mean-
ings, and emotions.
In addition, the communication process between con-
sumers and the involved authorities and industries is very
important for encouraging consumer acceptance. If consum-
ers trust the involved parties, they are much more likely to
accept the proposed infrastructure changes, especially when
they know little about them. Moreover, fair and transparent
procedures have been shown to increase acceptance levels,
thereby reducing the likelihood that lack of consumer accep-
tance will hinder the building of infrastructure needed for a
flexible energy system.
Compensation schemes may be used to strengthen public
acceptance of energy infrastructurefor example, by pro-
viding individual tax benefits or providing benefits for the
community that reduce local energy bills or create or expand
community facilities. Such communication and compensa-
tions schemes can be effective for increasing the accept-
ability of policies and energy infrastructure. However, there
may be some negative side effects if unintended messages
IEEE power & energy magazine 59
 Not only must customers become much more flexible in their
energy use, they must also be persuaded to adopt technologies
that facilitate greater reliance on renewable energy sources.
are signaled. For example, if compensating individuals or
communities financially is seen as a bribe or if an attempt
to engage the public is seen as insincere, people may not feel
that they are taken seriously.
Final Remarks
Consumers play an important role on different levels
in achieving flexibility in energy systems. In the first
instance, they can help make energy systems more flex-
ible by changing their energy consumption patterns, result-
ing in load shifting and reduction in energy demand. To
do this, the acceptance and use of smart technologies and
energy efficient measures by consumers are needed. Also,
more widely, consumers acceptance of energy infrastruc-
ture is necessary to facilitate a higher share of renewable
energy sources and transmission to ensure a more flexible
energy system.
The flexibility of the energy system (as well as other
energy research) is a complex problem that touches on
many different disciplines. There is an increasing amount of
research carried out in the disciplines of social science, eco-
nomics, and humanities on the factors that underlie consum-
ers energy behavior and acceptance and how these may be
changed via policy interventions and market design. Hence, a
more integrated approach is needed within the social sciences
and humanities, as well as with energy researchers in other dis-
ciplines, such as engineering, environmental science, and com-
puter science.
Future integrated energy research should focus on
understanding why consumers behave the way they do
and how policy instruments and market designs can help
to change consumers behavior to enhance flexibly. This is
needed to better understand issues such as the interaction
of policy instruments, the effect of transaction costs, and
the assumptions of technology adoption found in the engi-
neering literature. Currently, policies and markets strongly
favor the perspective of rational consumers. However,
there is a vast amount of research on irrational consumer
behavior that can and should be better used to inform pol-
icy making and market design to alter consumer behavior
more effectively.
Ultimately, consumers need to become more flexible
because that will give future generations a better chance to
have a sustainable, reliable, and affordable energy system.
Energy systems should be designed based on the needs and
60 IEEE power & energy magazine
desires of current and future society and energy consumers.
To realize this, designers of the energy system (e.g., engi-
neers, environmental scientists, and computer scientists)
should be better informed by the work of the social sci-
ences, economics, and humanities. In this way, a shift from
a technology-centered energy system toward a user-centered
energy system can be achieved.
Acknowledgment
This publication has emanated from research supported in
part by a research grant from the Science Foundation Ireland
(SFI) under the SFI Strategic Partnership Programme, grant
number SFI/15/SPP/E3125.
For Further Reading
H. Allcott and T. Rogers, The short-run and long-run ef-
fects of behavioral interventions: Experimental evidence
from energy conservation., Am. Econ. Rev., vol. 104, no.
10, pp. 30033037, 2014.
H. T. Haider, O. H. See, and W. Elmenreich, A review of
residential demand response of smart grid, Renew. Sustain.
Energy Rev., vol. 59, pp. 166178, June 2016.
M. A. R. Lopes, A. C. Henggeler, K. B. Janda, P. Peix-
oto, and N. Martins, The potential of energy behaviors
in a smart(er) grid: Policy implications from a Portuguese
exploratory study, Energy Pol., vol. 60, pp. 233245,
Mar. 2015.
N. Sintov and P. W. Schultz, Unlocking the potential
of smart grid technologies with behavioral science, Front.
Psychol., vol. 6, pp. 18, 2015.
B. K. Sovacool, What are we doing here? Analyzing
fifteen years of energy scholarship and proposing a social
science research agenda, Energy Res. Social Sci., vol. 1, no.
0, pp. 129, 2014.
L. Steg, G. Perlaviciute, and E. van der Werff, Under-
standing the human dimensions of a sustainable energy tran-
sition, Front. Psychol., vol. 6, pp. 17, 2015.
Biographies
Geertje Schuitema is with University College Dublin, Ireland.
Lisa Ryan is with University College Dublin, Ireland.
Claudia Aravena is with Heriot Watt University, United
Kingdom.
p&e
january/february 2017

Flexibility Challenges
for Energy Markets
By William Dhaeseleer, Laurens de Vries,
Chongqing Kang, and Erik Delarue
F
FOLLOWING THE FORMULATION OF CERTAIN
strategic policy goals, such as reducing greenhouse gases
(GHGs) and including more renewable sources (RES) as part
of the energy mix in several parts of the world, the practi-
cal translation and actual implementation of these goals have
led to the introduction of substantial volumes of intermittent
renewable electric sources. Because affordable bulk storage
for electricity is still lacking, demand and supply need to
be (instantaneously) balanced. The resulting challenge that
Digital Object Identifier 10.1109/MPE.2016.2629742
Date of publication: 2 February 2017
january/february 2017 1540-7977/172017IEEE
Fragmented
Policies and
Regulations
Lead to
Significant
Concerns
ISTOCKPHOTO.COM/AAGGRAPHICS
intermittent renewable power sources pose to the controlla-
bility of the electric power system requires greater flexibility
from other parts of the system, as well as flexibility through
interaction with other energy sectors such as the heating sec-
tor, the natural gas sector, and the transportation sector.
As a consequence, the overall energy system becomes
increasingly coupled, which requires appropriate communica-
tion within and among sectors and flexible adjustment and
collaboration capabilities, while certain technical, economic,
and consumer comfort constraints are still satisfied. Because
this coupling, which has a multitude of feedback options,
makes the system less predictable (due to unexpected choices
IEEE power & energy magazine 61
 and decisions by market participants but also due to intrinsic
nonlinear behavior), flexibility will undoubtably be key.
In this article, we address the influence of policy and
regulation on the efficient behavior of energy markets
and illustrate the extent to which implementation of some
well-intended, but possibly conflicting, policy choices may
result in inadequate or unexpected performance within
the overall energy system. We further highlight the im-
portance of flexibility and stress that more flexibility will
be required in nonideal markets to avoid unanticipated
side effects.
It is not our goal here to comment on or evaluate the legiti-
macy of certain strategic policy objectives in themselves; we
accept these as the prerogative of policy makers. Rather, we
focus on how these strategic objectives are translated into con-
crete implementation targets. The strong push for investments
in certain intermittent renewables impacts the performance of
other instruments (e.g., those aimed at CO2 emission reduc-
tion) or electricity wholesale markets. By moving too rapidly
(and, in so doing, ignoring system interactions), a variety of
simple, well-intended (local) targets may counteract or even
oppose each other with the result that, while some individual
targets may be reached, the larger strategic objective will be
compromised (or even missed) or only reached at an unneces-
sarily high cost.
We illustrate these issues using some typical examples
drawn mainly from Europe, although interesting system
interaction scenarios in the United States and China are
also mentioned. We next identify several possible attractive
avenues for fostering flexibility through robust policies and
markets with the goal of mitigating the current situation and
allowing forand even promotingbetter system integra-
tion in the future. Finally, we suggest some challenges, open
questions, and research issues for policy and regulation.
The State of Play:
The Need for Flexibility and Analysis
European Policy Measures
and Their Consequences
An example of European energy policies with substantial
side effects are the so-called 20-20-20 targets. The European
Energy and Climate Change Package of 2008 was based on
three main pillars, or targets, to be reached by 2020:
20% of overall consumed end energy to be from RES,
with a subtarget for the transportation sector of 10%,
mainly from biofuels
20% reduction of GHGs compared to 1990
20% more efficient energy consumption compared to
a (then undefined) benchmark evolution.
Of these targets, the first two are mandatory; the last was
implemented through a variety of individual binding direc-
tives, including, among others, one for combined heat and
power (CHP) and one for energy use in dwellings; but the
overall target for the European Union (EU) was not compul-
62 IEEE power & energy magazine
sory. Before addressing the interaction effects among them,
some comments on each of these targets are in order.
Renewable Energy Policy
The 20% RES requirement for the overall EU is distributed
across member states (or, more simply, countries). After
much effort on the part of the EUs administration services
to devise a partitioning based on the potential to produce
energy from RES, it turned out to be impossible to reach a
consensus; thus, a purely administrative partitioning was
determined, not at all related to potential. Starting from the
existing volume of RES in 2005 and taking into account a
slight bonus for early starters, the overall gap of 11 percent-
age points for the entire EU was filled by allocating half of
that to each member state (i.e., each country had to increase
its RES contribution by 5.5% of its end energy). The second
5.5 percentage points were redistributed across countries
based on gross domestic product (GDP) per capita, so that the
richer countries carried the heaviest burden. As a result, some
countries with substantial potential have a relatively small
target, while (relatively rich) countries with almost no poten-
tial face a very challenging target. Note that the EU Commis-
sion documents refer to a fair and effective distribution;
that it be efficient is not mentioned.
The three EU decision-making bodies (the Commission,
the Council, and the Parliament) could reach no agreement
on a mandatory European-wide renewables certificate trad-
ing system, by which green certificates would be exchange-
able per country. Instead, every country was allowed to set
up its own individual support scheme, resulting in local cer-
tificate systems in some countries (or even in parts of coun-
tries), feed-in tariffs, investment support, and tax breaks.
A number of cooperation mechanisms (statistical trans-
fers, joint projects, and joint support schemes) made it possi-
ble to both offer flexibility and meet part of a member states
target through the deployment of renewables in another mem-
ber state. However, all countries opted for their own targets.
Statistical transfers may be used to balance the books at
the end in 2020, such that a country not meeting its target by
domestic production can buy a transfer from another country
with a surplus. However, in the absence of a real market at the
moment and uncertainty as to whether countries will meet
their own targets, the use of these transfers is not actually
stimulated and, therefore, remains uncertain.
The overall renewables target refers to a fractional
requirement of 20% of overall end-energy consumption; the
subtarget within the transportation sector is a minimum of
10%. Given the limited options for deploying renewables in
the heating and cooling and transportation sectors, the elec-
tricity sector will need to compensate and so faces a much
higher target. Moreover, low-carbon options often involve
electrification, e.g., in the cases of electric vehicles and heat
pumps. Together, the policy targets as laid out in the Na-
tional Renewables Energy Actions Plans lead to an overall
requirement for the electricity system of 3334%, or about
january/february 2017
one-third of total European generation (in terms of electric
end energy) by 2020. The implications will be different for 12,000
different countries, depending on their individual required
targets and actual potential. But a simple calculation for av- 10,000

Net Demand (MW)

erage countries in Europe for the period 20082020 leads
to the following orders of magnitude observations, which
may seem obvious but do not appear to be fully recognized
by many policy makers.
With respect to hydro power (except, perhaps, for some
Eastern European countries), only small increases are
possible, largely due to environmental constraints.
As to biomass, the potential is difficult to predict be-
cause it competes with other types of land use and there
are competing applications for biomass, such as trans-
portation. In addition, its very environmental sustain-
ability is questioned. Consequently, there is consider-
able resistance to its use for electricity generation.
Wind, both onshore and offshore, is characterized
by an average effective number of operating hours
(ENOH) of about 2,200 h/a and about 4,000 h/a, re-
spectively.
Solar photovoltaics (PVs) has an average ENOH of
about 1,200 h/a.
All this leads to capacity factors (CFs) for intermittent
sources (wind plus PVs) as follows:
r
onshore and offshore wind/CFs: ~ 2545%
r
solar PVs/CFs: ~ 1314%.
To produce, say, 20 percentage points of the 34% elec-
tric end energy with technologies that operate only
1314% and/or 2540% of the time requires a large
volume of installed power-generation capacity.
If a good deal of wind and sun is available and demand
is low (e.g., during weekends), situations in which too
much electricity is produced will start to arise.
However, sometimes (as in the case of a cold spell,
such as the European winter of February 2012) with
temperature inversion, little wind, and dark skies
(hence, no PVs), at 17:0018:00 h when peak demand
arises in northwestern Europe, very little RES elec-
tricity will be produced, requiring classic thermal
backup (as long as electric storage is not available in
bulk quantities at affordable cost).
An example of what a residual load profile could look
like for different levels of wind and solar PVs is presented
in Figure 1.
CO2 Policy
The 20% GHG reduction target with respect to 1990 is reca-
librated as a 14% reduction compared to 2005 (because of
more complete and reliable numbers) and is then subdivided
in two distinct categories (see Figure 2). The primary CO2
policy instrument is the EU emission trading scheme (ETS),
which represents roughly half of CO2 emissions. It affects
energy-intensive industries and the electric power and heat
sector with an emissions cap that decreases by 1.74 percent-
8,000
6,000
4,000
2,000
0
0 20 40 60 80 100 120 140 160
Time (h)
0% Wind + Solar PV 20% Wind + Solar PV
10% Wind + Solar PV 40% Wind + Solar PV
figure 1. Net electrical power load during one week for vari-
ous fractions of annual renewable electrical energy generation.
[Figure based on extrapolated data from Belgian Transmission
System Operator Elia (2016), http://www.elia.be/.]
age points annually up to 2020. The remaining non-ETS sec-
tors (amounting to the other 50% of CO2 emissions) mainly
comprise transportation, the residential and service sector,
small-and-medium-size industries, and agriculture. They
work under a country allocation scheme that should lead to a
10% reduction compared to 2005. As for renewable energy,
the European target for the non-ETS sectors has been split
into individual member state targets (ranging from +20% to
20%), largely based on GDP per capita. It is important to
understand that both reduction categories are independent of
each other: for ETS, there is a cap-and-trade scheme among
companies in the designated sectors (a market-based mecha-
nism), while for the so-called reduction sharing effort in
the non-ETS sectors, the countries are responsible (in the
sectors mentioned).
Target:
20% Compared to 1990
14% Compared to 2005
EU ETS Non-ETS Sectors
21% Compared 10% Compared to 2005
to 2005
27 Member State Targets, Stretching from 20% to +20%
figure 2. EU GHG reduction targets following two separate
philosophies, via companies (ETS) and by countries (non-
ETS). (Source: Memo/08/34, Questions and answers on the
Commissions proposal for effort sharing, Brussels 2008.)

