7PMVNFt/VNCFSt+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
efficiency
> Distributed generation
impact
> Network-wide planning using
AMI/AMR data
> Key reliability assets
configuration 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.cooperpowereas.com
www.cyme.com
AD917001EN-2015
magazine
7PMVNFt/VNCFSt+BOVBSZ'FCSVBSZ
www.ieee.org/power
on the
cover
features
16 Unleashing the Flexibility of Gas
By Steve Heinen, Christian Hewicker,
Nick Jenkins, James McCalley, Mark O’Malley,
Sauro Pasini, and Simone Simoncini
Harnessing Flexibility from
Hot and Cold
By Juha Kiviluoma, Steve Heinen,
Hassan Qazi, Henrik Madsen, Goran Strbac,
Chongqing Kang, Ning Zhang,
Dieter Patteeuw, and Tobias Naegler
43 Unlocking Flexibility
By Emiliano Dall’Anese, Pierluigi Mancarella,
and Antonello Monti
53 The Consumer’s Role in
Flexible Energy Systems
By Geertje Schuitema, Lisa Ryan,
and Claudia Aravena
61 Flexibility Challenges for Energy Markets
By William D’haeseleer, Laurens de Vries,
Chongqing Kang, and Erik Delarue
34 The Triple Bottom Line for Efficiency
By Eoin Casey, Sara Beaini,
Sudeshna Pabi, Kent Zammit,
and Ammi Amarnath
72
columns &
departments
4
8
85
'SPNUIF&EJUPS
-FBEFST$PSOFS
(VFTU&EJUPSJBM
)JTUPSZ
4PDJFUZ/FXT
$BMFOEBS
*O.Z7JFX
DPOUFOUT
©ISTOCKPHOTO.COM/AURIS
©ISTOCKPHOTO.COM/LEOWOLFERT
43
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
mih.psat@gmail.com
445 Hoes Lane, Piscataway, NJ 08854 USA
www.ieee.org/magazines
Melvin I. Olken, Editor-in Chief Emeritus
Carl L. Sulzberger, Associate Editor Emeritus,
History
Geraldine Krolin-Taylor, Senior Managing Editor
Janet Dudar, Senior Art Director
Gail A. Schnitzer and Mark Morrissey, Associate Art Directors
Theresa L. Smith, Production Coordinator
Peter M. Tuohy, Production Director
Felicia Spagnoli, Advertising Production Manager
Dawn Melley, Editorial Director
Fran Zappulla, Staff Director, IEEE Publishing Operations
Editorial Board
J. Feltes, L. Goel,
N. Hatziargyriou,
Tao Hong, B. Johnson,
Ben Kroposki, P. Kundur, N. Lu,
A.P.S. Meliopoulos, M. Miller,
M. O’Malley, N. Ochoa,
C.E. Root, H. Rudnick,
M. Shahidehpour, G.B. Sheblé. C. Smith,
M. Thomas, E. Uzunovic,
S.S. Venkata, J. Wang
S. Widergren
Advertising
Erik Henson
Naylor Association Solutions
+1 352 333 3443, fax: +1 352 331 3525
ehenson@naylor.com
IEEE prohibits discrimination, harassment, and bullying. For more information,
visit http://www.ieee.org/web/aboutus/whatis/policies/p9-26.html.
Promoting Sustainable Forestry
SFI-01681
IEEE Power & Energy Magazine
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 for the
contents rests upon the authors and not upon the IEEE, the Society, or its members. IEEE Operations Center (for orders, subscriptions, address changes): 445 Hoes Lane, Piscataway, NJ 08854 USA. Telephone: +1 732 981 0060, +1 800 678 4333.
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 permissions: Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limits of U.S.
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,
Danvers, MA 01923 USA; 2) pre-1978 articles without fee. For other copying, reprint, or republication permission, write
Copyrights and Permissions Department, IEEE Operations Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. Copyright
© 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
IEEE POWER & ENERGY SOCIETY (PES)
Printed in U.S.A.
The IEEE Power & Energy Society is an organization of IEEE members whose principal interest is the advancement of the science and practice of electric 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
D. Novosel, President
S. Rahman, President-Elect
F. Lambert, Vice President, Chapters
M.P. Sanders, Vice President, Technical Activities
E. Uzunovic, Vice President, Education
M. Crow, Vice President, Publications
T. Mayne, Vice President, Meetings
J.C. Montero Q, Vice President,
Membership & Image
S. Bahramirad, Vice President, New Initiatives/
Outreach
C. Root, Treasurer
J. Bian, Secretary
M.M. Begovic, Past-President
IEEE Division VII Director
A. Rotz
IEEE Division VII Director-Elect
B. Meyer
Region Representatives
D. Diaz, B. Gwyn, R. Krepps,
N. Logic, P. Pabst, C. Wong,
United States
M. Armstrong, Canada
J. Milanovic, Europe, Middle East, & Africa
N. Segoshi, Latin America
D. Sharafi, Asia & Pacific
Standing Committee Chairs
H. Louie, Constitution & Bylaws
C. Root, Finance & Audit
M. Begovic, Nominations & Appointments
S. Brahma, Power & Energy Education
W.K. Reder & P. Sauer, Scholarship Plus
Chapter Representatives
F. Al Dhaheri, W. Almuhtadi, Z. Bo,
R. Cespedes, S. Cundeva, C. Diamond,
G. Gonzalez, J. Khan,
R. Nagaraja, N. Nair, M. Papic, I. Petruzela,
T. Salihy, G.N. Taranto, D. van Hertem,
M.C. Wong, Z. Zakaria
Chapter Committee Chairs
C. Diamond, Electronic Communications
E. Carlsen, Awards & Resources
Y. Chen, Distinguished Lecturer Program
C. Diamond, Chapters Web site
Membership & Image
Committee Chairs
Governing Board
Members-at-Large
V. Madani, Awards & Recognition
J. Hofman, Young Professionals
Open, Ambassadors
I. Riaño S., Social Media
J.C. Montero, Web Site Development
S. Ninalowo, Women in Power
W. Bishop, Marketing
J.C. Montero, Membership Development
C.Y. Chung, B. Enayati,
H. Koch, L. Ochoa
Technical Council
PES Executive Director
Patrick Ryan, +1 732 465 6618,
fax +1 732 562 3881, p.ryan@ieee.org
M.P. Sanders, Chair
F. Rahmatian, Vice Chair
V. Vittal, Secretary
J.H. Nelson, Past-Chair
D. Toland, Web Master
Technical Committee Chairs
A. Schneider, Analytical Methods for
Power Systems
K. Haran, Electric Machinery
W.T. Jewell, Energy Development & Power
Generation
C. Searles, Energy Storage & Stationary Battery
F. Frentzas, Insulated Conductors
T. Koshy, Nuclear Power Engineering
M. Dood, Power System Communications
& Cybersecurity
C. Canizares, Power System Dynamic Performance
J. McBride, Power System Instrumentation
& Measurements
H. Chen, Power System Operation Planning &
Economics
M. Pratap, Power System Relaying & Control
S. Chandler, Smart Building, Load &
Customer Systems
D. Watkins, Substations
R. Hotchkiss, Surge Protective Devices
P. Sullivan, Switchgear
S. Antosz, Transformers
D. Sabin, Transmission & Distribution
Technical Council
Coordinating Committees
D. Houseman, Intelligent Grid & Emerging
Technology
D. Alexander, Marine Systems
D. Lew, Wind & Solar Power
Technical Council
Standing Committees
J.H. Nelson, Awards
F. Rahmatian, Technical Sessions
V. Vittal, Organization & Procedures
T. Burse, Standards Coordination
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 extensions to existing ones. Most notably, further modelling flexibility 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 FLEXIBILity issue discusses creative ways that the
power system can reliably accommodate
the large-scale development of variable
resources and load that fluctuates considerably with weather, price, and behindthe-meter generation. This issue’s guest
editors, Mark O’Malley and Benjamin
Kroposki, collected six articles that discuss the importance of flexibility and how
it can be achieved in ways not always
considered by electric power engineers.
The “Guest Editorial” summarizes the
flexibility theme and introduces articles
that feature the following topics:
✔ the role of the consumer in energy systems
✔ the place of natural gas-fired generation and its fuel supply
✔ the interaction between water
operations and electric power
sectors
✔ the integration and opt i m i zat ion of multienergy system controls
✔ heating and cooling technologies
that provide storage
✔ policies and markets that promote
flexible electric system responses.
It is my hope you find these articles interesting and informative.
Reactors for the Roxy
At 6,000 seats, the Roxy Thea t r e i n
New York City was widely known.
The technology of stage lighting, howDigital Object Identifier 10.1109/MPE.2016.2627098
Date of publication: 2 February 2017
4
IEEE power & energy magazine
IMAGE LICENSED BY INGRAM PUBLISHING
ever, remained unknown to the general
public and most electrical engineers
until an article in this issue’s “History”
column on this topic. The use of reactance-type dimmers
utilizing imaginary
power in the imaginary
world of the stage is
truly captivating. A
special thanks to Tom
Blalock, one of our most
prolific history “History”
authors, who presents
the evolution of stage
lighting to us. The article is the last of Carl
Sulzberger’s remarkable
contributions to IEEE
Power & Energy Magazine as the associate editor for History.
Special Thanks
IEEE Power & Energy
Magazine saw 84 issues with Mel Olken as
editor-in-chief and Carl
Sulzberger as associate editor for the
“History” column. Their extraordinary
service to this magazine and to IEEE
is truly remarkable. Heartfelt thanks to
both of you on behalf
of the IEEE Power &
Energy Society, its officers, and IEEE staff.
On a more personal
note, I am deeply indebted to you for your
guidance and tutelage.
And a special thanks
in advance for the continued support of Mel
Olken, who always professionally and patiently
answers my questions
and provides true leadership as editor-inchief emeritus.
The
technology of
stage lighting
remained
unknown to
the general
public and
most electrical
engineers until
an article in
this issue’s
“History”
column.
Changing of the
Editorial Guard
This issue represents a
changing of the guard
for the magazine’s
january/february 2017
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 Küsnacht Zurich
Phone +41 44 914 36 66 Fax +41 44 991 19 71 www.neplan.ch info@neplan.ch
frick-partner.ch
Your NEPLAN experience
is just one click away
Editorial Board. A sincere thank you
for all the hard work by our outgoing
board members:
✔ Massoud Amin
✔ Andrew Hanson
✔ Stan Horowitz
✔ Ralph Masiello
✔ Keene Matsuda
✔ Arun Phadke
✔ Dick Piwko
✔ Pete Sauer
✔ Ben Shperling.
My appreciation for continued service on the Editorial Board by
✔ Lalit Goel
✔ Nickos Hatziargyriou
✔ Prabbha Kundar
✔ Sakis Meliopoulos
✔ Mark O’Malley
✔ Chris Root
✔ Hugh Rudnick
6
IEEE power & energy magazine
✔ Mohammad Shahidehpour
✔ Gerry Sheble
✔ Mani Venkata.
And welcome to our new Editorial
Board members:
✔ Jim Feltes
✔ Tao Hong
✔ Brian Johnson
✔ Ben Kroposki
✔ Ning Lu
✔ Mackay Miller
✔ Nando Ochoa
✔ Charlie Smith
✔ Mini Thomas
✔ Edvina Uzunovic
✔ Jianhui Wang
✔ Steve Widergren.
Enrique Tejera will serve as our liaison
with the newly formed Spanish version
of IEEE Power & Energy Magazine.
p&e
january/february 2017
leader’s corner
Jessica Bain
more power to the future
Internships and more
T
TWENTY-SEVEN YEARS AGO, I
joined the IEEE Power & Energy Society (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 often asked, “I’m a big fan of yours and
I’m 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 do—masters of their
craft—know that they wouldn’t have
achieved the success they have if it
weren’t for their mentors along the way.
Throughout my 27 years with PES,
from connecting with professors, publishing my papers in IEEE publications, 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. I’ve taken some time to reflect on the many interactions I’ve had
with members around the world. Every
member has different needs. The questions 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
an internship position?” My answer to
the first question is absolutely yes. My
answer to the second
question is “Get involved with your local
PES Chapters and Student Branch Chapters
when you first enroll
in college.” Students
n o t o n ly f i n d m o r e
mentors but also have
the opportunity to contribute and network
in person with PES
professionals.
As the PES secretary,
I want to build a stronger support network for student members. No one can
say they did it on their own. We’ve
done so much because someone gave
us strength and helped us along the
way. PES Chapters and Student Branch
Chapters are local volunteer organizations 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
activities and have programs for students
to grease the pipeline for future power
engineers. In addition to
socializing and networking, Chapter meetings
can include technology
demonstrations, invited
lectures, plant trips,
and tutorials. Websites
and e-newsletters keep
Chapter members up to
date with local happenings 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.
As the PES
secretary, I
want to build a
stronger support
network
for student
members.
p&e
january/february 2017
3K\VLFDODQG2SHUDWLRQDO0DUJLQV
320 6XLWHRQHVWRSVROXWLRQIRUDOO
\RXUSRZHUV\VWHPDQDO\VLVQHHGV
6PDUW7RROVIRUWKH6PDUW*ULG
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
8WLOL]LQJ308'DWDWRSUHGLFWV\VWHPLQVWDELOLW\DQGJHQHUDWHDODUPV
&RPSXWLQJSKDVHDQJOHOLPLWLQUHDOWLPHDQGRႇOLQHHQYLURQPHQWV
,PSURYLQJWKHWUDQVPLVVLRQJULGUHOLDELOLW\WRIDFLOLWDWHLQWHJUDWLRQRI
WKHUHQHZDEOHHQHUJ\
([HFXWLRQWLPHIRUDEXVFDVH
320
230
$&VROXWLRQIRURQHFRQWLQJHQF\VHF
$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
2ႇHUVIDVWDQGHDV\WRXVHWUDQVLHQWVWDELOLW\DQDO\VLV
3HUIRUPVIDVWIDXOWVFUHHQLQJWRGHWHUPLQHDQGUDQNPRVW
VHYHUHWKUHHSKDVHDQGXQEDODQFHGIDXOWV
2ႇHUVDERXQGDU\EDVHGVROXWLRQIDVWDQGDFFXUDWH
DVVHVVPHQWRIRSHUDWLQJOLPLWV
'HWHUPLQHVUHPHGLDODFWLRQVWRVWD\ZLWKLQV\VWHP
RSHUDWLQJFULWHULDDQGLQFUHDVHVYROWDJHVWDELOLW\PDUJLQV
$XWRPDWLFDOO\FRPSXWHVFULWLFDOFOHDULQJWLPH
'HWHUPLQHVDQGYLVXDOL]HVWKHG\QDPLFVHFXULW\UHJLRQ
3HUIRUPVFRQYHQWLRQDOK\EULGDQGOLQHDUVWDWHHVWLPDWLRQ
:HRႇHU6SHFLDOL]HG&RQVXOWLQJ6HUYLFHVDQG
3RZHU6\VWHP6WXGLHVEDVHGRQWKHVWDWHRIWKHDUW
VXLWHRISRZHUV\VWHPWRROV
LQIR#YUHQHUJ\FRP
ZZZYUHQHUJ\FRP
guest editorial
Mark O’Malley and Benjamin Kroposki
unlocking flexibility
energy systems integration
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 technologies such as wind and solar. There
is also much more use of information
and communications technologies to
enhance the visibility and controllability 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 variable generation. One way to unlock this
potential flexibility is to tap into other
energy domains. This concept of interconnecting 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 minimal impact on the environment. Integrating 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 including water, transport, and data and communications networks, which are an enabling technology for ESI.
The value of ESI is in coordinating
how energy systems produce and deliver
energy in all forms to reach reliable, economic, and/or environmental goals at apDigital Object Identifier 10.1109/MPE.2016.2629703
Date of publication: 2 February 2017
10
IEEE power & energy magazine
propriate scales. The analysis of integrated
energy systems can inform policy makers
and industry on the best strategies to accomplish these goals. The benefits of ESI
include the integration of higher levels of
variable renewables, an increased reliability 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 multiple energy systems at multiple spatiotemporal scales.
ESI, a multidisciplinary area ranging
from science, engineering, and technology to policy, economics, regulation,
and human behavior, is most valuable
at the physical, institutional, and spatial
interfaces, where there are new challenges and opportunities for research,
demonstration, and deployment to reap
its commercial and societal benefits.
The simultaneous focus on multiple disciplines and stakeholders makes ESI a
challenging and exciting area.
This issue of IEEE Power & Energy
Magazine contains six articles that examine the topic of flexibility in energy
systems. The articles reach outside the
normal domain of electricity and continue 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 electrical infrastructures have continued to
become more tightly coupled than ever
before and addresses the capability of
the gas system to meet electricity system flexibility requirements. Flexibility
of the system is analyzed in three parts:
from gas-powered generation, from gas
supply, and through multi-input, multioutput plants and appliances.
Gas-powered generation plants are typically 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 increasing levels of variable renewable energy 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 hydrogen 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 electricity for generating heat at times of excess
renewable electricity on the power grid,
and, vice versa, at times of peak electricity demand, they have the ability to switch
from electricity to gas.
