Volume 16 • Number 4 • July/August 2018

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.cyme.com www.cooperpowereas.com

AD917001EN-2015
 magazine

Volume 16 • Number 4 • July/August 2018

www.ieee.org/power

58
items on stage—images licensed by ingram publishing

©istockphoto.com/Jamesgdesign
electric lights— ©istockphoto.com/lapandr,
stage—©istockphoto.com/Vectorpocket,

on the
cover

features

contents
24 Electrify Everything? 69 Heat Electrification
By Philip Sterchele, Andreas Palzer, By Steve Heinen, Pierluigi Mancarella,
and Hans-Martin Henning Ciara O’Dwyer, and Mark O’Malley

34 An Electrified Future 79 Electrification and the Future

By Trieu Mai, Daniel Steinberg, Jeffrey Logan, of Electricity Markets
David Bielen, Kelly Eurek, and Colin McMillan By Ryan Jones, Ben Haley, Gabe Kwok,
Jeremy Hargreaves, and Jim Williams
48 Electrification in the United Kingdom
By Russell Fowler, Orlando Elmhirst, 90 An Electrified Nation
and Juliette Richards By Carla Frisch, Paul Donohoo-Vallett,
Caitlin Murphy, Elke Hodson,
58 On the Path to Decarbonization and Nathaniel Horner
By Amber Mahone, Zachary Subin, Ren Orans,
Mackay Miller, Lauren Regan, Mike Calviou, 99 System Restoration Readiness
Marcelo Saenz, and Nelson Bacalao By J.D. Willson and M.E. Long

columns &
108 departments
4 From the Editor 108 History
10 Letters to the Editor 122 Society News
12 Leader’s Corner 127 Calendar
20 Guest Editorial 132 In My View
robert lobenstein

Digital Object Identifier 10.1109/MPE.2017.2789000

july/august 2018 ieee power & energy magazine 1


Editor-in-Chief
Michael I. Henderson, mih.psat@gmail.com
Associate Editors
Hyde Merrill, Emertitus, History
John Paserba, History
Editorial Board
L. Barroso, A. Conejo, J. Feltes, N. Hatziargyriou,
T. Hong, B. Johnson, B. Kroposki, N. Lu,
B. Mather, M. Miller, M. O’Malley, N. Ochoa,
magazine
IEEE Periodicals/Magazines Department
445 Hoes Lane, Piscataway, NJ 08854 USA
www.ieee.org/magazines
J. Paserba, C.E. Root, H. Rudnick, M. Shahidehpour, Geraldine Krolin-Taylor, Senior Managing Editor
G.B. Sheblé, J.C. Smith, M. Thomas, E. Uzunovic, Janet Dudar, Senior Art Director
S.S. Venkata, J. Wang, S. Widergren Gail A. Schnitzer, Associate Art Director
Theresa L. Smith, Production Coordinator
Spanish Editorial Board Felicia Spagnoli, Advertising Production Manager
Enrique Tejera, Editor-in-Chief Peter M. Tuohy, Production Director
Editors: M. Baqueadano, G. Gonzalez, G. Valverde Kevin Lisankie, Editorial Services Director
Dawn M. Melley, Staff Director, IEEE Publishing
Advertising Operations
Erik Henson, Naylor Association Solutions
IEEE prohibits discrimination, harassment, and bullying. For more infor-
+1 352 333 3443, fax: +1 352 331 3525 mation, visit http://www.ieee.org/web/aboutus/whatis/policies/p9-26.html.
ehenson@naylor.com
Promoting Sustainable Forestry

ieee power & energy society (pes)

SFI-01681

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.

Governing Board I. Riaño S., Social Media Publications

S. Rahman, President G. González, Web Site Development Publications Board Chair, M. Crow
F. Lambert, President-Elect S. Ninalowo, Women in Power Editors-in-Chief
M. Armstrong, Vice President, Chapters J. Benedict, Marketing IEEE Electrification Magazine, I. Husain
F. Rahmatian, Vice President, Technical Activities M. Seif, Membership Development IEEE Power Engineering Letters, M. Fotuhi-
E. Uzunovic, Vice President, Education Firuzabad
M. Crow, Vice President, Publications Technical Council
F. Rahmatian, Chair, V. Vittal, Vice Chair IEEE Trans. on Energy Conversion, J. Jatskevich
T. Mayne, Vice President, Meetings IEEE Trans. on Power Delivery, W. Xu
J.C. Montero Q, Vice President, Membership H. Chen, Secretary, M. Sanders, Past-Chair
IEEE Trans. on Power Systems, N. Hatziargyriou
& Image Technical Committee Chairs IEEE Trans. on Smart Grid, J. Wang
S. Bahramirad, Vice President, New Initiatives/ A. Schneider, Analytical Methods for IEEE Trans. on Sustainable Energy, B. Chowdhury
Outreach Power Systems IEEE Power & Energy Technology Systems Journal,
N. Hadjsaid, Treasurer K. Chen, Electric Machinery S. Sudhoff
J. Bian, Secretary W.T. Jewell, Energy Development & Power IEEE Power & Energy Magazine, M. Henderson
D. Novosel, Past-President Generation eNewsletter, S. Fattah
D. Diaz, Region Rep., U.S. & Canada C. Searles, Energy Storage & Stationary Battery PES Representative to IEEE Press, Open
J. Milanovic, Region Rep., Europe, Middle East, E. Bascom, Insulated Conductors Website, Open
& Africa T. Koshy, Nuclear Power Engineering
S. Leon, Region Rep., Latin America M. Dood, Power System Communications Meetings
D. Sharafi, Region Rep., Asia & Pacific & Cybersecurity Committee Chairs
C.Y. Chung, Member-at-Large C. Canizares, Power System Dynamic Performance General Meeting Steering, D. Hall
B. Enayati, Member-at-Large J. McBride, Power System Instrumentation Joint Technical Committee Meeting Steering,
H. Koch, Member-at-Large & Measurements S. Ward
L. Ochoa, Member-at-Large L. Barroso, Power System Operation Planning Technically Cosponsored Conferences Steering,
& Economics J. Mitra
IEEE Division VII Director Transmission & Distribution Conference &
B. Meyer M. Pratap, Power System Relaying & Control
S. Chandler, Smart Building, Load & Exposition Committee for North America
Steering, C. Segneri
IEEE Division VII Director-Elect Customer Systems
Innovative Smart Grid Technology Conference–
Open D. Watkins, Substations
R. Hotchkiss, Surge Protective Devices Europe, D. Van Hertem
PES Executive Director T. Irwin, Switchgear Innovative Smart Grid Technology Conference–
Patrick Ryan, +1 732 465 6618, S. McNelly, Transformers Asia, TBD
fax +1 732 562 3881, p.ryan@ieee.org D. Sabin, Transmission & Distribution Website, Open

Standing Committee Chairs Technical Council Coordinating Committees Education

H. Louie, Constitution & Bylaws D. Houseman, Intelligent Grid & Emerging Committee Chairs
V. Kolluri, Fellows Technology Power and Energy Education, S. Brahma
N. Hadjsaid, Finance D. Alexander, Marine Systems Scholarship Plus, J. Peer
A. Apostolov, History D. Lew, Wind & Solar Power Website, Open
F. Lambert, Long-Range Planning
D. Novosel, Nominations & Appointments Technical Council Standing Committees New Initiatives and Outreach
M. Sanders, Awards Committee Chairs
Chapter Representatives V. Vittal, Technical Sessions IEEE Smart Village, R. Larsen, Chair
F. Al Dhaheri, Z. Bo, R. Cespedes, H. Chen, Organization & Procedures R. Podmore, Vice-Chair
S. Cundeva, C. Diamond, B. Djokic, T. Burse, Standards Coordination Industry Outreach, J. Giri, Chair
G. Gonzalez, J. Khan, R. Nagaraja, D. Toland, Webmaster Website, K. Anastasopoulos
N. Nair, M. Papic, I. Petruzela, Open, Industry Education
T. Salihy, G.N. Taranto, D. van Hertem,
M.C. Wong, Z. Zakaria
Chapter Committee Chairs IEEE Power & Energy Magazine
C. Diamond, Electronic Communications IEEE Power & Energy Magazine (ISSN 1540-7977) (IPEMCF) is published bimonthly by the Institute of Electrical and
E. Carlsen, Awards & Resources Electronics Engineers, Inc. Headquarters: 3 Park Avenue, 17th Floor, New York, NY 10016-5997 USA. Responsibility for  the
Y. Chen, Distinguished Lecturer Program contents rests upon the authors and not upon the IEEE, the Society, or its members. IEEE Operations Center (for orders, sub-
scriptions, address changes): 445 Hoes Lane, Piscataway, NJ 08854 USA. Telephone: +1 732 981 0060, +1 800 678 4333.
C. Diamond, Chapters Website Individual copies: IEEE members US$20.00 (first copy only), nonmembers US$82.00 per copy. Subscription Rates: Society members
Membership & Image Committee Chairs 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)
T. Doggett, VP Deputy 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
V. Madani, Awards & Recognition the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA; 2) pre-1978 articles without fee. For other copying,
J. Hofman, Young Professionals reprint, or republication permission, write Copyrights and Permissions Department, IEEE Operations Center, 445 Hoes Lane, Piscat-
Open, Ambassadors away, NJ 08854 USA. Copyright © 2018 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,
Digital Object Identifier 10.1109/MPE.2017.2789001 IEEE Operations Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. Canadian GST #125634188 printed in u.s.a.

2 ieee power & energy magazine july/august 2018

DIgSILENT

PowerFactory 2018
DIgSILENT has set standards and trends in power system modelling,
analysis and simulation for more than 25 years. The proven advantages
of the PowerFactory software are its overall functional integration, its
applicability to the modelling of generation, transmission, distribution and
industrial grids and the analysis of the interactions of these grids.

PowerFactory 2018 responds to the latest analysis strategies

for systems with a high proportion of renewable intermittent
generation by the introduction of Probabilistic Analysis. In addition
to the probabilistic AC/DC load flow and probabilistic optimal
power flow calculations, this major new development incorporates
powerful modelling capabilities, enabling efficient handling of large
stochastic data sets. The 2018 release also incorporates a variety
of new and enhanced models, as well as further improvements to
the algorithms and usability of existing modules. In addition, the
new version incorporates a new data model extension concept,
allowing highly flexible definition of user-defined attributes for
system components and data objects in general.

Selected Key Features

New module Probabilitic Analyis of AC/DC and optimal load
flow, incorporating powerful capabilities for managing stochastic
parameters based on distribution curves and their correlations
Remedial Action Schemes (RAS) for comprehensive post-fault
contingency analysis
New Frequency Response Analysis of dynamic models
Real-time data streaming for RMS simulations, based on IEEE C37.118
Save and restore time domain simulation state (snapshots) for
subsequent analysis
Enhanced Cable System Analysis with improved cable layout
modelling
Revision of DIgSILENT global library, with all protection devices
now incorporated
New and improved models including Multi-Core Cables, Thyristor
Controlled Series Capacitor and HVDC Line Commutated Converter
Data Model Extension concept for user-defined attributes,
opening up almost unlimited data handling possibilities
Powerful graphic search option for network elements in
diagrams, including geographic search

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

 from the editor
W
W h at a n exciti ng
time to be in the electric
power industry! as mem-
bers of the ieee, we stand
ready to deal with the rapid
growth of state-of-the-art
technologies, the major
transformation of the grid
to one with ever-more re -
newable resou rces, a nd
challenges arising from
interactions between the
bulk system grid and the
distribution system. given
all this activity, how do we
conduct long-term strategic
planning, ensure that com-
plex issues can be suitably
addressed, and meet tar-
geted reductions of car-
bon emissions?
Scenario analysis can identify many
of the key issues needed to inform pol-
icy makers and industry stakeholders.
the simulation of a variety of futures
can show the implications of potential
policies and the conceivable need for
changes in technologies and markets
to make the policies a reality. there
is real value to establishing common
frameworks for technical challenges
that must be addressed.
the reduction of carbon emissions
remains a primary policy worldwide as
a means of addressing climate change.
these policies have contributed to the
widespread growth of energy efficien-
Digital Object Identifier 10.1109/MPE.2018.2824098
Date of publication: 18 June 2018
4 ieee power & energy magazine
electric for all
scenario analyses for reducing carbon emissions
cy, wind and photovoltaic resources,
and, more recently, electric vehicle de-
velopment. although many established
power systems currently experience
low growth of demand, scenario analy-
sis shows increased electrification will
be required to significantly reduce the
total societal carbon emissions. One of
the most economical means of lower-
ing carbon emissions can be achieved
by converting a number of end users
of fossil fuels to electric loads supplied
by zero-emitting renewable resources.
targeted loads include transporta-
tion vehicles, heating plants, and some
industrial processes. Future power sys-
tems, however, could look very dif-
ferent with a considerable expansion
of the grid, storage technologies such
Michael Henderson
as batteries and electrically
produced fuels, and new de-
mand-response technolo-
gies all playing key roles in
meeting requirements for
ancillary services, peak shav-
ing, and valley filling.
In This Issue
this issue of IEEE Power &
Energy Magazine summa-
rizes the scopes, assump-
tions, methods, and results
of scenario analyses achiev-
ing low-carbon futures. the
©istockphoto.com/John1179
studies clarify several of
the challenges that could be
posed by various futures
and suggest ways they could
be successfully resolved. a
common theme is the need
for end-use electrification as a way of
reducing carbon emissions.
guest editor Mackay Miller com-
piled seven well-written articles that
summarize the means of conducting
scenario analyses and explore the tech-
nical and economic feasibility of “elec-
trifying everything.” as described in
the “guest editorial,” the issue provides
critical insights into studying power
system futures, including
✔ approaches for conducting sce-
nario analyses and understand-
ing counterintuitive results
✔ ways of evaluating and priorit-
itizing changes to the supply and
demand mix
✔ the identification of potential ad-
vantages of electrifying heat and
july/august 2018
 Analysis, planning and opamizaaon of networks
Fastest dynamic simulaaon
Asset Simulaaon and Forecasang

Electricity

Gas / Water / Heaang

NEPLAN relies on cuung-edge technology

> Modelling of any AC and DC networks with AC/DC converters
> State-of-the-art transient models for renewable energies, storage,
HVDC, VSC, wind turbines according to IEC 61970, etc.
> Calculaaon of very large transmission and distribuaon networks
> Automaac stress test for e-mobility and PV systems
> Combines electricit
electricity, gas and district heaang networks
> Risk analysis with asset management
Asset Management > Integrated protecaon device seeng and management in mula-user
environment
> High-end CIM Manager: Exchange of network data and models
according to IEC 61400

When it has to be fast, reliable and simple

NEPLAN AG Power Systems Engineering
Oberwachstgrasse 2
CH 8700 Kuesnacht, Zurich
Phone +41 44914 36 66
www.neplan.ch
Training info@neplan.ch
 transportation loads and the bar-
riers to achieving these goals
✔ differences and similarities of
achieving carbon-reduction goals
in different climatic regions
✔ the potential for demand to pro-
vide both energy and capac-
ity services
✔ the need for market reform and
changes to rate structures for
low-carbon futures
✔ a summary of several scenario
studies and the lessons learned
from studies of the United States.
the “in My View” column by Jim Wil-
liams discusses many of the factors
and policies that could reduce car-
bon emissions. he recognizes elec -
trification, energy efficiency, and
the retention of nuclear power as ma-
jor considerations.
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.
www.dsatools.com
VSAT
Contact us:
dsainfo@powertechlabs.com
6 ieee power & energy magazine
the issue also fea-
tures a standalone ar-
Scenario analysis solution techniques for
load flow and optimal
ticle, by J.D. Willson and can identify power flow. it should
M.e. Long, on system be of keen interest to
restoration read i ness many of the key both students and indus-
in north america over
the last 25 years. the
issues needed try practitioners.
authors acknowledge to inform PES Updates
important roles played the ieee plays an es-
by the ieee Power & policy makers sential role in prepar-
energy Society (PeS)
and the north ameri-
and industry i ng the next genera-
tion of electric power
can electric Reliabil- stakeholders. professionals and con-
ity corporation. tinuing their lifelong
education. as discussed
Book Review by Frank Lambert, PeS president-elect,
thanks to Ramu Ramanathan, our issue ieee Student Branch chapters have ex-
features a book review of Load Flow Op- perienced phenomenal growth. he en-
timization and Optimal Power, written courages local PeS chapters and Young
by J.c. Das. the book comprehensive- Professionals to show the true value of
ly covers system modeling and different continued PeS membership by providing
Advanced modeling and analysis features include:
• Comprehensive capabilities for voltage, transient, small signal, and
frequency stability assessment
• Integration with EMS/WAMS for on-line DSA to assess security
of real-time system conditions
• Computation of secure operation regions (stability limits)
• Recommendation of remedial actions
Experience and support:
• Prominent provider of the on-line DSA technology with over
40 installations worldwide including 7 out of 9 ISOs in North
America
• Consulting services for custom software and model development,
system studies, and training
TSAT SSAT
81022-0005
july/august 2018
 etap 18
®

Convergence of Power & Intelligence

● etapAPP™ Intelligent mobile app for data collection & synchronization

● etap iCE™ Intelligent Control & Remote Terminal Unit hardware
● StarZ™ T&D Protection & Coordination – distance relay, scheme logic, sliding fault
● ArcFault™ High Voltage Arc Flash for systems up to 800 kV – LG, LL, 3-phase
● eTraX™ AC & DC railway power analysis for traction & signaling systems
● DataX™ Data Exchange – Revit®, ArcGIS®, MultiSpeak®, CIM
● EMTP Electromagnetic Transient Program interface – PSCAD®, EMTP-RV
● ADMS Advanced Distribution Management – SCADA, DMS, OMS
● TDLF Time Domain Load Flow – AC, DC, time series

Learn What’s New

etap.com/18

sales@etap.com +1 949.900.1000
 speakers at Student
Branches and mentor-
A common slav M. Begovic and
Miriam P. Sanders.
company that quickly succumbed to
completion, which implemented other
ing individuals. theme is the in “Society news,” groundbreaking advances in the state-
we also recognize the of-the-art of those times.
Vote! need for recent passing of Mike
t h i s issue fe a t u r e s
end-use adibi, who made ma- Thanks
st a t e m e nt s by can- jor cont r ibut ions to a special note of appreciation to Mel
didates for the ieee electrification system restoration. Olken, who continues to provide gui-
Division Vii delegate- dance and tutelage, and to ieee pub-
ele ct /d i re ctor- ele ct as a way of History lications staff who make this publi-
2019 and ieee Divi-
sion Vii delegate/di-
reducing carbon in the “history” column,
Joseph J. cunningham
cation possible. thanks to the many
contributors to this issue, especially
rector 2020–2021 (the emissions. shares information on our guest editor, Mackay Milller, and
elected individual will the early history of the authors. a particular note of apprecia-
serve both as both delegate and direc- electrification of Lower Manhattan, tion to associate editor John Paserba
tor). Please see the “Society news” new York. he discusses the rapid rise and Robert c. henderson, who pro-
column for an introduction to the two of innovative technology developed by vides editorial assistance.
p&e
highly qualified candidates, Miro- L e o Da f t a nd t he excelsior Power

World Leader in Developing Powerful and

Field-Proven On-line and Off-line Tools
Bigwood for Power Grid On-line Monitoring,
Systems, Inc. Assessment, Enhancement,
www.bigwood-systems.com
sales@bigwood-systems.com and Optimization
t: +1.607.257.0915
f: +1.607.257.0237 On-line, Look-ahead, and Day-ahead Solutions
 Real-time and Look-ahead Static and Dynamic Security Assessment
Innovation prevails! (patented screening, ranking, and detailed analysis)
BSI has invented and holds the rights
to 18 patents in this area of expertise.  Actionable Control Recommendation Engine to eliminate violations and
mitigate unstable contingencies
 On-line System Operating Limit Engine: on-line operating limit determina-
tion for voltage stability, transient stability, and small-signal stability
 On-line Volt/VAR Assessment and Control, including on-line reactive re-
serve management
 Next-Generation State Estimation for transmission and distribution net-
works with a low number of measurements (patent pending)
 Global Optimal Power Flow for over 250,000 control variables with AC/
“Bigwood Systems’ Online Voltage
Stability product is simply the best in
DC power flow and security constraints, including feasibility check
the field.”
- CAISO Project Leader
 Advanced PMU-based Applications software portfolio

“BSI On-line Transient Stability Assess- World-Wide Customer Base

ment is a superb product.”
- TEPCO Manager
“The BSI VAR Management System is a
best practice tool for all utilities.”
- SERC Audit Staff
BSI has provided solutions and services to over 35 electrical utilities world-wide, includ-
ing long-standing customers such as California ISO, PJM Interconnection (the largest
system operator in the world), Tokyo Electric Power Company (TEPCO), Tennessee
Valley Authority, and United Kingdom Power Networks (England), among others. BSI’s
advanced technology, innovation, and satisfied customers are key differentiators.

8 ieee power & energy magazine july/august 2018

 letters to the editor
R
READERS ARE ENCOURAGED tO
share their views on issues affecting
the electric power engineering profes-
sion. Send your letters via e-mail to
Michael Henderson, editor-in-chief, mih
.psat@gmail.com. Letters may be ed-
ited for publication.
When Electrodes Erode
Your January/February 2018 issue has
a fine article about arc lights (R. D.
Barnett, “Arc lighting systems,” IEEE
Power & Energy Magazine, vol. 16, no. 1,
pp. 56–64, 2018). I learned that in a dc
arc, the positive electrode erodes in a
concave shape, thus casting light down-
ward along the axis of the carbon rods.
the article did not say how the elec-
trodes erode when excited by ac. Do
both electrodes develop hollow de-
pressions? Are there two light peaks
along each electrode? I seem to remem-
ber that movie projectors (probably ac)
mounted the rods or thogonal to the
line of sight to the screen, implying a
nondirectional arc.
Myron Kayton
Digital Object Identifier 10.1109/MPE.2018.2819027
Date of publication: 18 June 2018
10 ieee power & energy magazine
share your thoughts
send comments to mih.psat@gmail.com
Author’s Reply
In the ICS text, Electric Lighting and
Railways (Scranton, Pennsylvania,
1901, pp. 18:6–10), the unknown author
states that for the open
(no outer globe) “al-
ternating current arc, Your January/
both carbons become
pointed or have very
February 2018
small craters, so that issue has a fine
the light is thrown up-
wards much more than article about
with the direct-current
lamp.” Presumably, the
arc lights.
lamps that produce cra-
ters differ from those that produce points
in some unspecified way.
With a closed arc lamp, the light
distribution is similar to the open ac
lamp—light is also thrown upward but,
in this case (closed), both carbons have
flat ends. the author says that this is
a result of the arc’s tendency to “shift
around over the ends,” but he offers no
mechanism to explain the phenomenon.
In closed lamps, the erosion of the
carbons is greatly reduced because of
the partial exclusion of air, and perhaps
this affects the shape in some way. With
respect to the question regarding the two
light peaks, I believe the reader is imply-
ing that, because of the shifting polarity,
there could be two alternating sources
(top then bottom) of light. this may be
the case but for such a brief interval that
it would not be discernable to the hu-
man eye. It should be
emphasized that the ac
arc lamps were not in
widespread use. Pro-
jecting light up was not
of much use in street
lighting, the realm of
most arc lights, so not
much information was
provided for students
who would not fre-
quently encounter ac arc lights.
About the projector question,
pp. 18:89–90 from the same book deal
with arc lamps used for “photo-en-
graving work, blueprinting, search-
lights” and for “projection work.” In the
latter type, dc arc lamps were used be-
cause, in ac lamps, the arc hummed
loudly and moved around more than
for a dc arc. Focusing mirrors were used
with projector arcs, and the feed mecha-
nism was more precise to keep the arc at
the focal point of the mirrors. I hope this
answers your questions. Let me know if
you need anything more.
Bob Barnett
p&e
july/august 2018
 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:

• Insulation ccoordination: lightning and switching;

• Harmonics, ferroresonance, and power quality;
• HVDC and FACTS;
• Wind, solar, and distributed generation;
• Protection and relays;
• Equipment failure analysis.

Start your simulations today. Our technical support

team is ready to help.

info@pscad.com
pscad.com
Available in accessible formats upon request. Powered by Manitoba Hydro International Ltd.
 leader’s corner
Frank Lambert

today’s student members

they are tomorrow’s leaders

T
The Ieee has been a very Im-
portant part of my professional career
since my days as a student member
at Georgia Tech. after graduation in
1973, I began work at Georgia Power
Company as a distribution engineer
and started attending our local atlanta
Ieee Power & energy society (Pes)
Chapter meetings. These meetings were
very beneficial—they helped me grow
my professional network and ex-
pand areas of expertise for my career.
The electric utility industry in the
United states suffered a significant
downturn in the early 1990s and all
but stopped hiring new engineers. This
trend continued until 2011 when Pes
Digital Object Identifier 10.1109/MPE.2018.2823478
Date of publication: 18 June 2018
launched our scholarship+ Program to
attract students back to our profession.
Pes has been and continues to be
vital to our engineers working in in-
dustry. many engineers from my gen-
eration are retired or soon will be. We
need to work together with our local
Pes professional and student branch
(sb) Chapters to help bring in the next
generation of power engineering pro-
fessionals and get them involved in Pes
upon graduation.
Pes has experienced phenomenal
growth in our sb Chapters since we
held our first Pes student Congress in
ankara, Turkey, in august 2014. at the
beginning of 2014, Pes only had 134
sb Chapters. The second Pes student
Congress was held in Kuala Lumpur in
august 2016. by the end of that year,
the number of Pes sb Chapters had in-
creased to 240. Pes sb Chapters num-
bered 316 at the end of 2017, as shown
in Figure 1.
The growth exper ienced in sb
Chapters in regions 1–7 (the United
states and Canada) has been much less
than that enjoyed in other regions. To
bring the same level of excitement to
regions 1–7, a special one-time stu-
dent Congress was held in boston in
august 2017.
The Third Pes student Congress is
planned for 25–27 august 2018 in são
bernado do Campo, são Paulo, brazil,
to continue the growth. This congress
will have three pillars to prepare our fu-
ture leaders: academic, industry, and
entrepreneurship. Our total Pes member-
ship was 38,925 at the end of December

350

300 2014 2015 2016 2017

PES SB Chapters

250
200

150

100

50

0
R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 Total
2014 5 2 6 3 4 4 3 23 36 48 134
2015 10 2 8 3 4 8 3 25 49 63 175
2016 11 2 11 4 4 8 3 39 62 96 240
2017 12 2 11 5 5 11 5 57 90 118 316
IEEE Regions

figure 1. The growth of PES SB Chapters.

12 ieee power & energy magazine july/august 2018

THE TOTAL
PACKAGE

AFL offers electric utilities a wide range of custom-designed solutions that will help increase efficiency of
networking operation centers, power-generating facilities and substations. With more than 80 years of
industry expertise, our substation solutions are developed for applications up to 765 kV and available for
Swage, bolted or welded connections in bronze or aluminum materials. At AFL, we aren’t just one piece of the
solution, we are the solution.