january/february 2017 IEEE power & energy magazine 63

 Energy Conservation
Although for almost half a century there has been much
talk about energy conservation, energy savings, and energy
efficiency, these appear to be among the most difficult tar-
gets to achieve. After an evaluation in 2014, the policy was
adjusted, but it remains questionable whether the target will
be reached. It is also not entirely clear what the reference
baseline will be. In the future, it will become more impor-
tant to clarify what is meant by the consumption of pro-
sumers who avail of storage (which may have significant
efficiency losses of up to 2030%). Will only demand from
the grid be taken into account?
Furthermore, many support schemes are not cost reflec-
tive, such as net metering and feed-in tariffs. The ensuing
zero-marginal-cost electricity production by households
may stimulate frivolous electricity consumption. As a con-
sequence, the entire concept of energy efficiency may
lose its meaning. In the end, a more generalized concept of
resource efficiency (including investment cost, manufac-
turing and installation labor, fuel usage, if applicable, and so
forth) and flexibility in consumption may be called for. This
actually comes very close to the idea of economic efficiency,
which may be the only meaningful concept in this context.
Interactions Among EU Energy Policies
The philosophy behind the cap-and-trade EU ETS system
for GHG reduction is to achieve the GHG reduction target
in the most economically efficient way by reducing first
where it is cheapest. (Technically, this means that reduc-
tions take place first where the marginal abatement costs
are the lowest.) The significant deployment of renewables
(especially wind and PV solar, but also biomass) with sub-
stantial public support has not actually impacted Europes
CO2 emissions, as they are capped under the EU ETS. They
may have helped limit emissions to the cap, but at a cost
higher than what could have been optimally the case. These
RES will have reduced the CO2 emissions in particular
countries, but not in Europe as a whole. In other words, the
avoided CO2 emissions by these subsidized renewables will
be replaced by other CO2-emitting electricity-generating or
heating sources in industry that also belong to the EU ETS.
To put it bluntly, the subsidies for renewable energy have de
facto made it easier for burning more fossil fuels in indus-
try and coal for electricity generation.
A few comments are in order.
The same reasoning would apply to new nuclear
plants or to the enforced (premature) closure of coal-
fired plants with regard to the local versus overall Eu-
ropean CO2 emissions.
If the ETS had not existed, then the only alternative
might be a combination of measures such as RES
support and energy savings policies. Or, viewed posi-
tively, the contributions to reductions in CO2 emis-
sions due to RES may allow for a faster reduction of
the CO2 cap in the ETS.
64 IEEE power & energy magazine
At present, the CO2 cap in the EU ETS is hardly
reached, and a massive surplus of allowances exists,
so that it may seem that the previous reasoning is in-
correct. However, if there were fewer emissions than
emission allowances in the market, then the price of
these allowances should be zero. The nonzero price
means some actors withhold allowances from the
market so that the cap of actually available allow-
ances in the market is, indeed, be reached.
A few exceptions exist concerning the statement
made previously about CO2 price reduction. For ex-
ample, if locally produced PVs were to be dedicated
to feed heat pumps that effectively replace (small-
scale) CO2-emitting boilers, then overall CO2 emis-
sions would be reduced, not because of the PVs as
suchbecause that is still part of the electricity sec-
tor and thus the ETSbut because of the avoided
CO2 emissions for small-scale boilers, which belong
to the non-ETS sector, as shown on the right-hand
side of Figure 2. This last argument does not apply to
industrial heat pumps replacing large industrial boil-
ers because they are part of the capped ETS.
Another unintended policy effect is caused by the promo-
tion of CHP, as the deployment of small-scale CHP units
creates a shift of emissions from the ETS to the effort-shar-
ing (non-ETS) sectors. While a small-scale CHP usually
saves primary energy and replaces the emissions of a local
boiler, the total amount of CO2 emitted by the CHP is larger
than the boiler only (due to the additional electric power that
is generated). Because these local emissions are now part of
the residential sector (the right-hand side of Figure 2), which
is not under the CO2 ETS cap, emissions have increased
and are acting against reaching the country-specific target.
The fact that emissions under the ETS have declined is not
rewarded; it simply allows other facilities under the ETS to
emit more.
For large-scale CHP, the same reasoning as for the RES
deployment discussed earlier applies because the electric-
ity sector and the heating sector for large industries are all
part of the EU ETS (meaning that promoting large-scale
CHP does not reduce GHGs in Europe because of the cap).
Indeed, there are fewer emission certificates needed than
would have been the case for separate generation. To sum-
marize, it is almost certain that the promotion of CHP in
both cases will turn out to have a higher CO2 abatement cost
than if only one GHG reduction target had existed; these
cross-policy effects discourage an efficient route toward sat-
isfying CO2 targets.
The price evolution of EU allowances (EUA) the price
of CO2 for the power sector and large industrial facilities
is depicted in Figure 3. As can be observed, the CO2 price
has fluctuated considerably over the past years and is quite
low at the time of writing, around 5/ton CO2 in the fall of
2016. The behavior we see in Figure 3even with the sud-
den jumps and prices going to zerois perfectly explainable
january/february 2017
by the modalities and rules of the
setup of the EU ETS. This demon- 35

Carbon Price/Metric Ton

strates that the short-term behav- 30
Carbon Price /Mt

ior of the trading scheme is as it 25
should be. As an example, the 20
zero price at the end of 2007 is 15
due to the fact that the allowances
10
were not bankable and became
5
worthless as of 1 January 2008.
0
For later periods of the EU ETS,

05

06

07

08

09

10

11

12

13

14

15
the allowances were/are bank-

0
/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2
4

4
able, avoiding natural zero

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0

/0
29

29

29

29

29

29

29

29

29

29

29
prices but leading to a surplus of
allowances after 2012 (the formal figure 3. The price of EU ETS allowances between 2006 and 2015. (Graphic based
Kyoto period) as a consequence on data from Bloomberg and Sandbag, 2015.)
of the economic crisis, with much
lower CO2 emissions than origi-
nally anticipated and thus foreseen 25
by the cap.
The current low CO2 prices 20
result from many factors, some
US$/MBtu

15
of them straightforward conse-
quences of the design of a cap- 10
and-trade scheme, such as the
5
lasting economic crisis (char-
acterized by fewer emissions) 0
9

Ap 9
10

Ap 0
11

Ap 1
12

Ap 2
13

Ap 3
14

Ap 4
15

Ap 5
16
that started in 2008, the inflow
0

.0

.1

.1

.1

.1

.1

.1
r.

r.

r.

r.

r.

r.

r.

r.
ct

ct

ct

ct

ct

ct

ct
Ap

of international credits, and the

O

O
bankability of allowances that
Henry Hub NBP Japan LNG Contract
makes the surplus persistent over
Asian LNG Spot Brent
time. However, in addition (and
this is very important from a sys- Note: NBP = National Balancing Point (United Kingdom).
tems perspective) an unforeseen
factor is that the EUA prices have figure 4. The development of gas prices 20092016 in several parts of the world.
been pushed further downward Henry Hub represents U.S. prices, while NBP represents European prices. A com-
by the substantial injection of parison with Brent oil prices is given because many gas contracts in Europe are still
carbon-free renewables in the linked to the price of oil. Since April 2016, prices have increased in the United States
electric power sector. Indeed, the and worldwide. LNG: liquiefied natural gas. (Source: International Energy Agency,
introduction of CO2-free electric- Gas Medium-Term Market Report 2016, Organization for Economic Cooperation
and Development, Paris, 2016.)
ity has reduced the demand for
allowances, leading to even lower
prices (from which the other industries under the ETS had (and is still having) consequences on the merit order for
umbrella and, e.g., coal-fired units have been able to take electricity generation in the United States, where cheap gas
advantage). To put it in plain language, the high subsidies for has pushed coal-fired units out of the merit order, leading to
electricity-generating renewables have not only not impacted surplus coal on the world market and resulting, in turn, in
CO2 emissions on an EU level (because of the cap); they depressed world coal prices.
have affected the CO2 prices, making it cheaper for CO2- A further system effect in Europe, then, is that marginal
polluting units to generate electricity, while still meeting cost pricing in EU electricity markets is pushing efficient
the cap. combined-cycle gas-fired units out of the merit order as a
To further understand the effects in the European energy consequence of low world-market coal prices (due, as men-
markets, it is necessary to look as well at an important global tioned earlier, to the effects in the U.S. electricity market
interaction effect that, from an overall systems perspective, because of shale gas), the absence of a significant CO2 price
is of interest in its own right. The shale gas revolution in the signal, and the injection of zero-marginal-cost renewable
United Stateswith gas prices that have been, and still are, generation (see Figure 5). The green parties in the European
much lower than in other parts of the world (see Figure 4)has Parliament wanted to see coal-fired units pushed off the

january/february 2017 IEEE power & energy magazine 65

 wedge, which would have been the case if there had been no
shale gas effect on coal prices and if there had been a high
CO2 price in the EU ETS.
From March 2007 until the summer of 2008, the whole-
sale forward electricity prices in the Central West Euro-
pean (CWE) region increased from about 5055/MWh
to 90100/MWh, after which they gradually declined,
with a small upsurge in the spring of 2011 (because of the
Fukushima accident and related decisions in the German
market) to levels of about 2530/MWh in the late sum-
mer of 2016. The history from 2011 to 2016 is shown in
Figure 6 for the CWE countries France, Germany, and the
Benelux. This downward trend on wholesale prices makes
it hard for gas-fired units to make a profit. Many European
CCGTs are idle and being mothballed or kept as capac-
ity reserve through a capacity remuneration mechanism.
Discussions as to an appropriate market design (energy
only versus capacity remuneration schemes) are cur- the large differences in the generation portfolio among
rently ongoing.
As previously mentioned, zero-marginal-cost renew-
ables (with substantial installed capacity in many European
countries) contribute to the downward drive of the wholesale
electricity price when they are producing. In the absence of
subsidies, this would lower their own return on investment,
so they would effectively cannibalize themselves. Notwith-
standing the decreasing wholesale prices, ordinary custom-
ers see increasing retail prices, mainly as a consequence
of markups to recover the costs of the renewables support
schemes. This is illustrated in Figure 7 by the price evolu-
tion in Belgium for a typical end customer with annual con-
sumption of 3,500 kWh. Similar retail price increases have
occurred in Germany, with a steady increase from about
140/MWh in 2000 to a maximum of 291/MWh in 2014,
after which there was a slight decline in 2015 and 2016 to
about 287/MWh.
A final unintended effect of the rapid growth of renew-
ables in the European system is that the convergence of
cross-border electricity prices, which was a major goal of
the common electricity market, has suffered from mas-
sive renewables penetration. The reason is that the cross-
border high-voltage grid is currently not strong enough to
ensure price convergence (i.e., by being congested) given
countries. This is illustrated by the decoupling indicated
by the blue arrow in Figure 6 in the CWE market. The
cross-border market coupling is very weak in situations of
high wind and PV solar power production in this region. A
further issue facing the European market is the so-called
loop flows (or unidentified flows) in certain regions
such as CWE; these are also due to the lack of sufficient
internal and cross-border trans-
mission capacity.