The second article, “Harnessing
Flexibility from Hot and Cold” by Kiviluoma 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 controllability, they can also be valuable sources
of reserves that mitigate forecast errors
and faults. The article examines the pros
and cons of thermal-electricity integration and lays out generic principles and
characteristics related to thermal sector
flexibility as well as demonstrates its possibilities with specific examples.
A major potential for flexibility in the
heat sector is created by the low cost of
storing heat, which provides an opportunity 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
created by thermal loads in several countries around the world.
In “The Triple Bottom Line for Efficiency,” authors Casey et al., examine the
energy-water nexus and address efficiency 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, industrial processes, and end uses can conserve
electricity, hence reducing water use in
power plants. Water and wastewater treatment facilities are good candidates for
demand response because they are energy
intensive and have significant flexibility
through variable speed pumps and inher-
ent storage in water tanks. This operational flexibility can make these facilities
ideal partners for electric utilities seeking
to manage electric load through demandresponse programs.
“Unlocking Flexibility” by Dall’Anese
et al., examines integrated operational solutions that strike a balance among cost,
reliability, and environment while accounting 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 discusses the core challenges related to the
development of distributed control and
optimization algorithms that allow different parties to retain controllability of
their own energy assets while maintaining their individual performance and reliability objectives. The physical couplings
among systems is represented in mathematical terms through the so-called energy hub
approach that offers the flexibility to
tŽƌůĚ>ĞĂĚĞƌŝŶĞǀĞůŽƉŝŶŐ
ŝŐǁŽŽĚ
^LJƐƚĞŵƐ͕/ŶĐ͘
ĨŽƌWŽǁĞƌ'ƌŝĚKŶͲůŝŶĞƐƐĞƐƐŵĞŶƚ͕
ŶŚĂŶĐĞŵĞŶƚ͕ĂŶĚKƉƟŵŝnjĂƟŽŶ
ǁǁǁ͘ďŝŐǁŽŽĚͲƐLJƐƚĞŵƐ͘ĐŽŵ
ƚ͗нϭ͘ϲϬϳ͘Ϯϱϳ͘Ϭϵϭϱ
Ĩ͗нϭ͘ϲϬϳ͘Ϯϱϳ͘ϬϮϯϳ
,QQRYDWLRQSUHYDLOV
^/ŚĂƐŝŶǀĞŶƚĞĚĂŶĚŚŽůĚƐƚŚĞƌŝŐŚƚƐ
ƚŽϭϴƉĂƚĞŶƚƐŝŶƚŚŝƐĂƌĞĂŽĨĞdžƉĞƌƟƐĞ͘
KŶͲůŝŶĞ^ŽůƵƟŽŶƐĨŽƌŶĞƌŐLJŽŶƚƌŽůĞŶƚĞƌƐ
2QOLQH
6\VWHP2SHUDWLQJ
/LPLW 62/ (QJLQH
&$,62 /RRN$KHDGDQG
2QOLQH9ROW9$5
&RQWURO
8.31 ³%LJZRRG6\VWHPV¶2QOLQH9ROWDJH6WD
ELOLW\SURGXFWLVVLPSO\WKHEHVWLQWKH
ILHOG´
&$,623URMHFW/HDGHU
³%6,2QOLQH7UDQVLHQW6WDELOLW\$VVHVV
PHQWLVDVXSHUESURGXFW´
7(3&20DQDJHU
³%6,9$50DQDJHPHQW6\VWHPVLVD
EHVWSUDFWLFHWRROIRUDOOXWLOLWLHV´
6(5&$XGLW6WDII
12
IEEE power & energy magazine
2QOLQH9ROWDJH
7UDQVLHQW6WDELOLW\
3-07(3&2 1H[W*HQHUDWLRQ
6WDWH
(VWLPDWLRQIRU
7 '
2QOLQH5HDFWLYH5H
VHUYH 9$5 0DUJLQ
0RQLWRULQJ6\VWHP
79$ ƐĂůĞƐΛďŝŐǁŽŽĚͲ
ƐLJƐƚĞŵƐ͘ĐŽŵ
6XSHUDQG*OREDO
2SWLPDO3RZHU)ORZ
XSWRFRQ
WUROYDULDEOHV ([WHQVLYHVRIWZDUHSRUWIROLRIRUERWKWUDQVPLVVLRQDQGGLVWULEXWLRQV\VWHPV3OHDVH
YLVLWRXUZHEVLWHWROHDUQPRUHDERXWVXFFHVVIXOVWRULHVSURGXFWVDQGXWLOLW\FRQVXOWLQJ
ZZZELJZRRGV\VWHPVFRP
ƵƐƚŽŵĞƌƐ^/ŚĂƐƉƌŽǀŝĚĞĚƐŽůƵƟŽŶƐĂŶĚƐĞƌǀŝĐĞƐƚŽŽǀĞƌϯϱĞůĞĐƚƌŝĐĂůƵƟůŝƟĞƐǁŽƌůĚͲ
ǁŝĚĞ͕ŝŶĐůƵĚŝŶŐůŽŶŐƐƚĂŶĚŝŶŐĐƵƐƚŽŵĞƌƐ͕ĂůŝĨŽƌŶŝĂ/^K͕W:D/ŶƚĞƌĐŽŶŶĞĐƟŽŶ;>ĂƌŐĞƐƚ
ƐLJƐƚĞŵŽƉĞƌĂƚŽƌŝŶƚŚĞǁŽƌůĚͿ͕dŽŬLJŽůĞĐƚƌŝĐWŽǁĞƌŽŵƉĂŶLJ;dWKͿ͕dĞŶŶĞƐƐĞĞsĂůͲ
ůĞLJƵƚŚŽƌŝƚLJ͕hŶŝƚĞĚ<ŝŶŐĚŽŵWŽǁĞƌEĞƚǁŽƌŬƐ;ŶŐůĂŶĚͿ͕ĂŶĚEĞǁzŽƌŬ/^K͕ĂŵŽŶŐ
ŽƚŚĞƌƐ͘^/͛ƐĂĚǀĂŶĐĞĚƚĞĐŚŶŽůŽŐLJ͕ŝŶŶŽǀĂƟŽŶ͕ĂŶĚƐĂƟƐĮĞĚĐƵƐƚŽŵĞƌƐĂƌĞŬĞLJĚŝīĞƌĞŶƟͲ
ĂƚŽƌƐ͘
january/february 2017
capture energy-conversion factors at various spatial scales. The resultant energy
hub model involves an input–output matrix formulates the solution to multisystem level optimization problems. From an
operational perspective, the coordinated
control of various energy infrastructures
represents a significant change that favors 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 Consumer’s Role
in Flexible Energy Systems” by Schuitema
et al., examines the important role that consumers play on different levels in achieving flexibility in energy systems. They can
help make energy systems more flexible
by changing their energy-consumption
patterns, resulting in load shifting and reduction in energy demand. To do this,
customers must accept smart technologies and energy-efficiency measures. This
is something that electrical power system
operators don’t necessarily control or un-
derstand. The area of consumer acceptance
and energy use is a multifactor challenge
that typically needs the disciplines of social sciences, economics, and humanities
to understand. There are also factors such
as policy interventions and market design.
The article notes that a more integrated approach is needed within social sciences and
humanities as well as with energy researchers in other disciplines, such as engineering,
environmental science, and computer science, to design the best energy system of
the future.
The final article, “Flexibility Challenges for Energy Markets” by D’haeseleer 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 system 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 possible alternative routes for increasing system flexibility and promotes better energy
systems integration.
The issue concludes with the “In My
View” column, in which Mattias Anderson makes the case for the need of an
interdisciplinary approach to achieve appropriate levels of flexibility in our energy
system. He points out the interdisciplinary
opportunities within the developing area
of ESI. Anderson’s background in intellectual history and political science and
his many years of working with research
engineers gives him some interesting perspectives, 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 articles 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 electricity systems have become the backbone of modern energy
production. Both systems are closely interconnected due
to the vast deployment of efficient combined-cycle gas turbines (CCGTs) over the first decade of the 2000s, mainly in
Organization for Economic Cooperation and Development
countries. This increased interdependence and rapid penetration of variable renewable energy sources (varRE) make
the gas–electricity 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 systems 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 investment), due to the shale gas revolution and to natural gas’s
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 operation, 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 O’Malley, Sauro Pasini,
and Simone Simoncini
Digital Object Identifier 10.1109/MPE.2016.2621838
Date of publication: 2 February 2017
16
IEEE power & energy magazine
1540-7977/17©2017IEEE
january/february 2017
✔ Latin America is one of the world’s 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 investment 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 concerns about air pollution and the increasing availability of pipeline gas from Russia and central
Asia. Technically, China has the world’s largest recoverable shale gas resources; however, this
potential is constrained by geological complexity, shortages of water, and land access difficulties, as well as by limited industrial experience—all of which led the country to lower its
production targets. China’s 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 (2016–2020)
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 integration, 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 decarbonize 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 penetration of variable generation and electrification of
heat and transport will lead to increasingly variable operation of thermal dispatchable generators (this is already being observed in Europe).
The growing net-load variability affects not only
power stations but also networks and gas supply
systems (e.g., gas storage and LNG tanks).
In this article, we discuss the gas system’s
ability to meet the electricity system’s flexibility
requirements and also explore some of the technical, 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 generation—technology and electricity market design
✔ in gas supply—gas storage and gas/electricity market coordination
✔ through multiple-input, multiple-output plants and appliances.
Innovating Gas Systems to
Meet the Electricity System’s
Flexibility Requirements
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 renewable 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
18
IEEE power & energy magazine
2015
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
USD/MBtu
Gas turbine research and development (R&D) focuses on
improving the technical flexibility as well as economic profitability in electricity markets by minimizing startup times,
enhancing ramping capabilities, and reducing gas power stations’ minimum stable output. The R&D priorities for gas
power plants are as follows.
✔ Using advanced materials to minimize cycling impact and cost. For CCGTs in particular, this involves
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 equipUnited 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)
2003–2016. 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
life on steam turbines and HRSGs.
Open-cycle gas turbines (OCGTs) and aeroderivative gas
turbines (ADGTs) were always designed to provide flexibil✔ Maximizing load gradients during load changes. This
can be accomplished by using advanced materials and
ity, but CCGT power plants were initially designed to operate
real-time monitoring systems to minimize wear and
mid-load to base load. Over the last decade, European CCGTs
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, reciprocating engines have mainly been
used only for small, decentralized
applications because turbine efMinimum Turndown
Hot Startup
Ramping
ficiencies are considerably lower
OCGT
for these applications (<10 MW)
CCGT
and reciprocating engines can burn
ICE
a broader range of fuel composiICE CC
tions (pipeline quality gas and,
Hard Coal
e.g., synthetic natural gas, landfill
Lignite
gas, and biogas). However, recip0
4
6
0
10
20
30
20 40
60
0
2
rocating engines are now availFL/min (%)
FL (%)
Hours
able in sizes of up to 20 MW and
can be organized as banks of enfigure 2. The flexibility characteristics of thermal electricity generation plants.
gines to form a large power plant
Typical plant size (MW) is as follows: OCGT: 50–200; CCGT: 300–500; internal
(>200 MW). Today, in fact, many
combustion engine (ICE)/reciprocating engine: 20–200; ICE combined cycle (CC):
gas turbine manufacturers also
250–450; black (hard) coal: 500–1,000; lignite: 500–1,000. Note that nuclear plants
own reciprocating engine compaare excluded because they perform worse. FL: full load. (Sources: International
nies (Table 1). These plants proEnergy Agency, Energy Technology Perspectives 2014, Paris: OECD/IEA, 2014;
vide cost-effective N-1 reliability
German Institute for Economic Research, Current and Prospective Costs of Electricity
for islanded power systems due
Generation, Berlin, 2013; Verband der Elektrotechnik Elektronik Informationstechnik,
to the scalability of the cascadErneuerbare Energie braucht flexible Kraftwerke—Szenarien 2020, Frankfurt am
ing plants, which require only one
Main, 2012.)
18
16
14
12
10
8
6
4
2
0
january/february 2017
400
300
(MW)
additional reciprocating engine to meet the reliability standard. With regard to flexibility characteristics, reciprocating
engines could provide
✔ higher efficiency than OCGT and ADGT (up to 48%)
✔ higher part-load plant efficiency and very low minimum output, given that the plant (20–200 MW) is
based on small units (<20 MW)
✔ very quick startup time (a few minutes) and good ramping.
200
100
0
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
2005
2012
Electricity Market Design to Reward Flexibility
From an economic perspective, increasing levels of variable renewable energy penetration may reduce the running hours for these gas generators. Modeling based on the
United Kingdom’s 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 maintain 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
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 developed in some parts of the United States, notably in California and by the MISO system.
Determining the best type of gas plant to have in scenarios 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 perform against other flexibility sources in the electricity system
(e.g., interconnection, demand-side control, consumer interaction) and in the wider energy system (heat, water). The most
economic form of flexibility will be system specific. However,
6,000
90
78
80
5,100
Full Load Hours (h)
4,000
4,000
2,800
4,000
2,000
2,000
2,200
Startup Rates (Starts per TWh)
5,000
71
70
60
55
50
37
40
30
28
20
1,000
10
0
2007
2011 2012
(a)
2013
2014
0
2007
2011
2012
2013
2014
(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 flexibility will be essential to encouraging investment in flexibility and ensure reliability.
Flexibility in Gas Supply: Gas Storage
and Gas/Electricity Market Coordination
Gas Storage
In the past, gas network flexibility appeared abundant compared to electricity systems and was largely ignored in electricity 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 preparation 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, underground 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 generators may be called upon, which would lead to a large and
rapid decrease of the gas line pack. If this happens when
30
6h
338
Low Wind Case: 2020
336
334
332
High Wind: 2020
330
Low Wind: 2020
25
Power (GW)
Base Case: 2009
10 mcm
Line Pack (mcm)
340
peak gas and electricity demand coincide, the resulting pressure drop in the gas network could limit its ability to meet
rapid changes in gas demand (including gas for power generation) 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 linepack shortages, coordination between gas network and electricity system operators will be increasingly important.
Modeling results for the United Kingdom’s 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 facilities. Natural gas is injected into storage during the summer (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 flexibility 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 security 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
20
15
10
5
Base Case: 2009 High Wind Case: 2020
0
328
0
4
8
12 16 20 24 28 32 36 40 44 48
Time (h)
(a)
3.1 GW
0
4
8
12 16 20 24 28 32 36 40 44 48
Time (h)
(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 longterm 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 dayahead 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 isolation on different time frames throughout the day and have
often failed to create a homogenous structure. Among others, 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 calendar 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 continuous trading in the gas market
✔ different product definitions and mechanisms for allocation 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 operators 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 flexibility (such as line pack) in an optimal way.
As the deployment of varRE progresses, limited market coordination may lead to serious risks for flexibility,
such as the need for quickly ramping up generation by gasfired power plants. In recent years, the lack of coordination
between gas and electricity has already threatened reliability. 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 breakdown because the interdependence of the gas/electricity system 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 coordination 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 balancing (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 significant delay. This creates a dilemma for gas/electricity
market coordination as well as natural barriers for aligning
market opening/closing times. Regulators and system operators 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 levels 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
Need to maintain system frequency within
strict limits in real time
Maintenance of operating pressures within a certain
range due to line-pack capability
Balancing process
Close to real time (<1 s) power balance
Cumulative energy deviation over balancing time
frame or day
Balancing time frame
Focus on immediate action in the last
minutes/hour before real time
Focus on delayed actions (ex-post), typically
≥2 h
Adapted from DNV GL.
referred to as a multiple-input poly-generation conversion
system; a multiple-input,multiple-output conversion system;
energy hubs; or a hybrid energy-conversion system (HES).
The flexibility can be utilized for the electricity system, the
natural gas system, or both. Two kinds of flexibility can be
distinguished for HESs.
✔ Operational flexibility enables meeting highly variable net loads or maximizing operation at steady
state of certain HES appliances to minimize wear
and tear.
✔ Economic flexibility enables arbitrage between input
resources and output services, i.e., utilizing least-price
input resources while providing highest-price output services, subject to contractual and physical constraints. A traditional single-input, single-output power
plant may provide significant operational flexibility,
but it would not have economic flexibility.
HES. However, a large variety of alternative HES designs are
conceivable through the combining of different inputs (electricity, heat, fuels, and/or biomass) and outputs (electricity,
heating and cooling services, water, hydrogen, transportation
fuels, and/or commodity chemicals). The flexibility benefits
of HES deployment are exemplified here based on three different HES designs: 1) an advanced HES based on anaerobic
digestion (AD), 2) hybrid residential heaters, and 3) windelectrolyzer systems.
An advanced HES can be conceived around AD. As
illustrated in Figure 6, such an HES has three energy inputs (natural gas, biomass, and electricity), three energy
output services (biomethane, electricity, and cooling/heating), and three storage devices (biomass, heat, and biogas).
Additionally, it contains two heat sources: the combined
heat and power (CHP) unit and the low-temperature geothermal system with heat pump. The CHP and heat pump
serve both the anaerobic digester and the district energy
system. AD utilizes low-grade heat to support the digesExamples of Advanced
tion of organic materials (e.g., wasted food, plant clippings,
HES Designs
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 pipeline system. The district energy
system distributes heat obtained
Biomethane
Cleaning/
Biogas
from the CHP and heat pump
Injected
Upgrading
Storage
into Pipeline
systems to the demand; it could
also provide cooling, if an abNatural Gas
Combined
sorption chiller is included.