AFLglobal.com/IEEE
800-235-3423

table 1. PES student members by Region.
IEEE Region Geographical Area
1–7 United States and
Canada
8 Africa, Europe, and
Middle East
9 Latin America
10 Asia and Pacific
Total
2017, with almost 27% of that students.
see Table 1.
One of the biggest challenges facing
Pes today is the transition from stu-
dent member to member after gradu-
ation, and we need your help! This is
a particularly difficult time financially
for many new hires, and often they
don’t see room in their limited budget
14 ieee power & energy magazine
This can include providing speakers
for sb Chapter meetings, inviting sb
PES Student PES Graduate Chapter members to attend local Pes
Members Student Members
Chapter meetings free of charge, vol-
704 979 unteering to serve as mentors for stu-
dents, and sponsoring technical tours.
1,308 677 also encourage them to get involved
in some of the leading initiatives, like
1,682 236 Ieee smart village, Ieee smart Cit-
ies, and Ieee smart Grid, to experience
3,671 1,151
the value of membership while they are
7,365 3,043 still students.
all Pes graduate students are Pes
young Professionals (yPs) and can
for Pes membership. If they have not hold officer positions in local Pes
seen any value in Pes membership Chapters. We have success cases with
as a student, chances are they will not Pes graduate students who were sb
make the transition to Pes member af- Chapter chairs and then transitioned
ter graduation. to officers in their local Pes Chapter.
Consider inviting some past Pes sb
PES Chapters Chapter chairs and Pes graduate stu-
We encourage all of our Pes Chapters dents (yPs) to get involved as an of-
to adopt an sb Chapter in their area. ficer in your Chapter. This action can
july/august 2018

Rugged Test Equipment for the Power Industry
Single and 3-Phase
Transformer Ratiometers
10A and 200A
Contact Resistance
Power Factor Test Sets
Winding Resistance with
Core Demagnetization
Current Transformer Test Sets
Raytech USA
5 Years
What Does a 5-Year
Warranty Mean to You?
Our exclusive 5-year warranty provides value-added benefits to
each and every Raytech instrument.
• Less Repairs – lower overall cost of the instrument
• Less Downtime – higher productive hours
• More Value – each year of a warranty can be valued up to
10% of the total cost of the instrument
Our customers know that the Raytech standard 5-year warranty
means they will have years of trouble-free operation, without
the cost of replacing their instruments. If your current testing
equipment doesn’t come standard with a 5-year warranty,
contact Raytech USA.
Toll Free: 888 484 3779 | Phone 267 404 2676
118 South 2nd Street, Perkasie, PA 18944
www.RaytechUSA.com
 The First PES Student Congress shows the future leaders of tomorrow.

bring energy and excitement to your and help them start one. It only takes a viduals who have not been a Pes mem-
Pes Chapter! Pes faculty advisor and six Pes stu- ber before 2018. more information is
If your local university doesn’t have dent members. Pes has a “first-year” available at https://www.ieee-pes.org/
a Pes sb Chapter, reach out to them free student membership valid for indi- images/files/pdf/student_flyer.pdf.

Power Forward with OMICRON

We invest in the future of testing - nearly 20 % of our revenue
and more than 30 years of experience flow into R&D of testing
and monitoring solutions.

OMICRON instruments are designed and developed for reliable

condition diagnosis of your electrical equipment, such as power
and instrument transformers, circuit breakers, reclosers and relays.

Customers from more than 160 countries rely on OMICRON

products and services.
Jackie Peer
Head of Strategic
www.omicronenergy.com Business Development

16 ieee power & energy magazine

AD18072-General-AD-178x124mm-IEEEMagazine-ENU.indd 1 july/august
2018-05-11 2018
14:38:45
 The Second PES Student Congress was a success.
PES Young Professionals
We also encourage our yPs to volun-
teer as speakers for our sb Chapters
in their local area. experience has
shown that our student members very
much like to hear from recent gradu-
ates who are now in the workforce.
yPs can also serve as mentors and be
very helpful in providing advice for
career decisions. T he nea rest Pes
sb chapter can be found in the Pes
18 ieee power & energy magazine
Chapter locator; see https://www
.ieee-pes.org/pes-communities/chapters/
chapter-locator.
Other Ideas
If you or your Chapter has experience
with other approaches to ease the tran-
sition from student member to member,
please send me an e-mail at flambert@
ieee.org, and we will schedule a call
to discuss.
The Bottom Line
Pes student members need to expe-
rience the value of Pes membership
while they are still students. reach
out to them, and show them the ben-
efits of belonging to our Pes family.
Today’s Pes student members are
tomorrow’s leaders!
p&e
july/august 2018
Digital Object Identifier 10.1109/MPE.2018.2835845
 guest editorial
electrification
its role in deeply decarbonized energy systems
E
“ElEctrify EvErything” has in the first article, researchers from
become a common slogan, an easy-to- the fraunhofer institute for solar En-
understand pathway promising a clean- ergy systems describe pathways to a
er and more efficient energy system. nearly carbon dioxide (cO2)-free german
indeed, consensus is emerging that dra- energy system by 2050. the authors
matic growth in electrification across lay out a rigorous simu-
transportation, buildings, and even lation approach to eval-
some industrial uses will be a key strat- uate the technical and While the
egy for achieving deep decarbonization.
yet while the concept of electrification
economic feasibility of
achieving three pos-
concept of
is easy to understand, many of the key sible cO 2 em issions electrification
power system issues remain under- reductions targets by
appreciated and unresolved. the full 2050: 80, 90, and 95%. is easy to
electrification of transportation, for ex-
ample, would require massive changes
in all three futures, the
transformation of nearly
understand,
to grid design and operation, and full all economic sectors is many of the
electrification of heat networks in cold required. But large and
climates would place dramatic new sometimes counterin- key power
pressures on wintertime supply and de-
livery. in a highly electric economy with
tuitive differences also
emerge in meeting these
system issues
significant shares of variable wind and three decarbonization remain under-
solar electricity, systemic mismatches targets. By describing
between supply and demand would ne- and contrasting these appreciated and
cessitate vast amounts of storage as well
as entirely new paradigms for capital in-
differences, the authors
help us understand how
unresolved.
vestment and cost recovery. a seemingly small differ-
the goal of this issue of IEEE Power ence in emissions targets can have a large
& Energy Magazine is to provide a snap- system impact.
shot of state-of-the-art research fo- in the second article, researchers
cused on the role of electricity in deeply from the national renewable Energy
decarbonized energy systems. four of laboratory (nrEl) describe an almost
the articles take a geographic approach fully electrified U.s. energy system. sum-
and describe how national or regional marizing a recently published study, the
electric sectors might evolve in highly authors detail scenarios for high elec-
decarbonized futures. two of the arti- trification in the United states, with
cles explore topics that cut across many and without accelerated investment
regions: electrification of heat and mar- in energy efficiency and zero-carbon
ket design. the final article summarizes electricity. the authors explore how
a wide range of recent electrification each scenario would impact electric-
simulations and proposes principles for ity load shapes, generation capacity
advancing the quality of analysis. expansion, and cO 2 emissions. the
article also quantifies how much cO2
Digital Object Identifier 10.1109/MPE.2018.2824099
reduction can be expected from end-
Date of publication: 18 June 2018 use electrification on its own, as well
20 ieee power & energy magazine
Mackay Miller
as when electrification is paired with
energy efficiency, zero-carbon elec-
tricity, or both. in this way, the authors
illuminate issues that can help decision
makers identify priorities for cost-
effective cO2 abatement.
in the third article,
researchers from na-
tional grid in the Unit-
ed Kingdom describe
h o w ele ct r i f icat ion
contributes to achiev-
ing a deeply decarbon-
ized U.K. economy by
midcentury. Drawing
on their most recent
“future Energy scena-
rios” report, which has
been published annu-
ally since 2011, they lay
out what would need
to happen in three key
energy sectors (heat,
transport, and power)
to achieve the stated
national target of 80%
cO2 reduction by 2050. across heat
and transport, they find rapid growth in
end-use electrification and identify cus-
tomer adoption as a key barrier to the
pace needed for achieving the targets.
the fourth article compares and con-
trasts 2030 decarbonization pathways
in two distinct regions of the United
states with similar carbon reduction
goals: california and the seven-state
northeast region. researchers from En-
ergy and Environmental Economics, inc.
(E3), national grid U.s., and siemens
Power technologies international de-
scribe what would need to happen in
the two regions to achieve a 40% reduc-
tion by 2030. not surprisingly, given the
different climates, renewable resource
july/august 2018
Brilliant
HVDC-Diamond® System

Your trusted point-to-point connection to all facets of VSC HVDC

The leader in high-voltage DC applications is your local, single source for
Voltage-Sourced Converter (VSC)-based HVDC systems and equipment:

 Mitsubishi’s next generation X-Series IGBTs



 Local value-added services: from initial engineering studies to after-
sales maintenance and support
Learn more at www.meppi.com/hvdcdiamond or contact hvdc-sales@meppi.com
The IGBT X-Series: the hightest-rated
We control the enabling technology because we own it. Brilliant. voltage class with next generation chip

WWW.MEPPI.COM/HVDCDIAMOND
 cigrE SeSSiOn
August 26 - 31, 2018
paris - Palais des congrès
47 A TechnicAl
eXhiBiTiOn
*registration *Dernier délai oN 3 floors

ending soon! pour s’inscrire ! An All week

TEchNical
ProgrammE
register Bénéficiez
A unique
at reduced du tarif réduit OppOrTuniTy FOr
rate up to pour toute neTwOrking
31 July 2018 inscription reçue wiTh 6600
maNagErs aNd
*Registration closing au plus tard le ExPErTs from
on August 10, 2018
31 juillet 2018 ThE worldwidE
PowEr iNdusTry.
*Fermeture des inscriptions
le 10 août 2018

InternatIonal CounCIl on large eleCtrIC SyStemS

Conseil International des Grands Réseaux Électriques

Digital Object Identifier 10.1109/MPE.2018.2835846

endowments, and building stocks, the two pathways look quite
different, yet both rely heavily on electrification in achieving
the stated goals. these similarities and differences illuminate
how the unique features of different regions will powerfully
shape energy system evolution under decarbonization.
in the fifth article, a team of researchers from vector Un-
limited, University of Manchester, University college Dub-
lin, and nrEl explore heat electrification in depth. looking
across Europe, the team summarizes cutting-edge research
into how heating requirements might be anticipated, minimi-
zed, and managed. for example, they describe unique re-
search that simulates how household changes, such as relax-
ing thermostat settings by 2–3 °c, would have significant
system impacts as diverse as reducing generation capacity,
wind curtailment, and coal generator ramping. the article
reveals how important heat electrification will become in
energy-system planning in cold climates.
in the sixth article, researchers from Evolved Energy and
the University of san francisco explore the challenges that
electrification may pose for electricity market design. in par-
ticular, the authors describe the novel and under-appreciated
system challenges that would emerge in highly electrified,
highly renewable energy systems. they go on to analyze why
these challenges would likely pose existential questions for
existing electricity market designs and conclude by suggest-
ing concepts for future market design.
in the seventh article, researchers from the U.s. Depart-
ment of Energy survey the most high-profile U.s. electrifica-
tion studies of the past five years, identifying commonalities 1.800.231.6074
but also significant differences between the studies’ findings
and methodologies. for example, the scenarios forecast a substation
wide range of projected electricity growth, ranging from a
1.5× increase to a 3× increase by 2050. the authors then go
on to lay out a research program to improve the rigor, consis-
liners
tency, and relevance of electrification studies.
in the concluding “in My view” column, Jim Williams
of the University of san francisco notes that many coun-
tries have adopted science-based targets for greenhouse
gas reductions, and he then succinctly describes the mul-
tiple, overlapping paradigm shifts that would be required
to achieve these targets. the list of paradigm shifts spans
nearly all parts of the energy landscape: how energy effi-
ciency is promoted by policy, how struggling nuclear plants
are dealt with, how electricity markets are designed, and
how utilities are regulated.
together, this collection attempts to provide a useful
snapshot of activity in the electrification space. from the di-
versity of authors and varying geographies being studied, it
is apparent that electrification and its impacts have become a
vital discussion point within the industry. We hope you will
enjoy reading it and continue to engage with the power and
energy community to advance the state of the art in this cru- Transformer
cial area.
p&e
Bags
reefindustries.com
 By Philip Sterchele,
Andreas Palzer, and
Hans-Martin Henning
Electrify
Everything?
T
The firsT efforTs To limiT greenhouse gas
(ghg) emissions on an international scale date back to 1997
with the ratification of the Kyoto Protocol. all contracting
states agreed to reduce their emissions by 18% in relation
to 1990. on 12 september 2015, in the Conference of Par-
ties 21, a follow-up agreement was elaborated to assure an
ongoing pursuit of climate protection strategies. in total, 195
parties to the Convention agreed on common goals to further
reduce their ghg emissions. This climate agreement came
into force on 4 november 2016 and, as of november 2017,
was ratified by 170 states.
By the end of september 2010, the german federal govern-
ment had already developed a strategy for an environmentally
Digital Object Identifier 10.1109/MPE.2018.2824100
Date of publication: 18 June 2018
24 ieee power & energy magazine 1540-7977/18©2018IEEE
©istockphoto.com/tolokonov
friendly, reliable, and affordable energy supply. By doing so,
the first steps for transforming the german energy system
were taken. The primary goal of this transformation, called
“energiewende,” is to significantly reduce ghg emissions. fig-
ure 1 shows the historical emission values from 1990 to
2015 in germany divided into energy-related and other emis-
sions, such as from agriculture and specific industry pro-
cesses. The share of the energy-related emissions accounts
clearly for the majority of the total ghg emissions (approxi-
mately 85–90%). The blue line highlights the ghg emis-
sion reduction targets set by the german federal government:
−20% by 2010, −40% by 2020, −55% by 2030, −70% by
2040, and at least −80% in 2050 or −95% wherever possible
(relative to 1990).
To investigate what technologies are needed for a techni-
cally possible and economically feasible transformation of the
july/august 2018
 Exploring the Role of the Electric
Sector in a Nearly CO2-Neutral
National Energy System
german energy system in line
with the declared ghg reduction
targets, various studies by differ-
ent teams throughout germany
are being carried out. The energy
system models used as the basis
for those studies distinguish them-
selves by different features. The remix
model, for example, focuses heavily on
the electricity sector and its link to the
heating and mobility sector, considering, for
example, the import and export of electricity
with neighboring countries in a multiple-node
approach. The aim of the model is to assess future
energy-supply scenarios with high spatial and tempo-
ral resolution. like the remix model, the scope model is
a linear optimization program used to assess different energy-
supply scenarios. its typical study area is europe (the Balkan
regions are excluded), while the investigated period is defined
by target years until 2050. The remod-D model, on the other
hand, is a nonlinear optimization program to assess national
energy systems through a single-node approach. its focus is
identifying a cost-optimized national energy system in line
with set Co2 reduction targets as well as assessing interactions
between different energy carriers and consumption sectors
(see Palzer 2016 in the “for further reading” section for more
details). While these three models achieve a temporal resolu-
tion of 1 h, the Times Paneu model relies on typical days for
its calculations. The strength of this optimization model is the
assessment of interactions among the power, mobility, heating
and industry sectors, while simultaneously considering a high
geographical resolution.
methodological differences as well as a different param-
eterization of the models can lead to divergent results. That
is why, in a currently ongoing project called regmex, differ-
ent energy-system models are compared and used to analyze
the transformation toward a nearly Co2-neutral energy sys-
tem with the aim of identifying robust results, i.e., indispens-
able technologies to achieve such a transformation. While a
comparison of the individual models will not be discussed
here, some of the main findings will be presented to address
the following two research questions:
1) how can a cost-optimized transformation of the ger-
man energy system be achieved—with consideration
july/august 2018
of all energy carriers and consumer sectors—while
meeting the declared climate targets and ensuring a
secure energy supply at all times?
2) how does the resulting cost-optimized system con-
figuration change, especially concerning its degree of
electrification, depending on the investigated Co2
reduction targets?
Overview: The German Energy System
Trends in primary and final energy consumption in germany
have changed over time. Primary energy is defined as energy
that has not been subjected to any conversion or transforma-
tion process, for example, to heat or electricity. no conversion
or distribution losses are considered in calculating primary
energy. final energy, in contrast, is the useful, secondary
energy available to the final user, and its calculation includes
conversion or distribution losses. Primary energy consump-
tion per person in germany has decreased over the years from
185 kWh in 1990 to 158 kWh in 2015. By comparison, aver-
age consumption in the united states in 2015 was about 80%
higher, at 285 kWh/person. however, while primary energy
consumption in germany decreased, total final energy con-
sumption has not changed significantly over time. even though
energy is increasingly used more efficiently, growth, both
economic and in consumption, prevent an overall decrease in
energy consumption.
figure 2 shows the development of final energy con-
sumption divided into four consumption sectors: 1) low-tem-
perature heat, i.e., energy to supply heat for space heating
and domestic hot water supply; 2) process heat, for example
to dry paper or melt steel; 3) mobility, i.e., road traffic, ship-
ping, aviation and rail traffic; and 4) electricity, where all
electricity applications are included.
figure 2 shows that the german energy system relied
heavily on fossil fuels such as natural gas, oil, and coal, while
the total amount of the final energy consumption provided by
other fuel types, such as firewood, sewage sludge, mine gas,
and waste heat, amounted to approximately 15% in 2015. The
distribution of final energy consumption among the four sec-
tors in 2015 is roughly evenly distributed (each sector contrib-
uting between 18 and 30%). however, the distribution of the
energy-related Co2 emissions by sector shows a completely
different picture [figure 2(b)]. The electricity sector alone,
which accounts for 21% of the total final energy consumption,
ieee power & energy magazine 25
 resource share of gross electric-
ity generation in germany is con-
1,400
Other Energy-Related CO2 Emission stantly growing, starting from 6%
in 2000 to approximately 31.6%
GHG Emission, Million Ton CO2-eq.

1,200
in 2015.
1,000 still, even if the power-supply
–20 %
sector fundamentally transforms,
800 this alone will not be sufficient to
–40 %
achieve the ambitious climate pro-
600 tection targets shown in figure 1.
–55 %
The present situation shows that,
400 –70 % while the power sector is progres-
–80 % sively changing, the goals for other
200
sectors will most likely be missed
–95 % (for example, doubling the annual
0
renovation rate of buildings from 1
1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050
to 2%). To reduce ghg emissions
by at least 80% and, wherever
figure 1. GHG emissions in Germany from 1990 to 2015, and target values to 2050 possible, 95% below 1990 levels,
(blue dots). The green bars represent energy-related CO2 emissions and the red bars each one of the illustrated sectors
other GHG emissions. Percentage reductions refer to the 1990 value. (Source: (figure 2) will have to undergo
Henning and Palzer, 2015.) a substantial transformation itself,
i.e., follow a strict decarbonizing
is responsible for approximately 44% of the total energy-related strategy. To properly assess the transition of the energy sys-
Co2 emissions as power generation in germany relies heavily tem toward a climate friendly, reliable, and affordable supply
on fossil fuel-fired plants. in 2015, 71% of the gross electricity of energy, it is necessary to consider all sectors as well as
generation was provided by coal (42%), nuclear power (14%), their interactions.
natural gas (10%), and petroleum products and other fuels (5%),
while only 29% came from renewable energy resources. in Methodology
2015, the specific emission Co2 factor for electricity amounted The remod-D model is a bottom-up energy system model
to 535 g/kWh as a result of the extensive use of lignite and developed at the fraunhofer institute for solar energy systems
hard coal, with fuel Co2 emission factors of 337 g/kWhcoal and assesses the transformation of national energy systems to
and 407 g/kWhcoal, respectively, and overall power plant a more environmentally friendly and low-carbon economy.
efficiency below 40%. on the other hand, the share of final The fundamental idea behind the model is the identification
energy consumption for the supply of low-temperature heat of a cost-optimized system structure in line with the set Co2
in 2015 amounted to 30%, yet yielding only 18% of the total reduction targets by taking into account the highly complex
energy-related Co2 emissions. in this sector, gas boilers, not dependencies occurring between different energy carriers and
coal, are the dominant technology. The emission factor for sectors (sector coupling model). The objective is to dimension
natural gas amounts to 201 g/kWhgas, leading to lower emis- all system-relevant technologies in a cost-effective way, start-
sions when compared to the electricity sector. ing from 2015 until the target year 2050.
in this context, the changes in sources of electricity produc- The model calculations are based on a comprehensive
tion have been a central topic for the last few years. for exam- database in which weather data as well as technological and
ple, in July 2011, the german federal government declared a economic parameters are specified for each year from 1990
progressive reduction of nuclear power plants until a complete to 2050. The first part of the database, from 1990 to 2014,
shutdown of all reactors by 2022. similarly, in october 2017, describes technical parameters for all relevant components
germany’s environment minister declared that a withdrawal of the energy system. This includes installed capacities of
from coal-fired power is not a question of whether it should be power plants, storage facilities, heating technologies in build-
done but, rather, when and how it will be accomplished. While ings, building stock (as well as its physical characteristics),
the phase out from nuclear power is final, the retirement of number of vehicles, and transportation technologies. The sec-
coal-fired resources has not yet been legally embodied and is ond part, from 2015 to 2050, deals with assumptions regard-
the subject of controversial debate. at the same time, a further ing cost or performance projections of individual technolo-
increase in installed capacity of renewable energy resources is gies, technical restrictions (capacity limits per year), and other
expected. The government aims for a share of renewable elec- time-dependent characteristics. This information is essen-
tricity generation of 40–45% in 2025 and 55–60% in 2035. tial to determine the mix of generators, energy converters, and
The current development shows that the renewable energy consumers that meet environmental goals at minimal cost.

26 ieee power & energy magazine july/august 2018

 3,000 Distribution of Energy-Related CO2
Emissions in 2015
Final Energy Consumption (Twh)

Total = 754 Million Metric Tons

2,500
27%
28% 28% 29% 29% 30%
30% 29% 31% Low-
2,000
Temperature
17% Heat
21% 21% 21% 21% 21% 21%
1,500 21% 21% 18%
17% Electricity Industrial
1,000 18% 18% 17% 44% Processes
16% 19% 18% 18% 18%
17%
500
39% 33% 33% 34% 31% 31% 33% 30% 30%
Mobility
0 21%
1996 2008 2009 2010 2011 2012 2013 2014 2015
(a) (b)

Low-Temperature Heat Fossil Fuels Low-Temperature Heat Others Industrial Process Heat Fossil Fuels
Industrial Process Heat Others Electricity Mobility Fossil Fuels Mobility Others

figure 2. The (a) development and share of the final energy consumption by sectors and (b) distribution of energy-related
CO2 emissions by sector in 2015. (Source data: BMWI Energy Data, 2017.)

here, the model distinguishes between two kinds of different
structures: open and closed systems.
When considering the technology of an open system, the
model has the option to increase or decrease the specific
penetration of the technology as long as it doesn’t violate
the set expansion potential. for example, onshore wind tur-
bines can be installed as long as the total installed capacity
is below the cumulative technical potential (approximately
180 gWel in germany that, starting from 2015, is roughly
5 gWel per year). likewise, the model could also decide
not to install any onshore wind turbines, such that the total
amount of onshore wind power decreases over the years
through retirements. While the installed capacity of each
technology can increase or decrease in open systems (i.e.,
renewable energy resources, power plants, and thermal and
chemical storage facilities), in closed systems, the technol-
ogies are not limited by technical constraints but by others.
for example, the cumulative deployment of technologies
for space heating and domestic hot water supply is limited
by the total number of buildings in germany. here, the
assumption is that every building is equipped with one heat
generator. Whenever this technology reaches the end of its
expected lifetime, the model can decide whether to install
a heat generator of the same type or choose a different one.
This means that the model can optimize the share of each
technology available within the closed system, but it can’t
change the exogenously fixed number of technologies in
each closed system.
Conventional lignite and hard coal power plants, nuclear
power plants, oil-fired power plants, gas turbines (gTs), com-
bined heat and power (ChP) plants as well as gas-fired and
steam power plants are implemented as generators. renewable
energy can be supplied in the model using wind turbines (onshore
and offshore), photovoltaic systems, and hydropower plants.
Biomass can be used either directly or after conversion into a
secondary energy carrier. for example, wood can be burned
in boilers to provide process heat for industrial applications
and for the generation of low-temperature heat in the building
sector. Biogas systems (gasification systems with subsequent
synthesis into hydrogen, methane, or liquid fuels) and biodiesel
systems are implemented as possible systems for the conversion
of biomass. electrical energy-storage systems in the form of
stationary and mobile batteries (in vehicles) or pumped-storage
power plants are implemented as storage systems. hydrogen
and thermal hot water storage systems in different orders of
magnitudes are considered as well.
energy demand is divided into four groups, according to
the different fields of use: mobility, intrinsic electricity applica-
tions, heat for buildings (residential, nonresidential, and indus-
trial buildings), and process heat in the industry. The mobil-
ity sector is mapped in detail concerning passenger cars and
trucks, with seven vehicle classes each. The energy demand of
aviation, shipping, and fuel-based railway traffic is considered
in the balance, without temporal resolution. The basic elec-
tricity load is mapped using load profiles based on the data of
european transmission grid operators reduced by the weather-
related electric load for heating systems. The load for heating
systems is calculated model endogenously and is not included
in the basic load. Table 1 summarizes the main components of
the energy system considered in remod-D.
Comparative Findings and Discussion
in the following section, the results of three cost-optimized
scenarios are analyzed in detail. The scenarios differ from

july/august 2018 ieee power & energy magazine 27

 table 1. An overview of the most important system technologies considered in the model.

Plants for Conversion Mobility Concepts

Power Generation of Biomass Into Flexibility Options Heat Generator (Cars, Trucks)
• Photovoltaics • Hydrogen • Stationary batteries • Electric heat pumps • Battery electric vehicles
• Wind onshore • Methane • Batteries in vehicles (brine, air) • Fuel cell electric
• Wind offshore • Liquid fuels • Pumped storage plants • Gas heat pumps vehicles
• Run of river • Heat (boilers) • Power to H2 • Hybrid heat pumps • Internal combustion
(electrolysis) (electric + gas) engines (gas)
• Oil plant • Electricity
• Power to CH4 • Boilers (gas, oil, or • Internal combustion
• Lignite plant pellets) engines (liquid fuels)
• Hard coal plant • Power to liquid
(electrolysis + Fischer– • Small CHP units • Hybrid (battery +
• GT Tropsch synthesis) • Fuel cells fuel cell)
• Combined • Thermal storage (heat (hydrogen • Hybrid (battery + gas)
cycle GT pump, heating rod) or methane) • Hybrid (battery +
• Nuclear power • CHP units • Heating grid liquid fuels)
plant • Geothermal heat

each other in terms of the set Co2 reduction goals by 2050,
which are 80, 90, and 95% reductions in relation to 1990. it is
important to note that the energy system here is considered as
an isolated island system to improve comparability with other
involved energy system models. further assumptions will be
documented in the final report of the regmex project.
Power Generation
figure 3 shows the cost-optimized development of the installed
capacity for renewable energy sources, for each of the three
scenarios from 2020 until the year 2050. The results show that
a significant increase of renewable energy resources is needed to
achieve the scenario reductions of energy-related Co2 emissions.
in particular, photovoltaic systems and onshore wind turbines play
a key role when it comes to power generation. Their cumulative
technical potential for germany is assumed to be 320 gWel and
180 gWel, respectively, while it amounts to 40 gWel for offshore
wind turbines. This yields a total technical potential for the princi-
pal variable renewable energies of approximately 540 gWel.
in the first scenario (80% Co2 reduction), this potential is
exploited 61%; in the second (90% Co2 reduction), it is exploited
84%; and in the third scenario (95% Co2 reduction), it is exploited
completely. in concrete terms, starting from today’s installed capac-
ity of 91 gWel (nameplate wind and solar energy), this translates
into a cumulative capacity in 2050 of 323, 440, and 540 gWel
respectively. The simulation results suggest the following:
1) The rise of the cumulative installed capacity of re-
newable wind and photovoltaic resources is inversely
proportional to the total allowable Co2 emissions bud-
get, i.e., the lower the allowable Co2 emissions in the
energy system, the higher the necessary amount of en-
ergy production from renewable resources.
2) The relationship between the Co2 reduction target and
the required installed capacity of renewable resources
is not linear. moving from a 90% reduction to a 95%
reduction requires the same incremental increase in in-
stalled capacity (100 gWel) as moving from an 80% re-
duction to a 90% reduction.