The EU ETS Refurbished

Faced with low EUA prices in the
ETS market, with the awareness
that many market participants do
Win not foresee a long-term CO2 price
d PV GT
Coal CC and hesitate to make long-term in-
vestments, European policy mak-
Low nd ers have decided to reform the
ma
De current EU ETS. Via market in-
terventions (referred to as back-
loading and a market-stability
reserve), a volume of allowances
is being taken out of the market
with the possibility of reintro-
ETS
ETS ducing them later. Whether these
ETS
measures will alleviate the side ef-
fects of the EU energy and climate
figure 5. The limited load factors of combined-cycle gas turbines (CCGTs) in Europe. policies remains to be seen. The
Zero-marginal-cost renewables, together with low demand, push thermal generation volume of the backloading seemed
units out of the merit order (or off the ice wedge). With current European gas and not to be large enough to have a
coal prices, and a very low CO2 price penalty (via the ETS) represented by the baby
significant impact. Whether the
seals, CCGT plants risk being the first victims. For higher CO2 penalties, adult ETS
market-stability reserve will allevi-
sea lions would do the job of pulling coal-fired plants off the wedge first. For the
relative coal-to-gas prices in the U.S., CCGTs are currently more economic regard- ate the side effects of the EU ener-
less of a CO2 penalty, and coal generation is the prey. In China, the demand is still gy and climate policies remains to
sufficiently large so that coal and gas plants are called upon. (Image courtesy of be seen; moreover, the final volume
D. Patteeuw, KU Leuven, used with permission; adapted image inspired by http:// of allowances has not, in principle,
economicsforenergy.blogspot.be/2013/02/los-mecanismos-de-retribucion-de-la.html.) been altered.

66 IEEE power & energy magazine january/february 2017

European Targets Toward 2030
In the meantime (and as part of European promises for
the Paris COP21 Agreement), the EU has sharpened its
commitments toward 2030albeit in a different way, not
unimportant for system interaction. The firmest commit-
ment is the 40% reduction of GHGs with respect to 1990,
again split between an ETS part with an annual reduction
of 2.2 percentage points of the cap toward a 43% reduction
with respect to 2005 and an EU-wide non-ETS part of 30%
compared to 2005. For renewables, a 27% end-energy goal
by 2030 has been set, as well as an efficiency-improvement
goal of 30% compared to a baseline. For both the renewables
and efficiency goals, there are no binding national targets
but only an overall EU-wide objective.
Energy-System-Integration
Observations on Flexibility Challenges
in the United States and China
To illustrate that energy-system integration challenges are
global and require careful consideration internationally,
we include here a few examples from the United States and
China. Due to lack of space, some elements/cases are only
cited without a full discussion.
hesitating to accept the financial liability, as well as by lack
of public acceptance and complicated permitting), increased
flexibility on the electricity-generation side by means of
multifuel units (gas and oil), and demandresponse pro-
grams, among others.
On the other side of the United States, major energy-sys-
tem integration problems arose in 20002001 in California
(later referred to as the California electricity crisis). As
the first state to fully introduce liberalized markets (often
inappropriately called deregulation), California expe-
rienced a combination of events and circumstances that
led to the worlds richest countrys richest state having to
turn off the lights (actually, cut all powercalled roll-
ing blackouts) to keep the system from collapsing. There
are many reasons for this then unprecedented failure, but it
was clearly a mix of many interacting factors, conditions,
and regulations. We mention, among others, a regulated
65
60
55
50
/MWh

45
40
Energy-System-Integration Challenges in U.S. Markets 35
30 BE NL
In New England, the past few winters have seen some inter- 25 DE FR
esting issues concerning the interaction between gas and 20
1

6
electricity markets, whereby a stretched natural gas grid
01

01

01

01

01

01
/2

/2

/2

/2

/2

/2
has translated into very high gas and electricity prices. Gas
01

01

01

01

01

01
delivery occurs via pipelines from the south and the north
(Canada), including liquefied natural gas (LNG). Average gas figure 6. The decreasing tendency of forward electricity
capacity suffices during the winter, but the system becomes wholesale prices in the CWE market (France, Germany, and
stretched on particular peak-demand days. Because of market Benelux). [Source: Commission de Rgulation de llectricit
dynamics, which are a consequence of cheap shale gas in the et du Gaz (CREG), Belgium, Sept. 2016. The arrows have
been added by the authors.]
United States since 2010 and thus gas price differentials with
other world markets, LNG imports into New England have
seen a reduction; therefore, the winter supply of LNG from
Massachusettss Everett and Canadas Canaport has declined,
250
leading to winter spike prices. In addition, because of the low
average U.S. shale gas prices and the environmental drive to 200
/MWh

reduce CO2 emissions through the cap-and-trade system of 150

the Regional Greenhouse Gas Initiative (RGGI), over the last
few years natural-gas-fired electricity generation in New Eng- 100
land has increased, along with the retirement of other plants 50
(using nuclear, coal, and oil).
This increase in gas-fired electricity generation has,
7

6
0

indeed, led to lower average wholesale electricity prices,

20

20

20

20

20

20

20

20

20

20
/

/
01

01

01

01

01

01

01

01

01

01

but it transfers the physical stress in the gas network to fuel

adequacy/reliability concerns in electricity generation, giv-
ing rise to sometimes very high electricity peak prices. The
situation requires the full attention of system operators, who
need to take both long-term and short-term actions, such
as increasing gas-transmission capacity (although this is
hampered by insufficient interest among capital investors
desiring to see long-term commitments from shippers but
figure 7. Typical retail electricity prices in Belgium for av-
erage customers with annual consumption of 3,500 kWh/a.
The different curves are for the same supplier (Electrabel)
but different distribution-grid regions. The dip from April
2014 through September 2015 is somewhat artificial; it is
due to a temporary reduction of the VAT from 21% to 6%
and back (by different governments). (Source: CREG, 2016.)

january/february 2017 IEEE power & energy magazine 67

 cap on the retail price of electricity, while gas pricesand,
thus, wholesale electricity pricesbegan to rise; increas-
ing electricity demand; reduced imports of electricity from
other states; a slowdown in the pace of granting permits to
build new electric power plants, as a result of environmental
regulations; and market-power problems, including market
manipulation and even fraud. This crisis cost the California
economy (consumers, shareholders, taxpayers, and laid-off
employees, among others) several tens of billions of dol-
lars and led to the drastic scaling back the energy-market
liberalization philosophy in the state. Policy and regulation
were major factors in the development of this crisis. In this
case, system studies would certainly have pointed to the
Net Load31 March
28,000
26,000
24,000
22,000
20,000 2012 (Actual)
Megawatts
challenges and bottlenecks emerging from several types of
interactions; such studies would likely have shown where
greater regulatory flexibility could have alleviated, if not
entirely avoided, the dire consequences.
Fast-forward 15 years, and California is preparing for
another such energy-system challenge, with scientifictech-
nical discussions of system integration currently ongoing to
ensure that the necessary actions, preparations, and precautions
are taken on the technological side; the economic, financial, and
market environment; and an appropriate regulatory framework.
With the rapid growth of PV solar capacity and given no correc-
tive flexibility measures, it can be expected that, around 2020,
stiff ramping rates by the non-PV remainder of the generation
systemof the order of 13 GW in
3 hwill have to be coped with.
This is illustrated by the so-called
duck curve published by the Cali-
fornia independent system operator
CAISO, as shown in Figure 8.
In reaction to this duck curve,
studies have been initiated to

demonstrate that appropriate flex-

18,000 2013 (Actual) ibility measures, dealing with
2014 Increased
16,000 Potential both the total demand and the net
2015 Ramp
Overgeneration 2016 or residual demand, can teach
14,000 2018 2017
12,000 2019 the duck to flybasically show-
2020 ing that stiff ramping rates can be
10,000
avoided (see Figure 9). Whether
the measures currently suggested
0
12 a.m. 3 a.m. 6 a.m. 9 a.m. 12 p.m. 3 p.m. 6 p.m. 9 p.m. will actually work remains to be
Hour seen, but the fact that the discus-
sion has started is encouraging; by
figure 8. The duck curve illustrating the net or residual load to be covered the time needed, appropriate soft
by the non-PVs remainder of the electricity system in California. After the sun has solutions may be implemented.
set in the early evening, electricity demand increases for air-conditioning, light-
ing, and cooking needs, leading to big ramping rates to meet electricity demand. Energy-System-Integration
(Source:CAISO.) Challenges in China
China also faces considerable
challenges in terms of energy-
Duck Curve with All Ten Strategies Compared to Original Load
4,000
system inflexibility, especially
for the integration of renewable
3,500
energy. Renewable energy is
3,000 growing in a nonmarket environ-
2,500 ment, where wind and solar gen-
MW

2,000 eration is at the top of the merit

1,500 order. However, large amounts of
Post Strategies Total Load renewable energy are still being
1,000
Post Strategies Net Load curtailed or dumped, despite
500 Original Net Load the fact that it is scheduled with
0 first priority. This is because it is
1 2 3 4 5 6 7 8 9 1011 12 13 1415 16 17 18 19 2021 22 23 24
Hours obligatory for conventional gen-
erating units to offer flexibility
figure 9. Avoiding stiff ramping rates shown in Figure 8 (dotted curve) can be ac- (by reducing their output) but
complished by well-chosen strategies to level off the net load (green curve). (Source: without any financial compensa-
J. Lazar, Teaching the Duck to Fly, RAP, 2016.) tion. Because the flexibility is not