Biogas
Heat and
Electricity
Biomass
The heat storage (or accumuAnaerobic
Power
Mixing
Digester
lator) facility and the heat pump
provide that the HES meets heatHeat
ing and cooling demand while the
Heat
Heat
Heat
CHP meets the flexibility requireStorage
Heat
Biomass
ments from the electricity system;
Storage
Cooling
Low-Temp
alternatively (or additionally), the
District Energy
Services
Geothermal w/
heating demand may be controlled,
System
Electricity
Heating
Heat Pump
reducing the need for the heat pump
Services
or the heat storage.
The integration of AD into this
figure 6. An advanced HES design based on an anaerobic digester (three inputs and
HES is motivated in four ways.
three outputs).
22
IEEE power & energy magazine
january/february 2017
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 injection 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 hydrogen embrittles the pipeline. This HES enables storage of
excess renewable electricity in a gaseous fuel, thus providing access to the vast storage capabilities of the gas infrastructure. The gas network offers storage capabilities over
all time frames, from daily cycling as line pack to interseasonal storage underground, and is thus much more flexible
than other storage technologies. Therefore, this HES provides considerable operational flexibility, but its economic
potential depends largely on the price spread between wind
energy and hydrogen, as well as between hydrogen and natural gas. Also, the electrolyzer’s cost itself makes deployment 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
Conventional:
Single-Input,
Single-Output
Energy Converters
Output Side
Input Side
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
low grade, ranging from 30 to 38 °C, and can be extracted from a wide range of processes including heat
pumps and CHPs. This sets AD apart compared to processing 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 north–south “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 regions’ 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 facilities 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 landfill gas projects; only 247 are on farms and thus make
use of agricultural biomass.
✔ Third, if the input feedstock decomposed naturally, undergoing the same biological process as in AD, then it
would emit methane directly into the atmosphere. Considering 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 manifestation 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 residential 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 compared 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 costeffective 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
Input Side
✔ First, although AD requires heat, the necessary heat is
HES:
Multiple-Input,
Multiple-Output
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 natural 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 generation 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 capacities of 10–100 MW, however, could be connected directly to
the distribution substation rather than the distribution circuit,
which means that their DR potential is less limited. Distribution-substation connections of resources at this capacity maintain 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 benefits of load proximity with transmission accessibility.
HES deployment can potentially provide both operational 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 companies 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 typically start up quickly and provide excellent ramping capabilities 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 maximizing 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 operation of gas and electricity networks.
Further integrating the energy resources and energy services through HESs can increase both the operational and economic 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 increasing 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 Recherche, Luxembourg (project reference 6018454), and the CITIES 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. TRENTALOT3-015, Oct. 2012.
“Report on outages and curtailments during the Southwest cold weather event of February 1–5, 2011—Causes 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. 255–262, 2012.
M. Qadrdan, M. Chaudry, J. Wu, N. Jenkins, and J. B.
Ekanayake, “Impact of a large penetration of wind generation on the GB gas network,” Energy Policy, vol. 38, no. 1,
pp. 5684–5695, Jan. 2010.
N. Szarka, F. Scholwin, M. Trommler, H. Jacobi, M.
Eichhorn, A. Ortwein, and D. Thran, “A novel role for bioenergy: A flexible, demand-oriented power supply,” Energy,
vol. 61, no. 10, pp. 18–26, 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 O’Malley 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 benefit from demand flexibility. What is not as often mentioned is that electrification 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 heatsector flexibility and demonstrate
its possibilities using specific
examples. While we generally
use the word heat here, most of
the discussions also apply to cool,
which, after all, is just another
form of temperature difference.
A major potential for flexibility 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
©ISTOCKPHOTO.COM/MIMACZ, ROPE—IMAGE LICENSED BY INGRAM PUBLISHING
Heat Storage and
Hybrid Systems Can
Play a Major Role
Digital Object Identifier 10.1109/MPE.2016.2626618
Date of publication: 2 February 2017
january/february 2017
1540-7977/17©2017IEEE
IEEE power & energy magazine
25
Hybrid systems may take many different forms, from dual
heaters in buildings all the way to large district heating systems with combined heat-and-power (CHP) plants, fuel boilers, and electric heaters.
The Flexibility Potential
of Heating and Cooling
Together, heating and cooling represent a huge part of
energy consumption. However, the heat system is often not
considered as a single system, but rather—due to the historic
emphasis on energy supply—as subsystems of different supply sources (e.g., gas, coal, and electricity). Therefore, size
and flexibility potential are often overlooked. According to
2014 Eurostat figures, in the European Union (EU) around
Primary Energy Use
Final Energy Use
22%
30%
40%
45%
33%
29%
Heat
Transport
Electricity
400
350
300
250
200
150
100
50
0
Residential
Local Heating and Cooling
In
du
st
r
Tr
an y
sp
or
S
t
Ag erv
i
ce
ro
Sp /For s
e
ac
st
e
ry
W Hea
at
er ting
H
ea
tin
El
ec
g
tri Co
ok
c
Ap
in
g
pl
ia
nc
e
Li
gh s
tn
in
g
Energy Use (Mtoe)
figure 1. The primary and final energy use in the EU for
2014 (based on Eurostat figures) divided into three main
categories of energy end use.
Wood
Heat
Gas
Coal
Oil
Electricity
figure 2. The final EU-28 energy use for 2014 (based on
Eurostat figures). The residential sector is split into components 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.
26
30% of primary energy is used to produce heat, 30% is
used in the transport sector, and 40% is used for electricity,
including electricity to heat (see Figure 1).
The share of electricity in primary energy consumption
is higher than in final energy consumption because a large
proportion of energy is wasted in electricity generation processes using the Carnot cycle (Figure 1). The opposite is true
for heat, because nearly 100% of the energy in fuel is converted to useful heat. The large share of heat in the final
energy use translates to a large potential for power system
flexibility. For example, the value of surplus wind or solar
power is zero, but if that electricity can be used to replace
heating fuels, the value rises to the price of the fuel. The
value is affected by the conversion efficiency—the 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.
Heat is consumed in most end-use sectors, except transport,
but its use is very diverse. In residential and commercial buildings, heat is used for space and water heating. In the industrial sector, heat additionally provides process heat. In terms
of flexibility, some demands are more amenable to being controlled than others. Figure 2 divides final energy use (heat, in
particular) among the EU’s 28 member states (EU-28) into
several categories. Most heat in the EU is generated from natural gas, while coal, oil and biomass make up much of the rest.
In Europe, cooling demand is considerably lower
than heating demand. However, it is growing quickly with
increasing space-cooling requirements and also due to the
heavy urban development in warmer climates and new uses
needed in emerging services and industries, such as cooling
large data centers.
The heating/cooling usage types and demand profiles
will largely impact the flexibility potential, We use the following categories to describe the major parts of the heat/
cooling system:
✔ local heating and cooling of buildings
✔ district heating and cooling
✔ heat for industry.
IEEE power & energy magazine
Final energy use in residential buildings is dominated by
space and water heating demand—roughly 80% in Europe
and 60% in the United States. Worldwide, most residential
homes and commercial establishments produce heat locally
within the building using a variety of different heating/cooling technologies and are not connected to district heating
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, liquid evaporation for heat pumps, and Joule’s 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
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 planning boils down to managing supply and storage infrastructure, 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 portfolio in the long run.
Electric heating, if deployed in an uncoordinated manner, 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 inertia) to facilitate renewables integration and manage peak
loads. For example, some demand-side management programs 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 according to power system conditions, while also meeting the
building occupants’ comfort requirements. These examples 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 temperature difference between the outdoors and indoors during a
condensation/evaporation cycle. The heat cycle only requires
electricity to run the compressor and other auxiliary equipment, 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 performance coefficients of four to five. This minimizes generation 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 1.5–2.5 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 socalled “hybrid” heat pump) shows vast flexibility potential. 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-thermalcapacity materials, then the heaters are referred to as storage 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 underfloor 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. Temperature differences are smaller than in heating though; consequently, 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 conditioning. Conceivably, these chillers could also offer flexibility 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
14%
34%
12%
21%
20%
Space Heating and
Hot Water <100 °C
Process Heat <100 °C
Process Heat 100–500 °C
Process Heat 500–1,000 °C
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 building’s 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, district 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 others (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 further 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 comparison 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 viewpoint, they offer considerable economies of scale for heat
storage, the specific cost of which decreases nearly logarithmically with increasing size.
District cooling is much less common than district heating. 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 cooling is an aspiration, district cooling may have a more important 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 remaining 15% is due to space heating. Heat pumps can serve lowtemperature loads, while CHP units can serve somewhat
higher-temperature levels and still be able to produce
electricity. A large fraction of industrial heat loads, currently dominated by natural gas burners, requires higher
temperatures. However, electric heating technologies such
as resistance heating, electric arc heating, induction heating, 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 control. 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 connection 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 industrial 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 temperature. 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 temperatures 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 variation—mainly
january/february 2017
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 different heat uses:
✔ domestic refrigerator/freezer: 15 min–1 hour
✔ supermarket refrigeration systems: 15 min–3 hours
✔ thermal mass of buildings: 2–12 hours
✔ buildings with local hot water storage: 2–24 hours
✔ district heating pipelines: 1–5 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 55–95 °C offers
january/february 2017
22:00
20:00
18:00
16:00
14:00
12:00
8:00
10:00
6:00
4:00
2:00
0:00
Demand of Yearly Demand (%)
0.045
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
Large District Heating
System in Finland
Domestic Gas Demand
in Belgium
Large District Heating
System in China
Electric Space Heating
in Ireland
figure 4. Hourly heat profiles from a winter weekday.
1.00
Demand of Yearly Demand (%)
driven by longer-term ambient temperature variations. Inside
temperatures are kept nearly constant, even when occupants
are not present. By contrast, houses in Ireland leak more,
and the small share of buildings that rely on electric radiators use them mainly when occupants need the extra heat—
for example, in the morning and evening during weekdays.
When occupants are not present or they are sleeping, inside
temperatures are often allowed to drop. Despite the weather
being more moderate in Ireland than in Finland, the average
Irish living room is probably colder than its Finish counterpart due to different occupant expectations.
Annual profiles can also be quite different, although they
follow more closely the inverse of the ambient temperature.
Figure 5 shows that systems where the heat source also provides hot domestic water have some load during the summer.
In China, district heating systems can be shut down outside
the heating period.
While not shown in Figure 5, cooling could complement
the annual space heating profiles. In some climates where
heating and cooling needs are comparable, similar flexibility from cooling could complement flexibility from heating.
Wherever there are interconnected power grids spanning
across warm and cold climate zones, part of the variations,
depending on the relative strength of the interconnections,
can be smoothed at this continental scale. In either case,
heating and cooling could provide a rather stable source of
potential flexibility for the power sector. Furthermore, in
hot and sunny countries, cooling loads and PV generation
may complement one another well.
Industrial heat demand at the country level does not
exhibit strong seasonality and could, therefore, provide
year-round flexibility (e.g., the industrial heat demand from
Finland shown in Figure 5). Also, daily profiles, especially
in heavy industries, are typically relatively flat. In individual
industrial sites, the profiles can have more variation—for
example, lower demand for products can cut work shifts.
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1
51
101
151
201
251
301
351
Large District Heating
System in Finland
Large District Heating
System in China
Domestic Gas Demand
in Belgium
Electric Space Heating
Demand in Ireland
Industrial Heat Demand
in Finland
Heat Demand in
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 flexibility constrained by a limited time constant—although the
flexibility can still be valuable for the power system when
aggregated over millions of houses.
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
US$/kWh (Δ 50 K)
60
50
40
30
20
10
0
0
5
10
m3
(a)
15
20
US$/kWh (Δ 50 K)
10
8
Forecasting
6
4
2
0
0
5,000
10,000
15,000
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 Centre’s 2012 report.)
Normalized Load (–)
16 January 2015 08:00
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0
24
48
72
96
Horizon (Hours)
figure 7. The quality of heat forecasts 0–96 hours ahead for
the Sønderborg district heating system in Denmark (the blue
line). The 5% and 95% percentiles are shown in red, and the
actual load as it was observed later is shown in black. The
plot is generated using the adaptive heat-load forecasting
system PRESS (www.enfor.dk/products/press.aspx).
30
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. Furthermore, because the flexible loads consist of a number
of appliances dispersed across the system, reliability can
be statistically greater compared to an individual conventional power plant.
The building envelope itself also provides thermal inertia depending on the insulation level and thermal mass of
the building. In a well-insulated house, electric load can be
shifted (2–12 hours depending on building), while consumer
comfort is still met. Preheating or precooling increases flexibility but typically also increases energy usage (depending
on the insulation level). Thermal storage and building preheating enable considerable demand-shifting potential at a
comparably low cost.
IEEE power & energy magazine
Thermal-load forecasting is often used in district heating systems and for estimating electric heating loads. The uncertainty
of heat-load forecasts is important when trying to optimize
heating or cooling. Forecasting failures can lead to unwarranted costs or uncomfortable inside temperatures. For example, when a heat-load forecast error persists in one direction
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 0–96 hours ahead
for the Sønderborg district heating system in Denmark.
If heating and cooling loads are to be controlled in a
manner that provides flexibility to the power system, accurate forecasts will be important. Good forecasts will need
to include climatic variables, building characteristics, and
often predictions about user behavior. Building characteristics can be considered either through direct modeling or past
behavior. Occupancy behavior varies from family to family:
some families have a very systematic and easily predicted
load pattern, whereas others seem to have a very random and
less predictive load profile.
Similar to electric loads, it has been shown that the
aggregation of individual loads can decrease relative forecast errors and smooth out rapid variations. However,
when heating/cooling takes place in individual buildings,
the control algorithm cannot aggregate. On the other hand,
when several buildings are connected to a district heating
system, the forecasts for control can be aggregated and
relative forecast errors decrease. This is different from just
forecasting loads without control, where aggregation can
take place in all cases.
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 variation—mainly driven by ambient temperature variations.
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 presented 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 integration will be a challenge; consequently, there is strong interest 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 building’s heat demand by 1–10%. This
increase in electricity use poses a challenge in terms of compensating 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 natural gas for space and water heating, and this consumes
more than 1.5 times more energy than U.K. electricity consumption. 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 flexibility 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 renewjanuary/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 transition 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 system. 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
Saving in RES
Integration Cost (£/MWh)
Harnessing Heat Flexibility
12
10
Backup
Balancing (OPEX)
Balancing (CAPEX)
8
6
4
2
0
2030
2050
figure 8. The reduction in integration costs of renewable
generation (electricity sector only) enabled by the integrated 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 nonsynchronous generation, one option in providing upward primaryfrequency 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 100–400 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 €8.5–500 per heating appliance, depending on various factors (fuel costs, wind penetration, etc.), for Great
Britain’s 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 addition of 20 MW (35% of activated flexible load) to subsequent
reserve categories to allow for the load pickup.
Residential Heat–Electricity 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 leastcost 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 whenever low-price electricity is available. When compared to a
gas boiler alone, the results showed annual system-wide savings of €18–65 per household, depending on the gas price.
If a gas boiler is combined with a heat pump, they will operate mainly on electricity and shift to gas during periods of
low wind-power supply or high demand. Then, the annual
savings were €46–159 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 electricity, 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 curtailments. 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 interconnectors 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. Meanwhile CHP plants can be operated when there is not enough
low-price electricity available.
In residential buildings, smart thermostats can give functionality beyond temperature and time-of-use control. Communication with the Internet or an aggregator enables the utilization
of power prices and weather forecasts. Meanwhile, occupantmodeling intelligence can consider the actual needs of the occupants in the control scheme. For example, the model predictive
control (MPC) algorithm can make use of the additional information 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
support from the Fonds National de la Recherche, Luxembourg
(project reference 6018454) and CITIES (DSF 1305-00027B/
DSF). J. Kiviluoma acknowledges support from the FLEX-e
program funded by TEKES.
06:30 p.m.
06:00 p.m.
05:30 p.m.
05:00 p.m.
04:30 p.m.
Energy
Rush Hour
04:00 p.m.
03:30 p.m.
03:00 p.m.
02:30 p.m.
02:00 p.m.
1.0
0.8
0.6
0.4
0.2
0.0
01:30 p.m.
Fraction of Total Load
Precooling
Time
Actual ac Use
Projected ac Use Without Rush-Hour Rewards
figure 9. Employing heat flexibility in residential buildings
to avoid electricity demand when the electricity grid faces
high loads. (Source: nest.com.)
increased flexibility. MPC is applied by companies such as
BuildingIQ or QCoefficient when they exploit chiller efficiency
variations due to ambient temperature to achieve energy savings without overly affecting occupant comfort. An example of
smart-thermostat-based control is shown in Figure 9, where the
smart thermostat Nest performed large-scale peak shaving by
precooling American residential buildings.