600 600 600

Installed Capacity in Gwel

PV
500 500 500
Onshore Wind
400 Offshore Wind 400 400

300 300 300

200 200 200

100 100 100

0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)

figure 3. The annual development of the installed capacity of variable renewable energy resources. (a) 80% reduction,
(b) 90% reduction, and (c) 95% reduction.

28 ieee power & energy magazine july/august 2018

 140 140 140
Installed Capacity in Gwel

120 Coal Lignite CH4-GT 120 120

CH4-CCGT H2-GT
100 100 100
80 80 80
60 60 60
40 40 40
20 20 20
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)

figure 4. Temporal development of the installed capacity of thermal power plants. Combined cycle GTs (CCGT), hydro-
gen (H2), gas/methane (CH4). (a) 80% reduction, (b) 90% reduction, and (c) 95% reduction.

figure 4 shows the cost-optimized development of the required installed capacity of thermal power plants shown
installed capacity of thermal power plants for the three con- in figure 4.
sidered scenarios. for all three scenarios, a shift from higher figure 5 shows the cost-optimized development of differ-
ghg-emitting power plants (lignite and hard coal) to ent technologies for the conversion of electricity into syn-
lower emitting plants [Ch4-combined cycle gTs (CCgTs), thetic fuels, i.e., hydrogen, synthetic methane gas, and liquid
Ch4-gT, and h2-gT] becomes clear. major differences are fuels. These technologies are used when renewable energy sur-
evident, however, in the different levels of installed capacity. pluses occur, which is why the installed capacity of synthetic
The total cumulative installed capacity in the 95% reduction fuel technologies increases proportionally with the resulting
scenario in 2050 roughly doubles its initial value, while for expansion of renewable energies shown in figure 3.
the less-ambitious reduction targets (80 and 90%), the initial While the absolute value of the residual load for electric-
total installed capacity value doesn’t change as significantly ity surplus in 2050 amounts to 186 gWel, in the case of an
(+16%, +45%). 80% reduction target, it increases to 256 gWel when consid-
To better understand the illustrated development of the ering a reduction of 90% and to 328 gWel in the case of 95%
thermal power plant fleet, a closer look at the residual load is (Table 2). This nonlinear growth in residual load explains
necessary. The residual load describes the electric load of the why the cumulative capacity for power-conversion technolo-
energy system minus the nondispatchable power generation, gies increases nonlinearly with more restrictive Co2 reduc-
i.e., mainly variable renewable resources. This means that tion targets (figure 5).
whenever the residual load is positive, deficiency occurs and The illustrated power conversion technologies can be op-
power would need to be supplied from other sources. on the erated as flexible loads to reduce eventual electricity surplus
other hand, a negative value represents a surplus of renew- peaks, and they produce climate-friendly synthetic fuels.
able electricity. in either case, a range of different flexibility
options are required to balance the residual load.
in particular, the increasing use of variable renewable re- table 2. Maximum renewable power deficiencies
sources shown in figure 3 leads to higher peak values of the and surpluses in target years 2030 and 2050 for
three considered CO2 reduction targets.
residual load during the year and therefore to a higher need for The increase of the peak values of the residual
flexibility options. Technologies like power storage are effec- load corresponds with the increasing installed
tively used to balance small fluctuations of the residual load. capacity of variable energy resources.
When those fluctuations become more substantial, i.e., the power Power Power
requirement is higher, technologies such as gTs and CCgT Deficiency) Surplus
plants (for renewable power deficiency) or power conversion (in GWel) (in GWel)
technologies such as power-to-gas or power-to-heat (for renew- Reduction of
able power surplus) are mainly operated. Table 2 summarizes Energy-Related Emissions 2030 2050 2030 2050
the maximum and minimum values of the residual load in 2030 Scenario 1: −80% CO2 60 68 75 186
and 2050. (referred to 1990)
Table 2 shows that the maximum renewable power defi-
Scenario 2: −90% CO2 65 69 82 256
ciency (positive residual load) in 2050 for a 95% Co2 reduc- (referred to 1990)
tion target is almost 50% higher than the value of scenarios 1
Scenario 3: −95% CO2 62 101 134 328
and 2 (80% and 90%). This increase of renewable power (referred to 1990)
deficiency corresponds with the illustrated development of

july/august 2018 ieee power & energy magazine 29

 100 100 100
Power to Fuel
Installed Capacity in Gwel

80 Methanation 80 80
Electrolysis
60 60 60

40 40 40

20 20 20

0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)

figure 5. The annual development of the installed capacity of technologies for the production of synthetic fuels: (a) 80%
reduction, (b) 90% reduction, and (c) 95% reduction.

These fuels become essential for the decarbonization of those
energy consumers in the energy system, where a complete
technological reorientation may be hard to achieve under
today’s economic and political structures (for example, avia-
tion or some industrial processes).
Consumption Sectors
figure 6 illustrates the scenario development of heating
technologies for space heating and domestic hot water sup-
ply. While the total number of heating technologies is
directly linked to the exogenously set number of buildings
in germany (boundary condition), the share of each tech-
nology is a result of the cost optimization. note that every
heating technology can be supplemented by hot water stor-
age with solar heating and an electric heating rod.
While gas and oil boilers clearly dominate today’s heat-
ing technologies (approximately 70% of all installations),
their share decreases in all three scenarios. The resulting
level of reduction is directly linked to the considered Co2-
reduction target. Concretely, in 2050 the share of oil and gas
boilers amounts to 21% for an 80% Co2 reduction target,
9% for a 90% Co2 reduction target, and roughly 4% for 95%
Co2 reduction target.
electrical heat pumps exhibit the opposite trend. Their
number increases when considering more ambitious climate
protection goals. The results show that, in the third scenario
(−95% energy-related Co2 emissions), electrical heat pumps
(air or brine) account for almost half of the heating technolo-
gies in the energy system. These technologies bring two main
advantages for the overall energy system. first, their coeffi-
cient of performance allows an advantageous ratio between
the required energy for operation and the provided heat
(for detailed model analysis on heat pumps, see sterchele
et al. 2017 in the “for further reading” section). This benefit
in efficiency contributes to the decarbonization of the low-
temperature heat supply. second, electrical heat pumps can
be operated flexibly to balance the residual load in times of
renewable production surpluses (negative residual load), for
example, to charge thermal storage technologies. as shown
for other power conversion technologies, here the need for

30 30 30
Technologies in Millions

25 25 25
20 20 20
Heating

15 15 15
10 10 10
5 5 5
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)

CHP + Fuel Cell Gas Heat Pump Oil and Gas Boiler
Electric Heat Pump (Air, Brine) Biomass Boiler Heating Grid and Geothermal Heat

figure 6. The annual development of heating technologies for space heating and domestic hot water supply. Heating
grids are considered as a combination of CCGT, electrical heat pumps, gas boilers, thermal storage, and solar thermal
technology. (a) 80% reduction, (b) 90% reduction, and (c) 95% reduction.

30 ieee power & energy magazine july/august 2018

 60 60
50 50
Vehicles in Millions
40 40
30 30
20 20
10 10
0 0
2020 2030 2040 2050 2020
(a)
Battery Hybrid
figure 7. The annual development of vehicles (cars, trucks, and buses) for three considered scenarios: (a) 80, (b) 90, and
(c) 95% CO2 reduction targets. Boundary condition assumption: Trucks can’t be fully electrified.
flexibility options increases due to the increasing installed
capacity of variable energy resources. The number of con-
nections to a heating grid increases as well. heating grids
in the model are represented as a combination of different
technologies, which are CCgT, electrical heat pumps, peak
load gas boilers, and solar-assisted heat and hot water tanks
(including a heating rod). additionally, CCgT and micro-
ChP plants can be operated flexibly by generating power in
times of renewable power deficiency, while also charging hot
water tanks.
figure 7 shows the development of the road transportation
sector. similar to the heating technologies, the total num-
ber of vehicles is set exogenously, while the share of each
driving concept results from the cost optimization of the
model. The study assumed an upper limit for battery elec-
tric trucks of 5%. assumptions for driving technology and
range restrictions meant that only local electric transport for
trucks was considered in the study. higher penetrations of
electric transport could be considered under advanced tech-
nology scenarios. note that the number of vehicles in fig-
ure 7 includes both cars and trucks, and buses are included
in the category of trucks.
The development of the road transportation sector (fig-
ure 7) shows similarities to the development of the heating
technologies displayed in figure 6. in all three Co2 reduc-
tion scenarios, the share of today’s dominating transporta-
tion technology [internal combustion engines (iCes) based
on liquid fuels] decreases over the course of the years. a
faster phase out of this drive technology would be required
for more ambitious climate protection targets. simultane-
ously, like heating technologies, a shift to electric transport
technologies, such as battery electric vehicles, fuel cell elec-
tric vehicles, or hybrid vehicles (a combination of iCes with
battery), emerges. electric vehicles have a higher efficiency
than iCes, and battery-electric vehicles can be used to charge
and supply the electric distribution network based on the
residual load. Therefore, they represent an alternative flexibil-
ity option for the energy system.
july/august 2018
60
50
40
30
20
10
0
2030 2040 2050 2020 2030 2040 2050
(b) (c)
H2-FC ICE Gas ICE Fuel
methane-powered vehicles, iCes (gas), play a role in
each scenario. at present, this driving technology is less
expensive than electric vehicles and emits lower Co2 emis-
sions compared to vehicles based on liquid fuels. however,
the results show that gas-powered vehicles will persist until
2050 in the first scenario, while being a transitional tech-
nology in the other scenarios with more ambitious climate
protection targets.
figure 8 shows the technological development for
the provision of process heat in the industrial sector. The
parameterizing and calculation in the model distinguishes
different temperature levels, allowing, for example, the use
of electrical heat pumps only for temperature levels below
180 °C.
The development of technologies for the provision of pro-
cess heat shows a slower-acting switching behavior when
compared to other consumption sectors (space heating +
domestic hot water supply and road transportation). even
though gas and oil-based technologies decrease when consid-
ering more ambitious Co2 reduction targets, they are clearly
present in each scenario in the target year 2050. The share of
electric technologies, such as electric heat pumps and elec-
trode boilers, increases with increasing Co2 reduction tar-
gets, while remaining fairly constant in the case of an 80%
Co2 reduction target. at the same time, coal-fired boilers
phase out in the last years of the energy transition path (for a
95% Co2 reduction).
industrial ChP stations become less important over time,
beginning in the early simulated years. This trend partly arises
from the model-assumption that all heating technologies in the
industry sector are not operated responsively to grid condi-
tions, i.e., operation in the industry sector will, for business
reasons, remain independent from the residual load. This leads
to multiple hours during the year when those plants produce
electricity, although a surplus of electricity already occurs
(negative residual load). This leads to unfavorable operating
conditions for ChP stations, explaining why they are driven
out by other technologies.
ieee power & energy magazine 31
 Process Heat—Capacity in GWth
60 60
50 50
40 40
30 30
20 20
10 10
0 0
2020 2030 2040 2050 2020 2030
(a)
Coal Boiler H2 Boiler
Gas + Oil Boiler Heat Pump
figure 8. The annual development of technologies for the provision of process heat. (a) 80% reduction, (b) 90% reduc-
tion, and (c) 95% reduction.
Main Findings
To achieve the ambitious climate protection targets set by
the federal government, the german energy system must be
substantially transformed. To determine what such a trans-
formation could look like from a technical point of view, we
developed the remod-D model.
our analysis shows that the system composition in the
target year 2050 heavily depends on the allowed amount of
energy-related Co2 emissions. Concretely, the more ambi-
tious the climate protection targets are, the higher the required
installed capacity of renewable energy resources (mainly
solar plants and onshore wind turbines). starting from today’s
installed capacity in germany of approximately 100 gWel,
an increase of 230–440 gWel is needed to achieve a reduc-
tion of energy-related Co2 emissions between 80 and 95%,
respectively, by 2050 (figure 3). for a 95% reduction target,
this means a cumulative installed capacity of approximately
540 gWel, translating into a required increase of 8–14 gWel of
renewable energy resources per year. from today’s perspec-
tive, a sluggish expansion of the electric grid, public protests
against onshore wind turbines, and technological obstacles
may make the carbon-reduction target seem too ambitious.
however, it is important to note that the presented simulation
results do not take into account electricity imports or exports
from neighboring countries (in this study, the model used an
isolated island system). overall, this assumption of isolation
leads to higher required installed capacities of all considered
energy conversion and storage technologies.
a power supply dominated by renewable energy resources
leads to more intense fluctuations of the residual load, i.e.,
higher peak values of electricity surplus or deficiency. There-
fore, to harmonize power generation and consumption dur-
ing each hour, a greater variety of flexibility options will be
required. The main technologies in case of power deficiency
are power storages (stationary batteries, batteries in electric
vehicles, and pumped hydro storage) and thermal plants (gTs
32 ieee power & energy magazine
60
50
40
30
20
10
0
2040 2050 2020 2030 2040 2050
(b) (c)
CHP CH4 El. Boiler
Solar-Assisted Heat Biomass Boiler
and CCgT plants). alternatively, when electricity surplus
occurs, flexible loads are needed, such as power storage and
power technologies that convert electricity into synthetic fuels
or heat.
apart from the energy supply, to achieve the set Co2 reduc-
tion goals, the consumption sectors of the energy system need
to be transformed. our results show that the more ambitious
the climate protection targets are, the faster the required transi-
tion from a fossil fuel-based demand to a demand dominated
by electric technologies, i.e., the consumption sectors get pro-
gressively electrified. This behavior is shown in figure 9 for
three Co2 reduction targets: 80, 90, and 95% compared to
1990 levels.
The results suggest that, under given boundary conditions
to achieve a cost-optimal solution, the electrification of the
road transportation sector, i.e., cars and trucks as well as the
supply of low-temperature heat (space heating and domestic
hot water supply) should be prioritized over the industrial
sector. figure 9 shows that the transformation of these two
sectors should be addressed first, and their cost-optimal
level of electrification in 2050 is far above the level for the
industrial process heat supply in all three scenarios. While
the heating technologies in the industry sector are not oper-
ated responsively to grid conditions, battery-electric vehicles
as well as electric heating technologies for space heating
and domestic hot water supply can—to a certain extent—be
operated flexibly and therefore contribute in balancing the
residual load in times of power surplus or deficiency.
Ongoing Work
While the remod-D model considers many factors, the next
steps in its development concern extending its geographi-
cal scope (from the simulation of national to european energy
systems) as well as examining detailed flexibility options
within the energy system. This includes 1) a detailed descrip-
tion of vehicle-to-grid communication under consideration of
july/august 2018
 100 100 100
Degree of Electrification in %

Low-Temperature Heat
80 Road Traffic 80 80
Industrial Process Heat
60 60 60

40 40 40

20 20 20

0 0 0
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
(a) (b) (c)

figure 9. The degree of electrification of the consumption sectors: low-temperature heat (space heating and domestic
hot water supply), road transportation (cars, trucks, and buses), and industrial process heat for a reduction of energy-
related CO2 emissions of (a) 80, (b) 90, and (c) 95% below 1990 levels. The degree of electrification for the respective
sectors includes low-temperature heat: electric heat pumps; deep geothermic and heat rods road transportation: battery
electric vehicles, fuel cell electric vehicles; and partially hybrid vehicles industrial process heat: electric heat pumps
and electrode boilers.

driving profiles, different shares of vehicle batteries for flex- .forschungsjahrbuch-energie.de/downloads/forschungsjahrbuch-

ibility applications, and average discharge/loading capacities;
2) the assessment of the potential of smart communication
technologies with regard to heat generators, i.e., smart meter
rollout for heat pumps, heating rods, and ChP plants; and 3)
the detailed depiction of ramping behavior of thermal power
plants and power-to-gas technologies.
Acknowledgments
This work arose in the course of a Ph.D. thesis that is par-
tially funded by the reiner lemoine-stiftung foundation.
in addition, it has been partially supported by the german
federal ministry of economics and Technology (BmWi) in
the context of the project regmex (fkz: 0325874B): “model
experiments and Comparisons of Pathways leading to a
renewable energy-Based energy supply system” as a part
of the 6th energy research programme.
For Further Reading
h. m. henning and a. Palzer. (2015, nov.). What will the en-
ergy transformation cost?. fraunhofer institute for solar
energy systems, germany. [online]. available: https://www
.ise.fraunhofer.de/content/dam/ise/en/documents/publications/
studies/What-will-the-energy-transformation-cost.pdf
a. Palzer. (2016). Cross-sectoral modeling and optimiza-
tion of a future german energy system taking into account
energy efficiency measures in the building sector. Thesis,
fraunhofer institute for solar energy systems, germany,
2016. [online]. available: http://publica.fraunhofer.de/eprints/
urn_nbn_de_0011-n-408742-11.pdf
s. lechtenböhmer, T. Pregger, a. Palzer, h. C. gils, T.
Janßen, C. Krüger, D. schüwer, J. luhmann, m. Buddeke,
P. sterchele, C. Kost, and l. Brucker (2015). research pro-
gram description, regmex. [online]. available: http://www
2015.pdf
h. C. gils, “economic potential for future demand response
in germany: modelling approach and case study,” Applied En-
ergy, vol. 162, pp. 401–415, Jan. 2016.
g. norman. (2015). interaktion ee-strom, Wärme und
Verkehr5. Tech. rep. [online]. available: https://www.iee
.fraunhofer.de/content/dam/iwes-neu/energiesystemtechnik/
de/Dokumente/Veroeffentlichungen/2015/interaktion_
eestrom_Waerme_Verkehr_endbericht.pdf
J. Welsch, u. fahl, m. Blesl, and K. hufendiek. (2014).
modelling of storage processes and power-to-X-Technolo-
gies in Times-Paneu. Proc. Symposium Energieinnova-
tion. [online]. available: https://www.tugraz.at/fileadmin/
user_upload/events/eninnov2016/files/lf/session_a4/lf_
Welsch.pdf.
P. sterchele, a. Palzer, and h.-m. henning. (2017). The
role of heat pumps in the transformation of national energy
systems: example germany. Proc. 12th Int. Energy Agency
Heat Pump Conf. [online]. available: http://hpc2017.org/
wp-content/uploads/2017/05/Po.01-The-role-of-heat-pumps-
in-the-transformation-of-national-energy-systems-example-
germany.pdf
Biographies
Philip Sterchele is with the fraunhofer institute for solar
energy systems, Breisgau, germany.
Andreas Palzer is with the fraunhofer institute for solar
energy systems, Breisgau, germany.
Hans-Martin Henning is with the fraunhofer institute
for solar energy systems, Breisgau, germany.
p&e

july/august 2018 ieee power & energy magazine 33

 An By Trieu Mai,

Electrified
Daniel Steinberg,
Jeffrey Logan,
David Bielen,
Kelly Eurek, and

Future Colin McMillan

A
as an energy source, electricity
benefits from a number of desirable characteris-
tics: it can be transported at nearly the speed of light
with transmission infrastructure, it has zero end-
use emissions, it is highly flexible and controllable,
it is now storable at rapidly declining costs, and it can
offer improved service quality relative to conven-
tional fuels. as such, electrification—the conver-
sion of previously fossil-fueled end-use processes
to electricity—has been identified as a key pathway
to a clean, reliable, and secure energy future. elec-
tric vehicles are the most widely cited application
of electrification, but technology improvements in
electrically driven devices for buildings and indus-
trial end uses, including heat pumps for space and
water heating needs, induction stoves for cooking,
infrared or ultraviolet curing processes, and electric
arc furnaces for process heating, could lead to more
widespread electrification across these sectors.
in this article, we report results from a recently
published initial analysis conducted by the national
renewable energy laboratory (nrel) that simu-
lated widespread electrification from present day
through 2050 in the united states. the study focused
on 1) levels of technically achievable end-use electri-
fication, 2) power-sector capacity expansion needs
required to meet the growing demand for electricity

Digital Object Identifier 10.1109/MPE.2018.2820445

Date of publication: 18 June 2018

34 ieee power & energy magazine 1540-7977/18©2018IEEE july/august 2018

 Initial Scenarios and Future Research
for U.S. Energy and Electricity Systems
under an electrified future, 3) the impacts of electrification
on electricity demand load shapes, and 4) the implications for
economy-wide carbon dioxide (co2) emissions. compared to
a reference case, very high levels of electrification can lead
to over twice the total electricity demand by 2050, with most
of that in the transportation sector. such high levels of elec-
trification, when combined with additional improvements
in overall efficiency and grid decarbonization, can yield
economy-wide carbon emission reductions of 72–75% by
©istockphoto.com/metamorworks
july/august 2018
2050 compared to 2005 levels. ongoing work to analyze the
economic potential of electrification, the impact on electric-
ity demand load profiles, and the ability to operate the grid
reliably, given high levels of variable generation and smart
demand responsiveness, will add important new insights on
the potential future of electrification in the united states
and beyond.
Background and Motivation
the u.s. economy relies on a diverse range of fuel sources,
intermediate energy carriers, and end-use and supply tech-
nologies. in 2015, the u.s. energy system consumed 98 qua-
drillion British thermal units (quads) of primary energy, of
which 39% (38 quads) was used for electricity production
and the remainder consumed primarily through the direct
combustion of fossil fuels in the other sectors, such as trans-
portation, industry, and (commercial and residential) build-
ings, as shown in Figure 1. this electricity generation and
direct fossil consumption resulted in 5,259 million metric
tons (MMt) of co2 emissions from the 2015 energy sys-
tem, comprising a large majority of all u.s. greenhouse gas
(gHg) emissions. energy-related emissions represents about
80% of total gHg emissions on a co2-equivalent basis and
about 90% of net gHg emissions when both sources and
sinks are considered. in Figure 1, emissions from electricity
generation are allocated to the sectors and end uses using the
national and annual average emission factor.
the electrification of end-use services coupled with the
decarbonization of electricity generation has long been
identified as a means to achieve, among other things, a
low-carbon future. this low-carbon pathway would reduce
both energy and carbon intensities for the economy along
two primary dimensions. First, end-use electric technolo-
gies are often more energy efficient, requiring less energy
to provide the same service, than conventional fossil-based
options. electric vehicles, for example, are typically three to
four times more efficient than internal combustion engines
in converting on-board energy into power at the wheels,
although assessing differences in overall well-to-wheel
efficiencies requires accounting for factors such as the
electricity generation mix. similarly, an electric heat pump
can warm interiors with three to four times the efficiency
of a natural gas boiler. ultraviolet equipment for industrial
curing processes can be over 80% more efficient than pro-
cesses using combustion ovens. overcoming energy losses
at the power plant, where over half of potential energy is
lost as “waste heat” in conventional steam-turbine designs,
ieee power & energy magazine 35
 Transportation Industrial Residential Commercial
28 Quads Total 31 Quads Total 21 Quads Total 18 Quads Total
0 Quads Electricity 10 Quads Electricity 14 Quads Electricity 14 Quads Electricity
100

Nonelectricity
Other Other
Other
Water Heating
Other Water Heating
Commercial Light Trucks
Air Space Heating
Total 2015 Primary Energy Consumption (%)

Space Heating
75 Paper

Electricity
Mining
Freight Trucks

Refining Other
Other
50

Bulk Chemicals

Refrigeration
Light-Duty Vehicles
25 Lighting Lighting
Other
Space Cooling Space Cooling

Paper
Mining Water Heating Refrigeration
Refining
Bulk Chemicals Space Heating Space Heating
0
(a)

Transportation Industrial Residential Commercial

1,864 MMT Total 1,438 MMT Total 1,041 MMT Total 917 MMT Total
5 MMT Electricity 494 MMT Electricity 721 MMT Electricity 699 MMT Electricity
100 Other

Nonelectricity
Other Water Heating Other
Water Heating
Air Other
Total 2015 Energy-Related CO2 Emissions (%)

Space Heating
Commercial Light Trucks Space Heating
75

Freight Trucks
Paper Electricity
Mining

Refining
Other
Other
50

Bulk Chemicals

Light-Duty Vehicles Refrigeration

25 Lighting Lighting
Other

Space Cooling Space Cooling

Paper
Mining Water Heating Refrigeration
Refining
0 Bulk Chemicals Space Heating Space Heating
(b)

figure 1. The 2015 U.S. (a) primary energy consumption and (b) 2015 CO 2 emissions by sector and end use. [Data from
“Annual Energy Outlook 2017,” U.S. Energy Information Administration, and figure design courtesy of Jadun et al. (2017).]