68 IEEE power & energy magazine january/february 2017

priced in China, these conventional units do not have any
incentive to improve their ability to provide grid flexibility.
However, China has instituted many ongoing attempts
over recent years to meet the challenge of inflexibility. The
government has been making great efforts on pilot projects
and corresponding policies called the Energy Internet,
which consists of four major parts:
1) integrating multiple energy system (e.g., electricity,
heat, and gas)
2) establishing a cyber-physical system and making use
of big data toward a smarter energy system
3) deregulating the energy market
4) interconnecting power grids in multiple areas.
One of the most effective policies is the deregulation of
the electricity market. Take the ancillary market for peak
shaving in northeast China as an example. Conventional
units, which offer flexibility for wind power, are now paid
for by wind farms and other inflexible units. This policy has
achieved great success as more than 77% of conventional
units now offer a lower minimum output level than before
the policy was established.
In addition, China is trying to solve electric power system
flexibility issues by coordinating multiple kinds of energy. A
typical example is the inflexibility caused by the linked elec-
tricity and heat production from coal-fired CHP plants. Due to
the inflexible operation of the CHP units when forced to gen-
erate large amounts of heat, they also produce electricity so
that the room left for wind generation is small, leading to wind
curtailment. This also gives rise to the problem of severe air
pollution, as coal-fired CHPs are notorious for large emissions
of NOx, SO2, and other pollutants. To avoid so much wind
curtailment, as well as pollution, the government initiated a
program called heating by wind to coordinate the electric
power and heating systems. This program makes use of clean
wind energy to serve heating demand, thus leaving more room
for wind integration. However, it should be noted that heat-
ing by wind still needs special price policies in a nonmarket
environment because the current central heating price from
conventional CHP is only one-half to one-third the cost of
the electricity heating. The wind farms that participate in the
heating need to sacrifice some of their profit to make heating
by wind economically acceptable to heat consumers. Many
policy-related efforts are still expected to make wind power a
cheap source for heating.
Regulatory Encouragement for Flexibility
Flexibility Options: The Possibilities
In this context of a fragmented and imperfectly aligned set
of policy instruments, policy makers now face the challenge
of encouraging flexibility options to improve overall energy-
system integration and mitigate side effects. Flexibility means
exist within energy sectors (in particular, the electric power
sector), but it is also important to encourage interactions among
different energy-carrier sectors (such as electricity, gas, liquid
january/february 2017
fuels, heat, and cooling) and end-energy sectors (such as in-
dustry, residential and service sectors, and transportation). We
mention here the known flexibility possibilities (mainly origi-
nating in the electric power sector because of the absence of
massive cheap electricity storage).
The most straightforward flexibility options are
utilizing backup reserves from flexible dispatchable
thermal plants (upward and downward)
providing electric storage (short-term storage via mul-
tiple battery units; intermediate storage via pump-
hydro storage; and long-term, large-scale storage via,
perhaps, power to gas)
expanding transmission grids
encouraging active demand response or participation
by customers (industrial, commercial and service sec-
tors, and residential retail customers)
encouraging interaction with other carriers/sectors
(heating, transportation, etc.)
curtailing superfluous RES production (because high
power injection peaks can be avoided at the relatively
minor cost of a bit of curtailed energy), which means
that priority access for renewables should be reviewed
and become part of a system-wide perspective.
A major question still to be addressed is how market designs,
policies, and regulation affect these flexibility options.
Enabling Flexibility Options:
Challenges for Policy and Regulation
As has been illustrated previously, policy and regulation
often have unexpectedand, possibly. counterproductive
effects on overall system performance. It should, therefore,
be a part of good policy making to first study the overall
system by modeling its different parts, with much emphasis
on the interactions among the different subparts as well as
among different policies. As the behavior of the system
including the not-always-predictable behavior of customers
and other market actorswill be strongly nonlinear, care-
ful analysis is called for, well beyond the standard isolated
impact assessments.
First and foremost, policy makers should encourage cor-
rect system cost evaluation and, consequently, appropri-
ate pricing to guide consumers. As a general rule, market
requirements should provide sufficient freedom for market
participants to play their roles while eliminating any loop-
holes overly creative individuals or organizations can abuse;
this means that carefully considered boundary conditions
and/or justified constraints must apply. Also as a rule, vary-
ing prices can influence customer behavior, and real-time
pricing can steer markets in a desired direction. All custom-
ers connected to the electric grid need to contribute to its
costs. Following the principle of cost-reflectiveness, a grid-
connection tariff should be based (at least partly) on the con-
nection capacity or maximum annual capacity used (in kW)
rather than on energy consumption (in kWh). This applies,
in particular, to customers with much self-generation.
IEEE power & energy magazine 69
 An example of European energy policies
with substantial side effects are the so-called
20-20-20 targets.
A proper price signal is key for good, active customer-
demand participation; however, practical participation will
likely require the help of aggregators, who will need to be
allowed the freedom to act in the market and whose role
should be facilitated by distribution-grid companies (which
constitute a natural monopoly).
Particular attention should be paid to the challenging cir-
cumstances of a multitude of prosumers with rooftop PVs
(possibly) assisted by local battery storage. What will be
the appropriate pricing scheme for feeding back to the grid?
Guaranteed feed-in tariffs and net metering do not appear
sustainable in the long run. Also in this case, the interven-
tion of aggregators, perhaps also employing local storage
for grid ancillary services, may be called for. In this regard,
specifications on products for ancillary services should be
made as independent as possible of technologies, allowing
for an open competition among providers of such services
(coming from supply, demand, and/or storage).
One of the cheapest means to integrate intermittent
renewables over large geographical areas is by allowing
new high-voltage lines to be constructed (be it in open air
or as cables, as ac lines, or as high-voltage dc). The crucial
stumbling block of delayed or denied permits must be over-
come. This is a typical case in which the collective benefit
may supersede individual or personal desires (whereby the
enforcing authorities must appropriately compensate the
disadvantaged). The same applies to natural gas transmis-
sion grids. If there is insufficient grid capacity, there may
not be enough transport of gas during heavy winter con-
ditions (as in New England in the United States in 2014
2015) or because of geopolitically-inspired cuts (as on
New Years Day of 2006 and 2009 in Europe), with serious
consequences for electricity generation and heating. Gas-
compressor stations should operate bidirectionally where
doing so can improve security of supply; in addition, also
for gas-infrastructure projects, permits should be granted
in a timely fashion.
Policy makers should anticipate (or avoid) conflicting or
self-neutralizing targets, as we demonstrated in our discus-
sion of the 20-20-20 case in Europe. One should identify the
main problem (e.g., climate change and CO2 emissions) and
then impose one clear target. Because CO2 emissions lead to
external costs, these costs should be internalized, meaning
that some sort of CO2-related penalty on all CO2-emitting
sources may have to be considered (either by a simple CO2
tax or via a single cap-and-trade system, with perhaps a
CO2 budget for all emitting entities, even up to the level of
70 IEEE power & energy magazine
households). Renewables and CHP should take advantage of
that simple CO2-reduction scheme in a natural way, without
extra (likely distorting) support mechanisms.
Through the interaction of the electric power sector with
the thermal sector, ample attention is currently devoted
to thermal grids (of the third and fourth generation). But a
careful regulatory framework will be needed to guarantee a
return on investment for the thermal grid and for customer
satisfaction, especially in areas where natural gas distribution
networks are also available. Who will own the thermal grid?
Will it be a natural monopoly, with distribution/independent
system operator characteristics? Will customers be forced
to connect to thermal grids (and mothball efficient gas-fired
condensation boilers)? Will there still be the freedom to
install heat pumps and/or CHPs?
Transportation will likely see changes over the com-
ing years. Whether very efficient combustion engines will
survive or will be replaced by hybrid or battery electric
vehicles or by hydrogen-fed-fuel-cell vehiclesand over
what periodremains to be seen. It must be noted that in
many countries, car engines already pay a stiff (CO2) pen-
alty because of high excise taxes (especially in Europe),
meaning that cheaper options are available elsewhere in the
energy economy. Also, current, seemingly cheap electric
charging may change in many countries when authorities
start levying excise taxes on electric charging (to compen-
sate for missed revenues due to fewer fuel-consuming vehi-
cles). Will such excise taxes be charged for self-generated
electricity by prosumers? In the end, the revenue books of
governments must balance the budget; it is important that
energy-related taxes be imposed wisely without creating or
aggravating side effects.
Long-term storage of electrical energy still needs to be
resolved. A possible attractive candidate might be the so-
called power-to-gas route, whereby superfluous renewable
electricity is converted to hydrogen (via electrolysis) and then
made to react with CO2 (which, in turn, is captured some-
where) to produce renewable methane. It is technically pos-
sible, but the overall cost picture in a market environment (and
when all investment costs are appropriately accounted for) is
not yet fully clear. In any case, renewable methane will have
to compete in the common natural gas market.
As alluded to with regard to Figure 5, appropriate market
designs will have to be developed, opting for capacity-remu-
neration mechanisms or energy-only markets for dispatch-
able units or other means of flexibility to provide the required
balancing. If no satisfactory solution is found, there will be
january/february 2017
no firm means available. Conversely, remuneration for these
balancing means must be fair and not over-reward them.
Finally, careful thought will be necessary for proper
price-setting schemes in a (possibly) future all zero-mar-
ginal-cost generation environment, complemented by stor-
age with finite losses. Appropriate market rules should allow
prices to be based on the opportunity cost as seen by the
market. Artificial pricing schemes will likely lead to eco-
nomic inefficiencies.
Conclusions
The current set of energy policy instruments is characterized
by varying degrees of effectiveness, and policies sometimes
counteract one another. The full system costs of the resulting
tangle of incentives often are not fully accounted for, and they
are not well balanced and nontransparent, challenging system
efficiency. Indeed, because of interacting policy choices and
regulations, many energy markets are distorted by consider-
able hidden system costs, eventually to be paid by consum-
ers, taxpayers, or shareholders (which often include pension
funds). Because of a lack of economic bulk electricity storage,
interactions within the electricity system and with other sec-
tors, such as natural gas and heating, require increased flex-
ibility for smooth operation and cost-effective performance.
A well thought-through and consistent policy framework
is called for, ideally with some stability in the regulatory
framework (and certainly without retroactive measures).
Key will be clear, transparent, but comprehensive regulation,
whereby market players (such as aggregators and energy-
service companies) have the freedom to provide the services
requested by customers. Furthermore, targets and specifica-
tions should be set to be as independent as possible from
specific technologies, so markets can decide how to reach a
certain target or meet a certain specification (whereby both
supply- and demand-side actions, combined with storage,
can truly compete across system levels and borders).
In any case, because of the complexities we have described,
quick-and-dirty regulation will likely backfire, and even sim-
ple, positive-seeming measures may lead to unforeseen side
effects because of negative feedback and system interactions.
Policy makers are, therefore, advised to perform careful
system-wide studies to simulate and understand the systems
behavior and adjust draft legislation and/or regulation before
any rules are implemented.
For Further Reading
J. Delbeke and P. Vis, Eds. (2016). EU Climate Policy Ex-
plained. Brussels: European Union [Online]. Available:
http://ec.europa.eu/clima/publications/docs/eu_climate_
policy_explained_en.pdf
D. Helm, The Carbon Crunch; How Were Getting Cli-
mate Change Wrong: and How to Fix It. New Haven, CT:
Yale Univ. Press, 2012.
january/february 2017
M. O. Bettzge, D. Helm, and F. Roques. (2014). The cri-
sis of the European electricity system: Diagnosis and possi-
ble ways forward. Commissariat general la stratgie et la
prospective, Paris [Online]. Available: http://www.strategie
.gouv.fr/sites/strategie.gouv.fr/files/archives/CGSP_Report_
European_Electricity_System_030220141.pdf
International Energy Agency. (2016). Re-powering markets:
Msarket design and regulation during the transition to low-
carbon power systems. IEA/OECD, Paris [Online]. Available:
http://www.iea.org/publications/freepublications/publication/
re-powering-markets-market-design-and-regulation-during-
the-transition-to-low-carbon-power-systems.html
Energy Information Agency. (Feb. 7, 2014). High prices
how stresses in New England natural gas delivery system
[Online]. Available: https://www.eia.gov/naturalgas/review/
deliverysystem/2013/pdf/newengland_natgas.pdf
J. L. Sweeney, The California Electricity Crisis. Stan-
ford, CA: Hoover, 2002.
P. L. Joskow. (2001). Californias electricity crisis. Ox-
ford Rev. Econ. Policy [Online]. vol. 17, no. 3, pp. 365388.
Available: http://economics.mit.edu/files/1149
California ISO. What the duck curve tells us about
managing a green grid, Fast Facts, version CommPR/
2016. Available: https://www.caiso.com /Documents/
FlexibleResourcesHelpRenewables_FastFacts.pdf
J. Lazar. (Feb., 2016). Teaching the Duck to Fly (2nd
ed.). RAP. [Online]. Available: http://www.raponline.org/
wp-content/uploads/2016/05/rap-lazar-teachingtheduck2-
2016-feb-2.pdf
P. Denholm, M. OConnell, G. Brinkman, and J. Jorgen-
son. (2015). Overgeneration from Solar Energy in Cali-
fornia: A Field Guide to the Duck Chart, NREL, Golden,
CO, Tech. Rep. NREL/TP-6A20-65023 [Online]. Avail-
able: http://www.nrel.gov/docs/fy16osti/65023.pdf
C. Kang, X. Chen, Q. Xu, D. Ren, Y. Huang, Q. Xia,
W. Wang, C. Jiang, J. Liang, J. Xin, X. Chen, B. Peng, K.
Men, Z. Chen, X. Jin, H. Li, and J. Huang, Balance of
power: Toward a more environmentally friendly, efficient,
and effective integration of energy systems in China,
IEEE Power Energy Mag., vol. 11, no. 5, pp. 5664, Sept./
Oct. 2013.
Biographies
William Dhaeseleer is with the University of Leuven (KU
Leuven), Belgium.
Laurens de Vries is with the Delft University of Technology,
The Netherlands.
Chongqing Kang is with Tsinghua University, Beijing
China.
Erik Delarue is with the University of Leuven (KU Leuven),
Belgium.
p&e
IEEE power & energy magazine 71
 history
T
THE INTRODUCTION OF AC SER-
vices to the theater districts of large
cities during the early 20th century al-
lowed for the adoption of reactance-
type dimmers to control incandescent
stage lighting, in place of the hopelessly
inefficient resistance dimmers (rheo-
stats) that previously had to be used with
dc services. Then, in a relatively short
time, reactance dimmers were combined
with vacuum tubes to create the first truly
electronic type of lamp dimmer, which
allowed for a much greater flexibility of
control that was not improved upon until
the introduction of computer-type stage
lighting consoles during the latter part of
the century.
The Roxy Theatre
The grandiose Roxy Theatre was lo-
cated at 7th Avenue and 50th Street in
Manhattan, New York. It opened in
March 1927 and was the inspiration
of theatrical impresario, Samuel L.
(Roxy) Rothafel. The building was
constructed by developer Irwin Chanin,
who had built six other Broadway
theaters as well as the Chanin Build-
ing, which is still standing at Lexing-
ton Avenue and 42nd Street. The fan-
tastic Roxy Theatre was demolished
in 1960.
The stage lighting reactance-type
dimmer board for the Roxy was con-
structed by the former Hub Electric Com-
pany of Chicago, and it probably was
the largest such board ever built. It was
Digital Object Identifier 10.1109/MPE.2016.2620659
Date of publication: 2 February 2017
72 IEEE power & energy magazine
Thomas J. Blalock
reactors for the Roxy
evolution of ac stage lighting
Since the launch of IEEE Power & Energy Magazine in January 2003, Thomas
J. Blalock has authored 20 articles for the History column. His articles have
covered a wide variety of interesting topics on the history of electric power en-
gineering. This article, his 20th, deals with ac theatrical stage lighting, an avoca-
tion in which Tom has enjoyed a special interest and has achieved much.
He first became interested in stage lighting and rigging while working back-
stage during his high school years in Easton, Pennsylvania. He had little time to
pursue his interest while in college and during his early years at work for Gen-
eral Electric (GE) in Pittsfield, Massachusetts.
In the 1970s, Tom joined a local amateur theater group and spent much time
doing stage lighting and other backstage work. The theater group enjoyed the
use of a new facility with a large, fully equipped stage at the local community
college. Following the closure of the GE plant in 1987, Tom ran that theater
until 1999. During that time, his interest in stage lighting was rekindled with
particular emphasis on the history of theatrical lighting during the 20th century.
As to Toms educational and career background, he earned a B.S.E.E. degree
from Lafayette College and an M.E.E.E. degree from Rensselaer Polytechnic In-
stitute. His duties as a development engineer at the former GE High-Voltage En-
gineering Laboratory and later as a test engineer in the Transformer Test Depart-
ment, both in Pittsfield, included a broad range of activities, including lightning
protection and high-voltage switching surge studies. Since retiring from GE, Tom
has actively pursued his hobby of industrial archaeology, with particular em-
phasis on the exploration, preservation, and careful documentation of histori-
cally important and interesting electric power projects and equipment.
We are honored to welcome Tom back as our guest history author for this
issue of IEEE Power & Energy Magazine.
Carl Sulzberger
Associate Editor, History
22 ft (6.7 m) in length and contained to run it) had a view of the stage itself
hundreds of individual control levers. (see Figure 1).
It was installed in a separate room A very early attempt to use simple
off stage that was open ended so that series-connected reactors as theatrical
the operators (it took more than one dimmers was a failure. This occurred
1540-7977/172017IEEE january/february 2017
in Dalys Theatre (demolished in 1920),
which was located at 29th Street and
Broadway in Manhattan in 1888. AC
power was being supplied to that
a rea by the United Electr ic Light
a nd Power Compa ny ( U EL&P) in
competition with the dc power be-
ing supplied by the Edison Company.
UEL&P attempted to install reactors
with movable iron cores, to vary the
reactance, as stage lighting dimmers.
However, the design of these devices
left a lot to be desired, and it was found
that, in operation, they hummed like
a hive of angry bees. That, of course,
was completely unacceptable for the-
atrical use, and Daly told the UEL&P figure 1. The Hub Electric Company stage lighting dimmer board being de-
wiring men to just throw the whole livered to the Roxy theater. (Photo courtesy of the Theatre Historical Society of
America.)
thing out into the street. So ended
that experiment.
The lamp-dimming reactors used saturable core reactors, and they in- in contrast to the earlier attempt to re-
by Hub Electric for the Roxy Theatre cluded a second control winding duce reactance by means of a movable
board were series connected as well, through which dc was circulated. The core. These reactors had been supplied
but they were not just simple reactance dc served to saturate the magnetic core to Hub Electric by the Ward Leonard
coils. These devices were known as of the reactor to reduce its reactance, Electric Company.
 The magnitude of the dc in the con-
trol winding was, in turn, controlled by
means of a relatively small resistance dim-
mer operated by a lever on the board.
Of course, this rheostat did introduce
some energy loss but far less than
would be the case if resistance dimmers
were used to control the lamps directly.
Visit us at DistribuTECH,
Booth #1024
FULLY IEEE 1656 COMPLIANT
These levers, by the way, were capable
of being interconnected mechanically
to provide a crude form of master-
ing. However, the interconnection of
a large number of levers introduced a
fair amount of mechanical friction to be
overcome. It was not possible to provide
a long enough master lever to give an
It Fits
operator sufficient mechanical advan-
tage to accomplish this. Therefore, a
large wheel (seen at the center of the
board in Figure1) was provided for this
purpose. The wheel was geared to the
shafts inside the board that moved the
levers, so it provided the necessary me-
chanical advantage.
The operation of this wheel did not
allow for extremely rapid cues. The fast-
est cue possible was about 15 s, which
corresponded to about 30 revolutions
of the wheel. Extremely rapid cues (for
dramatic effect) had to be made by using
the control switches associated with each
dimmer. These, in turn, switched the main
lamp circuits by means of remote contac-
tors, and the switches themselves could
be connected to each other electrically to
provide a mastering function. The wheel,
however, was very useful for extremely
slow cues, such as sunrises or sunsets,
taking place over many minutes. Thus, it
was called the slow-motion wheel.
The Roxy Theatre was huge, with
nearly 6,000 seats, so its stage was cor-
respondingly large and required a lot
of lighting to cover it. The total electri-
cal load planned for the entire building
was 2,500 kW, with 780 kW allocated
for stage lighting via the Hub board.
A total of 23 watthour meters were
required for the various areas of the
building because different rates were ap-
plied to various uses (such as motors and
lights). However, another reason for such