In 2011, China initiated a series of pilot projects that
substitute electric boilers to for coal-fuelled CHP boilers
in the Jilin province. The electric boilers often use surplus
wind generation as their energy source. From 2015, the
project has expanded to all the northern provinces such as
Hebei, Liaoning, Inner Mongolia, and Xinjiang. The electric boilers are equipped with water tanks capable of providing 10~15 hours of storage.
For Further Reading
D. Patteeuw, K. Bruninx, A. Arteconi, E. Delarue, W.
D’haeseleer, and L Helsen, “Integrated modeling of active
demand response with electric heating systems coupled to
thermal energy storage systems,” Appl. Energy, vol. 151,
pp. 306–319, Aug. 2015.
S. Heinen, D. Burke, and M. O’Malley, “Electricity, gas,
heat integration via residential hybrid heating technologies:
An investment model assessment,” Energy, vol. 109, pp.
906–919, Aug. 2016.
K. MacLean R. Sansom, T. Watson, and R. Gross, “Comparing the impacts and costs of transitions in heat infrastructure,” Centre for Energy Policy and Technology, Imperial
College, London, U.K., Apr. 2016.
J. Kiviluoma, “Managing wind power variability and uncertainty through increased power system flexibility,” Ph.D.
dissertation, Aalto Univ., Dept. Applied Physics, VTT Sci 35,
Espoo, Finland, 2013, p. 77.
T. Naegler, S. Simon, M. Klein, and H. C. Gils, “Quantification
of the European industrial heat demand by branch and temperature level,” Int. J. Energy Res., vol. 39, pp. 2019–2030, Oct. 2015.
N. Zhang, X. Lu, C. P. Nielsen, M. B. McElroy, X. Chen,
Y. Deng, and C. Kang. “Reducing curtailment of wind electricity in China by employing electric boilers for heat and pumped
hydro for energy storage,” Appl. Energy, vol. 184, pp. 987–994,
Dec. 2016.
P. Bacher, H. Madsen, H. A. Nielsen, and B. Perers, “Shortterm heat load forecasting for single family houses,” Energy
Buildings, vol. 65, pp. 101–112, Oct. 2013.
Conclusions
Heating and cooling offer considerable flexibility potential
for power systems. Much of this could become cost-effective
as the share of variable and uncertain generation increases.
Simultaneously, electrification of heating offers a possibility
for heat sector decarbonization. However, the picture is not
entirely rosy. Seasonality in space-heating needs makes it a
less attractive source of flexibility. On the other hand, flexibility from heating could be partially complemented by flexibility from cooling or from more stable loads in the industrial
sector. At the same time, the industrial sector is very diverse
and will require elaborate research to understand the true
flexibility potential in heat-consuming industrial processes.
Acknowledgments
H. Madsen acknowledges support by CITIES (Center for ITIntelligent Energy Systems, DSF 1305-00027B). G. Strbac
acknowledges support from the U.K. Research Council-funded
project “Energy Storage for Low-Carbon Grids.” Tsinghua Unijanuary/february 2017
Biographies
Juha Kiviluoma is with the VTT Technical Research Centre, Espoo, Finland.
Steve Heinen is with University College Dublin, United
Kingdom.
Hassan Qazi is with University College Dublin, United
Kingdom.
Henrik Madsen is with the Danish Technical University,
Lyngby, Denmark.
Goran Strbac is with Imperial College London, United
Kingdom.
Chongqing Kang is with Tsinghua University, China.
Ning Zhang is with Tsinghua University, China.
Dieter Patteeuw is with Katholieke Universiteit Leuven,
Belgium.
Tobias Naegler is with the German Aerospace Center,
Stuttgart.
p&e
IEEE power & energy magazine
33
By Eoin Casey,
Sara Beaini,
Sudeshna Pabi,
Kent Zammit,
and Ammi Amarnath
E
Benefits of System Integration:
A Triple Bottom Line
Addressing the energy–water nexus invites a systemintegration approach (Figure 1). Today’s electric grid has
become the integrated energy network, which requires
versatility across power paths and adapting to the new
©ISTOCKPHOTO.COM/TCMAKE_PHOTO
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 energy–water
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 changing—becoming more flexible and resilient and
providing energy-efficiency gains—the 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 Internet 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.
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
1540-7977/17©2017IEEE
january/february 2017
Integrating Systems
Within Water and
Energy Networks
commercial buildings, with benefits for both customers and energy suppliers
✔ carbon reductions achieved through the electrification
of water systems, replacing fossil fuel point sources and
increasing efficiency for the pumping, heating, treatment, and transport of water and wastewater
✔ off-peak water treatment, water heating, and water
pumping to reduce both energy costs and generating-unit downturn
✔ distributed water treatment/WWT and local reuse,
providing greater system security and resiliency
✔ expanded availability of Energy Star and WaterSense appliances for water and energy savings, enabling easy ways to achieve customer savings and
satisfaction as well as to implement energy and water 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 resources and protects human health.
✔ Using efficient electric technologies for water treatment, 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 increasingly 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 integrated energy network (or “integrated network of resources”),
where further efficiencies may be gained through this integration 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
demand for systems such as water pumping to head
tanks, desalination, wastewater treatment (WWT), irrigation pumps, and water heaters in residential and
january/february 2017
figure 1. The energy–water nexus. Integrating the
energy grid and water network to create an integrated
energy network can drive more sustainable and efficient
resource utilization.
IEEE power & energy magazine
35
Today’s 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, flexibility, and resiliency, all of which effectively optimize their
efficiency, reliability, security, economics, and environmental performance. Numerous system integration technologies 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 behavior—is coming to the fore in the planning, design, and operation of the global energy system. ESI seeks to optimize
the energy system and other large-scale infrastructures—in
particular, water—by 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 system’s 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
Demand
Electric Power
Agricultural
Industrial
Municipal
figure 2. Resource management has become more
intricately integrated, specifically for water and energy
resources. Although not as obvious as energy consumption, water usage covers a variety of applications, from
power generation and industrial and agricultural applications 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 driving 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, particularly 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 Commission, refers to changes in electric usage by end-use customers “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 centrifuges. 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. Wastewater systems, on the other hand, may divert a portion of the
incoming sewage into holding cells or reduce aeration during peak electrical periods (Figure 4). Under the right circumstances, 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 capacity at about 700 U.S. water treatment and WWT facilities.
WWT systems struggle with significant flow changes,
particularly during rainfall events. Infiltration into the collection system represents a challenging problem for many systems, 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 manage peak electric demand by diverting wastewater to these
basins during peak electric periods. Fortunately, peak electric
Dams
Produce
Electricity
Energy Is Used for
Pumping Water
Power Plant
Cooling Uses
Water
Water Used
for Mining Fuels
Water Supply
Uses Energy
Energy Uses in
Water/Wastewater
Treatment
Water Flows
Energy Flows
Water and Energy Use
in the Home Are Related
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)
Screens
2)
Primary
Clarifier
3)
Activated Sludge
Reactors
4)
Secondary
Clarifier
Sand
Filter
Chlorination
Discharge
Mid-Atlantic WWTP2, Processes 7,570 m3/day
1)
Screens
2)
Primary
Clarifier
3)
Primary
Aeration Tank
Secondary
Aeration Tank
Secondary
Clarifier
4)
Multimedia
Filter
Chlorination
Discharge
Midwest WWTP1, Processes 1,363 m3/day
1)
2)
3)
Activated Sludge
Lagoons
Screens
Clarifiers
4)
Seasonal
Chlorinationa
Discharge
Midwest WWTP2, Processes 1,439 m3/day
1)
2)
Screens
Sequencing Batch
Reactor
Lagoon
Cell A
3)
Lagoon
Cell B
Lagoon
Cell C
Lagoon
Cell D
Lagoon
Cell E
Discharge
figure 4. WWT plants provide opportunities for electric system flexibility.
demand periods often coincide with hot and dry weather, giving plant managers some flexibility in keeping storage available 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 weatherrelated (rainfall), energy demand is not constant. Wastewater 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 equipment in return for rebates from the electrical utility. Load
shifting strategies in WWT facilities include pre-aeration,
38
IEEE power & energy magazine
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 curtailment has potential implications for WWT operations in
terms of water quality. There are some documented examples 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 wastewater. Another interesting approach is the concept of overoxygenating 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.
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 opportunity 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 flexibility. Thus, there is a need for real-time data analysis and
forecasting systems that will inform process-control strategies. 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 facilities 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 water–energy load management and storage
potential, there is a need to develop standardized communication 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 wastewater. 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
rnia
Arizona
New Mexico
igan
Lake
ch
Mi
Alaska
a
rid
Flo
Hawaii
Massachusetts
Mississippi
Colorado
Califo
on
Utah
Lake Ontario
New York
rie
n
Rhode Island
iga e E
Connecticut
k
a
Pennsylvania
L
New Jersey
Nebraska
District of Columbia
Ohio
Delaware
Illinois Indiana
West
Maryland
Virginia
Kansas
Virginia
Missouri
Kentucky
North
Carolina
Tennessee
Explanation
Oklahoma Arkansas
South
Water Withdrawals,
Carolina
in Million Gallons
Georgia
Per Day
Alabama
Texas
0–200
201–1,000
Louisiana
1,001–5,000
5,001–15,000
West-East
15,001–23,100
Puerto U.S.
Virgin
Division for
Rico Islands
This Report
Iowa
Nevada
ur
Wyoming
Wisconsin
H
South Dakota
e
pshir
Ham t
ine
n
o
Ma
Verm
ke
Idaho
La
North Dakota Minnesota
Oregon
Mich
Montana
New
Lake Superior
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
2,500–5,000
1,000–2,500
500–1,000
<500
This analysis estimates the methane generation potential of wastewater treatment plants using methodology from the EPA’s
inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2011 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. Environmental 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-efficient 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 conservation 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 municipal and industrial applications that currently use chemical treatment. End-use conservation would directly reduce water and
energy demand. Some examples of efficiency opportunities in
water treatment/WWT are discussed in the following.
40
IEEE power & energy magazine
Efficiency via Data Monitoring
and Process Control
As with any complex industrial process, the potential for
efficiency gains via computer-based monitoring and control in water and wastewater systems is significant. Monitoring and control technologies vary from simple devices
to advanced SCADA systems. SCADA systems are used
for precise control of key equipment and processes, including raw-water wells, water treatment, and distribution
pumping. Typically, these SCADA systems pull data from
field devices, such as programmable logic controllers, remote terminal units, and electric meters, and analyze/format 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 quality and reliability than on controlling systems to optimize
energy efficiency.
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 5–10% 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 5.4–10.9 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 possible, checking the system efficiency in real time, and then
notifying the operator when changes are required.
The most sophisticated control systems “learn” the characteristics 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 energyefficiency 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 supply 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 additional 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. Environmental Protection Agency and U.S. Bureau of Reclamation
data, the potential savings from advanced irrigation controls in residential and commercial applications is estimated
to be 1.5–3% of total electricity use in the public water supply. At a current electricity use rate of 39,000 million kWh
per year, this equates to potential savings of 0.5–1.2 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 challenging 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 example 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 industry 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 biological 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 processes—including, for example, reverse osmosis for desalination, advanced ionization for micro-pollutant removal,
and membrane bioreactors—are expected to continue to
be highly energy intensive. Some emerging developments
to address this problem include forward osmosis or membrane 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 commercial building applications, newer systems are under development 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, variablespeed drives, and advanced monitoring and process control 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 forecasting systems that will inform process-control strategies.
The benefits of such approaches include cost savings, flexibility, and improved control and decision support for planning plant upgrades.
Reconceptualizing the WWT Plant as an
Energy- and Resource-Generation Facility
Future WWT plants will no longer be just pure waste management facilities but rather recovery systems for clean
water, energy, and minerals. The chemical energy in wastewater is in the form of biodegradable and inert chemical
oxygen demand and reduced nitrogen (NH4+ or organic N).
Large-scale WWT plants recover 5–15% of this energy
through anaerobic digestion. However, because other important parts of the treatment plants are energy-intensive (aerobic biological oxidation requires 0.40–0.65 kWh/m3), it
follows that most treatment plants are net users of energy.
With the advent of new and improved treatment technologies, a net-energy-positive WWT plant is now considered
achievable. Emerging process technologies, including anammox, 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 thermal 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 typically recover only 5–20% of the chemical energy through
anaerobic digestion (the production of biogas and cogeneration). 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 setups. 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: residential wastewater can be treated to use for local irrigation, flushing toilets, etc.
✔ Decentralization can also have benefits in minimizing
the need to increase infrastructure for high-density infill 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 incentives for end users, and commercialization of new technologies with improved performance and lower costs. When
rallying the diverse stakeholders toward the triple bottom
line of efficiency, the dialog will revolve around the advantages of the system integration approach for the energy–
water nexus. The emerging system-integration case studies 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. Svoronos, “Control issues and challenges in wastewater treatment plants,” IEEE Control Syst. Mag., vol. 26, no. 4, pp.
63–69, 2006.
B. Sparn and R. Hunsberger, “Opportunities and challenges for water and wastewater industries to provide exchangeable services,” National Renewable Energy Laboratory, 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 Institute, Palo Alto, California.
Sudeshna Pabi is with the Electric Power Research Institute, Palo Alto, California.
Kent Zammit is with the Electric Power Research Institute, Palo Alto, California.
Ammi Amarnath is with the Electric Power Research Institute, Palo Alto, California.
p&e
january/february 2017
By Emiliano Dall’Anese,
Pierluigi Mancarella,
and Antonello Monti
Unlocking
Flexibility
©ISTOCKPHOTO.COM/LEOWOLFERT
Integrated
Optimization
and Control
of Multienergy
Systems
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 control of such systems at multiple spatiotemporal scales can bring significant
socioeconomic, operational efficiency, and environmental benefits. Accordingly, the concept of the multi-energy system is gaining considerable attention, with the overarching objectives of 1) uncovering fundamental gains (and
potential drawbacks) that emerge from the integrated operation of multiple
systems and 2) developing holistic yet computationally affordable optimization and control methods that maximize operational benefits, while 3)
acknowledging intrinsic interdependencies and quality-of-service requirements for each provider.
On a much broader scale, the main drivers for the integrated operation
of multiple infrastructures include the impetus toward a decarbonization of
various energy and transportation sectors and the potential for resolving the
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
1540-7977/17©2017IEEE
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 emissions 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)comfort, to name just a few.
Shifting Supply and Demand
Across Spatiotemporal Scales
From an operational perspective, the coordinated and seamless 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 renewable 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 multiple energy systems at multiple spatiotemporal scales. Flexibility can generally be seen as a system’s ability to provide
secure and economical supply-demand balance across spatial and temporal scales by leveraging and seamlessly coordinating various controllable assets. In the context of future
low-carbon power systems, major flexibility challenges are
associated with
✔ 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 networks—and, in doing so, across spatiotemporal scales—by 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 opportunities 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 overall 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 opportunities 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 mathematical terms through the so-called
“energy hub” formalism—a geAdjacent
ner ic, sca lable, a nd modula r
Systems Solar
modeling approach that offers the
Industrial
Area
flexibility to capture energy-conversion factors at various spatial
Residential
H1
H2
scales. In particular, the mathArea
ematical model associated with
Pumped Hydro
the energy hub shown in Figure 2
(Storage)
could offer representations of the
interdependencies among energy
carriers at the household, commercial, and energy-district levels and facilitate the formulation
H3
H4
Commercial
Wind
of (and solutions for) control/opArea
timization schemes to control
Water
Heating
Natural Gas
Electricity
co-generation and tri-generation
plants in distribution settings, as
well as gas-to-power and powerfigure 1. A complex system of systems providing basic electricity, heat/cooling, gas,
to-gas
and hydro facilities at the
and water services to end customers. Dark areas designated Hi (energy hubs) represent
transmission level.
locations and facilities where the various infrastructures are coupled.
44
IEEE power & energy magazine
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 systems, and TES in the form of, e.g., hot water storage and icethermal storage. Additional elements not explicitly represented
in Figure 2 could include reversible EHPs, absorption chillers, 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 hub’s operation and, hence,
facilitates the development of computationally affordable control/optimization schemes, while maintaining adequate accuracy in the representation of the underlying system physics. A
generic energy hub model of Figure 2 can be tailored to specific configurations (e.g., the unit, facility, plant, or geographical area to be modeled) by retaining relevant components and
conversion stages and incorporating different levels of simplifications/reductions. Overall, the resulting model involves
an input–output 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 minimize operational costs or to optimize the operation of an interconnected system of energy hubs.
This article provides an overview of possible joint control and optimization approaches for multi-energy systems;
it also elaborates on core challenges related to the development 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 interdependencies among energy subsystems.
Example of Transmission-Level
Modeling and Applications
Recent efforts have looked at the transmission-level infrastructure from a multi-energy system viewpoint, with the
objective of assessing the potentials for flexible and reliable 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 sectors 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 electricity 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 operational 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 simultaneous 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 delivery of heat (e.g., through EHPs and/or CHP), the evolving
electricity and heat sectors could bring about further new flexibility requirements in the gas network.
For instance, it might happen that there is not enough flexibility 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 provide 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 variations 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 intersector (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 heating 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
Water
Electricity
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 utilization 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 emerging 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 constraints. These constraints may relate, in particular, to regulatory 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, malfunctioning 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.