36 ieee power & energy magazine july/august 2018

is key to delivering the full benefits of electrification. sec-
ond, low-carbon options exist for electricity generation, with
multiple studies suggesting the feasibility of replacing con-
ventional electricity sources with low or zero-carbon alter-
natives such as renewable energy, nuclear, and fossil fuel
combined with carbon capture and sequestration. although
the cost of such a transition and composition of the least-
cost generation portfolio remains very much in debate,
this suite of available low-emissions technology options
provides alternatives to today’s electricity mix. other low-
emission alternatives to conventional technologies that are
not directly electricity based, such as hydrogen, biofuels,
and natural gas (such as natural gas-fueled vehicles), also
exist and are being developed.
Beyond potentially reducing energy-related emissions,
electrification can also offer other important social and eco-
nomic benefits such as reduced air pollution-related health
and environmental impacts, lower water demand, increased
productivity and resilience, increased flexibility and control-
lability of services like induction cooking or ductless air-
sourced heat pumps, reduced reliance on imported oil and natu-
ral gas, and a greater opportunity to export these increasingly
abundant fuels. of course, there are also potential negative
impacts such as higher capital costs, particularly associ-
ated with new infrastructure investments, greater expo-
sure to cybersecurity risks, and reduced economic activity
in the fuel production sectors. electrification can substan-
tially shape energy-system infrastructure development and
consumption, which would affect a variety of stakeholders
including electric utility planners, equipment manufacturers,
fuel producers, building developers, and regulatory decision
makers. Due to these potential widespread impacts as well as
the emissions motivations described earlier, further research
on the possibilities for and the costs, benefits, and impacts of
widespread electrification is warranted.
in this article we present a preliminary analysis that exam-
ines widespread electrification in the u.s. energy sector. our
analysis heavily leverages previous work published by research-
ers at the nrel in 2017 (presented in the “For Further read-
ing” section), which includes additional background material,
descriptions, and references. in this research, we apply an
exploratory scenario analysis approach under which electrifi-
cation is rapidly advanced across all end-use sectors and esti-
mate the impacts of achieving these electrification levels on
load patterns, the evolution of the u.s. power system, and co2
emissions. We also highlight our methodological approach, its
shortcomings, ongoing work to address these shortcomings,
and anticipated remaining research gaps.
Electrification Definition,
Opportunities, and Technologies
We define electrification narrowly as the adoption of elec-
tric-based technologies to replace technologies that are cur-
rently fueled by nonelectric sources, typically fossil fuels, but
also including other energy sources or carriers. this defini-
july/august 2018
tion excludes projected second-order increases in electricity
demands brought about by new or increasing service demands,
such as new plug loads, the expansion of data centers, or
increased indoor agriculture. instead, our definition focuses
on existing services, such as driving, heating, and materials
processing, which are currently being served by other energy
sources or carriers. our prospective analysis also examines the
electrification of projected increases in these service demands
through population growth and other factors.
using this definition and historical data for energy use
and emissions, we identify opportunities where electrifica-
tion can potentially have the biggest impact. Figure 1 shows
that 2015 energy consumption and emissions are not spread
uniformly across sectors and subsectors, and indicates the
extent to which reliance on electricity varies significantly
across sectors. in particular, the buildings sector is already
heavily reliant on electricity (78 and 70% for commercial
and residential buildings, respectively, on a primary energy
basis) while electricity serves less than 1% of transporta-
tion-related energy consumption. industry is intermediate
between these extremes (32% based on primary energy).
emissions follow qualitatively similar trends. in the absence
of cost considerations, these observations suggest that the
electrification potential, and corresponding emissions-reduction
potential, is greatest in transportation relative to the other
economic sectors. existing research on electrification has
applied a greater focus on vehicle electrification than any
other end use. in recent years, interest in transportation elec-
trification has also been motivated by declining battery costs
and increasing plug-in hybrid and fully electric vehicle sales
and development.
Within each sector, electrification opportunities also vary
by subsector and end use. For example, on-road vehicles
are the source of about 80% of total transportation-related
energy use and emissions. in fact, light-duty vehicles (includ-
ing cars and trucks) by themselves are responsible for the
majority of all transportation-related energy consumption
and emissions. For light-duty passenger vehicles, plug-in
hybrid and fully electric options are currently available with
increased sales, more model types, and heightened manu-
facturer interest in recent years. electric options for larger
commercial and freight uses are significantly more limited;
initial development in these sectors has also occurred, par-
ticularly in niche applications such as city buses and high-
traffic corridors.
although electricity is used to the greatest extent in the
buildings sector overall, several end uses in buildings, such
as space and water heating, currently rely much more heav-
ily on the direct combustion of natural gas or other fossil
fuels. recent improvements in electric air-source heat
pumps allow them to deliver heat efficiently even at rela-
tively low outdoor temperatures (approaching –18 °c) and
may offer opportunities for more widespread use. split sys-
tems that have one or more condensers outside serving mul-
tiple interior air handlers can overcome building renovation
ieee power & energy magazine 37
 4.5
4.0 Electricity
Final Energy Use (Quads)

3.5 Fuel Oil

3.0 Natural Gas
Propane
2.5
2.0
1.5
1.0
0.5
0.0
Space Heating Water Heating Air-Conditioning Refrigeration Other
End Use
(a)
2.0
Electricity
Final Energy Use (Quads)

Fuel Oil
1.5
Natural Gas
District Heat
1.0

0.5

0.0
Space Lighting Refrigeration Ventilation Space Cooking Water Computing Office Other
Heating Cooling Heating Equipment
End Use
(b)
4
Purchased Steam Coal (Excluding Coke and Breeze)
By-Products Natural Gas and Gas Liquids
3
Fuel Oil and Diesel Electricity
(Quads)

0
CHP and/or
Cogeneration Process

Conventional Boiler
Use

Process Heating

Machine Drive

Facility HVAC

Process Cooling and

Refrigeration

Other Process Use

Electrochemical
Processes

Facility Lighting

Other Facility Support

On-Site Transportation

Conventional Electricity
Generation

Other Nonprocess Use

End Use Not Reported

End Use
(c)

figure 2. Final energy consumption in (a) residential buildings, (b) commercial buildings, and (c) industry sectors by
end use and fuel. The data for industry exclude energy consumption as a nonfuel (feedstock). (Data from the EIA 2009
Residential Energy Consumption Survey, EIA 2012 Commercial Buildings Energy Consumption Survey, and EIA 2010
Manufacturing Energy Consumption Survey.)

38 ieee power & energy magazine july/august 2018

barriers associated with ducted systems. Heat pumps can
provide heating and cooling services from a single capital
investment and can be connected to water heating, clothes
drying, and other building energy needs. cooking, which
uses about as much energy as space cooling or water heating
in the commercial sector, also relies primarily on natural gas
burners. Highly efficient electric induction cooking is now
available and offers speed, control, and safety features not
available in gas burners. electric alternatives for heating
and cooking are used in many u.s. regions; however, the
adoption of efficient modern electric technologies is still
relatively nascent compared to many regions of the world.
the electrification potential for the industrial sector is
much more difficult to assess due to the diverse number of
end uses and industrial processes as well as the limited data
and modeling tools available. setting aside the technologi-
cal and economic challenges with electrification in indus-
try, direct industrial combustion for process heating offers
the greatest potential for emissions reductions in this sector.
examples of industrial electric technologies include electric
arc furnaces, infrared dryers, and heat pumps for process
heating and building heating, ventilation, and air-condition-
ing services. Figure 2 summarizes recent final energy con-
sumption by fuel and end use in the residential, commercial,
and industrial sectors.
Determining the Technical
Potential of Electrification
using the identified electrification opportunities from the
historical energy use and emissions data, we develop a set of
exploratory scenarios under which electrification is rapidly
advanced. We start with a reference scenario, which follows
a business-as-usual trajectory of electrification and service
demand growth and is based on current and enacted policy
and law. it does not include, for example, the clean Power
Plan. We include a high-electrification scenario, which
could be interpreted as reflecting technical potentials for
electrification although, as we’ll describe below, even in this
scenario electrification is not quite ubiquitous as we exclude
some subsectors and consider how stock turnover can slow
the pace of transport electrification. it is important to note
that the development of this near-complete electrification
scenario does not consider costs, consumer preferences, or
other potential barriers to electrification. this scenario also
does not capture potential major structural changes in the
economy or disruptive changes in behavior or technology
(e.g., autonomous vehicles). in other words, our scenarios do
not reflect economic or market potentials for electrification,
nor do they reflect predictions.
For transportation, we use a modified version of argonne
national laboratory’s Vision model to develop electrification
scenarios of nonmilitary, on-road transport. Modifications
include adjusting certain parameters and the addition of
electric vehicle options for medium- and heavy-duty trucks;
see steinberg et al. 2017 in the “For Further reading” section
july/august 2018
for more details. Vision is a stock-rollover accounting
tool for the u.s. vehicle fleet. using this tool, we generate
electrification scenarios for light-duty vehicles and heavy-
duty vehicles. in our high-electrification scenario, sales of
conventional light-duty vehicles, using internal combus-
tion engine drivetrains, are phased out completely by 2040.
instead, new passenger car and truck sales are a combination
of battery-electric, fuel cell, and, to a lesser extent, plug-in
hybrid electric vehicles. For light-duty trucks, we assume
greater sales shares for hydrogen-based fuel cell vehicles
to accommodate the heavier weight and larger loads. For
heavy-duty vehicles, we assume a phase-out of conventional
diesel and gasoline vehicle sales by 2050, at which point all
new purchases are battery-electric and fuel cell vehicles.
Battery-electric vehicles are used primarily for short-haul
applications, and we assume fuel-cell vehicles are needed
mostly for long-haul applications. Figure 3 summarizes the
distribution of fuels used to serve the projected vehicle miles
traveled across all transportation subsectors modeled for the
reference and high-electrification scenarios. only recently
have some analysts started considering viable electrified
applications in long-distance trucking, shipping, and even
air transport; future work is needed to assess these options
in greater detail.
our high-electrification scenario assumes 100% electri-
fication in all end uses in residential and commercial build-
ings. the full electrification technical potential is justified in
this scenario as electric technologies exist to serve nearly all
buildings end uses. in this scenario, space and water heating
services are served predominantly by heat pumps and a limited
share of electric resistance heaters. similarly, induction cook-
tops replace traditional stovetops. this scenario design does not
consider potential deployment challenges, including the slow
turnover of building stock or retrofit schedules, the economic
rationale for using heat pumps in cold climates where they are
less efficient, consumer preference, and other technical chal-
lenges with electrification for certain end uses or locations.
For industry, where electrification potentials are more
challenging to assess, we simply assume all conventional
boilers are replaced with electric technologies by 2050. as
shown by table 1, we assume nearly all process heating end
uses rely 100% on various electric technologies by 2050 with
the exception of process heating for iron and steel, which
we assume is only 21% electrified by 2050. this measure
of electrification refers to the fraction of total iron and steel
energy use that is purchased electricity. the share of steel
produced in electric arc furnaces in the united states is
much higher than the share produced in the combustion-
intensive blast furnace/basic oxygen furnace process. this
share of steel produced via electric arc furnaces is antici-
pated to grow under the scenario assumptions, but steel
production routes are not explicitly modeled in the analysis.
these assumptions are speculative and used to develop our
high-electrification scenarios only. More detailed research
is needed to consider whether electric technologies could
ieee power & energy magazine 39
 Reference High Electrification
5000 5000

4500 4500

4000 4000

3500 3500

VMT (Billion Miles)

VMT (Billion Miles)

HDV: Hydrogen
3000 3000 LDV: Hydrogen
HDV: Electricity
2500 2500
LDV: Electricity
2000 HDV: Oil/Gas/Bio
2000
LDV: Oil/Gas/Bio
1500 1500

1000 1000

500 500

0 0
2010 2020 2030 2040 2050 2010 2020 2030 2040 2050

figure 3. Vehicle miles traveled from 2010 to 2050 under the reference and high-electrification scenarios. LDV: light-duty
vehicles; HDV: heavy-duty vehicles.

table 1. The industrial end uses and electric technologies selected for analysis.
Percent Electrified
Electrotechnology Industry End-Use Service by 2050
Electrolytic reduction Nonferrous metals, excluding aluminum Process heating 100
Induction heating Metal fabrication Process heating 100
Electric boilers All manufacturing industries Conventional boiler use 100
Resistance heating and melting Glass Process heating 100
Direct arc melting Iron and steel Process heating 21
Industrial process heat pump Food, pulp and paper, and chemicals Process heating 100

cost-effectively serve processes with high-temperature and
large energy demands. in addition, our analysis does not
consider industrial equipment turnover, relative economics
of electric and nonelectric technologies, and nonenergy ben-
efits and impacts such as productivity and product quality.
in addition to the high-electrification view, we also con-
sider separate scenarios where significant increases in energy
efficiency are achieved, in combination with high electrifica-
tion, across all sectors. We refer to this as the high-combined
scenario, in which transportation efficiency (in terms of miles
per gallon of gasoline equivalent) is assumed to increase
beyond u.s. federal vehicle efficiency standards even after
the standards sunset. For buildings, we assume efficiency
improvements in electric end-use devices from 2017 to 2050
at various rates ranging from 0.5%/year to nearly 2%/year.
efficiency measures are assumed for services provided by
new electric technologies as well as in other end uses includ-
ing lighting, refrigeration, and other plug loads. For industry,
we simply assume an efficiency improvement of 0.8%/year.
this estimate was developed by weighting the current oppor-
tunity for state-of-the-art energy reductions identified by
u.s. Department of energy energy bandwidth studies by
their relative fraction of 2010 manufacturing primary energy
use, assuming these reductions would be achieved by 2050.
all remaining manufacturing industries not covered by the
energy bandwidth studies were assumed to reduce energy
use by 10% by 2050. these top-down sets of assumptions
do not specify the efficiency measures to achieve these rates
but, instead, rely on a wide range of measures such as vehicle
lightweighting, building insulation improvements, and pro-
cess improvements.
Figure 4 shows annual electricity consumption estimated
for the high-electrification and high-combined scenarios.
under high electrification, annual electricity demand grows
by 2.6%/year resulting in nearly 11,000 tWh of electricity
demand by 2050. this growth rate is roughly in line with

40 ieee power & energy magazine july/august 2018

 High Combined
High Electrification Electrification and Energy Efficiency
11,000 11,000
10,000 10,000
9,000 9,000
Electricity Demand (TWh)

Electricity Demand (TWh)

8,000 8,000
7,000 7,000
Transportation
6,000 6,000 Industrial
5,000 5,000 Buildings
Reference
4,000 4,000
3,000 3,000
2,000 2,000
1,000 1,000
0 0
2010 2020 2030 2040 2050 2010 2020 2030 2040 2050

figure 4. The reference and incremental electricity consumption from electrification by sector, 2010–2050.

historical growth rates during much of the 1900s and differs
significantly from more recent growth (flat) and projected
growth rate (0.7%/year) in our reference scenario (shown
in gray). as shown in Figure 4, most of the incremental elec-
tricity consumption growth is in the transportation sector
through the increased use of electric and fuel cell vehicles.
For the latter vehicle types, we assume, for simplicity, that
hydrogen production occurs only through electrolysis, which
because of the large share of fuel-cell vehicles and the low
efficiency of electrolysis-based hydrogen production results
in very large electricity requirements. in fact, in 2050 under
the high-electrification scenario, about half of the incre-
mental electricity consumption in transportation is used for
electrolysis-based hydrogen production, whereas the ratio of
vehicle miles traveled from direct electric vehicles to hydro-
gen-based fuel-cell vehicles is 3.2 to 1. electricity demand
growth in buildings and industry is also significant in the
high-electrification scenario. as a result, in this scenario, 2050
u.s. electricity consumption is more than twice 2050 demand
seasons, days, and hours play an important role for elec-
tric system resource adequacy and, ultimately, reliability.
resource adequacy refers to the ability to meet aggregate
electrical demand and is usually considered met when there
is sufficient firm capacity resources (primarily from the
supply side but also including demand-side resources) to
meet projected peak demands plus a reserve margin (~15%).
resource adequacy is only one component of reliability but
plays an important role in utility planning. changes in popu-
lation and energy use, including new electrical loads from
electrification, can impact the timing and magnitude of peak
demands and affect supply-side planning.
to demonstrate the effects of electrification on load shapes,
we calculate the ratio of national peak-to-average demand in our
scenarios (Figure 5). in the reference scenario, we assume limited
1.8
Ratio of Peak to Average Demand

in the reference scenario. 1.7

in the high-combined scenario, energy efficiency reduces
1.6
electricity demand growth to 1.9%/year. in the buildings
sector, the additional efficiency measures offset any incre- 1.5
mental demand from electrification, while the effects of effi- 1.4
ciency are less stark in industry and transportation. in total, 1.3
2050 electricity consumption in the high-combined scenario
1.2 Reference
is about 8,000 tWh, still approximately 3,000 tWh greater
High Electrification
than in the reference scenario. 1.1
High Combined
changes in electricity demand can have an impact on 1
electricity supply, including capacity deployment and other 2010 2020 2030 2040 2050
infrastructure evolution. But the growth in annual electric-
ity consumption reveals only part of the potential impacts; figure 5. The peak-to-average demand ratio for the refer-
changes in consumption patterns throughout the year, ence, high-electrification, and high-combined scenarios.

july/august 2018 ieee power & energy magazine 41

 We report results from a recently published initial analysis
conducted by NREL that simulated widespread electrification
from present day through 2050 in the United States.
changes to rate structures, end-use technologies, weather, and
other demand-side changes. as a result, the peak-to-average
demand ratio remains constant through time in this scenario,
with peak load 70% higher than average load (59% load factor).
in Figure 5, the “peak” is defined as the top 40 load hours.
this simple definition does not capture all implications, as
discussed later, but illustrates the large impact that electri-
fication can play on load shapes. in contrast, the peak-to-
average ratio decreases substantially through time in both
the high-electrification and high-combined scenarios. By
2050, peak demand is less than 9–13% higher than aver-
age demand in these scenarios (89–91% load factor). this
result is driven in large part by our bounding assumption that
electric vehicle charging and hydrogen production loads are
highly flexible. specifically, we assume zero charging dur-
ing the top 40 load hours of the year and that load associated
with hydrogen production is completely flexible within a day.
our implicit assumption is that rate structures incentivize
smart charging and the existence of sufficient infrastructure
to avoid vehicle charging during the peak hours and ade-
quate hydrogen storage and other infrastructure exist to shift
hydrogen production to nonpeak hours. given that load asso-
ciated with hydrogen production represents approximately
50% of total annual transportation electricity consumption
under this scenario (in 2050), the diurnal flexibility ascribed
to hydrogen production results in highly flexible load in the
vehicle sector. our analysis does not evaluate the costs to
achieve this level of flexibility, which is highlighted as a
future research need. Without intelligent policies and con-
sumer behavior, the peak-to-average demand ratio is likely
to remain higher than shown in the outer years of the analy-
sis, with important implications for build-out and operation
of the grid.
How electrification drives electricity consumption pat-
terns can have other impacts to electric system planning and
operations beyond resource adequacy. For example, the flat-
tening of demand profiles can have important implications for
the amount and type of supply-side resources; flatter demands
might drive the preference for technologies that provide energy
resources over capacity resources (e.g., resources designed for
efficient performance and low-cost energy but with higher
capital costs; these have traditionally been referred to as base-
load resources). in another example, changing demand profiles
can impact renewable integration challenges and opportuni-
ties. the capacity and energy values of wind and solar gen-
eration are closely tied to the correlation (or lack thereof)
42 ieee power & energy magazine
between variable generation profiles and demand. the amount
of renewable curtailment can also be affected by how electri-
fication impacts load profiles. Whether regulations, market
designs, and rates enable electrification to support efficient grid
evolution such as through greater demand-side participation is
an important research area.
Electricity Supply-Side Evolution
with Widespread Electrification
the impacts of u.s. electrification depend on how the
demand side might evolve but also on the supply-side future
of the electricity system. We develop various electricity sup-
ply-side scenarios—responding to the demand-side changes
envisioned—by employing the nrel regional energy
Deployment system (reeDs) model. the model simulates
the operation and expansion of the u.s. power system,
including power plants, transmission, and storage from pres-
ent day through 2050 by choosing the cost-optimal mix of
technologies. (utility stationary storage, as opposed to stor-
age in vehicles, is treated as a resource in the modeling.) the
least-cost solutions found by the model are constrained to
meet all regional electric power demand requirements (with
and without electrification), planning and operating reserve
requirements, technology resource constraints, and any pol-
icy requirements. in all scenarios, we use technology cost
and performance assumptions from the mid-cost projection
of the nrel 2016 annual technology Baseline and refer-
ence case fuel price assumptions from the eia 2016 annual
energy outlook.
Figure 6 shows the annual generation and installed
capacity from four scenarios modeled using reeDs. in the
reference scenario, which includes only current policies,
demand for new capacity resulting from business-as-usual
growth in load and end-of-life retirement of existing genera-
tors is met predominantly by new wind, solar, and natural
gas generation. total generation in 2050 under such a sce-
nario is 5,300 tWh, of which about 33% comes from wind
and solar (compared to about 7% in 2016), 28% from natural
gas, and 22% from nuclear, hydropower, and other renewable
technologies. While coal generation declines in the long run,
it still provides about 17% of 2050 generation in this refer-
ence scenario. on a capacity basis, we find 16 gW of average
net annual additions of wind and solar capacity from 2017 to
2050 and about 9 gW of net annual natural gas capacity addi-
tions, figures that are roughly in line with u.s. trends since
2010. in other words, without widespread electrification
july/august 2018
and new carbon policies, power sector changes observed in
recent years are projected to continue through 2050.
Figure 6, in the second column, shows the impact of high
electrification in isolation without any major targeted policy
changes on decarbonization or efficiency in the electric sec-
tor. not surprisingly, the incremental electrification-driven
demand growth results in a corresponding increase in total
generation, almost all of which comes from wind, solar, and
natural gas technologies. any price elasticities of demand
that might erode some of the incremental generation needs
of electrification are ignored here. combined wind and solar
generation is estimated to be 4,400 tWh (42%) in 2050,
exceeding current u.s. annual generation from all sources.
natural gas-based generation comprises a nearly equivalent
amount in 2050 (40%) with coal (8%) and other generation
(10%) comprising the remainder. to meet these penetration
levels, we estimate net average annual capacity additions of
over 45 gW for wind and solar technologies from 2017 to
2050 and even greater rates during the latter years. com-
bined with the 14 gW of net annual additions of natural
gas capacity, high electrification is found to offer a formi-
dable challenge, as well as opportunities, for infrastructure

Reference + High Electrification + Grid Decarbonization + High Efficiency

12,000

10,000
Annual Generation (TWh)

8,000

6,000

4,000

2,000

0
2010 2030 2050 2010 2030 2050 2010 2030 2050 2010 2030 2050
(a)
Reference + High Electrification + Grid Decarbonization + High Efficiency
4,000

3,500

3,000
Capacity (GW)

2,500

2,000

1,500

1,000

500

0
2010 2030 2050 2010 2030 2050 2010 2030 2050 2010 2030 2050
(b)

Nuclear Coal Natural Gas Hydropower Photovoltaic Storage

Carbon Capture and Sequestration Concentrating Solar Power Wind Other

figure 6. The (a) generation and (b) capacity mix over time for various supply- and demand-side scenarios. The first (far
left) panel includes both reference demand- and supply-side evolution. The second panel is end-use electrification only.
The third panel includes electrification and grid decarbonization (an 83% reduction from 2005 levels in 2050), and the
far-right panel combines electrification, grid decarbonization, and end-use energy efficiency.

july/august 2018 ieee power & energy magazine 43

 Beyond potentially reducing energy-related emissions,
electrification can also offer other important social
and economic benefits.
development in the u.s. electricity sector. these estimates
suggest growth rates in new capacity investments more aligned
with those currently found in large developing economies,
such as china and india, as opposed to the development
from recent history in the united states and europe.
combining high electrification with grid decarbonization
(the third column in Figure 6), the changes in generation
and capacity needs are even more pronounced. We define
grid decarbonization as 83% reductions in combustion-
related co2 emissions from 2005 levels by 2050 (for elec-
tricity production only) and model scenarios with a national
electric sector emissions cap achieving this reduction level.
specifically, we model a co2 combustion cap starting in
2022 based on emissions levels in the reference scenario and
decreasing linearly to the 83% reduction (of 2005 levels) by
2050. We include the cap in scenarios with and without high
levels of electrification. in this scenario, 2050 wind and solar
generation comprise nearly 7,400 tWh (65%), which would
likely introduce significant new challenges for grid integra-
tion. (We note that 600 tWh of this generation is from con-
centrating solar power with thermal energy storage.) total
stationary storage capacity from pumped hydro, batteries,
and compressed air energy storage sources exceeds 130 gW
by 2050 compared with about 30 gW in the reference sce-
nario. [storage nameplate capacity is shown in Figure 6(b).
the small amount of negative storage net generation is not
shown in the figure.] in the modeling, stationary storage
resources are treated as a utility resource and dispatched
optimally to meet system needs for energy, capacity, and
ancillary services.
renewable energy deployment and associated supply-
chain considerations would also be substantial under this
scenario with average net annual growth of about 78 gW.
With the stringent hypothetical carbon cap modeled in this
scenario, 2050 generation from fossil technologies without
carbon capture would be reduced dramatically (7% for natu-
ral gas and <1% for coal). However, we find carbon capture
and (underground) sequestration (ccs) technologies, pri-
marily using natural gas, become economically competitive
with nearly 17% of total 2050 generation from ccs, which
has an average capacity deployment rate of about 8 gW per
year through the full period but increases to over 25 gW per
year during the 2040s. other resources, including nuclear
and hydropower, comprise about 11% of total generation
in 2050 under this combined high-electrification and grid
decarbonization scenario.
44 ieee power & energy magazine
the last column of Figure 6 shows results from a scenario
that combines high electrification, grid decarbonization, and
high energy efficiency. the generation mix in this scenario
is similar to that from the previous one in percentage terms
(63% wind and solar, 13% ccs, less than 10% fossil with-
out ccs, and 14% other); however, the absolute magnitudes
are much lower for all technologies. For example, average
net wind and solar additions are estimated to be 57 gW per
year from 2017 to 2050 and lower than 5 gW per year of ccs
capacity is needed. this result suggests that efficiency can
help reduce infrastructure needs (and potentially bottlenecks)
that might arise from high electrification and stringent emis-
sions requirements.
CO2 Emissions Implications
of Widespread Electrification
the future of energy utilization and supply has obvious
implications for gHg emissions. our scenarios can be used
to estimate the individual and combined impacts of electrifi-
cation, grid decarbonization, and energy efficiency for direct
u.s. electricity and energy system emissions. Figure 7 shows
how electrification can lead to increases in direct co2 emis-
sions from the power system relative to the reference scenario
where emissions are estimated to remain largely flat or even
decline slightly in the long term. although electrification
increases absolute emissions from the power sector, emis-
sions intensities are lower in 2050 (23 MMtco2/MWh versus
28 MMtco2/MWh) as the incremental demand growth is
largely met by low-emitting renewable energy and natural
gas sources. these average emission intensities are about half
that of the current u.s. generation mix. More importantly,
electrification displaces emissions at the end use in the other
economic sectors, transportation, industry, and buildings,
resulting in energy systemwide emissions savings. We esti-
mate that electrification in isolation can avoid almost 1 billion
metric tons of co2 in 2050 relative to the reference scenario,
which is equivalent to a reduction of 20% in 2050 energy sys-
tem emissions.
Figure 7 shows that electrification alone has essentially
the same impact on total energy system emissions as grid
decarbonization alone (without electrification), both of which
result in measurable amounts of avoided emissions relative
to the reference case. in either case, however, total emissions
remain significantly higher than the levels some have esti-
mated are needed to reduce the probability of global tem-
perature increases above 2 °c. With either electrification or
july/august 2018
grid decarbonization, 2050 u.s.
energy system emissions are
6,000
estimated to decline by 36–38% Energy System Emissions
from 2005 levels. Refe
renc
e
We estimate that combining 5,000
Elec
electrification with grid decarbon- trific
ation
ization can lower energy system Grid
Dec
4,000 arbo

MMT CO2
emissions significantly. For co2, an niza
tion
d Single Measure
we estimate nearly 3.2 billion met- E D
ec le Impact
ar trc
ric tons fewer emissions in 2050 3,000 bo ific
+ n a
(66% lower than the reference sce- En iz tio
er at n
gy io
nario) when both these measures Ef
n
are achieved to the degree envi- 2,000 fic
ie
nc
sioned. Furthermore, implement- y
Combined Impact
ing high end-use energy efficiency 83% Reduction from 2005 Levels
1,000
would help avoid an additional
160 MMt. We find that even with
these combined measures—near- 0
complete electrification of trans- 2010 2020 2030 2040 2050
port and buildings, substantial
6,000
electrification of industry, and Power System Emissions
drastic decarbonization of electric-
ity generation—are insufficient to 5,000
achieve the 2050 emissions levels
indicated by climate scientists to
reduce the most-severe impacts 4,000
of climate change. u.s. energy
MMT CO2

system emissions in 2050 are esti- 3,000

mated to decline by 72% of 2005
levels with high electrification
and grid decarbonization and by 2,000
75% with additional high ener- Electrification
gy efficiency. Decarbonization
1,000