NO
When a single contact can bring down your
substation, installing precise t cover-up is
the best choice for mitigating your risk.
a large number of meters was that it was
extremely difficult to calibrate watthour
meters designed for currents in excess of

GAPS
Its a proven fact that cover-up works. Our 800 A (the total current corresponding
precise t covers are made to order based to the anticipated load would have been
on the exacting dimensions of the underlying about 20,000 A at 120 V). Presumably,
equipment it is the best product for
eliminating gaps. And, Greenjacket is fully
such calibration at that time required the

NO
compliant with the IEEE 1656 Guide test
parameters. Selecting Greenjacket ensures
you have the only precise t and most effective
protection available.
inclusion of the current transformer or
shunt to be used with the meter.
The building had a total of four elec-
tric services, two for ac and two for

ZAPS
When outage risks can have signicant dc. UEL&P supplied the ac services at
consequential damages for any utility or end 208Y/120 V, three phase, and the New
use customer, having the best protection is York Edison Company supplied the dc
your best choice.
services at 120/240 V, three wire. The
main reasons for multiple services were
www.greenjacketinc.com 1.866.464.7996 the high connected load and the need
for two different types of current. How-
ever, there also were requirements at

74 IEEE power & energy magazine january/february 2017

 Optimisation:
that time for theaters such as the Roxy
How many birds can your wires hold?
to have more than one service in the
event of the failure of one while an audi-
ence was in the building. Having power Knowing capacity and limits... knowing when and where
supplied by two different companies to best allocate investment...
further increased the reliability of ser-
vice in emergencies.
Just knowing...
DC was required for the control wind-
ings of the saturable core reactors used
in conjunction with the Hub control board. :KHQZHXQGHUVWDQGSXUSRVHDQGLQXHQFHVDQG
However, dc was also needed for the op- SHHULQWRWKHIXWXUHZHEHFRPHWUXO\HIIHFWLYHDQGFDQ
eration of arc lights, both in movie pro- SODQWRSURWIURPLWVIXOOSRWHQWLDO
jectors and for large follow spots from
the balconies. AC arc lights were avail-
able, but they tended to hum, just as the PLEXOS is the worlds leading energy simulation
reactors did in Dalys Theatre. software that delivers optimised solutions through
There were a total of 200 dimmers IRUHFDVWLQJ,WLVIXOO\FXVWRPLVDEOHDQGFRQJXUDEOHIRU
controllable from the Hub board. The ma- your needs.
jority of these were saturable core reactors
having capacities of up to 15 kW each.
Resistance dimmers for such large loads Because knowledge is power.
would have been horribly energy wasteful,
and their physical size would have made
them ponderous as well. Some of the le-
vers on the board did control resistance
dimmers for smaller, specialized lighting
loads up to 2,000 W. These probably op-
erated on ac as well since they were sup-
plying incandescent lighting loads.
Interestingly, the use of archaic re-
sistance dimmers continued for a very
long time during the 20th century in
the lighting of major Broadway shows
because the old Broadway theaters
were still supplied with dc, and the the-
ater owners were not inclined to pay for
new ac services. As a result, resistance
dimmers were the only devices that
could be used to control the stage
lighting. This situation did not begin
to change until public demand forced
the owners to provide air-conditioning
following the inclusion of such in the
new theaters at Lincoln Center built dur-
ing the early 1960s.
Consolidated Edisons main dc sup-
ply facility for the Times Square area
was the West 39th Street rotary con-
verter substation, and it was not able to
be retired until 1977.

The Thyratron-Reactor
Dimmer
The saturable core reactor dimmers
eliminated the huge energy losses (as
energyexemplar.com

figure 2. A typical thyratron-reactor dimmer schematic. (Image from Electric
Motors in Industry, by Shoults, Rife, and Johnson, Wiley, 1942.)
heat) associated with resistance dim-
mers, but they still required a fairly
large rheostatic control for their opera-
tion. The development of vacuum tubes
for use in radios during the 1920s led
to the possibility of similar tubes be-
76 IEEE power & energy magazine
ing used to supply adjustable dc for
the control winding of a saturable core
reactor, with the tube, in turn, then be-
ing controlled by means of a very low
power potentiometer type of device
in its grid circuitry.
OmniMax
How a Street Says Welcome
All the benefits of LED without sacrificing the light you love.
Its the right lightnot just the
brightest lightthat creates
inviting environments. OmniMax
was designed to mimic the HID
bulbs it replaces in both size and
performance. It screws directly
into any existing fixture, and takes
advantage of the lamps optics,
as they were designed.
It turned out that a type of tube re-
cently developed by Albert Hull, a sci-
entist working at the General Electric
Company (GE) plant in Schenectady,
New York, was perfect for this type of
application. This was the thyratron tube.
Besides being a scientist by occupation,
Hull was something of a Greek scholar.
Therefore, he decided to use the Greek
language to come up with the name for
this new tube. In Greek, the suffix tron
means an instrument, and the prefix
thyra (theta-upsilon-rho-alpha) means
a door, which referred to the gating ac-
tion of this three-element tube in control-
ling the flow of current through it.
Hull went on to name other recently
developed tubes in a similar manner.
The thyratron is a gas-filled triode, and
Hull decided to call a gas-filled diode
tube a phanotron, where phano (to
appear) was a reference to the appear-
ance of a bluish glow in the tube when
operating. Likewise, he named a high
TM
Experience smart technology,
guaranteed reliability, and the
best overall value in LED retrofits.
Special introductory pricing
available to select customers.
evluma.com/omnimax
january/february 2017
vacuum diode used for rectification advantage, however, resulted from the ing a single master controller that fed
a kenotron, where kenos means fact that the controlling device was a the individual dimmer controllers.
empty space, a reference to the high very small, very low power potenti- This presetting technique remained
vacuum in the tube. Finally, a common ometer. This meant that more than one in use for stage lighting control un-
three-element vacuum tube used exten- pot could be provided for each dim- til the development of computer-type
sively in radio circuitry was christened a mer so that the desired dimmer setting lighting boards during the late 20th cen-
pliotron, from pleion, which means could be preset for upcoming cues. The tury. That innovation meant that a huge
more, a reference to the tubes use as changeover from one cue to the next number of upcoming cues could be stored
an amplifier. could then be made simply by operat- in the computer memory. Manual preset
Figure 2 is a schematic of typical
thyratron-reactor dimmer circuitry from
this era. A resistive potentiometer (MCB)

making life visibly safer

is shown at the far left. Apparently, MCB
stands for main control board, which was
the location for this device since it was
the means by which the light board op-
erator controlled the dimmer.
A dual-plate kenotron tube (K) is
shown that acted as a differential recti-
fier. It combined the signal from the
pot with a feedback signal to derive
a resultant signal for the grid of the
SpanLite TM

thyratron tube (T). Grid capacitors (FI Self-Illuminated

and FB) introduced a time constant that Power Line Marker
determined how long into a half cycle
of the impressed ac it would take for Meets FAA Advisory
the thyratron to conduct and feed cur- Circular 70/7460-1L
rent to the control winding of the satu- (Dec. 2015)
rable core reactor. This, of course, was
a function of the position of the control Installs directly on
live lines up to 500 kV
potentiometer. The phanotron tube (P)
handled the inductive current in the
control winding during the alternate
half cycles when the thyratron was not
conducting. At the right are shown the
saturable reactor dimmer (SR), a line
contactor (LC), and the lamp load be-
ing dimmed (L).
The control board could be located
anywhere where it was convenient for
the operator to see the stage as he ran
the lighting cues. Most often this was
someplace in the house so that the op-
erator had the same view as the audi-
ence. This location usually had not
been possible with resistance dimmer
or simple reactance dimmer boards
Bird LED Obstruction Low Line
because of the plethora of high current
Diverters '%&0/ Flags & Markers
electrical connections to the boards. As
such, they were stuck backstage where
the operator had a very limited view of
the stage itself.
Flexibility in control board loca-
tion was a big advantage of thyratron- 