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 heating or cooling
✔ effective user flexibility in the net consumption/generation 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, providing 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 applications of multi-energy models to suggest opportunities for
increasing flexibility at the district, neighborhood, and residential levels.
Examples of Distribution-Level
Modeling and Applications
Flexibility from Distributed Multigeneration
Departing from conventional operational settings, the
main goal at the distribution level is to leverage the fastacting capabilities of power-electronics-interfaced DERs,
as well as the flexibility offered by a variety of other controllable assets, to enable sustainable capacity expansion,
respond to service requests precipitated by distribution
and transmission-systems operators, and achieve networklevel coordination—and 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 volatility 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 environmental 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
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 substantially reducing operational costs, thus offering enhanced
flexibility in provisioning the electricity grid.
An example of a general electricity-and-heat DMG structure is illustrated in Figure 4, where seven operational configurations 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
Boiler
haux
Eimp
Haux
chp
TES
Hs
EDS
αEchp EEHP
H
EHP EHP
HD
H
Eexp
EHP
ED
HEHP
EDS
ED
May
Aug.
Nov.
Feb.
Wind Generation Capability
Solar Generation Capability
Integrated ElectricityHeat OPF
Electrical Network Model
Date
Renewable and Conventional Generation Model
0
20
40
0
5
10
15
TES HO
Hs
EEHP
1800
2400
1200
Time
Regional Electricity and Heat
Demand Model +
Eimp
Heating Technology Scenario
0600
Industrial Heat Demand
Domestic Hot Water
Consumers
HD,ED
35
30
25
20
15
10
5
0
Use Clean Gas
for Heating or
Electrical
Generation
Gas Network Model
Hydrogen and
Synthetic Natural
Gas from Excess
Renewable
Electricity
(Power to Gas)
Integrated
Electricity-H2-Gas
Model
08 Mar. 15 Mar. 22 Mar. 29 Mar.
Renewables Curtailment
Actual Wind Output
Curtailed Wind Generation
Wind Generation Capacity
Use Gas Network as
a Means to Store
and Transport
H2-Gas Blend
Wind Generation (GW)
figure 3. An illustration of the methodology for assessing integrated electricity-heat-gas-hydrogen multi-energy systems, with specific application to Great Britain’s
electricity-gas transmission networks. HD: heat demand; ED: electric demand; EDS: electric distribution system.
Faux
CHP
h e, h t
(1-α)Echp
350
300
250
200
150
100
50
0
Fchp
System Heat
Demand (GW)
Domestic Heat Demand
Commercial Heat Demand
Generation Capability (GW)
january/february 2017
Consumers
HD,ED
IEEE power & energy magazine
47
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–
response incentives are not sufficient to make up for the extra
costs in moving from optimal set points to provide flexibility.
Thermal
Demand
Heat
Energy
Gas
Flow
Electricity
Electricity Market
Electricity (Net)
Demand
EHP
∑
Flexibility from Residential Neighborhoods
and Commercial Buildings
TES
∑
CHP
AB
∑
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
Buildings and homes are basic, spatially defined examples
of multi-energy hubs, given the heterogeneous setting that
supplies essential electricity, thermal, and manufacturing
needs. At the commercial level, HVAC units are prospective candidates for providing services to the power grid at
various time scales, especially because of the favorable flexibility 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 regulation services to the distribution grid, while simultaneously
minimizing the thermal discomfort of building occupants.
At a slower time scale, and taking advantage of the buildings’ 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 number of end customers can follow regulating commands dispatched by electrical distribution systems operators.
Costs and Benefits (μ )
8
Energy Cost Variation
Demand Response (DR)
Benefits
7
6
Electricity
Shifting
Potential
0.14
0.12
5
0.10
4
0.08
3
0.06
0.04
2
Non-Profitable
Region
1
0
0
0.02
DR Incentive (μ /kWhel)
Flexibility from Data Centers
10
50
20
30
40
Reduced Electricity Input from Grid (kWhel)
figure 5. An example of profitability map (in monetary
units μ) of the flexibility provided by a DMG plant. The thick
piecewise linear curve represents the energy cost increase
when moving away from the optimal set points for a given
load at a given time to provide flexibility to the grid; the
dashed lines represents the potential benefits, parameterized
with respect to different demand–response incentives.
48
IEEE power & energy magazine
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
as energy hubs that are characterized by
✔ local generation (more and more data centers are
equipped with their own local generation unit)
✔ electrical energy storage for reliability purposes,
which can also offer flexibility products
✔ thermal cooling processes
✔ a flexible load, provided by the possibility of scheduling computational power.
For this reason, recent research has been moving toward the
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.
Flexibility from Joint Water-Power
Optimization and Control
Operators of municipal water systems (MWSs) and wastewater systems (WWSs) have the core objective of providing
clean water and treated wastewater, according to well-defined
january/february 2017
water-quality requirements. MWSs
and WWSs are operated in a pricetaking setting, with tasks such as
pump scheduling aimed at minimizMin g (Power)
Min h (Gas)
ing operation costs based on the price
Min f (Water)
s.t Power
s.t Gas
s.t Water
of electricity. On the other hand, the
Constraints
Constraints
Constraints
power consumption of MWSs and
WWSs is used as an input to the optimal power-flow problems utilized by
Adjacent
power-systems operators.
Systems Solar
Industrial
However, it is increasingly recArea
ognized that a synergistic control
of power and water systems may
Residential
H1
H2
bring significant benefits from
Area
operational, reliability, and societal
Pumped Hydro
(Storage)
standpoints. In particular, controllable assets in MWSs and WWSs
can provide valuable services to the
power grid at multiple spatial scales
to enhance reliability and efficiency,
Commercial
H3
H4
Wind
as well as to cope with the volatility
Area
of distributed renewable-based genHeating
Natural Gas
Electricity
Water
eration; services include frequency
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 system’s 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 expansion, 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, thereHowever, 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 siga 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
among systems and operational constraints are acknowledged
Technical Challenges in the
(see Figure 7).
Optimization and Control
Inheriting the characteristics of, e.g., ac optimal power
of Integrated Multi-Energy Systems
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 computationtimal” 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 physExamples 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 multiple locally optimal solutions and a globally optimal solution
difficult to find and assess simultaneously) and, oftentimes,
nondeterministic polynomial-time-hard (NP-hard) programs that may be computationally prohibitive to solve even
for local optimization. Nonconvexity stems from the underlying 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 inputoutput 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 nonlinearity complexity) forms.
Nonconvexity from Flow Physics
Nonconvexity Due to Coupling Factors
The power-balance constraints in power systems are nonconvex due to nonlinear ac power-flow equations; additional sources of nonconvexity include constraints on the
minimum voltage service levels and on branch thermal
limits applied to power flows. In pump-scheduling problems and water-flow problems, nonconvex constraints naturally emerge from the relevant flow mathematical models
to capture the head losses or pressure losses due to friction
For modeling simplicity, the coupling factors among energy
carriers are often assumed to be constant. This is the case for
the fuel-to-power and fuel-to-heat conversion efficiencies of
CHP units, for example. However, a number of coupling factors are, in fact, nonlinear; examples include the efficiency
and power consumed by a variable-speed water pump, which
are nonlinear functions of the pump’s frequency. Further, the
operational region of some types of co-generation units or
absorption and compression chillers may be nonconvex.
Discrete Control Decisions
The binary and discrete decision variables required to consider 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 problem complexity. The dimension of the search space increases
exponentially with the number of binary variables (especially
if binary decisions are required over a prolonged optimization horizon), thus rendering solutions for the underlying
optimization problems computationally prohibitive.
Min f (Water) + g (Power) + h (Gas)
s.t Water Constraints
Power Constraints
Gas Constraints
+ Coupling Constraints
Tradeoff Between
Complexity and Flexibility
Setpoints
Adjacent
Systems
Industrial
Area
Solar
H1
Residential
Area
H2
Pumped Hydro
(Storage)
Commercial
Area
Electricity
H3
Heating
H4
Natural Gas
Wind
Water
figure 7. A global control problem optimizes a variety of performance objectives
and economic indicators, while acknowledging intrinsic interdependencies among
systems and operational constraints. Flexibility is naturally enabled by such a problem formulation, since the coupling constraints capture shifts in supply and demand
across energy vectors and networks.
50
IEEE power & energy magazine
Energy hubs and multigeneration
models based on so-called dispatch
factors, which are used to specify
how energy is split among conversion units and outputs, may introduce other sources of nonconvexity
via bilinear or trilinear terms appearing in equality constraints.
Nonlinearity could be, in this case,
bypassed either by introducing auxiliary variables or by simplify the
hub model. It is, thus, apparent that
tradeoffs between the complexity
of any modeling approach to the
optimization problems and achievable flexibility must be taken into
account in the system’s design and
operational processes.
Opportunities
Off-the-shelf solvers for mixedinteger nonlinear programs could,
in principle, be used to find solutions
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 exthe underlying nonconvexity and NP-hardness, suboptimal ist for how to leverage advances in convex relaxation and approxisolutions 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.
Key research issues involve the development of mechaits attraction basin). It is, however, worth pointing out that the
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 derivaespecially 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 nondynamics of system and ambient conditions. In the context of convex mixed-integer nonlinear programs.
future deregulated multi-energy markets, nonconvexity inherently 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 power-flow solutions. Sufficient condiMin 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
+ Consensus
+ Consensus
+ Consensus
programming and second-order
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
Industrial
methods for general topologies
Area
can be demonstrated by numerical
evidence. Further, powerful linear
Residential
H1
H2
Area
approximation methods enable
one to develop linear and quadratPumped 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 sectors, relaxations have, for instance,
H3
H4
Commercial
Wind
been recently proposed for waterArea
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 strategiAlthough 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 objectives, while acknowledging interdependencies among energy
subsystems. Although decomposition methods can conceivably 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 convergence properties.
Identifying constraints that couple energy assets and subnetworks managed by different entities is key to achieving these goals. For example, water-power coupling constraints pertain to the power/speed of pumps; coupling among
energy hubs may be via power lines, district heating/cooling 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 multipliers) can be leveraged to derive a relevant distributed optimization procedure.
In particular, these methods allow one to develop algorithms 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 landscapes, 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 direction is to ensure that the set points produced by the online
algorithm do not induce violations of operational and security limits. A cross-fertilization of control and online optimization tools is, therefore, the key to enable the synthesis
of distributed feedback control schemes that ensure satisfaction of physical and security limits while tracking solutions 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 perspectives by leveraging the flexibility of various controllable
52
IEEE power & energy magazine
assets to ensure a secure and economical supply–demand balance 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 computationally 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 strategically decompose the decision-making process across actors,
while offering fast time-scale flexibility options and steering 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 models 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. Peças-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.
1–22.
M. Geidl, G. Koeppel, P. Favre-Perrod, B. Klockl, G. Andersson, and K. Frohlich, “Energy hubs for the future,” IEEE
Power Energy Mag., vol. 5, no. 1, pp. 24–30, Jan./Feb. 2007.
S. Clegg and P. Mancarella, “Integrated electrical and
gas network flexibility assessment in low-carbon multi-energy systems,” IEEE Trans. Sustain. Energy, vol. 7, no. 2, pp.
718–731, 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. 1928–1938, Dec. 2013.
D. Müller, A. Monti, S. Stinner, T. Schlösser, T. Schütz,
P. Matthes, H. Wolisz, C. Molitor, H. Harb, and R. Streblow,
“Demand side management for city districts,” Build. Environ., vol. 91, pp. 283–293, Sept. 2015.
E. Dall’Anese and A. Simonetto, “Optimal power flow
pursuit,” IEEE Trans. Smart Grid, May 2016.
Biographies
Emiliano Dall’Anese is with the National Renewable Energy Laboratory, Golden, Colorado, United States.
Pierluigi Mancarella is with the University of Melbourne, Australia, and the University of Manchester, United
Kingdom.
Antonello Monti is with RWTH Aachen University,
Germany.
p&e
january/february 2017
The Consumer’s
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. Consumers 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 facilitate greater reliance on renewable energy sources.
Increasing consumers’ flexibility is complex,
and the solution lies in a combination of many disciplines, 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
1540-7977/17©2017IEEE
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 conducted 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 consumer’s role in a flexible energy system?
✔ How can consumers’ behavior be changed?
✔ What barriers are there for change?
✔ What influences consumers’ investment in flexibilityenhancing technologies?
✔ How can policies and market design be used to stimulate 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 integrated into energy research more generally.
The Consumer’s Role in
a Flexible Energy System
An increase in variable renewables (e.g., wind and solar
photovoltaic) in the energy system means that energy supply 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 basis—such as showering,
changing thermostat settings, and switching on/off lights or
appliances—and 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 vehicles, and it includes the adoption of technologies that facilitate 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 sometimes 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 one’s own energy, for example. Prosumers 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 flexibility also introduces new challenges in terms of major infrastructure 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 supply of energy is limited. Demand-side management (DSM)
is the term used to describe a range of measures for improving 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 consumers to change their energy consumption. Some of these DR
programs rely on changes in curtailment behavior—that 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 appropriate 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 consumption 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 one’s 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
have the ability to override this outside control and thus feel
personal control over the decision making.
Overall, consumers perceive the risks related to new technologies and DLC to be much higher than any benefits they
may provide. And, while a positive experience with one smart
technology may increase a consumer’s acceptance levels of
smart technologies generally, this is unlikely to occur if the
consumer’s experience is frustrating due to unrealistic expectations or a lack of policy and market support.
The characteristics of energy use partly depend on
household structures and socio-demographics. For example, energy consumption increases with the number of
inhabitants in the house, the age of the head of the household, and the number of unemployed household members. In
addition, homeowners usually have higher energy consumption as compared to renters. The extent to which interventions are effective to change consumers’ behavior depends,
to some extent, on such household characteristics.
Your Electricity Provider Can Remotely Limit the Use of
Certain Major Home Appliances and You Cannot Choose
to Reverse This Course of Action
29%
.m
.
m.
2a
1 a.
14 Day
Ce Ra
nt te
/kW
h
a.
3
.
8 p.
.m
m.
Peak Rate*
20 Cent/kWh
.
m
.
te
Ra h
ht
W
Nig ent/k
C
24
.
p.
m
m.
11 p.
p.m
9
Midnight
figure 1. The influence of consumers’ sense of personal
control on their likelihood of accepting programs allowing
utilities to control their appliances’ energy use. (Data derived
from Accenture, Understanding Consumer Preferences in
Energy Efficiency, 2010.)
10
4a
7 p.m
m.
5 a.
.
6 p.m.
6 a.m.
.
5 p.m
7 a.
m.
8a
m.
.
.m
.
.
m
.
.
1 p.m
2p
.m
a.
a.m
11 a.m.
.
p.
m
9
10
3
.
4p
Noon
january/february 2017
19%
% of Savings in Energy Bill
Time-of-Use Tariffs
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
(when energy supply is high and
demand is low) by setting lower
energy prices compared to highpeak hours. The overall objective is
to shift energy demand from highto low-peak hours, rather than to
reduce overall energy use. ToU
tariffs are designed to better reflect
the “true” energy generation costs
at a certain time of the day. However, this also implies that energy
prices will vary more, which may
cause uncertainty among consumers about how much energy will
cost. Generally, consumers prefer
certain outcomes over uncertain
outcomes; hence, they may initially object to ToU tariffs. However, in the longer term, if average
energy costs go down, resistance
may decrease.
ToU tariffs have already beenput 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
100
90
80
70
60
50
40
30
20
10
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
% of Sign Up
(Certainty + Probably)
Your Electricity Provider Can Remotely Limit the Use of
Certain Major Home Appliances and You Cannot Choose
to Reverse This Course of Action
Day
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 January–31 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.)
IEEE power & energy magazine
55
displays (IHDs) but also more static measures such as labels
relating to the energy performance of buildings and a host
of energy-using appliances. With or without IHDs, smart
meters facilitate large-scale flexibility by allowing control of
electricity use remotely.
Smart meters provide consumers with real-time information on their energy consumption via IHDs, websites, or
mobile apps. This is an advantage over traditional communication methods, where more generic information was provided via billing or other (paper-based) information. With
Information and Feedback
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 inforrequired 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 overload or redundancy, which would be
counterproductive in increasing consumers’ knowledge and awareness of
What Organizations Do You Trust to Inform You About Actions You Can Take
to Optimize Your Electricity Consumption?
their energy consumption.
Although the provision of inforDo Not Trust Neither Trust Nor Distrust Trust
mation can be effective in making
13
13
20
consumers’ energy behavior more
28
29
flexible, the effects are fairly small
49
51
53
(in the range of 5–15%) and tend to
be short term. It has been suggested
59
60
48
that to create a long-term learning
46
66
effect, consumers need help in inter43
36
preting their energy consumption
42
and frequent reminders, which can
28
27
26
25
be accomplished by adding infor14
11
8
7
mation on the consequences of their
consumption (e.g., financial consequences or carbon footprint). Also,
combining energy consumption with
feedback about one’s 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 information 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 informaSecurity Companies
tion may be trusted when it comes
from one source but not when it
comes from another. That is, inforfigure 3. Consumers’ level of trust in sources that provide information on actions
mation on how to optimize one’s
they can take to optimize their electricity consumption. (Data derived from Accenture, Understanding Consumer Preferences in Energy Efficiency, 2010.)
electricity consumption is deemed
56
IEEE power & energy magazine
Environmental Associations
Academics/Schools/Scientific Associations
Consumer Associations
Utilities/ Electricity Provides
Government/Governmental Organizations
Online Service Providers**
Retainers/Equipment Manufactures
Home Service Provides**
showed that the introduction of ToU tariffs could shift the
energy consumption at peak times to off-peak hours by 8%.