Future Work
and Remaining 0
Research Gaps 2010 2020 2030 2040 2050
the results presented here reflect
an initial attempt to explore the figure 7. Projected energy system and electricity system emissions.
potential implications of wide-
spread electrification of energy
services in the buildings, transportation, and industrial sec- recognizing the shortcomings of our preliminary analy-
tors in the future evolution of the u.s. electricity system. this anal- sis, nrel, in collaboration with other researchers, has initi-
ysis provides a quantitative but approximate indication of the ated a new study to employ higher fidelity modeling to better
impacts that electrification might have on u.s. emissions and understand the impacts of widespread electrification in the
power sector needs. given its exploratory nature, many gaps united states. the new study, referred to as the electrifica-
remain. the completed analysis only considers a limited set of tion Futures study (eFs) (www.nrel.gov/efs), is designed to
electrification, efficiency, and power sector scenarios, whereas employ three new modeling components missing from the
large uncertainties exist for many assumptions made for each existing study. First, the eFs will use a consistent cross-sector
scenario. in addition, the electrification potential estimates for framework (the energyPatHWays model) to track stock
each sector are based on relatively aggregated analysis and do turnover, costs, and performance for 366 end-use technolo-
not consider explicit technologies, adoption patterns, or other gies serving all major end uses. second, the eFs will develop
details that could ultimately affect anticipated electrification, a new bottom-up engineering model to project hourly elec-
future energy use, and consumption profiles, all of which would tricity consumption from future buildings, transportation,
have material impact on electricity supply and emissions. and industrial equipment stock. referred to as demand-side

july/august 2018 ieee power & energy magazine 45

 We also consider separate scenarios where significant
increases in energy efficiency are achieved, in combination
with high electrification, across all sectors.
grid (dsgrid), this new model will rely on multiple bottom-
up sector models to enable an unprecedented assessment of
future electricity load patterns. Finally, for the eFs, we plan to
employ unit commitment and economic dispatch modeling to
examine, at hourly and subhourly resolution, how power sys-
tems might operate with the new electrified load patterns and
greater shares of variable renewable electricity. this modeling
can also be used to assess the potential value of demand-side
resources, particularly from new electricity loads, to help sup-
port grid reliability. as with the study described in this article,
the eFs will also rely on the nrel reeDs model to develop
future electric sector expansion scenarios.
in addition to the more detailed modeling to be employed
for eFs, the new study will also include a wider range of
experts and a deeper exploration of available data, includ-
ing data for projected cost and performance of electric tech-
nologies. the models and data will be used to develop more
realistic scenarios than the “technical potential” scenarios
that are the focus of the results presented previously. We also
plan to consider more comprehensively system and house-
hold impacts in the eFs as well as a broader suite of envi-
ronmental and other impacts.
Despite ongoing and planned studies of electrification in
the united states, including the eFs, many research gaps
will remain. next, we present a list of key research needs for
a more complete understanding of electrification potential
and implications.
✔ End-use technology adoption modeling. consumer
adoption models exist for certain technologies, e.g.,
distributed photovoltaics and passenger vehicles, but
are less developed for other technologies and end-use
subsectors, including end uses in buildings, heavy-duty
vehicles, and in industry. a more specific analysis
of industry could benefit from expanding the account-
ing of physical equipment stocks and the development
of associated stock turnover models. advanced adoption
modeling considers multiple layers, including resource
and technology characteristics, demographics, behav-
ioral economics, capital investment of industrial firms,
and consumer preferences. the development of such a
model to span many end uses or the combination of mul-
tiple end-use specific models would illuminate the eco-
nomic and market potential for electrification as well as
highlight potential challenges and policy interventions.
✔ Electric technology road maps. technology road maps
are helpful to elucidate advancement opportunities for
46 ieee power & energy magazine
new technologies. they can be used to highlight where
research and development would be most effective. For
electric technologies, which are typically more capi-
tal intensive but with relatively lower operating costs,
technology road maps can indicate needed efficiency
improvements and cost reductions. these road maps
can be applied to electric technologies themselves as
well as enabling technologies, e.g., charging infrastruc-
ture or information and communication technologies to
support grid flexibility. road maps for industrial elec-
tric technologies, which recently have not typically re-
ceived as much attention as transportation and building
technologies, could fill a large research gap.
✔ Data on current energy use. For the united states,
the energy information administration and other
agencies provide valuable data on current and histori-
cal energy use. However, data gaps remain for many
subsectors, particularly when a complex array of fac-
tors can impact energy consumption and technology
alternatives. For example, the diversity and complex-
ity of the u.s. industrial sector makes any electrifi-
cation assessment difficult. Further, while energy
consumption surveys provide invaluable data on elec-
tricity consumption, they are updated only periodi-
cally and can be further disaggregated by technology,
geography, and end use.
✔ Cost and value of grid flexibility. our scenario analysis
highlights how flexible vehicle charging and hydrogen
production can reduce peak capacity needs; however, an
evaluation of the cost and benefits of this added flex-
ibility is also needed. More generally, research is needed
on the enablement costs associated with increased flex-
ibility, including inconvenience, rate structure design,
and electric charging and hydrogen infrastructure costs
and needs. similarly, greater high-fidelity transmission
and distribution system modeling is needed to assess the
value of this flexibility.
✔ Macroeconomic modeling and analysis. Models that
focus on broader economic impacts, such as comput-
able general equilibrium or integrated assessment
models, can be used to explore widespread electrifi-
cation, including potential rebound and leakage im-
pacts. typically, by themselves, these models lack the
technology and end-use fidelity needed to fully assess
electrification. Model improvements or efficient link-
ages between models would be needed.
july/august 2018
 ✔ Policy design and analysis. an assessment of differ-
ent policy interventions, including technology-push and
demand-pull policies, for electrification would inform
decision makers of the options to increase electrifica-
tion or reduce the negative impacts of electrification.
✔ Nonelectric options. although this article focuses on
electrification, an understanding of nonelectric techno-
logies or indirect electric technologies (e.g., power to
gas) is also needed to position electrification within a
broader suite of options. For carbon emissions mitiga-
tion, multiple options exist, and electrification may not
be the lowest-cost or most-preferred solution for certain
subsectors. an assessment of and investment in these
alternative pathways could be needed.
Summary
to a typical consumer, electricity can appear unlimited, low
cost, and available, literally, at the flip of a switch. From a system
operator’s perspective, planning for and maintaining an efficient
and reliable power system relies on a complex array of consider-
ations in power systems engineering, economics, and policy and
regulatory affairs. electrification—although a robust and likely
necessary pathway to a clean energy future—requires navigat-
ing the complexities of the power system along with the inter-
actions with multiple other complex systems, including energy
consumption and end-use service demand in the residential,
commercial, transportation, and industrial sectors.
a small but growing number of studies exploring the poten-
tial of widespread electrification in the u.s. energy system
identifies the benefits of electrification, including the potential
for substantial reductions in gHg and criteria pollutant emis-
sions. our scenario analysis, which focuses on a near-complete
electrification across all major economic sectors, finds that
electrification could indeed result in significant reductions in
energy-related co2 emissions; however, the magnitude of the
reductions is strongly sensitive to how the power system might
evolve. Without new electric sector policies, we find reductions
in systemwide energy-related emissions to 36% of 2005 levels
by 2050 with electrification only. similar energy-system emis-
sions are found by decarbonizing the power system to 17% of
2005 levels in 2050 alone (without electrification). the combi-
nation of electrification and a grid emissions cap would yield
greater emissions reductions (72–75% of 2005 levels by 2050),
especially if end-use energy efficiency is employed. these
energy systemwide reductions still do not meet the targets that
many estimate will be needed to reduce the worst impacts of
climate change, therefore highlighting that other means beyond
electrification (and grid decarbonization) would be needed if
the achievement of such goals is desired.
although electrification by itself is not a complete solu-
tion for eliminating all environmental hazards as indicated by
our analysis and others, it is difficult to imagine pathways to
achieve significant reductions in gHg emissions that do not
include increased electrification. the extent to which electrifi-
cation could be economically viable relative to other options is
july/august 2018
an open research question. Many other fundamental research
topics related to electrification—including consumer adop-
tion, electric technology evolution, macroeconomic impacts,
health and environmental impacts, bulk power system opera-
tions and reliability, and distribution system impacts—are
also needed. our analysis and ongoing studies, such as the
electrification Futures study, have started to reveal insights
in these areas but the herculean task requires a much greater
research endeavor. this research is needed to keep electricity
supply reliable and enable it to power wider swaths of a cleaner
u.s. economy.
For Further Reading
D. steinberg, D. Bielen, J. eichmann, K. eurek, J. logan, t.
Mai, c. McMillan, a. Parker, l. Vimmerstedt, and e. Wilson,
“electrification & decarbonization: a preliminary analysis
exploring u.s. energy use and greenhouse gas emissions in
scenarios with widespread end-use electrification and power
sector decarbonization,” national renewable energy labora-
tory tech rep. nrel/tP-6a20-68214, 2017.
“renewable electricity futures study,” National Renew-
able Energy Laboratory, national renewable energy lab-
oratory tech rep. nrel/tP-6a20-52409, 2012.
J. H. Williams, B. Haley, F. Kahrl, J. Moore, a. D. Jones,
M. s. torn, and H. McJeon, “Pathways to deep decarboniza-
tion in the united states,” the u.s. report of the Deep Decar-
bonization Pathways Project of the sustainable Development
solutions network and the institute for sustainable Develop-
ment and international relations, 2014.
“united states mid-century strategy for deep decarboniza-
tion,” the White House, 2016. https://unfccc.int/files/focus/long-
term_strategies/application/pdf/us_mid_century_strategy.pdf
K. Dennis, “environmentally beneficial electrification:
electricity as the end-use option,” Electricity J., vol. 28, no. 9,
pp. 100–112, nov. 2015.
P. Jadun, c. McMillan, D. steinberg, M. Muratori, l. Vim-
merstedt, and t. Mai, “electrification futures study: end-use
electric technology cost and performance projections through
2050,” national renewable energy laboratory tech. rep.
nrel/tP-6a20-70485, 2017.
Biographies
Trieu Mai is with the national renewable energy labora-
tory, golden, colorado.
Daniel Steinberg is with the national renewable energy
laboratory, golden, colorado.
Jeffrey Logan is with the national renewable energy
laboratory, golden, colorado.
David Bielen is with the national renewable energy
laboratory, golden, colorado.
Kelly Eurek is with the national renewable energy lab-
oratory, golden, colorado.
Colin McMillan is with the national renewable energy
laboratory, Washington, D.c.
p&e
ieee power & energy magazine 47

Electrification
in the
United Kingdom
F
For the United Kingdom, one oF the main
drivers for a green ambition is the Climate Change act of
2008. this forms the basis for the country’s approach
to responding to climate change and legally commits the
U.K. government to reducing greenhouse gas emissions
by at least 80% below 1990 levels by 2050. this level
of commitment to reducing emissions was further con-
firmed in the Paris agreement, the aim of which is to hold
the increase in global average temperature to well below
Digital Object Identifier 10.1109/MPE.2018.2822864
Date of publication: 18 June 2018
48 ieee power & energy magazine 1540-7977/18©2018IEEE
©istockphoto.com/traffic_analyzer
2 °C above preindustrial levels and to pursue efforts to limit
warming to 1.5 °C.
achieving these goals is only one potential outcome of
the energy future. Within national grid (United Kingdom),
we produce four credible future energy scenarios that look
out to 2050. these are contained in an annual publica-
tion, Future Energy Scenarios (FES). only one of the sce-
narios within this publication meets the 2050 target, two
degrees, which is named to reflect the target adopted by the
Paris agreement.
Within the carbon reduction arena, there are three areas
that will have the greatest impact: power generation, heating,
july/august 2018
 By Russell Fowler, Orlando Elmhirst,
and Juliette Richards

and transport. Currently, significant movement has been made in decarbonizing power genera-
tion, and some movement has been made in transport, but there has been little advancement in
residential heating. the two degrees scenario relies significantly (but not completely) on electri-
fication of the transport and heat sectors. in this article, we discuss what will need to happen for
the United Kingdom to meet its legally binding commitments and the corresponding implications
for the power sector.

What Are the Future Energy Scenarios?

national grid, as part of its role as great Britain’s system operator, produces in-depth analyses
for the future demand and supply of gas and electricity to the country. FES outlines a range of
credible pathways (the scenarios) for the future of energy.
these scenarios, which have been published each
year since 2011, describe the possible demands for and
A Case Study Based on sources of gas and electricity and their implications
for consumers and the energy industry. FES aims to
Future Energy Scenarios communicate the reality of each scenario in a clear
and concise way. all the scenarios have considered
energy demand and supply on a whole system basis,
incor porating gas and electricity across the great
Britain transmission and distribution systems. FES
does not constrain the levels of demand and supply because of network capabilities or oper-
ability issues.
Since FES is used by various stakeholders, it is used in a multitude of ways. Stakeholders have
said they use it as
✔ a foundation to build their own analyses and scenarios
✔ information for investment decisions
✔ a market view
✔ a reference point
✔ an academic source.
FES is also used as a basis for a range of further national grid analyses; for instance, it is the
starting point for regulated long-term network and operability planning. as each of the subse-
quent processes has its own requirements, further detailed analysis is undertaken, building on
FES. it is also used for activities that have shorter-term (up to five years) consequences, such as
the setting charges for customer use of the great Britain transmission system, and as a basis for
the electricity capacity report, which is used to support the government’s decision on the amount
of capacity to secure through the capacity market auctions.

How Are Scenarios Created?

the production of the scenarios is part of an annual development process that has three key stages:
stakeholder engagement, data and intelligence gathering, and analysis. at each stage, the energy
insight team applies its expertise and judgment to ensure that what is created is plausible and cred-
ible. Feedback from stakeholders, such as industry experts, is an essential part of the development
of FES. Following engagement with stakeholders, the scenario framework is created that includes
assumptions and levers.
in 2017, four future energy scenarios were developed: two degrees, Slow Progression,
Steady State, and Consumer Power. the process starts by using a 2 × 2 matrix with the axes

july/august 2018 ieee power & energy magazine 49

 of green ambition and prosperity; see Figure 1. they de -
scribe the political, economic, social, and technological
(PeSt) landscape.
the 2 × 2 matrix is based upon the U.K. government’s
energy trilemma of sustainability, affordability, and security
of supply. Security of supply is assumed to be met in all
scenarios. this leaves sustainability (green ambition) and
affordability (prosperity) for the models to flex depending
on the scenarios’ characteristics. green ambition reflects
the level at which society and policy engages with reduc-
ing carbon emissions and increasing sustainability. Prosper-
ity follows the amount of money available in the economy
for government expenditure, business investment, and con-
sumer spending.
the next layer of the framework is a more detailed consider-
ation of the PeSt categories, which become the assumptions.
to each of these assumptions, levers are applied depending on
the conditions applicable to each scenario. these assumptions
and levers are then applied to the demand models, of which
there are more than 50. how the model is structured depends
on how well the demand category is understood. if it is an es-
tablished demand source, then a modified regression analysis
is applied. if it is a new type of demand, then proxy models are
developed and applied.
For residential modeling, a bottom-up approach is used.
each appliance category is modeled individually, such as
space heating, water heating, and lighting. the outputs are
then combined to produce an overall demand.
For the industrial and commercial demand, a top-down
model is used, which was developed in conjunction with a
third party. it is a 24-sector model that uses historical govern- achieved. For example, in april 2017, the United Kingdom did
ment data for the basis of an autoregressive-moving-average
More Money
Available
Two Degrees
Consumer Power A World Where Environmental
Prosperity
A World That Is Relatively Sustainability Is Top Priority
Wealthy and Market Driven
Less Money
Available
Slow Progression
Steady State A World Focused on Long-Term
A World Focused on Security of Environmental Strategy
Supply and Short-Term Thinking
Less Focus Green Ambition
figure 1. The FES matrix.
50 ieee power & energy magazine
analysis. this model has, as an input, high-level economic
data supplied by oxford economics.
once all of the demand levels have been established, the sup-
ply models are produced using a similar process and matched to
meet the demand levels that were calculated first. details of the
different modeling methods used can be found in a FES meth-
ods document and the outputs in a FES Chart workbook (see
the “For Further reading” section). the models combined
then produce the scenario outputs for FES.
all the models within FES extend out to 2050. most of
them also contain an historical element going back to 2005.
this full history exists in the case of established catego-
ries of demand and supply, such as refrigeration demand,
but less so in the new and evolving demand sources, such as
electric vehicles (eVs) and smart meters. the inputs, models,
and outputs take account of stakeholder feedback and are
all subject to internal scrutiny and governance to ensure
their credibility and accuracy.
Decarbonizing Power
the decarbonization of power generation reduces emissions
directly and enables decarbonization in the transport and
heating sectors. the evolution of the power sector in the two
degrees scenario will be discussed in this section.
The Current State of Power Generation
in the United Kingdom
Considering the three sectors of power, heat, and transport in the
United Kingdom, power is seen as leading the way in decarbon-
ization. the past 12 months have seen a number of milestones
not use any coal for electricity generation for 24 h, the first time
this has happened since the indus-
trial revolution.
a number of renewable gen-
eration technologies have seen
sustained growth in the past five
years, notably solar and onshore
and offshore wind. Solar capac-
ity was almost nonexistent in the
United Kingdom at the start of the
decade. But since 2011, continued
reductions in the cost of photovol-
taic panels, alongside the intro-
duction of a feed-in tariff subsidy
regime, have led to huge growth
(see Figure 2).
Solar capacity was also eligible
for the renewables obligation (ro),
which required energy suppliers to
buy a proportion of their energy
f rom low-ca rbon generators.
More Focus
more recently, however, the ro
scheme closed to solar capacity
in 2015 (larger projects) and 2016
july/august 2018
 Within the carbon reduction arena, there are three
areas that will have the greatest impact: power
generation, heating, and transport.

(smaller projects). at the same time, feed-in tariffs rates were
decreased. this led to a “solar rush” in the United Kingdom
as developers tried to install and accredit projects before the
relevant closing dates for support, followed by a slowing
of growth after these points. this rush was largely unan-
ticipated by government and industry parties alike. By 2017,
solar capacity stood at over 12 gW, nearly achieving the gov-
ernment’s forecast of 13 gW by 2020 several years early.
Similarly to solar capacity, onshore wind generation in the
United Kingdom also saw large growth in the early parts of
this decade, with subsidy mechanisms supporting continued
installations and research and development. however, this
growth has slowed as support mechanisms are being with-
drawn between 2016 and 2019 because this technology is now
considered to be established by the U.K. government.
in contrast, offshore wind generators continue to receive
support via the contracts for difference (Cfd) mechanism,
which guarantees a power price to offshore wind generation.
in September 2017, an auction round saw future build off-
shore wind farms bid into the Cfd mechanism at a price
of electricity with an increasing number of market players.
decentralized generation (located not on the high-voltage
transmission network but, rather, on the lower-voltage dis-
tribution network) is changing the pattern of electricity
flows (see Figure 3). also, interconnectors are allowing
access to sources of supply from other countries. this is an
additional source of flexibility that is required as domestic
generation becomes more intermittent with the increase in
renewable capacity.
What Could the Future Look Like?
our two degrees scenario sees continued rapid decarbon-
ization of power generation achieved by significant growth
14
>25 MW
12
5 to ≤ 25 MW
10 50 kW to ≤ 5 MW
8 10 to ≤ 50 kW
TWh

equivalent to around £64/mWh (in 2017 prices) that is close 4 to ≤ 10 kW

6 0 to ≤ 4 kW
to the long-term wholesale price of power in the United
Kingdom. this, in turn, indicates that such projects are 4
coming close to running with minimal subsidies as costs fall 2
and efficiencies increase. of course, the United Kingdom,
-
as an island, is ideally situated to pursue offshore wind proj-
Jan. 2010
May 2010
Sept. 2010
Jan. 2011
May 2011
Sept. 2011
Jan. 2012
May 2012
Sept. 2012
Jan. 2013
May 2013
Sept. 2013
Jan. 2014
May 2014
Sept. 2014
Jan. 2015
May 2015
Sept. 2015
Jan. 2016
May 2016
Sept. 2016
ects across a number of sites. in 2017, Statoil launched the
world’s first floating large-scale offshore wind turbine off
the coast of Peterhead in Scotland. this is significant as such
turbines can operate in deeper waters, increasing the number
of potential sites for projects and capturing better wind con- figure 2. U.K. solar generation by system nameplate
ditions farther out to sea. capacity.
Some less-established renewable technologies, for exam-
ple marine generation, are also in the process of developing
projects, although tidal and wave generation projects have
Yesterday Today and into the Future
yet to receive support through the government’s Cfd mecha-
nism. Finally, the continued decrease in the cost of electric-
ity storage is leading to new opportunities as intermittent,
renewable generation is increasingly colocated with storage,
for example orsted energy’s Burbo bank project, located on
the merseyside coast, which has installed a 2-mW battery
alongside 90 mW of wind capacity.
the overall effect of growth in renewable, intermittent
electricity generation means that the U.K. power system is
no longer dominated by easily controllable, large thermal
plants. rather, there is a shift toward more complex flows figure 3. The changing shape of the network.

july/august 2018 ieee power & energy magazine 51

 impact of their energy usage, particularly at peak time, and
table 1. Data shifts from 2017 and 2050
in Two Degrees. this is assisted by improved information communication tech-
nology. the scale of these shifts is captured in table 1.
Today 2050
Great Britain population 66 million 75 million What Are The Challenges in Achieving This?
For this significant change to take place, the two degrees
Number of households 28.8 million 32.4 million
scenario makes a number of key assumptions, some of which
Assumed economic growth 2.1% may be challenging to implement on the actual system.
rate
First, it is assumed that there will be a supportive policy
Installed renewable 34% of 60% of and regulatory environment in several areas:
generation capacity installed installed
capacity capacity
✔ regulation and policy to encourage or mandate the
insulation of homes and greater appliance efficiency,
Carbon intensity of power 242 g/kWh 2 g/kWh including market support mechanisms for insulation,
generation
legislation for new buildings and new appliances, and
Great Britain gas demand 817 TWh 399 TWh scrappage programs
Power generation gas 249 TWh 34 TWh ✔ continued policy support for the roll out of smart me-
demand (including CCS) ters, resulting in all buildings having smart meters by
Industrial and commercial 235 TWh 205 TWh 2021 to encourage demand response
gas demand ✔ continued commitment by the government to main-
Residential gas demand 333 TWh 34 TWh
tain a carbon floor price and phase out all use of coal
by the mid-2020s
Transport gas demand 0 TWh 126 TWh ✔ continued support for new low-carbon generation ca-
Great Britain electricity 306 TWh 368 TWh pacity, for example, Cfd continuing for fewer estab-
demand lished technologies and other supportive regimes to
Industrial and commercial 191 TWh 217 TWh de-risk new investment such as cap and floor schemes
electricity demand ✔ regulation and policy to ensure the electrification of
Residential electricity 114 TWh 116 TWh heat and transport (this will be discussed later).
demand ✔ government backing to enable financial and political
support for nuclear power build out
Transport electricity 1 TWh 35 TWh
demand ✔ CCS technical developments encouraged by govern-
ment grants and incentives.
Installed residential smart 1.6 million 32 million
meters moreover, in the two degrees scenario, a high degree of
engagement from consumers is assumed. For example, the
majority of consumers will deliberately choose more effi-
in renewable generation (solar, wind, and some marine gen- cient appliances where possible, seek other ways of reducing
eration) and a large increase in installed nuclear capacity. energy demand, and choose to shift energy demand away
Flexible sources of electricity, to support a highly intermit- from peak times as a result of time-of-use tariffs.
tent system, are provided by increased interconnection to
other countries and just under 10 gW of storage from a vari- Decarbonizing Heat
ety of sources. in addition, this scenario sees carbon capture
and storage (CCS) being commercially available by 2030, The Current State of Heat Decarbonization
allowing for flexible electricity generation using natural gas in the United Kingdom
with minimal carbon emissions. By 2050, the carbon intensity natural gas currently provides the majority of space and
of electricity generation will have reduced to 2 g of Co2/KWh water heating in the United Kingdom. Since natural gas is a
(compared to an average of 242 g/KWh today). carbon-emitting fuel, heating is responsible for a large share
in terms of overall electricity demand, more electricity is of the United Kingdom’s carbon emissions. From a decar-
being used in the 2050 two degrees scenario than today and bonization perspective, heat is lagging behind in the journey
is accompanied by a fall in natural gas demand. a number of toward a low-carbon world.
factors bring about this shift. the electrification of transport Prior to the 1970s, the gas supplied to households in the United
and heat (discussed below) and a growth in industrial output Kingdom was derived from coal, so-called “town gas.” however,
due to this scenario’s high economic growth rate assumption in 1966, cleaner natural gas was discovered in the U.K. continen-
increase the demand for electricity. however, improved appli- tal shelf (UKCS), and the decision was made to convert all of the
ance and lighting efficiency and better home insulation reduce United Kingdom’s use of gas to natural gas through the use of this
the overall demand for energy. it is assumed that consumers indigenous resource. this involved refitting over 40 million appli-
are more engaged in trying to reduce the environmental ances, and the project took ten years to complete.