5

-.0#!&!,*
6709

reactor installations. Another huge

january/february 2017 IEEE power & energy magazine 77


figure 3. The GE dimmer board backstage at the Albany, New York, Palace
Theatre. (Photo courtesy of Thomas. J. Blalock.)
Advanced modeling and analysis features include:
Dynamic Security Assessment Software
Offering a simple, user-friendly interface
with extensive analysis options, DSAToolsTM
is a suite of software tools for power system
analysis. The software is designed for
applications in both off-line studies and
on-line dynamic security assessment.
Experience and support:
www.dsatools.com
VSAT
Contact us:
dsainfo@powertechlabs.com
78 IEEE power & energy magazine
lighting boards most often were man-
ufactured so as to allow only two, five,
or ten cues to be set in advance, since
there still were limits to the number of
potentiometers that could be provided.
The high current portions of a thy-
ratron-reactor dimmer circuit could be
located in any out-of-the-way space,
such as down in the basement near the
main electrical service. This included
the saturable core reactors themselves,
the remotely operated line contactors,
and the panels containing the thyra-
tron, kenotron and phanotron tubes.
General Electric
In the early 20th century, GE became in-
volved with the manufacture of switch-
boards for the control of stage lighting.
These indeed were just switchboards,
consisting of open knife switches and
fuse holders mounted on a thick slate
Comprehensive capabilities for voltage, transient, small signal, and
frequency stability assessment
Integration with EMS/WAMS for on-line DSA to assess security
of real-time system conditions
Computation of secure operation regions (stability limits)
Recommendation of remedial actions
Prominent provider of the on-line DSA technology with over
40 installations worldwide including 7 out of 9 ISOs in North
America
Consulting services for custom software and model development,
system studies, and training
TSAT SSAT
81022-0005
january/february 2017
panel. GE also manufactured resistance In 1931, GE installed a very modest of the possibility of having the control
dimmers, but these usually were sepa- nine-dimmer thyratron-reactor board at board located in the auditorium, and,
rately mounted above the board itself. the new Plaza Theatre in Schenect- instead, it was installed backstage just
These switchboards were constructed ady (demolished in 1964). In that same as the earlier manual-type boards had
exactly the same as industrial switch- year, a more significant 48-dimmer been. The reason for this is not known
boards of that era that were manufac- installation was made at the new Pal- (see Figure 3).
tured by GE at its Schenectady, New ace Theatre in Albany, New York. In The Palace Theatre still exists and is
York, plant. this case, the advantage was not taken still in use today but the thyratron-reactor
Such a stage switchboard still ex-
ists (but not in use) in the lobby of the
restored Colonial Theatre in Pittsfield,
Massachusetts. The board dates from
1912 and was refurbished for display
along with the restoration of the theater. Build Your
Both the board and the dimmers still
display GE nameplates. One-Line in the Field
In the mid-1920s, GE acquired
the Trumbull Electric Manufactur- as You Collect the Data
ing Company of Plainville, Connecticut,
which manufactured auto-transformer
EasyPower OnSite is the first mobile data
dimmer boards for stage lighting con- EasyPower
Power
OnSite
collection tool designed from the ground up to
trol. The auto-transformer was a logi- TM
use a touch interface for collecting power
cal device to adapt for use as a stage system data and building one-lines in the field.
lighting dimmer. It was simple and
rugged and did not require an auxil- OnSite simplifies your
iary source of dc for its operation as data collection process
with the saturable core reactor dimmer. and improves produc-
It was, however, still quite large and tivity by dramatically
heavy due to the need for a laminated reducing the time it takes
to collect data and create
steel core.
the corresponding
As a result of the work of Albert Hull,
one-line models.
GE held the early patents on the thyra-
tron tube. Consequently, GE decided to
enter into the development of thyratron-
reactor dimmers. Significant early de-
velopment work also had occurred at
the Schenectady GE plant regarding the
use of saturable core reactors for various
industrial control applications.
The early development of thyratron-
reactor dimmers was carried out at the Simplify the Process
Schenectady plant by research engi- Organize Data Build One-Line
Collect Data Input Data
neers Allen Bailey and Harold LaRoque System Analysis
and, later, by Dudley Chambers and
Elbert Schneider. All four eventually BEFORE
received patents for their contributions.
Ultimately, GE only made ten installa- Collect Data &
Build One-Line System Analysis
tions of thyratron-reactor lighting con- Transfer Data and One-Line
trol in theaters, during the late 1920s AFTER
and through the mid-1930s. The first
such installation was for the Chicago
Civic Opera House in 1929. This was a
relatively simple two-scene preset board
Learn more at

www.EasyPower.com/OnSitePE
controlling 140 dimmers and having a Power made easy.

total load capacity of 1,250 kW.

january/february 2017 IEEE power & energy magazine 79

 IF YOU HAVE
A CHALLENGE
WE HAVE A SOLUTION

For over 100 years Slacan Industries has been Canadas largest
designer and manufacturer of Transmission, Distribution, and
Communication products. We also design and manufacture
OEM products, Substations and Structural Steelwork for the
Utility Market. We distinguish ourselves by providing customer
satisfaction, innovative design and industry experience. figure 4. The control board in the auditorium of the Earl
Carroll Theatre. [Photo courtesy of the Museum of Innova-
tion and Science (miSci) Archives, Schenectady, New York.]
SLACAN
Industries Inc.
145 Roy Blvd, P.O. Box 2500
Brantford, Ontario N3T 6E3
519.758.8888 slacan.com

The Society of Corporate Compliance & Ethics

UTILITIES & ENERGY

COMPLIANCE & ETHICS CONFERENCE
NEW LOCATION IN

2017
FEBRUARY 1922, 2017
WASHINGTON DC

Questions: katie.burk@corporatecompliance.org SAME GREAT

EDUCATION AND
The utilities and energy NETWORKING
industries are highly regulated.

Take advantage of the

opportunity to discuss your
concerns and enjoy sharing
and exchanging ideas with
others facing the same
regulations.
figure 5. The basement dimmer racks in the Earl Carroll
Theatre with a saturable core reactor at upper left and tube
panels at the center. [Photo courtesy of the Museum of Inno-
vation and Science (miSci) Archives, Schenectady, New York.]

Learn more and register at board is now long gone. The author did have the opportunity to
corporatecompliance.org/utilities see it back in the 1970s but, by then, it was no longer in use.
Today, unfortunately, the entire installation is gone.

80 IEEE power & energy magazine january/february 2017

 Also in 1931, a GE instal- (routinely referred to as the
lation was made at the new electrician in theatrical par-
Earl Carroll Theatre, located lance) was called the color
across 50th Street from the conductor, and he appeared
Roxy Theatre in Manhattan. in full evening dress.
Earl Carroll was a theatrical This was not a huge the-
impresario along the lines of ater (or stage), so a modest
Roxy Rothafel. His female 50 dimmers were installed,
revue called Vanities was and almost half of them were
based on Florence Ziegfelds devoted to the auditorium
highly popular Follies. lighting, which was really a
The Earl Carroll Theatre part of the show. The total
(actually the second theater connected lamp load was
of this name on this site) was 367 kW, and the thyratron-
an Art Deco extravaganza. reactor control required a to-
The auditorium lighting was tal of only 200 W (just 4 W
almost as elaborate as that per dimmer). The arrange-
for the stage itself. Light- ment of the board was as a
ing fixtures colored in red, two-scene preset control (see
green, and blue were capable figure 6. Earl Carroll explaining the operation of power Figure 4). The saturable core
of being additively mixed distribution panels to three of his employees. [Photo reactors and the tube panels
to create any color of the courtesy of the Museum of Innovation and Science (miSci) were installed in racks lo-
rainbow (and white). This Archives, Schenectady, New York.] cated in the basement (see
concept had been developed Figures 5 and 6).
by GE and was called colorama. The actor dimmers was located in a pit out- This elaborate theater existed as
control board for the GE thyratron-re- side of the orchestra pit. The operator such for only eight years. It was then
 OIL
Introducing Dry-Type
Series Resonant Test
Sets with low PD
levels for testing MV
cables or generators.

figure 7. The Radio City Music Hall control console outside of the orchestra
pit. [Photo courtesy of the Museum of Innovation and Science (miSci) Archives,
Schenectady, New York.]

figure 8. Operating the Radio City Music Hall control console. (Photo from a
Rating-30kV 200kVA Patent Pending 1947 Radio City Music Hall brochure.)

Please contact us for all
your High Voltage AC, DC,
Impulse Test, Measurement
Equipment and Service
Requirements.
EHVTEST.COM/DRY-TYPE
renovated to become retail space, and
some of the ornate auditorium decor
remained in place above a false ceiling
in a Woolworths store until the build-
ing was demolished in 1990.
The Westinghouse Electric Manufac-
turing Company of Pittsburgh, Pennsyl-
vania, entered the electronic light dim-
mer field in the early 1930s by teaming
up with the Ward Leonard Electric
Company to introduce a rival device that
circumvented GEs patent on the use of
the thyratron tube in conjunction with
a saturable core reactor. This involved
using a small saturable reactor that, in
turn, controlled the main lamp-dimming
reactor. A phanotron (rectifier) tube was
used in conjunction with this small re-
actor, and the combination was referred
to (for reasons not now apparent) as a
hysterset. A thyratron tube was not nec-
essary in this control scheme.

82 IEEE power & energy magazine january/february 2017

 In 1932, a large installation of this
type was made by Westinghouse to
control the stage lighting in the Center
Theatre (now long gone), which was a
part of the mammoth Rockefeller Cen-
ter complex being constructed in Man-
hattan at that time. However, in the
same year, GE installed what would be
their largest thyratron-reactor control
in the world-renowned theater that is
still in operation today, Radio City Mu-
sic Hall, also in Rockefeller Center.
As in the Earl Carroll Theatre, the
control board for Radio City Music Hall
was installed in a smaller pit outside of
the large orchestra pit. Again, of course,
the large racks containing the saturable
core reactors and the control tubes were
located out of the way in the basement.
This control board used a new type of
dimmer controlling device that was re-
ferred to as a solenoid potentiometer.
It replaced the simple resistive potenti-
ometers used previously to control the
signal to the grid of the thyratron tube.
FOR ALL LINES
Complete system
of rating
 Overhead powerlines, underground
cables, and substation equipment
 Real-time dynamic line ratings
 View powerlines, substations and weather
stations in Google Map
ELECTROTECH www.LineAmps.com
ENTERPRISES LLC For further information please contact sales@LineAmps.com +1-703-369-1466
2016, Electrotech Enterprises LLC. All Rights Reserved | Windows, Excel and Google Logo are registered trademarks of Microsoft and Google Inc. respectively.
This new device consisted of a coil of
wire with a movable iron core inside.
Ironically, this basically is the same con-
figuration used at Dalys Theatre back in
the 1880s, which was a total failure. In
this case, however, the solenoid was not
being used to control the lamps directly.
The solenoid potentiometer concept
was developed by a research engineer
at the Schenectady GE plant by the
name of C. (Chauncey) Guy Suits, who
was later awarded a patent for its use
in this regard. He eventually became
the director of the famed GE Research
Laboratory following W.D. Coolidge,
the inventor of the X-ray tube.
Suits also was instrumental in the de-
velopment of a saturable core reactor
control scheme used on the nearly 200-
ft (61-m)-long General Electric sign
at the plant in Schenectady. He used cir-
cuitry of his own invention, which was
capable of creating a timed brightening
and dimming of lamps with no mechani-
cal components. Various portions of the
for PC Windows
sign (with a total of 1,400 lamps) were
illuminated in a sequence that was re-
peated over and over. This sign is still in
place on top of Building 37 of the plant
but no longer exhibits this interesting se-
quence. It is simply lit up at night.
The Radio City Music Hall installa-
tion controlled a total of about 320 dim-
mers, ranging from 2 to 16 kW capacity
each. Sixty or so of these dimmers were
used to control elaborate lighting in the
auditorium (as in the Earl Carroll The-
atre) with lighting fixtures colored in
amber, red, green and blue (the additive
mixing of these approximating white
light). The total connected lighting load
was about 3,300 kW, and the board was
arranged for five-scene preset control
(see Figure 7).
Following this, a GE thyratron-reactor
installation was made at the old Metropol-
itan Opera House located on 39th Street
in Manhattan prior to the construction
of Lincoln Center. This controlled 170
dimmers, ranging from 2 to 12 kW each,
and the total connected lighting load was
836 kW. Interestingly, this installation,
in 1933, replaced the original somewhat
Rube Goldberg resistance dimmer in-
stallation that dated from the construction
of the building in 1903. The final four
(less impressive) GE thyratron-reactor
installations were in theaters in Philadel-
phia, Pennsylvania; Iowa City, Iowa; To-
ronto, Ontario; and Mexico City, Mexico.
Epilogue
Undoubtedly, all of these thyratron re-
actor installations are long gone (as is
the Roxy). The Radio City Music Hall
installation did remain in use until a
major renovation of the massive theater
in 1999. At that time, all of the reactor/
tube racks were scrapped, but the con-
trol board itself was saved, its pit in the
auditorium simply covered over; as of
April 2016, it is still there.
During the 1940s, thyratron tubes
were developed of sufficient capacity
that the tubes themselves could be used
to control the dimming of incandes-
cent lamp loads. Two tubes were used
in an inverse-parallel connection so
that each tube conducted on opposite
half cycles of the impressed ac. By the
1960s, however, the solid-state version
of this concept had come into use. De-
vices known as silicon-controlled rec-
tifiers were connected in pairs, just as
large thyratron tubes had been, to dim
large lamp loads. Stage lighting control
today consists of new generations of
such solid-state devices used as dim-
mers, along with computerized control
to replace the old manual preset control
scheme used for boards such as that in
Radio City Music Hall (see Figure 8).
Acknowledgment
The author wishes to thank Marc
Grimshaw of the International Alli-
ance of Theatre and Stage Employees,
Local No. 1, for information regarding
the present status of the original Radio
City Music Hall lighting control board.
For Further Reading
The Roxy theatre, Marquee, vol. 11,
pp. 130, no. 1, 1979.
J. R. Manheimer and T. H. Joseph,
Electronic tube control for theatre
lighting, presented at the fall meet-
ing of the Society of Motion Picture
Engineers, 1934.
J. E. Rubin, The technical devel-
opment of stage lighting apparatus in
the United States, 19001950, Ph.D.
dissertation, Stanford Univ., Stanford,
CA, 1959.
E. D. Schneider, Thyratron-reactor
lighting control, AIEE Trans., vol. 57,
pp. 328334, June 1938.
L. C. Brenneman, Radio city music
hall: A technical discussion, Marquee,
vol. 31, no. 3, pp. 1823, 1999.
T. J. Blalock, Edisons direct cur-
rent influenced Broadway show light-
ing, IEEE Power Eng. Rev., vol. 22,
no. 10, pp. 3637, Oct. 2002.
p&e
society news