However, average consumption was found to reduce little
or remain the same. If energy efficiency is also a goal, ToU
tariffs should be combined with the provision of information on energy consumption and costs (e.g., feedback information through paper bills and technology displays or
smart meters).
january/february 2017
Other Drivers of Behavioral Change
Economists see economic and information policy instruments as means to overcome market failures such as the presence 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 flexibility, 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 importance 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 laundry, or being warm are more important than “energy” and
“energy demand” in themselves. Nudging is another strategy
to change consumers’ behavior toward becoming more flexible. 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
% People Saying “yes” to Smart
Meter with Automated Control
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 organizations) than when it comes from sources that are not
typically seen to have consumers’ interest as a main priority (e.g., industry or government).
It is well known that, when the provision of information
enhances transparency on energy use and utility billing, consumers 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 important 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.
100
90
80
70
60
50
40
30
20
10
0
Opt-In
Denmark
Opt-Out
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 consumers’ adoption of smart meters with remote control.
(Source: M. Broman Toft, G. Schuitema, and J.Thøgersen,
“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. 113–123.)
has been accomplished, for example, by suggesting a different 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 control), consumer uptake often increases.
Technology Adoption
and Investment Behavior
Consumers can facilitate a flexible energy system by adopting new technologies and investing in them. Such investments 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 technolNo Based On Country’s Assessment
ogy before they will consider doNo Decision Yet
ing so too.
Selective
FI
That social influence is imporSE
tant for the uptake of new technologies is observed, for example, in the
EE
installation of solar photovoltaic
LV
DK
panels. Solar contagion is a term
IE
LT
that is used for the phenomenon
UK
that, in neighborhoods where soNL
PL
DE
BE
lar panels are visible, more solar
LU
CZ
panels appear. This is probably
SK
FR
because the visibility of solar panAT
HU
els triggers others to adopt them,
SI
RO
HR
which is likely to be connected to
PT
the symbolic functions of technolES
IT
BG
ogy and the desire to express one’s
identity, for example as an innoGR
CY
vator or a green consumer.
In the deployment of new techMT
nologies, grants and subsidies are
often proposed as a way to encourage or kick start the market by
figure 5. The EU smart meter rollout in 2020. (Source: Briefing European Parliament,
spreading
the adoption of technoloSmart Electricity Grids and Meters in the EU Member States, Brussels: European Parliagies
through
social networks. Once
ment, Sept. 2015; image reproduced with permission from the European Parliament.)
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 benefiback time. There may be distribution issues, however, where ciary of the reduced energy bills.
Another way to kick start the market is to encourage
low-income groups are the most adversely affected. As an
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
approach is more likely to engage and involve consumers,
gies that enhance the flexibility of energy systems.
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 ownertechnologies 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.
Finally, the benefits of community projects may be
their investments will pay off and whether the adoption of
flexibility-enhancing technologies (such as smart meters) larger because, in contrast to individual consumers, the total
energy demand of communities is larger, which makes it
will really reduce their energy bills.
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 individual’s 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. Upscaling 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 necessary to achieve more flexible energy demand from consumers, such as installing smart appliances and smart meters
and negotiating contracts that include provisions for automated 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 appliances 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 cost–benefit 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 (installing 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 cost–benefit 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 cost–benefit 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 technologies (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 technologies such as heat pumps (in the former case) and lightemitting 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 performance of buildings and a host of energy-using appliances.
DSM programs have been introduced in many countries
to facilitate more flexible consumer energy-using behavior. While the majority of existing DSM programs are utility-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 integrated, 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 occur—for example, through utility DR
and DSM programs or other intermediaries such as aggregators. There is also increasing recognition that DR involving 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 consumers’ role in flexible energy systems is the public acceptance 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. Consequently, 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 infrastructure 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 underlying psychological factors such as values and lifestyles as well
as attachment to place, personal identities, symbolic meanings, and emotions.
In addition, the communication process between consumers and the involved authorities and industries is very
important for encouraging consumer acceptance. If consumers 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 acceptance will hinder the building of infrastructure needed for a
flexible energy system.
Compensation schemes may be used to strengthen public
acceptance of energy infrastructure—for example, by providing individual tax benefits or providing benefits for the
community that reduce local energy bills or create or expand
community facilities. Such communication and compensations schemes can be effective for increasing the acceptability 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 flexible by changing their energy consumption patterns, resulting 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 infrastructure 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, economics, and humanities on the factors that underlie consumers’ 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 disciplines, such as engineering, environmental science, and computer 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 engineering 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 policy 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., engineers, environmental scientists, and computer scientists)
should be better informed by the work of the social sciences, 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 effects of behavioral interventions: Experimental evidence
from energy conservation.,” Am. Econ. Rev., vol. 104, no.
10, pp. 3003–3037, 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. 166–178, June 2016.
M. A. R. Lopes, A. C. Henggeler, K. B. Janda, P. Peixoto, 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. 233–245,
Mar. 2015.
N. Sintov and P. W. Schultz, “Unlocking the potential
of smart grid technologies with behavioral science,” Front.
Psychol., vol. 6, pp. 1–8, 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. 1–29, 2014.
L. Steg, G. Perlaviciute, and E. van der Werff, “Understanding the human dimensions of a sustainable energy transition,” Front. Psychol., vol. 6, pp. 1–7, 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
©ISTOCKPHOTO.COM/AAGGRAPHICS
Fragmented
Policies and
Regulations
Lead to
Significant
Concerns
Flexibility Challenges
for Energy Markets
By William D’haeseleer, 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 practical 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
intermittent renewable power sources pose to the controllability 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 sector, the natural gas sector, and the transportation sector.
As a consequence, the overall energy system becomes
increasingly coupled, which requires appropriate communication 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
1540-7977/17©2017IEEE
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 importance 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 legitimacy 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 concrete implementation targets. The strong push for investments
in certain intermittent renewables impacts the performance of
other instruments (e.g., those aimed at CO2 emission reduction) 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 unnecessarily 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 for—and even promoting—better system integration 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 “directives,” 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 compul62
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 EU’s 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 percentage 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 potential face a very challenging target. Note that the EU Commission 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 trading system, by which green certificates would be exchangeable per country. Instead, every country was allowed to set
up its own individual support scheme, resulting in local certificate systems in some countries (or even in parts of countries), feed-in tariffs, investment support, and tax breaks.
A number of cooperation mechanisms (statistical transfers, joint projects, and joint support schemes) made it possible to both offer flexibility and meet part of a member state’s
target through the deployment of renewables in another member 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 electricity 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 National Renewables Energy Actions Plans lead to an overall
requirement for the electricity system of 33–34%, or about
january/february 2017
CO2 Policy
The 20% GHG reduction target with respect to 1990 is recalibrated 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 percentjanuary/february 2017
12,000
10,000
Net Demand (MW)
one-third of total European generation (in terms of electric
end energy) by 2020. The implications will be different for
different countries, depending on their individual required
targets and actual potential. But a simple calculation for “average” countries in Europe for the period 2008–2020 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 because it competes with other types of land use and there
are competing applications for biomass, such as transportation. In addition, its very environmental sustainability is questioned. Consequently, there is considerable 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, respectively.
✔ 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:
ronshore and offshore wind/CFs: ~ 25–45%
rsolar PVs/CFs: ~ 13–14%.
✔ To produce, say, 20 percentage points of the 34% electric end energy with technologies that operate only
13–14% and/or 25–40% 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 17:00–18:00 h when peak demand
arises in northwestern Europe, very little RES electricity 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.
8,000
6,000
4,000
2,000
0
0
20
40
60
80 100 120 140 160
Time (h)
0% Wind + Solar PV
10% Wind + Solar PV
20% Wind + Solar PV
40% Wind + Solar PV
figure 1. Net electrical power load during one week for various 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 sectors (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 mechanism), 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
–21% Compared
to 2005
Non-ETS Sectors
–10% Compared 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 (nonETS). (Source: Memo/08/34, “Questions and answers on the
Commission’s proposal for effort sharing,” Brussels 2008.)
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 targets 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 important to clarify what is meant by the “consumption” of “prosumers” who avail of storage (which may have significant
efficiency losses of up to 20–30%). Will only demand from
the grid be taken into account?
Furthermore, many support schemes are not cost reflective, such as net metering and feed-in tariffs. The ensuing
zero-marginal-cost electricity production by households
may stimulate frivolous electricity consumption. As a consequence, the entire concept of “energy efficiency” may
lose its meaning. In the end, a more generalized concept of
“resource efficiency” (including investment cost, manufacturing 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 reductions 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 substantial public support has not actually impacted Europe’s
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 industry 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 coalfired plants with regard to the local versus overall European 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 positively, the contributions to reductions in CO2 emissions 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 incorrect. 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” allowances in the market is, indeed, be reached.
✔ A few exceptions exist concerning the statement
made previously about CO2 price reduction. For example, if locally produced PVs were to be dedicated
to feed heat pumps that effectively replace (smallscale) CO2-emitting boilers, then overall CO2 emissions would be reduced, not because of the PVs as
such—because that is still part of the electricity sector and thus the ETS—but 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 boilers because they are part of the capped ETS.
Another unintended policy effect is caused by the promotion of CHP, as the deployment of small-scale CHP units
creates a shift of emissions from the ETS to the effort-sharing (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 electricity 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 summarize, 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 satisfying 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 3—even with the sudden jumps and prices going to zero—is perfectly explainable
january/february 2017
january/february 2017
15
/2
0
14
29
/0
4
/2
0
13
29
/0
4
/2
0
12
29
/0
4
/2
0
11
29
/0
4
/2
0
10
29
/0
4
/2
0
09
29
/0
4
/2
0
08
29
/0
4
/2
0
07
29
/0
4
/2
0
06
/0
4
/2
0
29
/0
4
29
Ap 9
r.
10
O
ct
.1
Ap 0
r.
11
O
ct
.1
Ap 1
r.
12
O
ct
.1
Ap 2
r.
13
O
ct
.1
Ap 3
r.
14
O
ct
.1
Ap 4
r.
15
O
ct
.1
Ap 5
r.
16
.0
ct
O
Ap
r.
0
9
US$/MBtu
29
/0
4
/2
0
05
Carbon Price/Metric Ton
by the modalities and rules of the
€ 35
setup of the EU ETS. This demonstrates 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,
the allowances were/are “bankable,” avoiding “natural” zero
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 originally anticipated and thus foreseen
25
by the cap.
20
The current low CO2 prices
result from many factors, some
15
of them straightforward consequences of the design of a cap10
and-trade scheme, such as the
5
lasting economic crisis (characterized by fewer emissions)
0
that started in 2008, the inflow
of international credits, and the
bankability of allowances that
Henry Hub
NBP
Japan LNG Contract
makes the surplus persistent over
Brent
Asian LNG Spot
time. However, in addition (and
Note: NBP = National Balancing Point (United Kingdom).
this is very important from a systems perspective) an unforeseen
factor is that the EUA prices have figure 4. The development of gas prices 2009–2016 in several parts of the world.
been pushed further downward Henry Hub represents U.S. prices, while NBP represents European prices. A comby 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.
A further system effect in Europe, then, is that marginal
have affected the CO2 prices, making it cheaper for CO2polluting units to generate electricity, while still meeting cost pricing in EU electricity markets is pushing efficient
combined-cycle gas-fired units out of the merit order as a
the cap.
To further understand the effects in the European energy consequence of low world-market coal prices (due, as menmarkets, 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 States—with 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
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 wholesale forward electricity prices in the Central West European (CWE) region increased from about €50–55/MWh
to €90–100/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 €25–30/MWh in the late summer 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 capacity reserve through a capacity remuneration mechanism.
Discussions as to an appropriate market design (“energy
only” versus “capacity remuneration schemes”) are currently ongoing.
As previously mentioned, zero-marginal-cost renewables (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 customers 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 evolution in Belgium for a typical end customer with annual consumption 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 renewables 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 massive renewables penetration. The reason is that the crossborder high-voltage grid is currently not strong enough to
ensure price convergence (i.e., by being congested) given
the large differences in the generation portfolio among
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 transmission capacity.
The EU ETS Refurbished
Win
d
PV
Coal
GT
CC
Low nd
ma
De
ETS
ETS
ETS
figure 5. The limited load factors of combined-cycle gas turbines (CCGTs) in Europe.
Zero-marginal-cost renewables, together with low demand, push thermal generation
units out of the merit order (or off the ice wedge). With current European gas and
coal prices, and a very low CO2 price penalty (via the ETS) represented by the baby
seals, CCGT plants risk being the first victims. For higher CO2 penalties, adult ETS
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 regardless of a CO2 penalty, and coal generation is the prey. In China, the demand is still
sufficiently large so that coal and gas plants are called upon. (Image courtesy of
D. Patteeuw, KU Leuven, used with permission; adapted image inspired by http://
economicsforenergy.blogspot.be/2013/02/los-mecanismos-de-retribucion-de-la.html.)
66
IEEE power & energy magazine
Faced with low EUA prices in the
ETS market, with the awareness
that many market participants do
not foresee a long-term CO2 price
and hesitate to make long-term investments, European policy makers have decided to “reform” the
current EU ETS. Via market interventions (referred to as “backloading” and a “market-stability
reserve”), a volume of allowances
is being taken out of the market
with the possibility of reintroducing them later. Whether these
measures will alleviate the side effects of the EU energy and climate
policies remains to be seen. The
volume of the backloading seemed
not to be large enough to have a
significant impact. Whether the
market-stability reserve will alleviate the side effects of the EU energy and climate policies remains to
be seen; moreover, the final volume
of allowances has not, in principle,
been altered.
january/february 2017
january/february 2017
6
01
/2
01
5
01
/2
01
4
01
01
/2
01
/2
01
01
/2
01
3
NL
FR
2
€/MWh
1
BE
DE
01
figure 6. The decreasing tendency of forward electricity
wholesale prices in the CWE market (France, Germany, and
Benelux). [Source: Commission de Régulation de l’Électricité
et du Gaz (CREG), Belgium, Sept. 2016. The arrows have
been added by the authors.]
250
200
150
100
6
20
1
5
01
/
01
/
20
1
4
20
1
3
01
/
20
1
2
01
/
01
/
20
1
1
20
1
0
20
1
01
/
9
01
/
20
0
8
01
/
20
0
7
50
01
/
In New England, the past few winters have seen some interesting issues concerning the interaction between gas and
electricity markets, whereby a stretched natural gas grid
has translated into very high gas and electricity prices. Gas
delivery occurs via pipelines from the south and the north
(Canada), including liquefied natural gas (LNG). Average gas
capacity suffices during the winter, but the system becomes
stretched on particular peak-demand days. Because of market
dynamics, which are a consequence of cheap shale gas in the
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
Massachusetts’s Everett and Canada’s Canaport has declined,
leading to winter spike prices. In addition, because of the low
average U.S. shale gas prices and the environmental drive to
reduce CO2 emissions through the cap-and-trade system of
the Regional Greenhouse Gas Initiative (RGGI), over the last
few years natural-gas-fired electricity generation in New England has increased, along with the retirement of other plants
(using nuclear, coal, and oil).
This increase in gas-fired electricity generation has,
indeed, led to lower average wholesale electricity prices,
but it transfers the physical stress in the gas network to fuel
adequacy/reliability concerns in electricity generation, giving 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
/2
Energy-System-Integration Challenges in U.S. Markets
65
60
55
50
45
40
35
30
25
20
01
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.
€/MWh
Energy-System-Integration
Observations on Flexibility Challenges
in the United States and China
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 demand–response programs, among others.
On the other side of the United States, major energy-system integration problems arose in 2000–2001 in California
(later referred to as the “California electricity crisis”). As
the first state to fully introduce liberalized markets (often
inappropriately called “deregulation”), California experienced a combination of events and circumstances that
led to the world’s richest country’s richest state having to
“turn off the lights” (actually, cut all power—called rolling 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
20
0
In the meantime (and as part of European promises for
the Paris COP21 Agreement), the EU has “sharpened” its
commitments toward 2030—albeit in a different way, not
unimportant for system interaction. The firmest commitment 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.
01
/
European Targets Toward 2030
figure 7. Typical retail electricity prices in Belgium for average 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.)
IEEE power & energy magazine
67
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 scientific–technical 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 corrective flexibility measures, it can be expected that, around 2020,
stiff ramping rates by the non-PV remainder of the generation
system—of the order of 13 GW in
3 h—will have to be coped with.
This is illustrated by the so-called
Net Load–31 March
28,000
“duck curve” published by the Cali26,000
fornia independent system operator
CAISO, as shown in Figure 8.