52 ieee power & energy magazine july/august 2018

 The decarbonization of power generation reduces
emissions directly and enables decarbonization
in the transport and heating sectors.
From the 1970s until early in this century, the United King-
dom was self-sufficient in natural gas, which it was able to
extract from the UKCS in the north Sea. however, since the
early 2000s, gas extraction from the UKCS has been declining.
today, about 80% of homes have gas heating, and all
together these households use 332.6 tWh of gas a year for
heating and cooking, representing 36% of U.K. annual gas
consumption, the largest gas demand category. Commer-
cial heating accounts for 48 tWh per year, or about 4% of
total annual energy consumption. industry uses 187 tWh or
approximately 17% of total energy consumption, predomi-
nately for high-temperature industrial processing.
of the 29 million residential homes in the United King-
dom, around 150,000 (~0.5%) have low-carbon heating
installations such as heat pumps, biomass boilers, and micro
combined heat and power systems. this relatively low adop-
tion is despite government attempts to support their installa-
tions in the form of the following:
✔ the domestic renewable heat incentive, a govern-
ment financial incentive to promote the use of renew-
able heat. People who join the program and stick to its
rules receive quarterly payments for seven years for
the estimated amount of clean, green renewable heat
their system produces.
✔ the green deal, a government financial loan program
to make energy-saving improvements to one’s home.
this would typically include measures such as wall
insulation, draught-proofing, and double glazing. the
original program ran from 2013 to mid-2015 and is
scheduled to be relaunched in 2018.
Larger schemes such as district heating account for fewer
than 500,000 buildings. Such programs provide around 2%
of heating demand in the United Kingdom.
however, this is not to say that the heat sector is devoid
of progress toward decarbonization goals. Currently in
the United Kingdom, research is underway to consider
a number of low-carbon heating solutions, including elec-
tric heating (ground and air source heat pumps), the use of
hydrogen instead of natural gas, and district heating pro-
grams. the Leeds h21 Citygate project is a study with the
aim of determining the feasibility of converting an entire
city’s natural gas network to 100% hydrogen. at the same
time Cadent, the United Kingdom’s largest gas distribution
company, is researching the introduction of hydrogen into
the gas network (as a blend with natural gas). the con-
cepts being explored would use gas, in conjunction with
CCS, to generate hydrogen.
july/august 2018
What Needs to Happen?
to reach the 2050 carbon reduction target in a cost-effec-
tive manner, the two degrees scenario includes a signifi-
cant drive toward the electrification of heating. this drive
is unlikely to occur spontaneously, and it will require incen-
tives to persuade households to move from gas and at the
same time make their homes much more energy efficient.
if the drive was a simple switch from gas boilers to elec-
tric resistance heat, then the amount of electricity required
to heat the housing stock would be such that costly network
infrastructural changes would be required. this would be the
case not only for the generation requirements but also for the
transmission and distribution systems reinforcement, which
would have to deliver the energy. FES does not attach a cost
to its scenarios, as the intention is to look at an unconstrained
world, except where this is obviously unreasonable. Subse-
quent financial analysis is carried out by dependent model-
ing and the outcomes are reported in the network options
assessment. therefore, for the most cost-optimal system solu-
tion, the two degrees scenario sees a need for the demand to
drop. this drop will be in the face of a rising population that
will result in 4 million additional homes being built.
the two degrees scenario forecasts a large-scale
move away from gas boilers: from today’s 21 million to
approximately 7 million by 2050, largely (but not exclusively)
through conversion to air-source heat pumps (aShPs). the
journey toward electrified, or low-carbon, heating takes
place in three phases, the product of an optimization model
that is applied solely to the two degrees scenario. Phase 1
sees some rollout of aShPs by those who have the properties
that can accommodate and benefit from such heating. Phase 2
sees hybrid heat pumps becoming more popular, especially
in properties that are not particularly energy efficient and will
still require the extra heat boost that a hybrid accommodates.
Based on stakeholder feedback, hybrid systems are only a step-
ping stone away from full gas heating and, therefore, more
attractive for consumers. in the third phase, aShPs dominate
once again, reflecting an assumption of significant techno-
logical improvements that, in part, drive up their efficiency.
From now to 2050 is effectively the timescale for two
boiler replacements. a persistent challenge with conversion
is that boiler replacements are often distress purchases, so
there needs to be both a clear cost advantage and a simple
switch-over process in place to make such changes the pre-
ferred option, possibly in advance of product breakdowns.
the two degrees scenario also includes other forms of
heating such as fuel cells, district heating, and micro heat and
ieee power & energy magazine 53
 power units, but these will be in the minority. the growth of
all heating technologies is detailed in Figure 4. the develop-
ment of hydrogen as a large-scale solution to decarbonizing
heating is still problematic because there are many techni-
cal issues to be overcome such as appliance switching, CCS
attachment to gas-supplied hydrogen generation plants, and
regional-scale switchovers. We await more concrete devel-
opments before fully integrating this into our scenarios, but
we have included a sensitivity in this year’s FES that looks at
the possibility of hydrogen networks.
the other required element of the two degrees scenario is
significant improvement in the thermal efficiency of homes.
two degrees assumes the housing stock will be 30% more
efficient at retaining heat by 2050. this will result in a drop
in both electricity and gas usage in terms of both annual and
peak demands. this will be no mean achievement, and it is
important that the new green deal, and any complementary
policies that may evolve, will be much more successful than
the previous green deal.
25
20
15
Millions
What Could This Look Like?
the effect of the above-mentioned activities will produce a
drop in both commercial and residential natural gas use, as
shown in Figure 5. the resultant reduction in consumption
for heating is dramatic and possible if effective incentives
are introduced. By 2050, the demand for residential natural
gas will fall dramatically by 300 tWh, or nearly 90%.
as can be seen in Figure 4, heat pumps of one type or
another will dominate. the amount of electricity that will be
required to drive them is in the region of 60 tWh by 2050.
the peak demand for heat pumps follows a similar curve to
the number of installations. today the demand is about 0.1 gW;
by 2030 it is 3.8 gW and reaches 10.7 gW in 2050.
the dramatic drop in the demand for natural gas for heating
will not directly translate into an equivalent rise in electric-
ity demand. With appliance efficiency savings and improved
thermal efficiency of the housing stock, the energy required is
reduced to a fraction of its current value. this will mean that a
radical overhaul of the electricity system will not be required
to accommodate the heating load expected of it. this makes
this model one of the best overall solutions for the consumer.
as an aside, this relatively low annual electricity demand
for heating also has been associated with a relatively low peak
demand. electricity peak is usually at 17:30 h on a winter’s week-
day, which is when residential heating would be required, thus

adding to an already high usage period. however, this level of

10
additional peak load (11 gW, or 15% of the total) is manageable.
5 Commercial heating will broadly follow the residential
sector’s path and, like that sector, the demand will be grow-
0
2015 2020 2025 2030 2035 2040 2045 2050 ing out to 2050. Consequently, the reduction in natural gas
utilization will be offset by the growth in the commercial
ASHP Biomass
Gas Heat Pump
heating sector, and by 2050 it will be using almost the same
Micro CHP
Ground-Source Heat Pump Fuel Cell amount of gas as today.
Hybrid Heat Pump Gas Boiler the industrial sector’s use of natural gas for heating will be
driven more by the requirements for process heating. Unlike
figure 4. Installed low-carbon heating technologies in the the commercial sector, the industrial sector continues its his-
Two Degrees scenario. torical decline. By 2050, gas demand will be down by 30 tWh.
the annual demand of heat pumps, as projected in FES
2017, is significantly greater than what was estimated in FES
350 2013. the estimates for 2030, in these two documents, are
24 tWh and 14 tWh, respectively. the reason for this change is
300 that new assumptions were applied to the model. the one with
250 the most dramatic effect was a decision to shift from an assump-
200 tion of fewer installations in average homes to more installa-
TWh

tions in larger homes. While these larger homes would require

150
almost double the energy input to heat them, their owners were
100 assumed to be more able to bear the cost of installation.
50
0
Decarbonizing Transport
2015 2020 2025 2030 2035 2040 2045 2050
Gas: Industrial Gas: Commercial
Current State of Play
Gas: Residential Electricity: Heat Pumps the decarbonization of transport, and in particular cars, has,
until recently, been relatively muted within the United Kingdom.
figure 5. Annual demands for gas, by sector, and heat the adoption of plug-in eVs has been predominately led by plug-
pumps’ electricity in the Two Degrees scenario. in hybrid eVs (PheVs) with pure eVs (PeVs) taking second

54 ieee power & energy magazine july/august 2018

place. their combined figures only
represent around 0.3% of the regis- Non EVs
tered cars in the United Kingdom 90,000 35,000,000 PHEV
PHEV
(see Figure 6). 80,000 PEV
PEV 30,000,000
Because there is a relatively 70,000
small number of eVs, their impact 60,000 25,000,000
on the electricity networks has been 50,000 20,000,000
minimal. most of the vehicles are
40,000
charging from domestic chargers 15,000,000
30,000
with either 3.5 or 7 kW chargers.
20,000 10,000,000
the use of fast charging stations
of 50 kW or higher is beginning to 10,000 5,000,000 EVs = 0.3%
develop. Such chargers take about 0
0
30 min to charge today’s average
2010

2011

2012

2013

2014

2015

2016
2010 2012 2014 2016
eV. there are more powerful char-
(a) (b)
gers available, of above 100 kW,
but they are the exception. recent
announcements have indicated figure 6. (a) EV numbers and (b) in the context of all car numbers.
plans to build a network of at least
400 rapid eV chargers across europe. these are likely to be how much traveling that is undertaken varies in each sce-
150-kW units and may be capable of supporting 350-kW char- nario. in the United Kingdom, individuals’ annual average
gers in the future. trip distances have decreased over recent years. this trend
Some policy impetus has been given by recent U.K. gov- will continue in two degrees, driven by more working at
ernment announcements. the Secretary of State for environ- home and less traveling to meetings. this will come about
ment, Food and rural affairs has said that the government through information and technology communications im-
has an ambition to end the sale of new conventional petrol provements as the United Kingdom moves toward a more
and diesel cars and vans by 2040. in the same press release, service-driven economy.
the transport secretary goes on to say that the expectation is two degrees sees shared autonomous vehicle use growing
for nearly every car and van on U.K. roads to be zero emis- from today’s 0% to 50% by 2050. With the adoption of auton-
sion by 2050. omous vehicles, an individual’s overall mileage per car will be
these announcements have been supported by £80 million reduced as sharing will become commonplace in this greener
to improve charging infrastructure for eVs and £246 million society; in other FES scenarios, mileage will increase. these
to assist businesses in designing, developing, and manufactur- vehicles will predominately charge at central depots rather
ing batteries for the electrification of vehicles. additionally, than at domestic sites. this also means fewer cars on the road,
the 2017 autumn budget included £400 million for electric potentially 25 million by 2050, which is about 5 million fewer
car charging points across the nation, an extra £100 million than today, in spite of a growing population. importantly,
for plug-in car grants to help people to buy eVs, and £40 mil- the energy requirements of PeVs will improve from today’s
lion for eV charging research and development. added to this 0.266 kWh/mi to 0.123 kWh/mi by 2050, brought about by
is the ever-declining cost of batteries, which constitutes the more efficient engines and a move toward smaller vehicles, as
most expensive element of an eV. the combination of gov- opposed to larger SUV types.
ernment policy and falling prices should enable accelerated no one means of charging will be universal in any of the
uptake of eVs. scenarios; there will be a mosaic of solutions, with adop-
tion depending on the car owner’s circumstances. there will
What Needs to Happen? be a place for residential trickle chargers and, at the other
For low-carbon transport to play its required role while lim- extreme, a national network of superfast chargers (250 kW
iting the disruptive impact it would have on the network, a or more). nondomestic charging will be important. in 2011,
number of steps are required. these steps were assumed for a nationwide survey by the U.K. department of transport
two degrees prior to the government’s announcements, and, established that only 57% of households had access to off-
fortunately, they are aligned. street parking. assuming that 43% do not, this will equate to
Currently, PheVs represent 83% of the eVs on the road, 8.6 million vehicles by 2050.
and this will steadily decline to 43% in 2030, and by 2040 it will as discussed earlier, time-of-use tariffs will be in opera-
be 9%. in 2040, new PheVs sales will be phased out. PheVs tion, and 85% of car owners will take advantage of the dif-
currently use either diesel or petrol engines and, therefore, ferential charging rates. as a result, they will move away
contribute to carbon emissions. as a result, in our two degrees from charging at peak times. two degrees assumes that
scenario, all eV sales will be PeVs by 2045. smart chargers will become standard post-2020, partly as a

july/august 2018 ieee power & energy magazine 55

 result of a successful smart meter rollout and partly as a nec-
essary means to keep peak demand down. a smart charger,
at its most basic level, is a device that is programmable and
has a two-way communication facility.
it is envisaged that these devices will be able to carry out
functionality such as
✔ responding to dynamic pricing signals
✔ integration into a domestic electricity niche, being
able to “understand” all the demands and generation
of a household, rank the eV’s capabilities in an order
of merit, and respond appropriately to maximize the
eV’s potential
✔ integration, both as an input and an output, into the
wider electricity ecology, whether through a local, re-
gional, or virtual community
✔ predictive learning of what demand is required, there-
by ensuring the eV battery is used in the most optimal
manner for the consumer
✔ over r ide facilities to let consumers opt out of
functionalities.
at a bare minimum, we firmly believe that smart char-
gers must be able to facilitate and enable consumers to move
away from peak-time charging. if this is not achieved, then
the stresses on the electricity networks will be significant,
incurring substantial network reinforcement and incremen-
tal generation capacity. in two degrees, smart charging
mitigates these stresses, resulting in only about 6 gW of
incremental peak demand by 2030, an additional 10% above
today’s peak.
What Could This Look Like?
Figure 7 illustrates the annual and peak demand required
by eVs (and heat pumps) from today to 2050. in 2050, with
about 25 million eVs on the road and assuming an average
power rating of a modest 30 kW, there is 750 gW of poten-
tial capacity. if only a fraction of this could be harnessed as
a balancing resource, the eV fleet could contribute signifi-
cantly toward managing a system dominated by large-scale
intermittent renewable generation.
Using eVs for more than just powering a vehicle is begin-
ning to be explored. in denmark, a vehicle-to-grid (V2g)
trial has already been set up, generating up to €1,400 per car
per year by entering the ancillary services market. V2g is a
rapidly developing area, and there are a number of research
projects being conducted on future opportunities. innovation
United Kingdom has recently awarded almost £6 million to
the “e4Future” V2g demonstration project. this project has
been proposed by a consortium of seven interested parties,
including national grid, with nissan as the project lead. it
aims to demonstrate the possibilities that can be accrued,
as well technical characteristics and power system impacts,
from 1,000 V2g installations.
Less well developed is the concept of aggregating eVs
so that they could act as a virtual power station. this would
mean that, rather than offering only ancillary services, they
56 ieee power & energy magazine
would be able to offer balancing services, particularly at
peak demand and times of low renewable generation.
Some level of distribution reinforcement will likely be
required. recent pilot projects such as my electric avenue
have identified potential issues with scaling up charging at
the distribution level. in one example, voltage issues were
identified when five 3.5-kW chargers connected to a network
cluster (with 134 dwellings) were charging at the same time.
the project concluded that across Britain, 32% of low-volt-
age circuits (312,000) will require reinforcing when 40–70%
of customers have eVs using 3.5-kW chargers.
Summary
across the two degrees scenario, significant interventions
will be required to transform not only the power sector but
also adoption trends in both the transport and building mar-
kets. today, 30% of all natural gas consumed in the United
Kingdom is for power generation. the decarbonization of
power is significantly advanced, and a number of renewable
energy milestones have been reached.
heating is the least developed in its progress toward decar-
bonization. Currently, 80% of U.K. homes rely on gas for
heating, and this figure has changed little in recent years. it
accounts for about one-third of natural gas usage in the United
Kingdom. natural gas is still the preferred choice of consum-
ers. however, for the 2050 target to be met, a majority of heat-
ing must be electrified and the housing stock will require sig-
nificant (30%) thermal efficiency improvement.
the electrification of transport is still in its infancy,
although the price of batteries has been dropping and will
act as a spur. the recent government ambition to end the sale
of conventional petrol and diesel cars and vans by 2040 will
focus minds on the need for accelerated vehicle adoption and
required infrastructure development.
Network and Operability Considerations
the FES analysis does not look at network requirements for
its scenarios—it takes an unconstrained view. however, if
two degrees is to become a reality, there are a number of
challenges to be addressed.
✔ renewable generation is mostly intermittent, so alter-
native sources of generation may need to be in place.
What is the most cost-effective means to achieve this?
✔ Some renewable sources do not produce reactive
power, so compensation must be available. What will
these look like, and where will they be situated?
✔ there is a geographical shift in the sources of genera-
tion and demand that will require network investment.
how can this be done, and how can we mitigate the
risk of creating stranded assets?
✔ Peak loads could increase as a result of the new loads.
how can the adoption and use of smart appliances
be normalized?
✔ Will the current distribution system be fit for purpose in a
world where eVs and electrified heating are the norm?
july/august 2018
 12
Current GB Peak
Demand Is 61 GW
10
Peak Demand (GW)

4
Engaged Consumers Charge Their
Lower-Demanding EVs at Off-Peak Time
2
as Do Shared Autonomous Vehicles
(85% of Consumers Are Engaged)
0
2015 2020 2025 2030 2035 2040 2045 2050
70 More-Efficient ASHPs Replace
Current GB Annual Hybrids, but Less Heat Required
Demand Is 306 TWh Due to Thermal Efficiency of
60 Houses

50
Annual Demand (TWh)

Rise of Hybrid HPs

Thermal Efficiency of
40 Houses Improves

30

Increased Take Up of ASHP

20 Slower Rate of Increase as PEVs
Replace the Less-Efficient PHEVs

10
All Cars Sold Are PEVs Post 2040
0
2015 2020 2025 2030 2035 2040 2045 2050

EVs Heat Pumps

figure 7. Additional annual and peak demands from EVs and heat pumps.

given the sizable role of electricity in the two degrees scenario,
these and many more questions will need to be addressed
in the future as society moves toward a less carbon-inten-
sive future.
For Further Reading
national grid. (July, 2017). Future energy scenarios 2017.
[online]. available: http://fes.nationalgrid.com/fes-document/
national grid. (July, 2017). Future energy scenarios mod-
eling methods. [online]. available: http://fes.nationalgrid
.com/fes-document/
national grid. (July, 2017). Future energy scenarios chart
workbook. [online]. available: http://fes.nationalgrid.com/
fes-document/
northern gas networks. (July, 2016). h21 leeds city gate.
[online]. available: https://www.northerngasnetworks
.co.uk/wp-content/uploads/2017/04/h21-report-interactive-
PdF-July-2016.compressed.pdf
national grid. (Jan., 2018). network options assessment.
[online]. available: https://www.nationalgrid.com/uk/
publications/network-options-assessment-noa
Biographies
Russell Fowler is with national grid, United Kingdom.
Orlando Elmhirst is with national grid, United Kingdom.
Juliette Richards is with national grid, United Kingdom.
p&e

july/august 2018 ieee power & energy magazine 57

 On the Path to
Decarbonization
By Amber Mahone, Zachary Subin, Ren Orans,
Mackay Miller, Lauren Regan, Mike Calviou,
Marcelo Saenz, and Nelson Bacalao

C
Climate Change threatens our quality of life and the
habitability of planet earth for many species. the intergovernmental Panel
on Climate Change estimates that, to reduce the risk that global temperature
increases more than 2 °C above preindustrial levels, greenhouse gas (ghg)
emissions in developed countries must fall by approximately 80% below 1990
levels by 2050. a number of states and regions in the united states have com-
mitted to reducing long-term ghg emissions by this level, including Califor-
nia, new york, and new england. several studies have evaluated the u.s. and
state-specific long-term paths to achieving an 80% reduction in ghg emis-
sions by 2050. less attention, however, has been given to the midterm transi-
tion challenges associated with meeting these long-term climate goals.
this article compares the midterm (2030) electricity-sector challenges
and implications of two recent studies exploring decarbonization strategies
in California and the northeast united states. (in this article, the northeast
refers to new york, massachusetts, Connecticut, rhode island, Vermont, new

Digital Object Identifier 10.1109/MPE.2018.2822865

Date of publication: 18 June 2018

58 ieee power & energy magazine 1540-7977/18©2018IEEE july/august 2018

hampshire, and maine.) these decarbonization studies focus on achieving a 40% reduction in economy-wide ghg emissions
by 2030 (40 × 30), on the way to achieving an 80% reduction in ghgs by 2050 (80 × 50) relative to 1990 levels. although the
scope of ghg reduction goals in both regions is economy wide, the focus of this article is on impacts and implications for the
electricity sector, which will play a critical role in meeting these ghg targets.
in both regions, energy efficiency (ee) and renewable electricity (re) have been the focus of ghg reduction efforts to
date and already resulted in significant reductions relative to historic levels. however, ee and re on their own are insuf-
ficient to meet the 2030 and 2050 goals. significant reductions in transport and heat emissions will be necessary, with
electrification in transportation and fuel
switching for other end uses likely to play

Electrification and Renewables a pivotal role.

there are low-carbon alternatives

in California and the to electrification, including biofuels and

hydrogen fuel produced from low-carbon

Northeast United States electricity. likewise, alternatives to re

include nuclear power or carbon capture
and sequestration. these alternative tech-
nologies are not the focus of this article
but have been investigated in other deep
decarbonization scenario analyses. in general, in the 2030 time frame, these alternatives appear to face either significant supply
constraints (in the case of biofuels) or cost challenges, in the case of hydrogen, nuclear, and carbon capture and sequestration.
against this backdrop, both studies focused on the power system implications of widespread electrification, with the goal
of understanding whether power system constraints could limit achievement of the 2030 ghg targets. We evaluate 2030
power system constraints using five metrics:
✔ Incremental electric energy requirements (gigawatthours): total electric energy requirements, and, in particular,
the growth attributable to new transportation and building loads, provide insights into the bulk energy system’s ability
to meet total demand.
✔ Incremental peak demand requirements (gigawatts): Peak demand growth, a measure of the impact of new elec-
trification loads on maximum capacity requirements for resource adequacy, also provides a useful proxy for potential
impacts on system reliability metrics. it is important to note that only average temperature conditions are tested. the
peak demand impacts of more extreme cold temperatures were not tested in this analysis and warrant further research.
✔ Curtailment (percent of RE): Curtailment occurs when electricity supply exceeds demand in any given hour, either
due to emergency conditions or because variable re (Vre) cannot be cost-effectively integrated onto the grid, and
renewable output must be reduced. quantifying the evolution of curtailment over time provides a useful proxy for
operational and economic challenges of grid integration.
✔ GHG emissions [metric tons carbon dioxide (CO 2 )]: to achieve ghg reduction goals, electrification must contrib-
ute to a net reduction in economy-wide ghg emissions, even if electrification causes electricity sector emissions to
increase on the margin.
✔ Electricity cost (average retail rate): electricity costs indicate whether the low-carbon energy system pathway is likely
to be sustainable from an electricity consumer perspective. Customers that adopt electric vehicles (eVs) and electrifica-
tion in their buildings could see total energy bills that decrease, due to avoided fossil fuel costs. however, if electricity
rates increase significantly for nonadopting customers, this could cause a backlash against an electrification strategy.
each metric provides insights into how to manage the electricity sector transformation on the road to a low-
carbon future. While our studies used different research methods and models to test the impacts
of electrification, we reach similar conclusions based on these five metrics. the sys-
temwide challenges of a highly electrified, highly renewable future appear
to be manageable in the 2030 time frame. if managed appropriately,
the electrification of transportation and building loads may

©istockphoto.com/JamesGdesiGn

july/august 2018 ieee power & energy magazine 59


450
Fossil Fuel CO2 Emissions
400
(Million Metric Tons)
350
300 Historical GHG
Emissions 2030 GHG
250
Target
200
150
100 California
Northeast (NE + NY) 2050 GHG
50
Target
0 are modeled, achieving 50 mmtCo2 by 2050 (an 87% reduc-
1990 2000 2010 2020 2030 2040 2050
Year
figure 1. In-state fossil-fuel CO2 emissions in California
and the U.S. Northeast, historical trends, and future goals.
help to both lower retail electricity costs and to mitigate grid
integration challenges relative to a low electrification future.
We also identify three key challenges warranting further
research: 1) accelerating customer adoption of building and
transportation electrification, which is lagging the pace needed
to achieve stated targets; 2) reforming electricity market de-
sign, especially to ensure long-term resource adequacy; and
3) managing distribution-level grid impacts, including larger
instantaneous loads, multidirectional flows, and increased
voltage fluctuations.
this article first describes the sectoral emissions land-
scape in both regions. then we describe the study methods
for both areas and results for California and the northeast
region, respectively. finally, we conclude with a discussion
of findings and areas for further research.
A Comparison of U.S. Northeast
and California Emissions
ghg emissions from fossil fuel combustion in the seven-state
u.s. northeast region and California are of a similar magni-
tude. according to the u.s. energy information administra-
tion (eia), 1990 Co2 emissions from fossil fuel combustion
(excluding imported electricity) were 382 million metric
tons Co2 (mmtCo2) in the northeast and 364 mmtCo2
in California. for consistency across states and regions, emis-
sions data are based on the u.s. eia’s state carbon dioxide
emissions data, which excludes emissions associated with
imported electricity as well as non-fossil fuel combustion Co2
and other ghgs.
fossil fuel emissions in both regions peaked around 2004.
in the northeast, by 2015, annual Co2 emissions had fallen
by approximately 16% from 1990 levels, driven mainly by
a shift away from coal- and petroleum-fueled electricity
to natural gas-fired generation. reduced production from
heavy, energy-intensive industries and ee investment played
a role as well.
California’s fossil fuel emissions have fallen less dramati-
cally and are now approximately equal to 1990 levels. the state’s
60 ieee power & energy magazine
emissions reductions have not been as rapid as the northeast’s
over the past decade in part because it relied less to begin with
on coal- and petroleum-fueled generation and, as a result, has not
benefited as much from the trend toward natural gas generation.
the 80 × 50 target implies an annual 2050 fossil fuel
emissions budget of 77 mmtCo2 in the northeast and 73
mmtCo2 in California, assuming commensurate reduc-
tions in non-fossil fuel ghg emissions (such as methane and
fluorinated gases) are achieved (figure 1). in the California
analysis, deeper than 80% reductions in fossil fuel emissions
tion from 1990 levels). deeper reductions in the energy sector
are modeled due to anticipated limitations on the mitigation
potential for methane and other agricultural ghg emissions.
meeting these goals will require a step change in the
pace of emissions reductions in both regions. in California,
the pace of ghg reductions needed to meet the 2030 ghg
goal is even faster than the already ambitious rate of change
required in the northeast, requiring a more ambitious suite
of emissions reduction measures.
Emissions by Sector and the Need for
Transportation and Building Electrification
in both the northeast and California, the electric power sec-
tor is broadly on track for meeting 80 × 50 goals, but the
transport and heat sectors are not. transportation emissions
are the largest single component in both regions, exceeding
42% of emissions in both new york and new england and
representing 37% of California’s total emissions. moreover,
transportation emissions are higher today than in 1990 in
both regions.
in the northeast, buildings are the second largest source
of direct ghg emissions, representing over 30% of total
ghg emissions. ee programs and fuel conversions from oil
to natural gas have reduced heating ghg emissions from
buildings by 12% from 1990 levels. still, delivered fuels
(mainly oil) retain a 30% market share. in contrast, direct
ghg emissions from buildings in California represent only
about 9% of total emissions, due to the milder climate and
higher reliance (over 90%) on natural gas heat. figure 2 com-
pares in-state Co2 emissions from fossil fuel combustion, by
sector, for both regions.
given this current landscape, transportation electrification
is widely viewed as a central decarbonization strategy in both
regions. in the northeast, support for building electrification
is growing due to the region’s high reliance on heating oil
and higher share of emissions from buildings. in California,
building electrification has not been policy priority to date, as
buildings represent a much smaller share of total emissions
and because the prevalence of low-cost natural gas for heating
presents greater economic hurdles to switching.
Methods
this section describes the analytical methods and scenarios
applied in both regions.
july/august 2018
 In both the Northeast and California, the electric power sector
is broadly on track for meeting 80 × 50 goals, but the
transport and heat sectors are not.
California Methodology
to simulate the impact of ghg reduction strategies on Cali-
fornia’s electric sector and energy infrastructure needs,
energy and environmental economics, inc. (e3) used the
California PathWays model. the model is an economy-
wide, scenario-based energy technology model representing
California’s energy system. this model incorporates infor-
mation about the lifetime and rate of turnover of ghg-emit-
ting equipment and technologies (e.g., buildings, vehicles,
water heaters, and light bulbs), allowing users to specify the
rates of sales of new equipment, the re buildout, and other
electric sector assumptions, as well as biofuel use, other ee
improvements, and changes in noncombustion ghg emis-
sions. the model calculates total economy-wide ghg emis-
sions, energy supply and demand, incremental capital costs
associated with the sale of new technology and equipment,
and fuel costs associated with each scenario.
using this tool, we estimate the hourly electricity demand for
each year between 2017 and 2050, reflecting changes over time
to end-use energy demands due to economic and population
growth, ee, and electrification. dispatchable electric generation
is dispatched against the hourly electric loads net of renewable
generation, allowing the model to calculate hours of renewable
overgeneration (i.e., when renewable supply exceeds demand in
any given hour). renewable integration solutions, such as flex-
ible eV charging, energy storage, and flexible loads in build-
ings, are added to the scenario to reduce renewable curtailment.
California’s electric system is modeled as a single zone, with
import and export constraints imposed at the borders.
the annual electricity demands, broken out by sector and
major end-use category [e.g., light-duty vehicle (ldV) elec-
trification and building electrification] were then fed into a
second electric-sector-only model as a check on the electricity
dispatch results and to evaluate optimal electric sector capacity
expansion and renewable integration solutions. this electric-
sector model, resolVe, minimizes total electric sector capi-
tal expenditures and fuel costs in California over the planning
horizon, subject to an electric sector ghg constraint. the elec-
tric sector ghg constraint is derived from the economy-wide
pathways scenario, creating a set of assumptions for the elec-
tric sector that are internally consistent with meeting the state’s
economy-wide ghg reduction targets in 2030 and 2050.
Northeast Methodology
the northeast study, carried out by national grid and sie-
mens Power technology international, used iterative auro-
july/august 2018
raXmP and gas Pipeline Competition model (gPCm) simu-
lations to simulate investment and operational impacts across
the region. auroraXmP is an electricity dispatch and capac-
ity expansion model of the electric grid. gPCm is a model
that incorporates natural gas supply, demand, and infrastruc-
ture inputs, including pipelines, to solve for expected prices
and flows throughout north america.
electrification load forecasts were developed exogenously
from the simulation based on projections of end-use perfor-
mance and stock turnover of key technologies. the study
team created hourly, temperature-dependent, heating load
shapes based on the most up-to-date available data, includ-
ing a 2016 massachusetts and rhode island study that gen-
erated field-validated, air-source heat pump operating data.
hourly ldV electrification load shapes were likewise based
on field-validated data from the u.s. department of energy
“eV Project” from 2011 to 2013.
eVs and electric heating equipment adoption trends, and
the resulting electric load profiles, were created for simu-
lation in auroraXmP, which performed capacity expansion
and hourly, zonal economic dispatch spanning the new
york independent system operator (nyiso) and iso new
england (iso-ne) footprints from 2017 to 2030, while
maintaining minimum (15%) reserve margins. the study
also modeled monthly natural gas fuel price dynamics based
400
(Million Metric Tons CO2)
GHG Emissions
300
200
100
0
1990 2015 1990 2015
CA NE
Residential Commercial
Industrial Transportation
Electric Power—In State
figure 2. In-state fossil-fuel CO2 emissions in California
and the U.S. Northeast by sector, 1990 and 2015. (Data
source: U.S. EIA State Carbon Dioxide Emissions Data,
released October 2017.)
ieee power & energy magazine 61
 on existing and announced pipeline infrastructure. iterations
between auroraXmP and gPCm ensured that gas prices
and power demand were in sync.
Scenarios and Results
in both studies, scenarios were developed to meet the 40 ×
30 economy-wide ghg reduction goals, while the Califor-
nia study also extended these scenarios further to achieve
the 80 × 50 ghg reduction goal.
California Study Scenarios
in the California study, ten mitigation scenarios were tested
in the Pathways model; each mitigation scenario was con-
strained to achieve both the 40 × 30 and 80 × 50 goals. in
contrast, the reference scenario reflects a business-as-
usual pathway with policies frozen at pre-2015 levels. the
reference scenario includes a 33% renewable portfolio stan-
dard (rPs) through 2030, historical levels for ee savings
goals, modest levels of transportation electrification, and no
new building electrification.
the high-electrification scenario is one of these ten
mitigation scenarios and includes a doubling of ee above
historical goals, additional high levels of industrial effi-
ciency, and an increase in renewable generation to about
a 70% rPs by 2030 (equivalent to approximately 60%
renewable generation, excluding large hydroelectric gen-
eration). the high-electrification scenario also includes
the increased electrification of buildings, cars, trucks,
buses and off-road equipment, as well as improvements in
the fuel economy of vehicles and reductions in per capita
vehicle miles traveled.
500
450
400
350
300
TWh
250
200
150
100
50
0 the impacts of electric ee. as a result, by 2050, total electric-
2015 2020 2025 2030 2035 2040 2045 2050
Hydrogen Production
Other Transportation Electrification
Light-Duty EVs
Incremental Building Electrification
Buildings
Industrial
Agriculture and Other
figure 3. California electric loads by sector and selected
end use in the high-electrification scenario (TWh, 2015–2050).
62 ieee power & energy magazine
in the high-electrification scenario, 60% of new sales (and
20% of total stock) of ldVs are assumed to be zero-emission
vehicles (ZeV) by 2030. this translates to 5 million light-
duty plug-in hybrid eVs (PheVs) and all-eVs, with an addi-
tional 800,000 ldV fuel-cell vehicles by 2030. this scenario
assumes 38,000 battery electric (3% of total) medium-duty
trucks. in addition, 50% of new sales (approximately 20%
of total stock) of water heaters and heating, ventilation, and
air-conditioning (hVaC) systems are assumed to be high-
efficiency electric heat pumps by 2030. By 2040, 100% of new
sales of ldVs and building hVaC systems and water heaters
are assumed to be electric.
scenarios that excluded building electrification were
also tested. these scenarios achieved the state’s 2030 and
2050 ghg reduction goals only when relatively high levels
of out-of-state biomethane, hydrogen, or synthetic carbon-
neutral methane were available. using current cost assump-
tions, these electrification alternatives are expected to have
a higher total societal cost than the building electrification
scenario. however, future costs for hydrogen, biomethane,
and building retrofit costs to convert to electric end uses are
not well understood. as a result, these scenario cost projec-
tions are highly uncertain.
in contrast, it does not appear to be feasible to design a
scenario that meets California’s ghg reduction goals with-
out nearly complete adoption of zero-emission on-road cars
and trucks as well as significant electrification of other forms
of transportation. While biofuels represent an alternative
option to electrification in the transportation sector, the sup-
ply of sustainable biofuels does not appear to be sufficient to
replace a majority of transportation fuels.
California Study Results
Incremental Electric Energy
and Peak Demand Requirements
in California’s high-electrification scenario, total electricity
demand remains relatively flat between present day and 2030
due to the assumed doubling of building ee savings by 2030,
offsetting the impacts of transport and building electrifica-
tion. after 2030, electricity demand increases dramatically as
higher levels of electrification and new electricity demands for
hydrogen production for fuel-cell vehicles begin to outweigh
ity demand in the high-electrification scenario is 60% higher
than current levels (figure 3). in the long term, the largest
new sources of electricity demand are from the transportation
sector, including electrified light-duty eVs, electrified trucks,
freight, and other transportation vehicles, which represent 20%
of total electricity demand in 2050.
another new source of electricity demand in this sce-
nario is grid-connected electrolysis for hydrogen production.
hydrogen produced using low-carbon electricity reduces the
need for renewable curtailment and/or energy storage and
represents a low-carbon energy carrier to help decarbonize
july/august 2018
sectors of the economy that are otherwise difficult to elec-
trify, such as heavy-duty trucks and some industrial applica-
tions. in the high-electrification scenario, new electric loads
from hydrogen production represent 2% of total electricity
loads in 2030, growing to 10% of total electric demand by
2050. similar dynamics could occur in the northeast, though
that study did not explore hydrogen pathways.
Electric Generation
California’s electric generation mix is currently about 50%
natural gas including imports and combined heat and power,
40% zero-carbon renewable generation [including utility-
scale renewables, rooftop solar photovoltaic (PV) systems,
and hydroelectric generation], with the remainder coming
from nuclear generation and out-of-state coal. By 2030, in
the high-electrification scenario, gas generation is cut in half,
falling to less than 25% of total generation. Coal imports
disappear while zero-carbon generation resources represent
over 70% of the generation mix in 2030. By 2050, the genera-
which could result in either higher or lower peak demand
impacts, depending on the timing of the vehicle charging.
Curtailment
in California, curtailment increases from less than 1% of
annual Vre generation in the 2030 reference case to less than
4% of annual Vre generation in the high-electrification sce-
nario, even with over 60% of electricity sales coming from
renewable resources. the electrification of buildings in Cali-
fornia has little effect on renewable curtailment through 2030
because electric loads from space heating and water heating
largely occur during the mornings, evenings, and in the win-
ter, when solar generation is less plentiful. in our analysis,
600
500