2016 General Meeting

paving the way for grid modernization in Boston

M
MORE THAN 3,200 IEEE POWER
& Energy Society (PES) members and
energy industry enthusiasts converged
in Boston on 1721 July 2016 to experi-
ence the next chapter in the Societys
long history of hosting exceptional
meetings and events. Similar to the
2015 General Meeting (GM) in Denver,
the expectation from attendees was for
an outstanding opening plenary session
with prestigious speakers, including
Damir Novosel, PES president and
president of Quanta Technology
Miroslav Begovic, PES past pres-
ident and head professor of the
Texas A&M Department of Elec-
trical and Computer Engineering The opening reception was held at the Boston Public Library.
Babak Enayati, chair of the 2016
IEEE PES GM Local Organiz- tendees had the opportunity to be pres-
ing Committee and lead R&D ent for additional panel sessions cover-
engineer with National Grid ing topics such as transformer resiliency
Cheryl LaFleur, commissioner and physical security, the impact of en-
at the Federal Energy Regula- vironmental regulations on power mar-
tory Commission kets, power grid resilience, and so much
Marcy Reed, president of National more. Paper and poster sessions are also
Grid (Massachusetts jurisdiction) a popular attraction for GM attendees.
David L. Geier, vice president of From an overview of the best confer-
electric transmission and system ence papers on power system stability
engineering with San Diego Gas and protection to an emerging technolo-
& Electric gies poster session, the prospects to get
Juan de Bedout, chief technol- PES President Damir Novosel welcomes involved were numerous. And that was
ogy officer for General Electrics everyone at the opening session. just the first couple days. Days two and
Energy Connections business. three were full of panels and poster ses-
PES President Damir Novosel gave various nominations and appointments sions on microgrids, grid planning, DER,
a rousing PES member meeting open- for the upcoming election. Afterward, big data, sustainable energy, smart build-
ing session and provided the crowd Babak Enayati introduced the plenary ings, smart cities, flexible energy sys-
with a state of the Society address. Past session panelists, each of whom pro- tems, and energy policy.
President Miroslav Begovic then pro- vided detailed industry presentations To cap the event, PES announced the
vided support and introductions for the followed by question-and-answer ses- recipients of its Society-level awards, all
sions for the attending assembly. of which were recognized and honored
Digital Object Identifier 10.1109/MPE.2016.2620638
Following the GMs theme, Paving during a formal ceremony in Boston.
Date of publication: 2 February 2017 the Way for Grid Modernization, at- These Society-level awards recognize

january/february 2017 IEEE power & energy magazine 85

 s Meets FAA Specifications!
Color Size Shape!
International Orange
s Tested and approved by
major power companies!
s Thousands still in service
after 40 years
s Universal attaching! Fits any
wire .1" to 1"!
s Installs in 5 minutes!
s Withstands hail! Attendees listen intently to the speakers during the plenary session.
s No maintenance! Does not
slip, oscillate, chafe, cause
electrolysis or harmonic
vibration.
s Ships in halves nested.
9, 12, 20, 24, 30, and 36" balls
& special sizes available

Call now 573-796-3812

ext. 2001
Fax 573-796-3770
www.tanawiremarker.com
TANA WIRE MARKERS
P.O. Box 370, California, MO 65018

The awards dinner was well attended.

A history
that spans
the past century.

There were many discussions during the poster session.

and credit important technical, educational, and service contributions by

the global power and energy community. PES members can nominate
their colleagues for these awards by visiting http://www.ieee-pes.org/
A vision pes-communities/awards.
that spans As another GM came to a close, PES members and many others
looked forward to the 2017 IEEE PES GM in Chicago, scheduled for
the next century. 1620 July. The theme for the 2017 GM is Energizing a More Secure,
Resilient, and Adaptable Grid. Visit http://pes-gm.org/2017/ for more in-
formation and to participate in a call for papers and other opportunities.

www.DeltaStar.com p&e

1-800-368-3017
calendar
T
THE IEEE POWER & ENERGY
Societys (PESs) website (http://www
.ieee-pes.org) features a meetings sec-
tion that includes calls for papers and
additional information about each of
the PES-sponsored meetings.
May 2017
IEEE International Conference on
Electrical Machines and Drives (IEM-
DC 2017), 2124 May, Miami, Florida,
United States, contact Dr. Dan Ionel,
dan.ionel@ieee.org, http://iemdc.org/
June 2017
IEEE PowerTech Manchester (Power
Tech 2017), 1822 June, Manchester,
United Kingdom, contact Prof. Jovica
Milanovic, milanovic@manchester
.ac.uk, http://ieee-powertech.org/
IEEE PES PowerAfrica (PowerAfri-
ca 2017), 2530 June, Accra, Ghana,
contact Dr. Eric Kuada, dr.eric.kuada@
ieee.org
IEEE Second International Confer-
ence on DC Microgrids (ICDCM
2017), 2729 June, Nuremberg, Ger-
many, contact Prof. Roger Dougal,
dougal@cec.sc.edu, http://www.icdcm.co/
Digital Object Identifier 10.1109/MPE.2016.2620678
Date of publication: 2 February 2017
january/february 2017
PES meetings
for more information, www.ieee-pes.org
July 2017
IEEE PES General Meeting (GM
2017), 1620 July, Chicago, Illinois,
United States, contact Joseph Sva-
chula, joseph.svachula@ComEd.com,
http://pes-gm.org/2017/
August 2017
IEEE Electric Ship Technologies
Symposium (ESTS 2017), 1517 Au-
gust, Washington, D.C., United States,
contact Dr. Scott Sudhoff, sudhoff@
purdue.edu, http://ests17.mit.edu/
September 2017
IEEE PES Innovative Smart Grid
Technologies LA (ISGT LA 2017),
2022 September, Quito, Ecuador,
contact Gabriel Arguello, garguello@
cenace.org.ec, http://ieee-isgt-latam.org/
IEEE PES Innovative Smart Grid
Technologies Europe (ISGT Europe
2017), 2629 September, Torino, Ita-
ly, contact Prof. Gianfranco Chicco,
gianfranco.chicco@polito.it
December 2017
IEEE PES Innovative Smart Grid
Technologies Asia (ISGT Asia), 36
December, Auckland, New Zealand,
contact Dr. Ramesh Rayudu, Ramesh
.Rayudu@vuw.ac.nz
January 2018
IEEE PES 2018 Joint Technical
Committee Meeting (JTCM 2018),
711 January, Jacksonville, Florida,
United States, contact Solveig Ward,
sward@quanta-technology.com, www
.pestechnical.org
April 2018
IEEE PES Transmission and Distri-
bution Conference and Exposition
(T&D 2018), 1619 April, Denver, Col-
orado, United States, contact Tommy
Mayne, mayne25@charter.net, http://
www.ieeet-d.org/
August 2018
IEEE PES General Meeting (GM
2018), 510 August, Portland, Or-
egon, United States, contact Don Hall,
donald.hall@pepcoholdings.com
October 2018
IEEE PES Asia-Pacific Power & En-
ergy Engineering Conference (AP-
PEEC 2018), 710 October, Sabah,
Malaysia, contact Dr. Zuhaina Zakaria,
zuhaina@ieee.org
IEEE PES Innovative Smart Grid
Technologies Europe (ISGT Europe
2018), 2125 October, Sarajevo, Bos-
nia and Herzegovina, contact Prof.
Senad Huseinbegovic, shuseinbegovic@
etf.unsa.ba
p&e
IEEE power & energy magazine 87
 Correction
In the November/December 2016 issue of IEEE Power & Energy Magazine, Figure 2 in [1] shows earthquake damage at
Sylmar that occurred on 9 February 1971. The source of the photos is an IEEE presentation by Wayne Litzenberger and not
Hydro-Quebec.

Reference
[1] R. B. Wadele, Tales of power system failures, IEEE Power Energy Mag., vol. 14, no. 6, pp. 1823, Nov./Dec. 2016.

Digital Object Identifier 10.1109/MPE.2016.2640103

Date of publication: 2 February 2017

2016

REGISTER NOW for the 2017 IEEE PES Innovative

Smart Grid Technologies Conference (ISGT 2017)
April 23-26, 2017 | Crystal Gateway Marriot, Arlington, VA (DC Metro Area)
Experts around the world gather annually at the ISGT North America
Conference to discuss state-of-the-art innovations in smart grid technologies.
The program, as always, will feature special sessions and tutorials on
wide-ranging topics related to grid modernization.

For More Information, visit ieee-isgt.org

Digital Object Identifier 10.1109/MPE.2017.2654738

88 IEEE power & energy magazine january/february 2017

 Sheraton Chicago Hotel and Towers

Energizing a More Secure,

Resilient & Adaptable Grid
Register Now for the 2017 PES General Meeting in Chicago, IL

The 2017 IEEE Power & Energy Society General Meeting will be held from July 16-20, 2017 at
Sheraton Chicago Hotel and Towers. Registration is now open!

The PES General Meeting attracts over 3,400 professionals from every
segment of the electric power and energy industries. It features a
comprehensive technical program, including super sessions, panel
sessions, tutorials, a student program, companion activities, and more!

As always, IEEE PES has put together an outstanding program,

with Super Sessions addressing such topics as:
8
620%,%
4%-21
!-$
0)$
%1)+)%-#7
8
-%0'7
2.0!'%
8
7"%0
!-$
(71)#!+
%#30)27
8
)120)"32%$
%-%0!2).-

%'3+!2).-
-')-%%0)-'
.$%+)-'

and Impacts

.-2
,)11
2()1
1/%#2!#3+!0
%4%-2
,!*%
7.30
/+!-1
-.5
2.
!22%-$
%
+..*
&.05!0$
2.
1%%)-'
7.3
)-

Chicago!

For more information visit: pes-gm.org/2017

Digital Object Identifier 10.1109/MPE.2017.2654739

january/february 2017 IEEE power & energy magazine 89

 in my view (continued from p. 92)

as a question of the systems physical
properties. Just look at the trend toward
prosumers, cooperatives and companies
investing in clean energy.
The level of a systems physical inte-
gration will also depend on who owns
it and the interest in integration. When
asking what an optimal energy system
looks like, we have to give equal con-
sideration to how much biomass can
contribute in determining what kind of
ownership should be promoted. That
opens up a discussion where the social
scientist isnt the strange kid in the cor-
ner getting a project work package to
solve the issue with that human factor.
Second, as renewable technologies
such as wind and solar have matured,
it has become increasingly evident that
some of the main barriers to increasing
their share of the energy market are non-
technical. The European Technology and
Innovation Platform for Wind Energy
(ETIPWind) recently published its stra-
tegic research and innovation agenda,
and one of the five priority pillars iden-
tified is industrialization. While indus-
trialization requires technical solutions
for the production, logistics, and main-
tenance of wind turbines, it has a sub-
stantial organizational component.
Or take the discussions of energy
democracy, which promotes the local
ownership of energy production. This
idea holds significant appeal from a
social science perspective. However,
local energy communities risk creating
the problem of prioritizing the local so-
lution over the global one, which could
ultimately make the energy system less
sustainable and less flexible. Hence,
you also need an engineer in the room.
Last, but not least, ESI is inherently
technology agnostic; it doesnt favor
any particular technology, and, as a
consequence, the social scientist does
not easily become the servant of a par-
ticularly technological interest group,
be it district heating, biomass, solar, or
batteries. The discussion is taken out of
the technology-specific context and put
into a much broader field where orga-
nizational, sociological, political, and
technical levels of expertise are brought
into play.
To achieve a higher level of system
flexibility at the lowest possible cost, the
integration of the different disciplines
consequently needs to be pursued more
vigorously to enable the best possible in-
tegration of the different components of
the energy system.
p&e