24,000
In reaction to this duck curve,
22,000
studies have been initiated to
20,000
2012 (Actual)
demonstrate that appropriate flex2013 (Actual)
18,000
ibility measures, dealing with
Increased
2014
Potential
16,000
both the total demand and the net
2015
Ramp
Overgeneration
2016
or residual demand, can “teach
14,000
2017
2018
the duck to fly”—basically show2019
12,000
2020
ing that stiff ramping rates can be
10,000
avoided (see Figure 9). Whether
the measures currently suggested
0
will actually work remains to be
12 a.m. 3 a.m. 6 a.m. 9 a.m. 12 p.m. 3 p.m. 6 p.m. 9 p.m.
Hour
seen, but the fact that the discussion has started is encouraging; by
the time needed, appropriate soft
figure 8. The duck curve illustrating the “net” or “residual” load to be covered
by the non-PVs remainder of the electricity system in California. After the sun has
solutions may be implemented.
Megawatts
cap on the retail price of electricity, while gas prices—and,
thus, wholesale electricity prices—began to rise; increasing 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 dollars 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
set in the early evening, electricity demand increases for air-conditioning, lighting, and cooking needs, leading to big ramping rates to meet electricity demand.
(Source: CAISO.)
Duck Curve with All Ten Strategies Compared to Original Load
4,000
3,500
3,000
MW
2,500
2,000
1,500
1,000
500
0
Post Strategies Total Load
Post Strategies Net Load
Original Net Load
1 2 3 4 5 6 7 8 9 1011 12 13 1415 16 17 18 19 2021 22 23 24
Hours
figure 9. Avoiding stiff ramping rates shown in Figure 8 (dotted curve) can be accomplished by well-chosen strategies to level off the net load (green curve). (Source:
J. Lazar, Teaching the “Duck” to Fly, RAP, 2016.)
68
IEEE power & energy magazine
Energy-System-Integration
Challenges in China
China also faces considerable
challenges in terms of energysystem inflexibility, especially
for the integration of renewable
energy. Renewable energy is
growing in a nonmarket environment, where wind and solar generation is at the top of the merit
order. However, large amounts of
renewable energy are still being
curtailed or dumped, despite
the fact that it is scheduled with
first priority. This is because it is
obligatory for conventional generating units to offer flexibility
(by reducing their output) but
without any financial compensation. Because the flexibility is not
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 electricity and heat production from coal-fired CHP plants. Due to
the inflexible operation of the CHP units when forced to generate 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 “heating 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 energysystem 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 industry, residential and service sectors, and transportation). We
mention here the known flexibility possibilities (mainly originating 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 multiple battery units; intermediate storage via pumphydro 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 sectors, 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 unexpected—and, 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 actors—will be strongly nonlinear, careful analysis is called for, well beyond the standard isolated
“impact assessments.”
First and foremost, policy makers should encourage correct system cost evaluation and, consequently, appropriate pricing to guide consumers. As a general rule, market
requirements should provide sufficient freedom for market
participants to play their roles while eliminating any loopholes overly creative individuals or organizations can abuse;
this means that carefully considered boundary conditions
and/or justified constraints must apply. Also as a rule, varying prices can influence customer behavior, and real-time
pricing can steer markets in a desired direction. All customers connected to the electric grid need to contribute to its
costs. Following the principle of cost-reflectiveness, a gridconnection tariff should be based (at least partly) on the connection 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 customerdemand 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 circumstances 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 intervention 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 overcome. 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 transmission grids. If there is insufficient grid capacity, there may
not be enough transport of gas during heavy winter conditions (as in New England in the United States in 2014–
2015) or because of geopolitically-inspired cuts (as on
New Year’s Day of 2006 and 2009 in Europe), with serious
consequences for electricity generation and heating. Gascompressor 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 discussion 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 coming years. Whether very efficient combustion engines will
survive or will be replaced by hybrid or battery electric
vehicles or by hydrogen-fed-fuel-cell vehicles—and over
what period—remains to be seen. It must be noted that in
many countries, car engines already pay a stiff (CO2) penalty 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 compensate for missed revenues due to fewer fuel-consuming vehicles). 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 socalled 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 somewhere) to produce “renewable methane.” It is technically possible, 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-remuneration mechanisms or energy-only markets for dispatchable 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-marginal-cost generation environment, complemented by storage 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 economic 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 considerable hidden system costs, eventually to be paid by consumers, 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 sectors, such as natural gas and heating, require increased flexibility 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 energyservice companies) have the freedom to provide the services
requested by customers. Furthermore, targets and specifications 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 simple, 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 system’s
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 Explained. Brussels: European Union [Online]. Available:
http://ec.europa.eu/clima/publications/docs/eu_climate_
policy_explained_en.pdf
D. Helm, The Carbon Crunch; How We’re Getting Climate Change Wrong: and How to Fix It. New Haven, CT:
Yale Univ. Press, 2012.
january/february 2017
M. O. Bettzüge, D. Helm, and F. Roques. (2014). The crisis of the European electricity system: Diagnosis and possible ways forward. Commissariat general à la stratégie et à 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 lowcarbon power systems. IEA/OECD, Paris [Online]. Available:
http://www.iea.org/publications/freepublications/publication/
re-powering-markets-market-design-and-regulation-duringthe-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. Stanford, CA: Hoover, 2002.
P. L. Joskow. (2001). California’s electricity crisis. Oxford Rev. Econ. Policy [Online]. vol. 17, no. 3, pp. 365–388.
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-teachingtheduck22016-feb-2.pdf
P. Denholm, M. O’Connell, G. Brinkman, and J. Jorgenson. (2015). Overgeneration from Solar Energy in California: A Field Guide to the Duck Chart, NREL, Golden,
CO, Tech. Rep. NREL/TP-6A20-65023 [Online]. Available: 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. 56–64, Sept./
Oct. 2013.
Biographies
William D’haeseleer 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
Thomas J. Blalock
reactors for the Roxy
evolution of ac stage lighting
T
THE INTRODUCTION OF AC SERvices to the “theater districts” of large
cities during the early 20th century allowed for the adoption of reactancetype dimmers to control incandescent
stage lighting, in place of the hopelessly
inefficient resistance dimmers (rheostats) 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 located 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 Building, which is still standing at Lexington Avenue and 42nd Street. The fantastic Roxy Theatre was demolished
in 1960.
The stage lighting reactance-type
dimmer board for the Roxy was constructed by the former Hub Electric Company 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
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 engineering. This article, his 20th, deals with ac theatrical stage lighting, an avocation in which Tom has enjoyed a special interest and has achieved much.
He first became interested in stage lighting and rigging while working backstage 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 General 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 Tom’s 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 Institute. His duties as a development engineer at the former GE High-Voltage Engineering Laboratory and later as a test engineer in the Transformer Test Department, 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 emphasis on the exploration, preservation, and careful documentation of historically 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
hundreds of individual control levers.
It was installed in a separate room
off stage that was open ended so that
the operators (it took more than one
1540-7977/17©2017IEEE
to run it) had a view of the stage itself
(see Figure 1).
A very early attempt to use simple
series-connected reactors as theatrical
dimmers was a failure. This occurred
january/february 2017
in Daly’s 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 being 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 theatrical use, and Daly told the UEL&P
wiring men to just “throw the whole
thing out into the street.” So ended
that experiment.
The lamp-dimming reactors used
by Hub Electric for the Roxy Theatre
board were series connected as well,
but they were not just simple reactance
coils. These devices were known as
figure 1. The Hub Electric Company stage lighting dimmer board being delivered to the Roxy theater. (Photo courtesy of the Theatre Historical Society of
America.)
saturable core reactors, and they included a second control winding
through which dc was circulated. The
dc served to saturate the magnetic core
of the reactor to reduce its reactance,
in contrast to the earlier attempt to reduce reactance by means of a movable
core. These reactors had been supplied
to Hub Electric by the Ward Leonard
Electric Company.
The magnitude of the dc in the control winding was, in turn, controlled by
means of a relatively small resistance dimmer 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
These levers, by the way, were capable
of being interconnected mechanically
to provide a crude form of mastering. 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”
FULLY IEEE 1656 COMPLIANT
NO
GAPS
NO
ZAPS
www.greenjacketinc.com
74
When a single contact can bring down your
substation, installing precise fit cover-up is
the best choice for mitigating your risk.
It’s a proven fact that cover-up works. Our
precise fit covers are made to order based
on the exacting dimensions of the underlying
equipment – it is the best product for
eliminating gaps. And, Greenjacket is fully
compliant with the IEEE 1656 Guide test
parameters. Selecting Greenjacket ensures
you have the only precise fit and most effective
protection available.
When outage risks can have significant
consequential damages for any utility or end
use customer, having the best protection is
your best choice.
1.866.464.7996
IEEE power & energy magazine
operator sufficient mechanical advantage to accomplish this. Therefore, a
large wheel (seen at the center of the
board in Figure 1) was provided for this
purpose. The wheel was geared to the
shafts inside the board that moved the
levers, so it provided the necessary mechanical advantage.
The operation of this wheel did not
allow for extremely rapid cues. The fastest 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 contactors, 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 correspondingly large and required a lot
of lighting to cover it. The total electrical 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 applied to various uses (such as motors and
lights). However, another reason for such
a large number of meters was that it was
extremely difficult to calibrate watthour
meters designed for currents in excess of
800 A (the total current corresponding
to the anticipated load would have been
about 20,000 A at 120 V). Presumably,
such calibration at that time required the
inclusion of the current transformer or
shunt to be used with the meter.
The building had a total of four electric services, two for ac and two for
dc. UEL&P supplied the ac services at
208Y/120 V, three phase, and the New
York Edison Company supplied the dc
services at 120/240 V, three wire. The
main reasons for multiple services were
the high connected load and the need
for two different types of current. However, there also were requirements at
january/february 2017
that time for theaters such as the Roxy
to have more than one service in the
event of the failure of one while an audience was in the building. Having power
supplied by two different companies
further increased the reliability of service in emergencies.
DC was required for the control windings of the saturable core reactors used
in conjunction with the Hub control board.
However, dc was also needed for the operation of arc lights, both in movie projectors and for large follow spots from
the balconies. AC arc lights were available, but they tended to hum, just as the
reactors did in Daly’s Theatre.
There were a total of 200 dimmers
controllable from the Hub board. The majority of these were saturable core reactors
having capacities of up to 15 kW each.
Resistance dimmers for such large loads
would have been horribly energy wasteful,
and their physical size would have made
them ponderous as well. Some of the levers on the board did control resistance
dimmers for smaller, specialized lighting
loads up to 2,000 W. These probably operated on ac as well since they were supplying incandescent lighting loads.
Interestingly, the use of archaic resistance 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 theater 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 during the early 1960s.
Consolidated Edison’s main dc supply facility for the Times Square area
was the West 39th Street rotary converter substation, and it was not able to
be retired until 1977.
Optimisation:
How many birds can your wires hold?
Knowing capacity and limits... knowing when and where
to best allocate investment...
Just knowing...
:KHQZHXQGHUVWDQGSXUSRVHDQGLQçXHQFHVDQG
SHHULQWRWKHIXWXUHZHEHFRPHWUXO\HIIHFWLYHDQGFDQ
SODQWRSURæWIURPLWVIXOOSRWHQWLDO
PLEXOS is the world’s leading energy simulation
software that delivers optimised solutions through
IRUHFDVWLQJ,WLVIXOO\FXVWRPLVDEOHDQGFRQæJXUDEOHIRU
your needs.
Because knowledge is power.
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 dimmers, but they still required a fairly
large rheostatic control for their operation. The development of vacuum tubes
for use in radios during the 1920s led
to the possibility of similar tubes be-
ing used to supply adjustable dc for
the control winding of a saturable core
reactor, with the tube, in turn, then being controlled by means of a very low
power “potentiometer” type of device
in its grid circuitry.
It turned out that a type of tube recently developed by Albert Hull, a scientist 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 action of this three-element tube in controlling 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 appearance of a bluish glow in the tube when
operating. Likewise, he named a high
OmniMax
TM
How a Street Says Welcome
All the benefits of LED without sacrificing the light you love.
It’s the right light—not just the
brightest light—that 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 lamp’s optics,
as they were designed.
76
IEEE power & energy magazine
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
january/february 2017
advantage, however, resulted from the
fact that the controlling device was a
very small, very low power potentiometer. This meant that more than one
“pot” could be provided for each dimmer so that the desired dimmer setting
could be preset for upcoming cues. The
changeover from one cue to the next
could then be made simply by operat-
ing a single master controller that fed
the individual dimmer controllers.
This presetting technique remained
in use for stage lighting control until the development of computer-type
lighting boards during the late 20th century. That innovation meant that a huge
number of upcoming cues could be stored
in the computer memory. Manual preset
making life visibly safer
SpanLite
TM
Self-Illuminated
Power Line Marker
Meets FAA Advisory
Circular 70/7460-1L
(Dec. 2015)
Installs directly on
live lines up to 500 kV
Bird
Diverters
LED Obstruction
'%&0/
Low Line
Flags & Markers
5-.0#!&!,*
6709
vacuum diode used for rectification
a “kenotron,” where “kenos” means
empty space, a reference to the high
vacuum in the tube. Finally, a common
three-element vacuum tube used extensively in radio circuitry was christened a
“pliotron,” from “pleion,” which means
more, a reference to the tube’s use as
an amplifier.
Figure 2 is a schematic of typical
thyratron-reactor dimmer circuitry from
this era. A resistive potentiometer (MCB)
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 operator controlled the dimmer.
A dual-plate kenotron tube (K) is
shown that acted as a differential rectifier. It combined the signal from the
“pot” with a feedback signal to derive
a resultant signal for the grid of the
thyratron tube (T). Grid capacitors (FI
and FB) introduced a time constant that
determined how long into a half cycle
of the impressed ac it would take for
the thyratron to conduct and feed current to the control winding of the saturable core reactor. This, of course, was
a function of the position of the control
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 being 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 operator had the same view as the audience. This location usually had not
been possible with resistance dimmer
or simple reactance dimmer boards
because of the plethora of high current
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 location was a big advantage of thyratronreactor installations. Another huge
IEEE power & energy magazine
77
lighting boards most often were manufactured 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 thyratron-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 thyratron, kenotron and phanotron tubes.
General Electric
figure 3. The GE dimmer board backstage at the Albany, New York, Palace
Theatre. (Photo courtesy of Thomas. J. Blalock.)
In the early 20th century, GE became involved with the manufacture of switchboards for the control of stage lighting.
These indeed were just switchboards,
consisting of open knife switches and
fuse holders mounted on a thick slate
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.
œ Comprehensive capabilities for voltage, transient, small signal, and
frequency stability assessment
œIntegration with EMS/WAMS for on-line DSA to assess security
of real-time system conditions
œComputation of secure operation regions (stability limits)
œRecommendation of remedial actions
Experience and support:
www.dsatools.com
œProminent provider of the on-line DSA technology with over
40 installations worldwide including 7 out of 9 ISOs in North
America
œConsulting services for custom software and model development,
system studies, and training
VSAT
Contact us:
dsainfo@powertechlabs.com
78
IEEE power & energy magazine
TSAT
SSAT
81022-0005
january/february 2017
panel. GE also manufactured resistance
dimmers, but these usually were separately mounted above the board itself.
These switchboards were constructed
exactly the same as industrial switchboards of that era that were manufactured by GE at its Schenectady, New
York, plant.
Such a stage switchboard still exists (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.
Both the board and the dimmers still
display GE nameplates.
In the mid-1920s, GE acquired
the Trumbull Electric Manufacturing Company of Plainville, Connecticut,
which manufactured auto-transformer
dimmer boards for stage lighting control. The auto-transformer was a logical device to adapt for use as a stage
lighting dimmer. It was simple and
rugged and did not require an auxiliary source of dc for its operation as
with the saturable core reactor dimmer.
It was, however, still quite large and
heavy due to the need for a laminated
steel core.
As a result of the work of Albert Hull,
GE held the early patents on the thyratron tube. Consequently, GE decided to
enter into the development of thyratronreactor dimmers. Significant early development 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 thyratronreactor dimmers was carried out at the
Schenectady plant by research engineers Allen Bailey and Harold LaRoque
and, later, by Dudley Chambers and
Elbert Schneider. All four eventually
received patents for their contributions.
Ultimately, GE only made ten installations of thyratron-reactor lighting control in theaters, during the late 1920s
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
controlling 140 dimmers and having a
total load capacity of 1,250 kW.
january/february 2017
of the possibility of having the control
board located in the auditorium, and,
instead, it was installed backstage just
as the earlier manual-type boards had
been. The reason for this is not known
(see Figure 3).
The Palace Theatre still exists and is
still in use today but the thyratron-reactor
In 1931, GE installed a very modest
nine-dimmer thyratron-reactor board at
the new Plaza Theatre in Schenectady (demolished in 1964). In that same
year, a more significant 48-dimmer
installation was made at the new Palace Theatre in Albany, New York. In
this case, the advantage was not taken
Build Your
One-Line in the Field
as You Collect the Data
EasyPower
Power
OnSite
TM
EasyPower OnSite is the first mobile data
collection tool designed from the ground up to
use a touch interface for collecting power
system data and building one-lines in the field.
OnSite simplifies your
data collection process
and improves productivity by dramatically
reducing the time it takes
to collect data and create
the corresponding
one-line models.