Generation (TWh)
tion mix is 95% zero carbon (figure 4). 400
the renewable capacity build-out required to achieve these
300
levels of zero-carbon generation is impressive, nearly tripling in
fewer than 15 years. in the high-electrification scenario, utility- 200
scale solar PVs increase from fewer than 10 gW today to more
than 30 gW in 2030, accompanied by an expansion in rooftop 100
solar PVs from 5 gW today to 18 gW by 2030. in-state wind
resources stay relatively constant between 8 and 9 gW through 0
2015
2018
2021
2024
2027
2030
2033
2036
2039
2042
2045
2048
2030, while out-of-state wind development is only assumed
to occur after 2030 due to challenges with building interstate
CHP Nuclear Natural Gas Imports Hydro
transmission lines to deliver the wind to loads.
Biomass Geothermal Solar Rooftop PV Wind

Peak Demand and Capacity Needs

in the high-electrification scenario, the 2030 system peak figure 4. California electricity generation mix and percent
demand is only slightly higher than today’s level and actu- zero-carbon electricity in the high-electrification scenario
ally lower in 2030 compared to the reference scenario, due to (2015–2050).
high levels of ee and flexible loads. By 2050, even with total
electricity loads doubled, the system peak demand is barely
higher than in the reference scenario due to the aggressive
System Peak Demand (GW)

100
implementation of energy efficiency, controlled eV charg-
ing, and flexible loads in buildings. Without the benefit of 90
higher ee (reference efficiency) or flexible loads in build- 80
ings or controlled charging of eVs (no flexible loads), sys-
tem peak demands in the high-electrification scenario could 70
be 10–20% higher than in the reference scenario by 2050 60
(figure 5). California’s new renewable and energy-stor-
50
age capacity, in combination with existing levels of gas and 2015 2020 2025 2030 2035 2040 2045 2050
hydroelectric capacity, is expected to be sufficient to meet
Reference
the state’s 2030 planning reserve margin, although more High-Electrification Scenario
detailed studies are needed. High-Electrification Scenario
this finding emphasizes the importance of flexible build- with Reference Efficiency
ing loads and controlled eV charging for managing future High-Electrification Scenario
system peak demands. these scenarios assume widespread with No Flexible Loads
adoption of lower voltage (240 V) eV charging during day-
time periods. none of the scenarios evaluated the impact of figure 5. California system peak demand by scenario (in
dc fast charging (480 V) of eVs on system peak demand, gigawatts, net of behind-the-meter generation).

july/august 2018 ieee power & energy magazine 63

 new building electrification loads are modest through 2030,
and after 2030, out-of-state wind is assumed to be available to
provide a winter and nighttime source of Vre.
the flexible and controlled charging of eVs presents
an important opportunity to integrate higher levels of
solar generation onto the grid. access to workplace charg-
ing for eVs, incentives to charge eVs during the middle
of the day, plus time-of-use rates for building loads are
assumed to reduce renewable curtailment to below 4% of
total renewable generation hours in 2030 in the high-elec-
trification scenario.
GHG Emissions
California’s electricity sector ghg emissions decrease
from about 84 mmtCo2 today (including emissions from
imported electricity) to just over 30 mmtCo2 by 2030 in
the high-electrification scenario, a decrease of over 60%,
even as total electricity demand remains relatively flat.
meanwhile, ghg emissions across the economy as a whole
decline by just over 40% relative to today.
By 2050, electricity sector ghg emissions fall to around
8 mmtCo 2 , a reduction of over 90% from 1990 levels,
while total economy-wide ghg emissions fall 80% by 2050.
reducing emissions from the electric power sector by 90%,
through higher adoption of renewable generation and ee, is
expected to be both lower cost and more technically achiev-
able than achieving proportional reductions in other sectors.
Cost
California’s electricity rates are presently among some of the
highest in the country, at approximately us$0.16/kWh, while
the state’s average electricity bills remain lower than the na-
tional average, due to strong ee codes and programs and a
relatively mild climate. in the high-electrification scenario,
by 2030, on average, total electricity rates would be about
us$.03/kWh higher, or about 14% higher than the refer-
ence scenario in 2030, before inflation. of this cost increase,
approximately us$.01/kWh is expected in the reference sce-
nario, while us$.02/kWh is attributable to new renewables
and energy storage as well as higher transmission and dis-
tribution costs. electrification, if anything, appears likely to
slightly reduce the average cost of electricity by increasing
the load factor on existing transmission and distribution (the
load factor is the ratio of the average energy demand and
the peak demand). a higher load factor means that there is a
higher utilization rate of transmission and distribution assets,
allowing these fixed costs to be spread over more sales, thus
lowering the average rate. furthermore, higher electricity
bills due to electrification are, in part, offset by fossil fuels
savings on natural gas, gasoline, and diesel.
Northeast Study Scenarios
the northeast study developed and compared a high re
scenario to a 40 × 30 scenario. the main difference was that
the high re scenario assumed little electrification while the
64 ieee power & energy magazine
40 × 30 scenario also assumed dramatically reduced gaso-
line and heating oil consumption, predominantly through
electrification. fuel economy standards, existing ee targets,
and other baseline assumptions were held constant between
the scenarios. state-level re standards were enforced in
both scenarios, although both were allowed to exceed these
targets. the region achieves 35% renewable generation by
2030, excluding large hydroelectric generation (51% includ-
ing it). the scenarios aimed to test the operational and emis-
sions differences between a pathway relying mainly on re
deployment versus a ghg reduction pathway spanning the
electricity, transport, and building sectors.
for each scenario, the study developed estimates of 40 ×
30 compliant housing and light-duty truck fleet stock turnover.
Battery eVs (BeVs) and PheVs were both considered in the
stock turnover forecast to meet the 40 × 30 target. other forms
of transportation electrification were not included in the sce-
narios. all eV types were assumed to become more efficient
as new models become available, at a rate of 2.5% per year.
Vehicle miles traveled assumptions were held constant across
the scenarios. We exogenously developed a 40 × 30 compliant
home-heating stock turnover forecast, nearly eliminating oil-
based heat, primarily through electrification and secondarily
through conversion to natural gas.
reference load growth and ee assumptions were based
on iso-ne and nyiso forecasts, which only extended
through 2026, and were extrapolated to 2030 at straight-
line rates. in both scenarios, distributed generation was
forecast exogenously to comprise about 10% of all Vre ca-
pacity additions.
the 40 × 30 scenario enforced a steep climb to 50% elec-
trification of ldV miles traveled by 2030. this penetration
far exceeds adoption forecasts, requiring 100% BeV/PheV
market share of annual sales by 2028. the 40 × 30 scenario
assumed electrification of approximately 25% of all residential
space heat as well as an increase in oil-to-gas conversions.
these levels of fuel switching in the transport and heating
sectors were estimated to be necessary to meet the 40 × 30 ghg
goals, barring additional ghg mitigation from higher levels of
ee, re, or biofuels. in this sense, the northeast 40 × 30 scenario
represents an upper-end “stress case” on the amount of re addi-
tions and end-use electrification that could be required to meet
the region’s ghg reductions goals.
Northeast Study Results
Incremental Electric Energy
and Peak Demand Requirements
in the northeast, peak and annual load remain relatively
flat through approximately 2023, as ee and electrification
effectively balance each other, and then peak and annual
load begin to increase as the uptake of eVs and heat pumps
accelerates. By 2030, peak load is approximately 10% higher
in the 40 × 30 scenario than the high re scenario, while
annual load is roughly 13% higher (figure 6).
july/august 2018
 load factor increases moderately in the 40 × 30 scenario. driven by new electrification load. While still ending lower
load factor is a function of two competing trends: increased than 2017 levels, gas generation in the 40 × 30 scenario is
peak load from electrification, which reduces the load factor, about 20 tWh higher than the high re scenario.
versus significant off-peak load building through both eV the high re scenario added approximately 32 gW of
and heat pump load, which increases it. despite assuming renewable capacity by 2030, while the 40 × 30 scenario
modest levels of intelligent eV charging (approximately half added another 3 gW of capacity (35 gW total) to meet
of charging was assumed to be price responsive by 2030), incremental electrification load (figure 8). Both scenarios
the 40 × 30 scenario found eV adoption to improve load selected mainly wind and solar with marginal amounts of
factor overall. small hydro. offshore wind was selected only to achieve
the northeast has been a summer peaking system since compliance with policy mandates. notably, while the 40 ×
1989; however, higher transport and heat electrification was 30 simulation allowed new natural gas generation capacity
found to cause winter peak loads to begin to reconverge to be added to meet incremental electric demand, the model
toward summer peak by 2030. extending these trends to predominantly selected new re capacity in the outer years,
2050, a winter peaking system would become likely. the driven by highly competitive capital costs.
mitigation of winter peak loads would depend upon smart-
vehicle charging, accelerated ee investment, and flexible
electric and thermal loads.
350,000
Electric Generation
the northeast’s installed capacity currently totals approxi- 300,000
mately 67 gW, and the electricity supply mix is about 40% Annual Generation (GWh)
250,000
natural gas, 30% nuclear, 13% hydro, and 6% wind and solar
generation, with the remainder coming from oil and imports. 200,000
in both the 40 × 30 and high re scenarios, the renewable
150,000
generation share increases to approximately 51%, inclusive
of all hydro resources (figure 7). 100,000
gas generation in the 40 × 30 scenario falls during the
first half of the period, driven by Vre additions and in- 50,000
creased hydro imports enabled by new transmission, but -
begins to rebound modestly in the latter half of the period, 2016 2018 2020 2022 2024 2026 2028 2030
Offshore Wind Solar Coal
Onshore Wind Biomass Gas Other
340,000 70 Rooftop PV Oil Gas Combined Cycle
Hydro Nuclear
330,000 68

320,000 66 figure 7. Northeast electricity generation in the 40 × 30

Annual Energy (GWh)

Peak Power (GW)

scenario.
310,000 64

300,000 62

290,000 60
40
280,000 58 35 Offshore Wind
30 Onshore Wind
270,000 56 25 Solar
GW

20
260,000 54 15
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030

10
5
High RE Scenario - Annual 0
High RE

40 × 30

High RE

40 × 30

High RE

40 × 30

40 × 30 Scenario - Annual
High RE Scenario - Peak
40 × 30 Scenario - Peak
2020 2025 2030

figure 6. The Northeast’s annual and peak loads under

40 × 30 and high RE scenarios. figure 8. Northeast cumulative renewable additions.

july/august 2018 ieee power & energy magazine 65

 GHG Emissions
in the high re scenario, electric sector emissions fall roughly
40% below 2016 levels, from roughly 56 mmtCo2 to 33
mmtCo2 in 2030. this represents a nearly 70% reduction
below 1990 levels. in contrast, 40 × 30 scenario electricity
emissions fall less, only 29% below 2016 levels to 40 mmtCo2,
roughly 63% below 1990 levels.
trends in economy-wide emissions are different, however. in
the high re scenario, economy-wide emissions fall 30% below
1990 levels, from 382 to 266 mmtCo2. the 40 × 30 scenario
falls 40% below 1990 levels, to 229 mmtCo2, roughly 36 mmt
further than the high re scenario (figure 9). in the 40 × 30 sce-
nario, renewable generation, transport electrification, and build-
ing fuel substitution away from petroleum to natural gas and
electricity all support achievement of the 40 × 30 target.

Curtailment
450 in the 40 × 30 scenario, curtailment is estimated to
1990–2015
High RE Scenario reach 4% by 2030 but reaches 7% in the high re scenario,
400
40 × 30 Scenario despite similar levels of Vre penetration. the general find-
40% Below 1990 ing is that higher levels of Vre adoption in both regions
MMTCO2e

350
may be complemented by new electric loads, particularly if
300 there is some flexibility to eV charging. in both scenarios,
increased transmission would be required, although the
250 limited spatial resolution of the simulation does not provide
specific estimates.
200
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2018
2020
2022
2024
2026
2028
2030

Cost
northeast electricity rates today are currently higher than the
national average, although widespread ee investment means that
Figure 9. Northeast economy-wide emissions in the high total annual electricity costs per customer are near the national
RE and 40 × 30 scenarios. average. Both the 40 × 30 and high re scenarios were found
to raise electricity supply costs
by approximately $0.03/kWh, or
table 1. The 40 × 30 scenario assumptions for California approximately 20% above today’s
and the Northeast in 2030, compared to present day.
rates, before inflation. While the
Today 2030 40 × 30 scenario triggered higher
Category California Northeast California Northeast capacity additions, as these incre-
mental costs are spread over more
Solar, including behind the meter 13% <1% 38% 16% demand, the cost impacts were
(generation)
on par with the high re scenario.
Wind (generation) 9% 2% 11% 18% from a customer point of view,
Total RE generation, excluding 29% 5% 60% 35% the 40 × 30 cost increases would
large hydro (generation) be partially or wholly mitigated
Total RE generation, including 39% 21% 70% 51% by reduced expenditures on petro-
large hydro (generation) leum products.
Light-duty ZEV adoption 2% <2% 64% 100%
(annual sales) Regional Synthesis
ZEV penetration (light-duty fleet) <1% <2% 20% 50%
of Study Results
Key scenario assumptions are
ZEV penetration (medium- 0% 0% 3% 0% summarized in table 1, includ-
duty fleet)
ing a comparison between today
ZEV penetration (heavy- 0% 0% 4% 0% and 2030 for California and the
duty fleet)
northeast. By 2030, the Cali-
Other off-road and hydrogen 0% 0% 5% 0% fornia scenario includes more
production electric loads (load) renewable and large hydroelec-
Total transportation electrification 0% 0% 9% 8% tric generation than in the north-
loads (load) east scenario (70% versus 51%).
Electric heat penetration (heating 4% <2% 7% 25% the northeast 40 × 30 scenario
demand in all buildings) includes more aggressive light-
Total electric heat loads (load) 3% 2% 4% 7% duty eV adoption assumptions but
does not assume the electrification

66 ieee power & energy magazine july/august 2018

of medium or heavy-duty trucking, or other transportation addition of new grid-connected battery storage. Beyond 2030,
sectors, as the California scenario does. as a result, by 2030, the California study suggests that hydrogen production could
the total new transportation electrification loads in both sce- emerge as an important energy storage medium and could help
narios are comparable, representing between 8 and 9% of total reduce curtailment.
loads in California and the northeast. natural gas demand declines in both regions between
the northeast scenario assumes relatively higher levels today and 2030, although the decrease is more pronounced
of building electrification and lower ee improvements than in California than in the northeast due to the higher lev-
the California scenario. in addition, since building heating els of renewable generation and ee in the California sce-
demands are higher in the northeast than in California, nario. California’s demand for natural gas from electric
these electrification assumptions result in a higher share of generation decreases by nearly 40% over this period, while
total loads coming from electric heat. in the northeast sce- nonelectric gas demand declines by 15%. in the northeast,
nario, nearly 25% of building heat demand is met by heat electric generation gas demand decreases by 20%, while
pumps by 2030, compared to only a 7% increase in Califor- nonelectric gas demand declines by only 1%. since the
nia. in the northeast scenarios, total electric heating loads northeast has a higher reliance on higher emitting fuel oils
are nearly as large as transportation electrification loads by for building heating than California, significant ghg sav-
2030 (table 1). ings can be achieved by switching to both natural gas and
a comparison of the two studies’ results across the electricity for heat.
five metrics established at the outset of this article illustrates some
of the key differences and similarities between the decar- Summary
bonization strategies simulated in each region (table 2). the long-term decarbonization of our energy economies is
the California scenario simulates a more diversified approach expected to require not just high levels of ee and low-carbon
to decarbonization, with a heavy reliance on ee, renew- electricity but also significant transportation and end-use elec-
able energy, biofuels, electrification, and flexible loads. the trification by 2050. this article provides greater clarity on the
northeast scenario maintains steady growth in ee but accel- changes to the California and northeast energy systems on
erates re development and electrification to simulate how the that path to 2050. the study results suggest that while exist-
system would operate in a high-re, highly electrified future. ing energy-policy mandates achieve significant reductions in
despite the differences in sce-
nario design between the study
regions, it is useful to compare the table 2. The 40 × 30 scenario results for California
and the Northeast in 2030, compared to present day.
regional results. in the Califor-
nia scenario, electricity demand Today 2030
hardly increases by 2030 in the
high-electrification scenario, due Category California Northeast California Northeast
to high levels of ee offsetting new Total electricity demand (TWh) 288 291 295 334
electrification loads, while peak Electricity demand due to 1 1 30 50
demand slightly declines, again electrification (TWh)
due to high levels of ee, as well
Peak electricity demand (GW) 61 58 59 68
as additional demand response
(dr), flexible loads and daytime RE Curtailment (RE generation) 0% <1% 4% 5%
controlled charging of eVs. in Natural gas demand—electric 1179 863 669 703
contrast, in the northeast scenario, generation (bcf/yr)
less ee, dr, and flexible charging Natural gas demand— 1616 1336 1370 1319
is assumed, resulting in significant nonelectric generation (bcf/yr)
increases in both total electricity Carbon intensity of power 231 227 99 128
demand and peak demand. in both generation (g CO2/kWh)
studies, the load factor is observed
Carbon intensity of nonelectric 5307 5307 4563* 5307
to increase. generation natural gas (g CO2/
total levels of renewable energy therm)
curtailment are similar across both Total electricity sector GHG 72 56 32 40
regions, despite the higher preva- emissions (MMTCO2e)
lence of solar in California. in the
Average delivered cost of $0.16 $.16 $0.19 $.19
California scenario, overgeneration electricity (US$/kWh)
of solar is avoided, in part, through
*The California 2030 high-electrification scenario includes biomethane blended into
flexible loads and daytime charg- the natural gas pipeline, which reduces the carbon intensity of gas.
ing of eVs but also through the

july/august 2018 ieee power & energy magazine 67

 ghg emissions, in neither case are they sufficient to achieve
2030, let alone 2050, ghg mitigation goals.
market transformation will be pivotal. California and
northeast policy makers have recognized the importance
of transportation electrification in meeting regional ghg
reduction goals through participation in initiatives such
as the ZeV mandate. California and most northeast states
offer some form of incentive for eVs. most northeast states
also offer partial rebates for high-efficiency heat pumps,
but uptake remains limited, and incentivizing fuel switch-
ing is rarely allowed under ee program rules. in California,
the energy Commission has approved an optional building
code that would enable new all-electric buildings to qualify;
however, the state currently does not offer incentives for
fuel switching to high-efficiency electric end uses. in both
regions, technology cost reductions and vibrant customer
markets will be essential, and these can be supported by
stronger, more harmonized policy and regulatory support.
the coincidence of variable renewable generation and
new electrification loads will emerge as an increasingly
important planning metric. Peak demand impacts were
found to be highly dependent on the effectiveness of con-
trolled charging of eVs as well as flexible loads in build-
ings. in the northeast, due to its colder climate, building
electrification is expected to have a larger and earlier impact
on both peak and total demand than in California. in both
regions, building electrification was found to increase peak
demand, while widespread building electrification would
push both regions from historically summer-peaking sys-
tems to winter-peaking systems over the long run. in both
regions, a smarter grid and flexible resources, on both the
supply and demand side, are needed to integrate higher
levels of Vre. to the extent feasible, renewable resource
diversity can help to maintain coincidence between new
loads and new generation.
While both studies find that the bulk grid appears capable
of accommodating aggressive 2030 electrification levels, it
remains uncertain what distribution-system upgrades will be
needed to enable higher levels of electrification. it is likely
that a mix of local grid reinforcement, storage, and network
intelligence will be required to ensure reliability at low-
voltage levels to accommodate high levels of electrification.
these issues will require new planning processes and new
markets for procuring distributed solutions.
California and the northeast are poised to become the
learning laboratories for deep decarbonization and are com-
mitted to making this transformation as effective and efficient
as possible. these studies can help inform this transformation.
For Further Reading
s. yeh, C. yang, m. gibbs, d. roland-holst, J. greenblatt,
a. mahone, d. Wei, g. Brinkman, J. Cunningham, a. egg-
68 ieee power & energy magazine
ert, B. haley, e. hart, and J. Williams, “a modeling compar-
ison of deep greenhouse gas emissions reduction scenarios
by 2030 in California,” Energy Strategy Rev., vol. 13–14, pp.
169–180, oct. 2016.
J. Williams, B. haley, f. Kahrl, J. moore, a. Jones,
m. torn, and h. mcJeon. (2015). Pathways to deep decar-
bonization 2015 report sdsn: iddri, deep decarboniza-
tion pathways project (ddPP). [online]. available: http://
deepdecarbonization.org/wp-content/uploads/2016/03/
ddPP_2015_rePort.pdf
a. olson, a. mahone, a. mileva, g. de moor, and n.
schlag. (2016) Volume iV of the Caiso senate Bill 350 study.
[online]. available: https://www.caiso.com/documents/
sB350study-Volume4renewableenergyPort folioanalysis
.pdf
d. Korn, J. Walczyk, a. Jackson, a. machado, J. Kon-
goletos, and e. Pfann. (2016). 2016 ductless mini‐split heat
pump impact evaluation in massachusetts and rhode is-
land. [online]. available: http://ma-eeac.org/wordpress/wp-
content/uploads/ductless-mini-split-heat-Pump-impact-
evaluation.pdf
J. h. Williams, a. deBenedictis, r. ghanadan, a. ma-
hone, J. moore, W. r. morrow, iii, s. Price, and m. s. torn,
“the technology path to deep greenhouse gas emissions cuts
by 2050: the pivotal role of electricity,” Science, vol. 335,
no. 53, pp. 53–59, 2012.
a. mahone, e. hart, B. haley, J. Williams, s. Borge-
son, n. ryan, and s. Price. (2015). California pathways:
ghg scenario results. energy and environmental eco-
nomics, inc.. [online]. available: http://www.ethree.com/
wp-content/uploads/2017/02/e3_PathWays_ghg_
scenarios_updated_april2015.pdf
Biographies
Amber Mahone is with energy and environmental eco-
nomics (e3), san francisco, California.
Zachary Subin is with energy and environmental eco-
nomics (e3), san francisco, California.
Ren Orans is with energy and environmental econom-
ics (e3), san francisco, California.
Mackay Miller is with national grid u.s., Waltham,
massachusetts.
Lauren Regan is with national grid u.s., Waltham,
massachusetts.
Mike Calviou is with national grid u.s., Waltham, mas-
sachusetts.
Marcelo Saenz is with siemens energy Business advi-
sory, houston, texas.
Nelson Bacalao is with siemens Power technologies,
international, houston, texas.
p&e
july/august 2018
Heat
Electrification
©istockphoto.com/-strizh-