Holcombe Department of Electrical and Computer Engineering

Faculty Search for Power and Energy Faculty
at the Charleston Innovation Campus in N. Charleston, SC and the Clemson University main campus in Clemson, SC.
Applications and nominations are sought for multiple faculty positions in electrical power engineering at two locations: Clemson Universitys new Zucker Family Graduate
Education Center at its Charleston Innovation Campus in North Charleston, SC (http://www.clemson.edu/restoration/) and the main Clemson campus in Clemson, SC
KWWSZZZFOHPVRQHGXHFH(DFKSRVLWLRQZLOOEHOOHGZLWKDQLQGLYLGXDOZKRPHULWVWKHUDQNRI$VVRFLDWHRU)XOO3URIHVVRUZLWKWHQXUHDQGZKRKDVDQDWLRQDOO\UHFRJQL]HG
UHFRUGRIRXWVWDQGLQJVFKRODUVKLS$SSURSULDWHO\TXDOLHGFDQGLGDWHVPD\EHFRQVLGHUHGIRUD7LWOHG3URIHVVRUVKLS7KHFDQGLGDWHVWHDFKLQJDQGUHVHDUFKVKRXOGHQFRPSDVVD
broad range of topics related to power systems, electric machines and drives, power electronics, energy storage, energy analytics, and wind and solar power integration. A solid
XQGHUVWDQGLQJRIWKHF\EHULQIUDVWUXFWXUHUHODWHGDUHDVRISRZHUV\VWHPVVXFKDVHPEHGGHGV\VWHPVF\EHUVHFXULW\QHWZRUNLQJDQGUHPRWHVHQVLQJLVDOVRLPSRUWDQW7KHSHUVRQV
OOLQJWKHSRVLWLRQVLQERWK1&KDUOHVWRQ6&DQG&OHPVRQ6&ZLOOKROGDIDFXOW\SRVLWLRQLQWKH+ROFRPEH'HSDUWPHQWRI(OHFWULFDODQG&RPSXWHU(QJLQHHULQJDW&OHPVRQ
University.
7KH+ROFRPEH'HSDUWPHQWRI(OHFWULFDODQG&RPSXWHU(QJLQHHULQJLVRQHRIWKHODUJHVWDQGPRVWDFWLYHGHSDUWPHQWVDW&OHPVRQZLWKIDFXOW\PHPEHUVDSSUR[LPDWHO\
XQGHUJUDGXDWHVDQGJUDGXDWHVWXGHQWV7KH&KDUOHVWRQ&DPSXVKRXVHVD0SRZHUIDFLOLW\LQLWLDWHGLQIURPD0'HSDUWPHQWRI(QHUJ\JUDQW7KHIDFLOLW\
includes the SCE&G Energy Innovation Center which contains the worlds most-advanced wind-turbine drivetrain testing facility capable of full-scale highly accelerated
PHFKDQLFDODQGHOHFWULFDOWHVWLQJRIDGYDQFHGGULYHWUDLQV\VWHPVIRUZLQGWXUELQHV7KH6&( *(QHUJ\,QQRYDWLRQ&HQWHUDOVRKRXVHVWKH'XNH(QHUJ\(OHFWULFDO*ULG5HVHDUFK
,QQRYDWLRQ DQG 'HYHORSPHQW &HQWHU D IDFLOLW\ ZLWK UHDOWLPH VLPXODWLRQ DQG 09$ KDUGZDUHLQWKHORRS FDSDELOLW\ 7KH PDLQ FDPSXV LQFOXGHV VWDWHRIWKHDUW UHDOWLPH
simulation facilities for research in intelligent control of the electric grid, a modern power-electronics laboratory, and a strong undergraduate and graduate emphasis in power
systems.
&OHPVRQ8QLYHUVLW\LVWKHODQGJUDQWLQVWLWXWLRQIRUWKH6WDWHRI6RXWK&DUROLQDHQUROOLQJDSSUR[LPDWHO\XQGHUJUDGXDWHVDQGJUDGXDWHVWXGHQWV)LYHLQWHUGLVFLSOLQDU\
FROOHJHVKRXVHVWURQJSURJUDPVLQDUFKLWHFWXUHHQJLQHHULQJVFLHQFHDJULFXOWXUHEXVLQHVVVRFLDOVFLHQFHVDUWVDQGHGXFDWLRQ$IDFXOW\RIDQGVWDIIRIVXSSRUW
XQGHUJUDGXDWHGHJUHHRIIHULQJVPDVWHUVGHJUHHSURJUDPVDQG3K'SURJUDPV$QDQQXDORSHUDWLQJEXGJHWRIDSSUR[LPDWHO\PLOOLRQDQGDQHQGRZPHQWRI
PLOOLRQIXQGSURJUDPVDQGRSHUDWLRQV7KH8QLYHUVLW\KDVH[WHUQDOO\IXQGHGUHVHDUFKH[SHQGLWXUHVRIPLOOLRQSHU\HDU5HVHDUFKDQGHFRQRPLFGHYHORSPHQWDFWLYLWLHVDUH
HQKDQFHGE\SXEOLFSULYDWHSDUWQHUVKLSVDWLQQRYDWLRQFDPSXVHVDQGUHVHDUFKDQGHGXFDWLRQFHQWHUVORFDWHGWKURXJKRXW6RXWK&DUROLQD&OHPVRQ8QLYHUVLW\LVUDQNHGUG
DPRQJQDWLRQDOSXEOLFXQLYHUVLWLHVE\861HZV :RUOG5HSRUW
$SSOLFDQWV PXVW KDYH DQ HDUQHG GRFWRUDWH LQ (OHFWULFDO (QJLQHHULQJ RU D FORVHO\ UHODWHG HOG$SSOLFDQWV VKRXOG VXEPLW D FXUUHQW FXUULFXOXP YLWDH VWDWHPHQWV RI UHVHDUFK
DQG WHDFKLQJ VWUDWHJ\ DQG D PLQLPXP RI YH UHIHUHQFHV ZLWK IXOO FRQWDFW LQIRUPDWLRQ$SSOLFDWLRQ PDWHULDO VKRXOG EH VXEPLWWHG HOHFWURQLFDOO\ DW WKH IROORZLQJ:HE OLQN
KWWSVDSSO\LQWHUIROLRFRP
7KHFRYHUOHWWHUVKRXOGLQGLFDWHZKHWKHUWKH&KDUOHVWRQRU&OHPVRQORFDWLRQRUERWKLVRILQWHUHVWWRWKHDSSOLFDQW7RHQVXUHIXOOFRQVLGHUDWLRQDSSOLFDQWVPXVWDSSO\E\
)HEUXDU\KRZHYHUWKHVHDUFKZLOOUHPDLQRSHQXQWLOWKHSRVLWLRQLVOOHG
Clemson University is an AA/EEO employer and does not discriminate against any person or group on the basis of age, color, disability, gender, pregnancy, national origin, race, religion, sexual orientation,
veteran status or genetic information. Clemson University is building a culturally diverse faculty and staff committed to working in a multicultural environment and encourages applications from minorities
and women.

90 IEEE power & energy magazine january/february 2017

advertisers index
The Advertisers Index contained in this issue is compiled as a service to our readers and advertisers: the
publisher is not liable for errors or omissions although every effort is made to ensure its accuracy. Be sure
to let our advertisers know you found them through IEEE Power & Energy Magazine.

Company page# URL Phone

ASPEN, Inc. CVR 4 www.aspeninc.com +1 650 347 3997

Battcon 2017 84 www.battcon.com

Bigwood Systems, Inc. 12 www.bigwood-systems.com +1 607 257 0915

Cantega Technologies, Inc. 74 www.greenjacketinc.com +1 866 464 7996

Clemson University 90 www.clemson.edu

CYME CVR 2 www.cyme.com +1 800 361 3627

Delta Star 86 www.DeltaStar.com +1 800 368 3017

DIgSILENT GmbH 3 www.digsilent.com

EasyPower LLC 79 www.EasyPower.com/OnSitePE +1 503 655 5059

Electrotech Enterprises LLC 83 www.LineAmps.com +1 703 369 1466

Energy Exemplar 75 energyexemplar.com

ETAP 7 www.etap.com +1 800 477 ETAP

Evergreen High Voltage 82 ehvtest.com/dry-type

Evluma LED Lighting 76 www.evluma.com/omnimax

Manitoba HVDC Research Centre 11 www.pscad.com +1 204 989 1240

Metglas, Inc. 73 www.metglas.com +1 800 581 7654

Neplan AG 5 www.neplan.ch +41 44 914 36 66

NR Electric Co., Ltd 81 www.nrec.com +86 25 87178888

P & R Technologies 77 www.pr-tech.com +1 800 722 8078

Power Engineers 15 powereng.com/distribution

Powertech Labs Inc. 78 www.dsatools.com +1 604 590 7500

PowerWorld Corporation 14 www.powerworld.com +1 217 384 6330

RTDS Technologies, Inc. 6 www.rtds.com +1 204 989 9700

Siemens Power Technologies International CVR 3 usa.siemens.com/360-transmission-

planning

SKM Systems Analysis, Inc. 13 www.skm.com +1 800 500 4SKM

Slacan Industries, Inc. 80 slacan.com +1 519 758 8888

Tana Wire Marker 86 www.tanawiremarker.com +1 573 796 3812

The Society of Corporate Compliance & Ethics 80 corporatecompliance.org/utilities

V & R Energy Systems Research 9 www.vrenergy.com +1 888 449 8881

445 Hoes Lane, Piscataway, NJ 08854

IEEE power & energy magazine representative

Erik Henson
Naylor Association Solutions
+1 352 333 3443, fax: +1 352 331 3525
ehenson@naylor.com
Digital Object Identifier 10.1109/MPE.2016.2651218

january/february 2017 IEEE power & energy magazine 91

 in my view
F
FLEXIBILITY IS KEY FOR A SYS-
tem under strain. When the city metro
breaks down, you require other means
of transportation. Before a big confer-
ence, you might need staff to work out-
side regular office hours. And when
the production of electricity is depen-
dent on how the wind blows or the
sun shines, something or someone has
to adapt.
One of the best ways to increase
your flexibility is to use the entire port-
folio of available resources. The metro
is down? Walk; take the bus, the train,
your bike, or your car. Short on staff for
next weeks conference? Ask HR, com-
munication, planning, or someone else
if you can use some of their people;
youll pay it back later. Are we getting
too much electricity from wind? Use it
for heating or get people to turn on ap-
pliances or change to electric cars.
Flexibility is largely about systems,
so its no surprise that, in recent years,
theres been increasing interest in ap-
plying systems thinking to energy re-
search. As a policy adviser, Ive been
involved first hand in the transition
toward a renewable-based energy sys-
tem with its large share of intermittent
energy sources, such as wind and solar.
The articles in this issue of IEEE
Power & Energy Magazine cover dif-
ferent aspects of the energy system,
including heating, gas, and electricity,
all the way to the consumer and public
acceptance.
Digital Object Identifier 10.1109/MPE.2016.2637118
Date of publication: 2 February 2017
92 IEEE power & energy magazine
flexibility
its more than an engineering challenge
The problem with the system and its
parts is always to ensure that the sum is
greater than the parts, not vice versa. As
Dhaeseleer and colleagues note in their
article in this issue Flexibility Chal-
lenges for Energy Markets, by ignor-
ing systems interactions, a variety of
some well-meant (local) simple targets
counteract and even oppose each other
so that some individual targets may per-
haps be reached, while the overall strate-
gic objective is compromised.
In a slogan format, you could say
that the challenge is to avoid a system
where you have optimal subsystems but
a suboptimal system. The challenge is
partly an engineering one and partly a
political and social one. And although
that isnt how it is usually discussed
and presented (including in the articles
in this issue of the magazine), one
could even (provocatively) argue that
the engineers contribution to enabling
the transition to a sustainable energy
system and society is rather peripheral
compared to the Herculean political
and social efforts needed to enable the
system and its inhabitants to change.
You think power to gas is a challenge?
Try getting a globally binding agree-
ment on CO2 emission reduction.
A real challenge in increasing the
flexibility of the energy system is to
establish a collaboration between the en-
gineers and others involved in the hard
sciences with those involved in policy
studies, social science, and humani-
ties. Policy makers, funding agencies,
and some groups in the research com-
munities have realized this. In Europe,
Mattias Andersson
several of the joint programs under the
European Energy Research Alliance
have initiatives in this area, including
the Joint Programme for Energy Sys-
tems Integration.
Multidisciplinary research sounds
good in theory, but it is difficult to imple-
ment in practice. In reality, research proj-
ects tend to be divided into either techni-
cal projects with limited attention to social
acceptance and other human factors or
social science projects with limited atten-
tion to technical or physical limitations.
A first step to overcoming this di-
vide would be to create a playing field
where all sciences could contribute.
Energy systems integration (ESI) could
be a new field of research that would
enable a deeper integration of the sci-
ences, hard and soft. The white paper
on ESI from the International Institute
for Energy Systems Integration (iiesi.
org) clearly recognizes the importance
of this, even if the list of authors is pri-
marily from the engineering and hard
science side of the aisle.
There are at least three reasons why
ESI can contribute to overcoming this
divide. First, renewable ESI has ar-
rived during a time of transition from a
centralized to a decentralized energy
system. New stakeholders, including
private citizens, cooperatives, compa-
nies and municipalities, and regions and
states have different configurations of
ownership, especially for renewable en-
ergy resources. Thus, ESI is as much a
question of the ownership of the system
(continued on p. 90)
january/february 2017
A 360 View of
Transmission Planning
Capabilities of the new PSS Product Suite

The Siemens PTI PSS Product Suite now provides users with a full
spectrum of customizable engineering tools that span across the entire
transmission planning value chain.
Transmission solution capabilities:
) Enhance PSSE base software with a variety of add-on modules
) Integrate transmission and distribution by importing PSSE network
models into PSSSINCAL
) Grid Data Management: Simplify the management of your network
model data across multiple systems

usa.siemens.com/360-transmission-planning
ASPEN
OneLiner
Short circuit and relay coordination program
trusted by utilities worldwide

ASPEN OneLiner
Powerful, fast, intuitive and easy to use

Applications Advanced features

PRC compliance Simulates breaker-failure conditions
Fault location Models wind and solar plants and VSC converters
Circuit breaker rating Time-distance diagram
Relay setting and testing Vendor-specific distance relay models
System-wide coordination studies Relay coordination checking with stepped events

ASPEN 49 N. San Mateo Dr., San Mateo, CA 94401 USA

650-347-3997 info@aspeninc.com www.aspeninc.com