Simplify the Process
Collect Data
Organize Data
Input Data
Build One-Line
System Analysis
BEFORE
Collect Data &
Build One-Line
System Analysis
Transfer Data and One-Line
AFTER
®
Power made easy.
Learn more at
www.EasyPower.com/OnSitePE
IEEE power & energy magazine
79
IF YOU HAVE
A CHALLENGE –
WE HAVE A SOLUTION
For over 100 years Slacan Industries has been Canada’s 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.
SLACAN
figure 4. The control board in the auditorium of the Earl
Carroll Theatre. [Photo courtesy of the Museum of Innovation and Science (miSci) Archives, Schenectady, New York.]
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
FEBRUARY 19–22, 2017
WASHINGTON DC
Questions: katie.burk@corporatecompliance.org
The utilities and energy
industries are highly regulated.
NEW LOCATION IN
2017
SAME GREAT
EDUCATION AND
NETWORKING
Take advantage of the
opportunity to discuss your
concerns and enjoy sharing
and exchanging ideas with
others facing the same
regulations.
Learn more and register at
corporatecompliance.org/utilities
80
IEEE power & energy magazine
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 Innovation and Science (miSci) Archives, Schenectady, New York.]
board is now long gone. The author did have the opportunity to
see it back in the 1970s but, by then, it was no longer in use.
Today, unfortunately, the entire installation is gone.
january/february 2017
(routinely referred to as the
Also in 1931, a GE instalelectrician in theatrical parlation was made at the new
lance) was called the color
Earl Carroll Theatre, located
conductor, and he appeared
across 50th Street from the
in full evening dress.
Roxy Theatre in Manhattan.
This was not a huge theEarl Carroll was a theatrical
ater (or stage), so a modest
impresario along the lines of
50 dimmers were installed,
“Roxy” Rothafel. His female
and almost half of them were
revue called “Vanities” was
devoted to the auditorium
based on Florence Ziegfeld’s
lighting, which was really a
highly popular “Follies.”
part of the show. The total
The Earl Carroll Theatre
connected lamp load was
(actually the second theater
367 kW, and the thyratronof this name on this site) was
reactor control required a toan Art Deco extravaganza.
tal of only 200 W (just 4 W
The auditorium lighting was
per dimmer). The arrangealmost as elaborate as that
ment of the board was as a
for the stage itself. Lighttwo-scene preset control (see
ing fixtures colored in red,
Figure 4). The saturable core
green, and blue were capable figure 6. Earl Carroll explaining the operation of power
reactors and the tube panels
of being additively mixed distribution panels to three of his employees. [Photo
were installed in racks loto create any color of the courtesy of the Museum of Innovation and Science (miSci)
cated in the basement (see
rainbow (and white). This Archives, Schenectady, New York.]
Figures 5 and 6).
concept had been developed
This elaborate theater existed as
by GE and was called colorama. The actor dimmers was located in a pit outcontrol 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.]
Rating-30kV 200kVA — Patent Pending
Please contact us for all
your High Voltage AC, DC,
Impulse Test, Measurement
Equipment and Service
Requirements.
EHVTEST.COM/DRY-TYPE
82
IEEE power & energy magazine
figure 8. Operating the Radio City Music Hall control console. (Photo from a
1947 Radio City Music Hall brochure.)
renovated to become retail space, and
some of the ornate auditorium decor
remained in place above a false ceiling
in a Woolworth’s store until the building was demolished in 1990.
The Westinghouse Electric Manufacturing Company of Pittsburgh, Pennsylvania, entered the electronic light dimmer field in the early 1930s by teaming
up with the Ward Leonard Electric
Company to introduce a rival device that
circumvented GE’s 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 reactor, and the combination was referred
to (for reasons not now apparent) as a
hysterset. A thyratron tube was not necessary in this control scheme.
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 Center complex being constructed in Manhattan 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 Music 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 referred to as a solenoid potentiometer.
It replaced the simple resistive potentiometers used previously to control the
signal to the grid of the thyratron tube.
This new device consisted of a coil of
wire with a movable iron core inside.
Ironically, this basically is the same configuration used at Daly’s 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 development of a saturable core reactor
control scheme used on the nearly 200ft (61-m)-long “General Electric” sign
at the plant in Schenectady. He used circuitry of his own invention, which was
capable of creating a timed brightening
and dimming of lamps with no mechanical components. Various portions of the
sign (with a total of 1,400 lamps) were
illuminated in a sequence that was repeated over and over. This sign is still in
place on top of Building 37 of the plant
but no longer exhibits this interesting sequence. It is simply lit up at night.
The Radio City Music Hall installation controlled a total of about 320 dimmers, 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 Theatre) 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 Metropolitan 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,
FOR ALL LINES
Complete system
of rating
for PC Windows
Overhead powerlines, underground
cables, and substation equipment
Real-time dynamic line ratings
View powerlines, substations and weather
stations in Google Map
ELECTROTECH
ENTERPRISES LLC
www.LineAmps.com
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.
and the total connected lighting load was
836 kW. Interestingly, this installation,
in 1933, replaced the original somewhat
“Rube Goldberg” resistance dimmer installation that dated from the construction
of the building in 1903. The final four
(less impressive) GE thyratron-reactor
installations were in theaters in Philadelphia, Pennsylvania; Iowa City, Iowa; Toronto, Ontario; and Mexico City, Mexico.
Epilogue
Undoubtedly, all of these thyratron reactor 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 control 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 incandescent 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. Devices known as silicon-controlled rectifiers 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 dimmers, 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 Alliance 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. 1–30, no. 1, 1979.
J. R. Manheimer and T. H. Joseph,
“Electronic tube control for theatre
lighting,” presented at the fall meeting of the Society of Motion Picture
Engineers, 1934.
J. E. Rubin, “The technical development of stage lighting apparatus in
the United States, 1900–1950,” Ph.D.
dissertation, Stanford Univ., Stanford,
CA, 1959.
E. D. Schneider, “Thyratron-reactor
lighting control,” AIEE Trans., vol. 57,
pp. 328–334, June 1938.
L. C. Brenneman, “Radio city music
hall: A technical discussion,” Marquee,
vol. 31, no. 3, pp. 18–23, 1999.
T. J. Blalock, “Edison’s direct current influenced Broadway show lighting,” IEEE Power Eng. Rev., vol. 22,
no. 10, pp. 36–37, 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 17–21 July 2016 to experience the next chapter in the Society’s
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 president and head professor of the
Texas A&M Department of Electrical and Computer Engineering
✔ Babak Enayati, chair of the 2016
IEEE PES GM Local Organizing Committee and lead R&D
engineer with National Grid
✔ Cheryl LaFleur, commissioner
at the Federal Energy Regulatory Commission
✔ Marcy Reed, president of National
Grid (Massachusetts jurisdiction)
✔ David L. Geier, vice president of
electric transmission and system
engineering with San Diego Gas
& Electric
✔ Juan de Bedout, chief technology officer for General Electric’s
Energy Connections business.
PES President Damir Novosel gave
a rousing PES member meeting opening session and provided the crowd
with a state of the Society address. Past
President Miroslav Begovic then provided support and introductions for the
Digital Object Identifier 10.1109/MPE.2016.2620638
Date of publication: 2 February 2017
january/february 2017
The opening reception was held at the Boston Public Library.
PES President Damir Novosel welcomes
everyone at the opening session.
various nominations and appointments
for the upcoming election. Afterward,
Babak Enayati introduced the plenary
session panelists, each of whom provided detailed industry presentations
followed by question-and-answer sessions for the attending assembly.
Following the GM’s theme, “Paving
the Way for Grid Modernization,” at-
tendees had the opportunity to be present for additional panel sessions covering topics such as transformer resiliency
and physical security, the impact of environmental regulations on power markets, power grid resilience, and so much
more. Paper and poster sessions are also
a popular attraction for GM attendees.
From an overview of the best conference papers on power system stability
and protection to an emerging technologies poster session, the prospects to get
involved were numerous. And that was
just the first couple days. Days two and
three were full of panels and poster sessions on microgrids, grid planning, DER,
big data, sustainable energy, smart buildings, smart cities, flexible energy systems, and energy policy.
To cap the event, PES announced the
recipients of its Society-level awards, all
of which were recognized and honored
during a formal ceremony in Boston.
These Society-level awards recognize
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!
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
Attendees listen intently to the speakers during the plenary session.
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.
A vision
that spans
the next century.
www.DeltaStar.com
1-800-368-3017
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/
pes-communities/awards.
As another GM came to a close, PES members and many others
looked forward to the 2017 IEEE PES GM in Chicago, scheduled for
16–20 July. The theme for the 2017 GM is “Energizing a More Secure,
Resilient, and Adaptable Grid.” Visit http://pes-gm.org/2017/ for more information and to participate in a call for papers and other opportunities.
p&e
calendar
PES meetings
for more information, www.ieee-pes.org
T
THE IEEE POWER & ENERGY
Society’s (PES’s) website (http://www
.ieee-pes.org) features a meetings section that includes calls for papers and
additional information about each of
the PES-sponsored meetings.
July 2017
January 2018
IEEE PES General Meeting (GM
2017), 16–20 July, Chicago, Illinois,
United States, contact Joseph Svachula, joseph.svachula@ComEd.com,
http://pes-gm.org/2017/
IEEE PES 2018 Joint Technical
Committee Meeting (JTCM 2018),
7–11 January, Jacksonville, Florida,
United States, contact Solveig Ward,
sward@quanta-technology.com, www
.pestechnical.org
May 2017
August 2017
IEEE International Conference on
Electrical Machines and Drives (IEMDC 2017), 21–24 May, Miami, Florida,
United States, contact Dr. Dan Ionel,
dan.ionel@ieee.org, http://iemdc.org/
IEEE Electric Ship Technologies
Symposium (ESTS 2017), 15–17 August, Washington, D.C., United States,
contact Dr. Scott Sudhoff, sudhoff@
purdue.edu, http://ests17.mit.edu/
June 2017
September 2017
IEEE PowerTech Manchester (Power
Tech 2017), 18–22 June, Manchester,
United Kingdom, contact Prof. Jovica
Milanovic, milanovic@manchester
.ac.uk, http://ieee-powertech.org/
IEEE PES Innovative Smart Grid
Technologies LA (ISGT LA 2017),
20–22 September, Quito, Ecuador,
contact Gabriel Arguello, garguello@
cenace.org.ec, http://ieee-isgt-latam.org/
IEEE PES PowerAfrica (PowerAfrica 2017), 25–30 June, Accra, Ghana,
contact Dr. Eric Kuada, dr.eric.kuada@
ieee.org
IEEE PES Innovative Smart Grid
Technologies Europe (ISGT Europe
2017), 26–29 September, Torino, Italy, contact Prof. Gianfranco Chicco,
gianfranco.chicco@polito.it
IEEE Second International Conference on DC Microgrids (ICDCM
2017), 27–29 June, Nuremberg, Germany, 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
December 2017
IEEE PES Innovative Smart Grid
Technologies Asia (ISGT Asia), 3–6
December, Auckland, New Zealand,
contact Dr. Ramesh Rayudu, Ramesh
.Rayudu@vuw.ac.nz
April 2018
IEEE PES Transmission and Distribution Conference and Exposition
(T&D 2018), 16–19 April, Denver, Colorado, United States, contact Tommy
Mayne, mayne25@charter.net, http://
www.ieeet-d.org/
August 2018
IEEE PES General Meeting (GM
2018), 5–10 August, Portland, Oregon, United States, contact Don Hall,
donald.hall@pepcoholdings.com
October 2018
IEEE PES Asia-Pacific Power & Energy Engineering Conference (APPEEC 2018), 7–10 October, Sabah,
Malaysia, contact Dr. Zuhaina Zakaria,
zuhaina@ieee.org
IEEE PES Innovative Smart Grid
Technologies Europe (ISGT Europe
2018), 21–25 October, Sarajevo, Bosnia 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. 18–23, 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:
8620%,%4%-21!-$0)$%1)+)%-#7
8-%0'72.0!'%
87"%0!-$(71)#!+%#30)27
8)120)"32%$%-%0!2).-%'3+!2).--')-%%0)-'.$%+)-'
and Impacts
.-2,)112()11/%#2!#3+!0%4%-2,!*%7.30/+!-1-.52.!22%-$ %+..*&.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 system’s physical
properties. Just look at the trend toward
prosumers, cooperatives and companies
investing in clean energy.
The level of a system’s physical integration 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 consideration to how much biomass can
contribute in determining what kind of
ownership should be promoted. That
opens up a discussion where the social
scientist isn’t the strange kid in the corner 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 strategic research and innovation agenda,
and one of the five priority pillars identified is industrialization. While industrialization requires technical solutions
for the production, logistics, and maintenance of wind turbines, it has a substantial 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 solution 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 doesn’t favor
any particular technology, and, as a
consequence, the social scientist does
not easily become the servant of a particularly 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 organizational, 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 integration 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 University’s 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 (DFKSRVLWLRQZLOOEH¿OOHGZLWKDQLQGLYLGXDOZKRPHULWVWKHUDQNRI$VVRFLDWHRU)XOO3URIHVVRUZLWKWHQXUHDQGZKRKDVDQDWLRQDOO\UHFRJQL]HG
UHFRUGRIRXWVWDQGLQJVFKRODUVKLS$SSURSULDWHO\TXDOL¿HGFDQGLGDWHVPD\EHFRQVLGHUHGIRUD7LWOHG3URIHVVRUVKLS7KHFDQGLGDWH¶VWHDFKLQJDQGUHVHDUFKVKRXOGHQFRPSDVVD
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 world’s 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$IDFXOW\RIDQGVWDIIRIVXSSRUW
XQGHUJUDGXDWHGHJUHHRIIHULQJVPDVWHU¶VGHJUHHSURJUDPVDQG3K'SURJUDPV$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&OHPVRQORFDWLRQ RUERWK LVRILQWHUHVWWRWKHDSSOLFDQW7RHQVXUHIXOOFRQVLGHUDWLRQDSSOLFDQWVPXVWDSSO\E\
)HEUXDU\KRZHYHUWKHVHDUFKZLOOUHPDLQRSHQXQWLOWKHSRVLWLRQLV¿OOHG
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#
ASPEN, Inc.
CVR 4
URL
Phone
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
Delta Star
DIgSILENT GmbH
www.cyme.com
+1 800 361 3627
86
www.DeltaStar.com
+1 800 368 3017
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
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
Neplan AG
Siemens Power Technologies International
CVR 3
usa.siemens.com/360-transmissionplanning
SKM Systems Analysis, Inc.
13
www.skm.com
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
+1 800 500 4SKM
+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
Mattias Andersson
flexibility
it’s more than an engineering challenge
F
FLEXIBILITY IS KEY FOR A SYStem under strain. When the city metro
breaks down, you require other means
of transportation. Before a big conference, you might need staff to work outside regular office hours. And when
the production of electricity is dependent 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 portfolio of available resources. The metro
is down? Walk; take the bus, the train,
your bike, or your car. Short on staff for
next week’s conference? Ask HR, communication, planning, or someone else
if you can use some of their people;
you’ll pay it back later. Are we getting
too much electricity from wind? Use it
for heating or get people to turn on appliances or change to electric cars.
Flexibility is largely about systems,
so it’s no surprise that, in recent years,
there’s been increasing interest in applying systems thinking to energy research. As a policy adviser, I’ve been
involved first hand in the transition
toward a renewable-based energy system with its large share of intermittent
energy sources, such as wind and solar.
The articles in this issue of IEEE
Power & Energy Magazine cover different 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
The problem with the system and its
parts is always to ensure that the sum is
greater than the parts, not vice versa. As
D’haeseleer and colleagues note in their
article in this issue “Flexibility Challenges for Energy Markets,” by “ignoring systems interactions, a variety of
some well-meant (local) simple targets
counteract and even oppose each other
so that some individual targets may perhaps be reached, while the overall strategic 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 isn’t 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 agreement on CO2 emission reduction.
A real challenge in increasing the
flexibility of the energy system is to
establish a collaboration between the engineers and others involved in the “hard”
sciences with those involved in policy
studies, social science, and humanities. Policy makers, funding agencies,
and some groups in the research communities have realized this. In Europe,
several of the joint programs under the
European Energy Research Alliance
have initiatives in this area, including
the Joint Programme for Energy Systems Integration.
Multidisciplinary research sounds
good in theory, but it is difficult to implement in practice. In reality, research projects tend to be divided into either technical projects with limited attention to social
acceptance and “other human factors” or
social science projects with limited attention to technical or physical limitations.
A first step to overcoming this divide 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 sciences, 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 primarily 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 arrived during a time of transition from a
centralized to a decentralized energy
system. New stakeholders, including
private citizens, cooperatives, companies and municipalities, and regions and
states have different configurations of
ownership, especially for renewable energy 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 PSS®E base software with a variety of add-on modules
) Integrate transmission and distribution by importing PSS®E network
models into PSS®SINCAL
) 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
™
Applications
PRC compliance
Fault location
Circuit breaker rating
Relay setting and testing
System-wide coordination studies
Powerful, fast, intuitive and easy to use
Advanced features
Simulates breaker-failure conditions
Models wind and solar plants and VSC converters
Time-distance diagram
Vendor-specific distance relay models
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