The Latest Research in Europe

H
Heat represents rougHly 50% of tHe euro-
pean union’s (eu) final energy demand (figure 1). But in
the past, heat was largely absent in the energy debate, given
the traditional focus on energy supply data, which only
shows heat fuels, mainly fossil fuels and a small fraction of
electricity. the potential synergies between heat and elec-
tricity in the transition to a clean energy system have been
recognized in the eu’s 2016 heat strategy. the increasingly
decarbonized electricity system powered by renewable elec-
tricity can provide clean heat supply, while the flexibility of
heat demand can support electricity peak management and
the integration of variable renewable energies.
Heat electrification represents a growth area for the elec-
tricity industry, but balancing challenges are a major con-
cern for short-term operations as well as long-term capacity
planning. the high seasonality of heat demand in many
countries, coupled with the increased supply variability due
to the parallel deployment of variable renewable generation,
is increasing the need for flexible operation and planning
to ensure supply and demand balance. electric heating, if
deployed in an uncoordinated manner, results in propor-
tionally stronger winter peak growth than average demand
growth and could further decrease asset utilization. How-
ever, heat can be stored more efficiently and economically
than electricity, which offers new opportunities for energy
system integration solutions. an intelligent or controlled
integration of electric heat can draw on the flexibility of the
heating sector (thermal storage and inertia) to facilitate the
integration of renewables and manage peak loads. the use of
information and communication technology (ICt) in electric

By Steve Heinen, Pierluigi Mancarella,

Ciara O’Dwyer, and Mark O’Malley
Digital Object Identifier 10.1109/MPE.2018.2822867
Date of publication: 18 June 2018

july/august 2018 1540-7977/18©2018IEEE ieee power & energy magazine 69

 and release the heat during the
Hot Water, 4% Other Heating, 1% day. Due to concerns over primary
Process Cooling,
energy consumption, in 2009, the
1% parliament ordered for the removal
Space Cooling, of all those heaters by 2019 or at
1% the end of their economic lifetime.
Space However, in the following years,
Non-H/C, Heating and Heating,
Cooling, the heaters were identified as a
50% 27% Process
50% Heating,
valuable source for flexibility to
16% support the integration of variable
renewable energy, and the parlia-
ment reversed its decision in 2013.
the french and german examples
illustrate the strong interaction
figure 1. Heat demand in the EU, 2015. [Source: Heat Roadmap Europe (2017),
Heating and Cooling—Facts and Figures.] of residential electric heating and
the power sector but also raise the
question on how integration of elec-
heaters could therefore provide the option to shift demand tric heating should be done to provide flexibility.
loads according to power system conditions, while also meet- this article will present latest research on heat electrifi-
ing the building occupant’s heat requirement. cation in europe, with a focus on heat demand, heat technol-
In france, where the majority of residential heating is ogies required to electrify, and system integration challenges
based on unmanaged electric resistance heating, the power and business models.
system is already experiencing the strong impact of heating.
electric resistance heating was widely deployed in the 1980s Heat Demand
in france to guarantee demand when the country built its Half of the european union (eu28)’s heating demand today is
nuclear generation fleet. today, the temperature sensitivity of used for space heating (figure 1). the largest heating demand
the french power demand (2,300 MW/°C) accounts for nearly in the residential and commercial sectors is space heating,
50% of the total temperature sensitivity in the eu. It is a major and in the industrial sector it is process heating (figure 2).
driver for extreme peak loads and security of supply. some Heat is not a homogeneous product like electricity and needs
demand-side management programs are being carried out as be differentiated based on the delivery temperature. Indoor
ad hoc measures to improve flexibility. spaces can be heated using low-grade heat (30–100 °C) that
In germany, electric overnight storage heaters were can be equally well produced from electricity and fossil fuels
quite popular in the past. these heaters use electricity dur- using a variety of mature conversion technologies. Industrial
ing night, store heat in materials with high thermal capacity, processes require high-grade heat with temperatures between
100 and 500 °C (30% of industrial heat demand) and higher
(42% of industrial heat demand) for applications such as iron
smelting. Conversion technologies for high-process heat are
3,500 under development and not as commercially mature or uneco-
Final Energy Demand (TWh)

3,000 nomical depending on the application.

the potential of large-scale electrification of heat in the
2,500
next decades is mainly in low-grade heat. this article will
2,000 focus on residential applications due to the large heat demand
1,500 of the sector as well as data availability, but the findings can
also be applied to the commercial sector.
1,000
In the residential sector, space and water heating demand
500 represents roughly 80% of final energy use in europe and
0 60% in the united states. If heat is electrified, electricity
Industry Residential Services demand for it could increase considerably (also depending
Other Heating Hot Water on the parallel deployment of building efficiency measures
Process Cooling Process Heating in new and existing buildings). the potential impacts of heat
Space Cooling Space Heating electrification on the electricity system will also depend on
the flexibility of the heat system to decouple electricity use
figure 2. The final energy demand in industrial, residen- and heat demand by storing heat. as described in the ger-
tial, and commercial sectors, 2015. [Source: Heat Roadmap man example, this could also benefit the integration of vari-
Europe (2017), Heating and Cooling—Facts and Figures.] able renewable energy.

70 ieee power & energy magazine july/august 2018

 residential heating demand has strong seasonal varia-
tions. While demand for domestic hot water remains through- 350
out the year, space-heating demand falls away in the summer Heat
300 Electricity

Heat/Electricity (GW)
months (figure 3). space-heating demand is highest in the
250
winter months, and on the coldest days the opportunities to
provide power-system flexibility may be limited due to the 200
high energy demands and the reduced potential to shift the 150
timing of the demand. During the shoulder seasons of spring 100
and autumn, heating loads are reduced, and opportunities for
50
heat load shifting and storage in thermal stores are increased.
0

Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Determinants of Heat Demand
residential space-heating demand varies considerably on a
regional basis. While heating demand is heavily dependent figure 3. The hourly electricity demand and low-grade
on climatic conditions (heating degree days), other signifi- heat demand in the United Kingdom for 2012. [Source:
cant factors include typical dwelling size, household size DECC (2012). The Future of Heating: A Strategic Frame-
(number of occupants), average indoor temperatures, build- work for Low Carbon Heat in the UK. U.K. Department of
ing insulation levels, and the type and efficiency of adopted Energy and Climate Change.]
heating technologies. for future heating demand estimation,
these trends need to be examined.
Changes in population will have a large impact on the total
demand for heat in the residential sector. While different regions 1
are displaying opposing population trends, for the eu overall, 0.9
the population is slowly increasing. While more stringent build-
0.8
ing regulations are driving down demand for heat, increasing
Coincidence Factor

0.7
average dwelling sizes, declining average household sizes (i.e.,
fewer occupants per household), and behavioral changes with 0.6
increased requirements for thermal comfort (average indoor 0.5
temperature) are moving in the opposite direction. While build- 0.4
ing regulations still have an enormous impact on space heating 0.3
demand for new buildings that only use a fraction of the heating 0.2
demand of older dwellings, overall residential heating demand 0.1
will depend heavily on the uptake of retrofit activities. retrofits 0
are being widely promoted in the eu given that the majority 0 100 200 300 400
Number of Houses
of homes that will be present by 2050 have already been built.
(a)
aggregate water heating demand is influenced by pop-
ulation size. However, there are also strong regional varia-
tions in water heating demand per capita. typical end uses, 1
heating technology, and consumer behavior drive these dif- 0.9 Gas Boiler
Old Building
ferences. for example, frequent sauna bathing in finland 0.8
Modern Building
Coincidence Factor

drives a particularly high energy use for water heating 0.7

New Building
per person. 0.6
0.5
Diversity 0.4
there are important differences between individual and
0.3
aggregated demand profiles and their impact on electric-
0.2
ity system planning across different aggregation levels, from
0.1
local low-voltage (lV) feeders to national wholesale markets.
load aggregation diversifies demand, which means that the 0
0 100 200 300 400
peak of the aggregated demand (after-diversity maximum Number of Houses
demand) is much lower than the sum of individual load peaks (b)
due to the stochastic nature of demand. therefore, the coin-
cidence factor, which is the ratio of simultaneous maximum figure 4. The coincidence factor in a (a) typical and (b)
demand relative to the sum of individual maximum demands, peak U.K. winter day for semidetached houses of various
decreases with higher aggregation [figure 4(a)]. In practice, age with air source HPs (gas boiler case for comparison).

july/august 2018 ieee power & energy magazine 71

 this means that a peak demand increase will have stronger
impacts on local lV grids than high-voltage transmission
lines and wholesale markets. note that on the day when the
systemwide peak heat demand occurs, demand diversity will
be less pronounced than on a typical day because end users
will tend to use heat for longer times and more simultane-
ously [figure 4(b)].
Data Time Resolution
along with the aggregation level, the time resolution also
impacts the peak load level observed. at higher resolution
load data, the observed peak demand tends to be flatter as
the peak is smoothened over a wider time interval. the opti-
mal temporal granularity of heat data to capture peak demand
also depends on the level of aggregation. However, high-gran-
ularity heat data is often not available because many coun-
tries lack experience with electric heating. High-resolution
bottom-up models (figure 5) are able to represent the heat
demand while also capturing the probabilistic nature of cus-
tomer behavior (such as occupancy patterns) and building
characteristics. supported by statistical inputs such as time-
of-use surveys, heat demand has been modeled to study the
impact of heat electrification on electricity systems. Depend-
ing on the grid-level analyzed and the requirement to capture
diversified peak load, different time resolutions may be neces-
sary, but for most cases 5–10-min resolution data can capture
diversified demand and limit computational cost. High-qual-
ity demand data and models are essential building blocks to
properly assess the impact and flexibility of high heat electri-
fication on different levels of the electricity system.
Heat Technologies
the heating system characteristics and level of integration
determine the flexibility of residential heating systems.
Heater
electric heating systems are mostly based on heat pumps
(Hps) and resistance-based heaters. Hps are increasingly
promoted in low-carbon strategies due to their high efficien-
cies. they make use of the natural temperature difference
between a source (such as underground or ambient air) and
indoor in a condensation/evaporation cycle, which is why
it is often classified as renewable heat. the heat cycle only
requires electricity to run the compressor and other auxil-
iary equipment, therefore producing two to five units of heat
for air-source Hps (asHps) [and potentially more for ground
source Hps (gsHps)] for each unit of electricity consumed.
the higher efficiency of Hps compared to resistance heaters
(which create one unit of heat for each unit of electricity con-
sumed) results in lower electric loads, which minimizes gen-
eration requirements and peak load. It is important to note
that the Hp coefficient of performance (Cop) is dependent
on the temperature difference between source and sink. this
has two implications for electricity demand: 1) the Cop of
asHps is lower during cold temperature spells when heat
is most needed and 2) the Cop is higher if the heat delivery
temperature indoors is low, favoring the use of air-based dis-
tribution systems over hydronic systems. In europe, however
most houses use hydronic (water-based) distribution systems
because they are more compact. In modern low-temp heat-
ers or underfloor heaters, delivery temperature can be as
low as 35 °C.
Direct electrical heating solutions include radiators, fan
heaters, panel heaters, and electric storage heaters (discussed
in the next section). the main advantage of these heating
technologies is a low investment cost. However, while the
heaters are highly efficient at point of use (almost 100%
of the electricity is converted to heat), when you consider the
losses incurred in the electricity generation and transmission,
these heaters do not compare favorably in terms of primary
energy use and typical operating costs to, for example, an effi-
cient gas boiler. operating costs tend to be high compared to
other local heating solutions. another disadvantage of direct
electrical heating is that the electrical power demand is closely
coupled with the heating demand, with very limited flexibility
in power scheduling.

Thermal Storage
thermal storage in the building ena-
10 2,500
9 bles the optimization of the elec-
8 2,000 tricity consumption and charging
7 based on electricity market condi-
Load (kW)

Load (kW)

6 1,500
5
tions while still providing thermal
4 1,000 comfort to the user. temporal decou-
3 pling of electrical power demand
2 500
1 from the heating demand can be
0 0 achieved when electrical heating
with a sufficient storage capacity
0
0
0
0
0

1, 0
1, 0
1, 0
0

0
0
0
0
0

1, 0
1, 0
1, 0
0
18
36
54
72
90
08
26
44

18
36
54
72
90
08
26
44

is deployed.
(a) (b)
electrical space heaters with
internal storage have been widely
figure 5. (a) Individual and (b) aggregated electricity profiles for 1,000 customers used for decades in many regions
with electric heating and domestic hot water, over 24 h, with 1-min resolution. with temperate climates, including

72 ieee power & energy magazine july/august 2018

 Electricity is also widely used in the residential sector
for water heating, which is typically combined with
thermal storage (a water cylinder).
Ireland, the united Kingdom, germany, france, and new
Zealand. traditional nighttime storage heaters were used to
reduce systemwide peak loads and fill in the night “valley”
and allowed customers to use less-expensive off-peak elec-
tricity tariffs to meet their heating demand. storage heaters
contain a core of high thermal capacity bricks, heated
by a resistive heating element and surrounded by a highly
insulated enclosure. they are designed to be used on a 24-h
cycle (i.e., charged at night and heat released during the day),
although older storage heaters could not always keep suf-
ficient charge levels to maintain thermal comfort on colder
winter days. Modern storage heaters, through improved insu-
lation and controls (the release of the heat during a 24-h cycle
can be more accurately controlled with the use of a variable
speed circulation fan), offer a higher degree of comfort to end
users. Moreover, with adequate controls and communication,
smart electrical thermal storage (sets) can participate in
active demand-side management (DsM) and enhance power
system flexibility. the eu-funded Horizon 2020 realValue
project is trialing sets systems at a household level, provid-
ing a range of system services while maintaining the thermal
comfort of customers.
electricity is also widely used in the residential sector
for water heating, which is typically combined with thermal
storage (a water cylinder). as with space heating, this stor-
age capacity decouples the power demand from the heating
demand, and when equipped with communication and con-
trol architecture, electrical water heaters can also participate
in DsM.
the building envelope itself also provides thermal inertia
depending on the insulation level. In a well-insulated house,
electric load can be shifted (around 5–12 h), depending on
the building insulation level, while consumer comfort is still
met. preheating increases flexibility but typically increases
energy usage, depending on the insulation level. thermal
storage and building preheating enables large demand shift-
ing potential at very low cost, especially when compared to
electricity storage.
Hybridization
Hybrid heaters combine different heating appliances in one sys-
tem and can switch between those appliances during opera-
tion, providing a very flexible demand. possible configura-
tions include Hp-gas boiler (Hp-B) and gas boiler-resistance
heaters. Hp-B systems have been available commercially
for several years and were developed to lower investment
july/august 2018
compared to stand-alone Hp systems. However, hybrid heat-
ers could offer electricity system benefits if connected in a
smart manner that enables the electricity system to access
the flexibility of the gas system by switching from the Hp to
the gas boiler whenever the electricity system is under stress,
which can include an extended period over several days.
Hybrid gas boiler-resistance heater systems could also pro-
vide the option to use excess renewable electricity by switch-
ing from gas to electricity.
System Integration
Widespread electrical heating can have a significant impact
on the overall system energy demand as well as on the shape
of that demand. High shares of direct electrical space heating
increase the winter system peak demand. on the other hand,
when combined with thermal storage, the electrical demand
can be shifted to off-peak hours, with minimal impact on
the peak, and with the additional advantage of increasing
baseload plant utilization in the off-peak hours. However,
as shares of variable renewable energy increase and the net
load becomes more variable and harder to forecast, smarter
controls and two-way communication are required to fully
take advantage of the thermal energy storage capacity.
Coupling Small-Scale Electric Heating
and Thermal Storage
Cooptimizing heating (combined with storage) and power
system scheduling could reduce overall system costs and
increase shares of variable renewable generation that can
be integrated. as discussed earlier, sets for both space
and water heating (i.e., resistive electrical heating com-
bined with thermal storage) is currently being trialed in
the eu Horizon 2020 realValue project. a combination of
physical demonstrations in three winter peaking systems
(Ireland, germany, and latvia) along with detailed build-
ing and power system modeling will demonstrate how local
small-scale energy storage with advanced ICt, optimized
across the eu energy system, could bring benefits to all
market participants.
enormous benefits can be realized from the integration
of the power and heat energy systems. However, to ade-
quately assess the scale and value of the flexibility that can be
accessed from the heat sector, detailed modeling is required
to consider the impact of any load shifting and demand
response activities on the end users. Cooptimized building-
to-grid models have been developed and refined as part of
ieee power & energy magazine 73
 the realValue project and can be used to explore cost-opti-
mum scheduling solutions for the combined systems, while
also ensuring the thermal comfort of the end users. Interim
results for the Irish system indicate that significant system
cost savings can be achieved when sets is the chosen tech-
nology for electrically heated homes (approximately 7% of
homes for space heating and 17% for water heating), com-
pared to direct resistive heating. these savings are achieved
through load shifting alone with significant further cost sav-
ings achievable when ancillary services are also provided by
the devices. the results shown in figure 6 compare the base
case (i.e., inflexible heating load) to three different penetra-
tions of sets: 0, 50, and 100%. for sets-0 there are no
sets devices, but the thermal inertia of the building is uti-
lized to increase the flexibility of the heating load and modest
system cost reduction are achieved. Cost reductions increase
when the maximum indoor temperature limit is increased,
although energy consumption also increases [figure 6(b)],
as preheating results in additional heat losses to the sur-
roundings. as such, incentives and tariffs that incentivize
this behavior are essential. the advantage of a higher maxi-
mum indoor temperature is no longer seen for the sets100
case (all electrically heated homes assumed to have sets)
as the active thermal storage is more energy efficient than
the passive storage alternative. for the sets100 case, system
generation cost savings in excess of 1% (€10.1  million) are
achieved. figure 6(c) and (d) highlights some of the drivers of wind speeds in Ireland and great Britain. With increasingly
these cost reductions, with large reductions in both generator
start-up costs and wind curtailment seen as the penetration of
sets devices increases.
the aggregate load shape for electrical heating depends on
many factors, including technology type (direct/storage/Hp/
810 7.6
806 7.4
802
798 7.2
794 7.0
BC SETS 0 SETS 50 SETS 100 BC SETS 0 SETS 50 SETS 100
(a)
410
3.1
2.9
350
2.7
2.5 290
2.3 230
BC SETS 0 SETS 50 SETS 100 BC SETS 0 SETS 50 SETS 100
(c)
Max. Temp. (°C) 22 24
figure 6. The annual analysis for different storage capabilities on the Irish system.
(a) System generation cost (M€), (b) space heating use per house (MWh), (c) start-up
cost (M€), and (d) wind curtailment (GWh).
74 ieee power & energy magazine
hybrid), occupancy patterns, building construction, incentives
(e.g., time-of-use or real-time pricing) and control strategies
(e.g., direct load control by an aggregator). another signifi-
cant influencing factor is the participation in the provision
of additional services. for example, when electrical ther-
mal storage is scheduled and cooptimized within the power
system, charging will typically take place at times of rela-
tively low system demand (and low prices). However, when
the devices are also scheduled to provide ancillary services
such as reserve, the aggregate charging schedule for a fleet
of devices will be spread out, covering hours of the day with
relatively higher system demand in order to enable the provi-
sion of upward reserve.
Weather Dependence for an
Integrated Heat-Electricity System
Heat demand is strongly correlated to weather, in particular
temperature. Heat electrification increases the weather sen-
sitivity of electricity demand, while the parallel integration of
renewable energy increases the weather sensitivity of the sup-
ply side. Coincidental weather impact could therefore stress
system adequacy by decreasing supply at times of increasing
demand. for example, a climate phenomenon known as the
north atlantic oscillation, which relates to pressure differ-
ences between a location near Iceland and a location around the
azores, has been shown to influence both air temperature and
high wind penetration and heat electrification rates, such an
event could lead to high net load peaks. an hourly simulation
for eight different weather years for an Irish system with 25%
of buildings electrified and 40% wind generation over a year
illustrates how both ambient temperature and wind capacity
factors decrease as demand nears
peak demand (figure 7). Demand
response measures could be consid-
ered to ensure adequacy in such an
event. [note that the increase in wind
capacity factor toward the right in
figure 7(b) is inconclusive as the
averaging sample becomes smaller
and may only reflect outliers.]
(b)
Benefits of Hybrid
Heating Devices
and Thermal Storage
as described in the technology
section, hybrid heaters that are
fuel ed by both electricity and
(d) natural gas offer the flexibility
to switch between different fuels.
a range of commercial hybrid
Hp-B products are available, but
those are locally controlled based
on ambient outside temperature
only in a way that gas boilers boost
july/august 2018
the Hp during cold temperatures. similarly, many Hps are Thermal Inertia
de facto equipped with electric resistance heaters to provide the building envelope has a thermal mass that adds inertia
peak load. to a building’s thermodynamics since it takes time to heat up
on a planning scale, hybrid heaters could enable the mini- and cool down. this inherent thermal storage capacity could
mization of electric peaks and avoid generation expansion to be utilized to preheat the building and store heat at times of
meet electric heating demand. the heater itself can be opti- high energy supply and is available at no extra capital cost, as
mized to take advantage of com-
plementarities in the cost structure
of different technologies: Hps have
6 10 0.6 10
high capital costs but relatively low
operational costs, while gas boilers

Average Ambient Air Temperature (°C)

and resistance heaters have lower
8 8

Hours with Greater Demand (%)

Hours with Greater Demand (%)

capital costs and are more expen- 4
sive to run. a systemwide planning

Average Wind CF (/)

analysis (see Heinen et al., 2017 in 0.4
the “for further reading” section) 6 6
for a future Irish system with 40% 2
wind generation and 25% heat
electrification found that, despite 4 4
large natural gas savings, the sys- 0.3
temwide cost of Hp electrification 0
tends to outweigh the benefits due 2 2
to power-system investment and
operational cost (figure 8). Hybrid
solutions considerably improve –2 0 0.2 0
0.75 0.8 0.85 0.9 0.95 1 0.75 0.8 0.85 0.9 0.95 1
the cost-benefit analysis. Hybrid Normalized Demand (/) Normalized Demand (/)
Hp-resistance (Hp-erH) heaters (a) (b)
reduce capital costs for the cus-
tomer. Hybrid gas boiler-electric figure 7. The average ambient air temperature and wind capacity factor as a func-
resistance (B-r) heaters provide tion of normalized demand over eight years of data. (a) Peak demand and tempera-
an efficient solution to maximize ture and (b) peak demand and wind. (Source: Heinen et al., 2017.)
wind integration, but the majority
of the heating is done using natural
gas, and this solution also favors
low marginal cost coal generation, 500
therefore potentially increasing 450
the carbon emissions compared 400
to pure gas boilers. Hybrid Hp-B 350
Million €/Year

deployment leads to the lowest 300

system cost due to considerable 250
fuel savings from the Hp usage, a 200
reduction of generation expansion 150
needs and a reduction of custom- 100
er capital cost for the heater itself 50
(figure 8). also, as opposed to 0
Cost
Sav

Cost
Sav

Cost
Sav

Cost
Sav

Cost
Sav

Cost
Sav

Cost
Sav

Cost
Sav

other solutions, the hybrid Hp-B

does not use a water storage tank,
HP-B B-ERH HP-ERH HP HP-B B-ERH HP-ERH HP
which highlights the high opera-
tional flexibility of this technology. Gas Price: 8.5 €/GJ Gas Price: 12.75 €/GJ
In this particular study, distribu-
Power: Investment Power: Fuel and CO2 Heat: Investment
tion network reinforcement costs
Heat: Gas and CO2 Net Saving Net Cost
related to the different heating
technologies were not analyzed
but could tip the balance further figure 8. Electricity and heat-related costs and savings for different technologies
toward hybrid systems. (HP-B, HP, HP-B, and HP-ERH) relative to gas boiler. (Source: Heinen et al., 2016.)

july/august 2018 ieee power & energy magazine 75

 indoor temperatures and distribu-
24 tion temperatures, diversity, Hp
20
22
type, Hp performance, auxiliary
15 heater type, and heat emitter type.
10 20 as mentioned previously, some of
the impacts of heat electrification

Indoor Temperature (°C)

5 18
are most evident in distribution
Heat Output (kW)

0 16
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 networks and particularly at the
(a) lV level where most such technol-
24 ogies are connected.
20
22
In general, electricity demand
15 exhibits a relatively high degree
20 of diversity that makes the after-
10
5 18 diversity maximum demand quite
smaller than the sums of the indi-
0 16
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 vidual peaks, with a coincidence
(b) factor that is relatively low and on
the order of 15–20% for the united
HP Indoor Temperature (°C)
Kingdom. However, heat demand
has a comparatively lower diver-
figure 9. The standard and flexible operation of different heaters and building sity with a coincidence factor of up
indoor air temperature for the windiest week of an average year. (a) HP and (b) HP to 35% (figure 4). this has impli-
Flex. (Source: Heinen et al., 2017.) cations in terms of network impact
and design, as it effectively means
opposed to other storage technologies. If real-time or smart that larger infrastructure is required compared to a situation
control is enabled, then an Hp or another electric heater where the peaks were driven by nonheating consumption.
could be operated in a flexible manner according to power Because of this loss of diversity due to heating electrifica-
system conditions without impacting the thermal comfort of tion, the impact due to an increasing penetration of electrified
building occupants. heating could manifest itself on the lV network even at low
a systemwide cost study for Ireland (see Heinen et al., 2016 systemwide penetration levels. In practice, lV transformers
in the “for further reading” section) has shown that utiliza- specifically could be affected first by this loss of diversity,
tion of thermal inertia enables cost-effective heat electrification due to thermal issues (i.e., sheer increase in after-diversity
using Hps. Hp operation during the windiest week exempli- maximum demand) (figure 10), followed by lV conductors
fies the peak load reduction benefits of flexible operation ver- also driven by thermal issues. Voltage drop impacts are likely
sus standard operation. In standard operation, the Hp is only to appear only at higher penetration levels (see also navarro-
allowed to operate as indicated by the user thermostat settings, espinosa and Mancarella, 2014 in the “for further read-
which in Ireland is typically a few hours in the morning and ing” section). In figure 10, reinforcement will be required to
after working hours until late evening. this creates high morn- integrate electric Hps when transformer utilization exceeds
ing and afternoon peaks as the heater needs to bring the building 100% (or earlier if a safety margin is considered).
indoor temperature to the desired indoor temperature, assumed given the role of peak demand growth on potential lV network
to be 20 °C. In the flexible operation mode, the Hp can oper- impacts, thermal storage can reduce network impact, in partic-
ate at any time as long as the indoor temperature is 20 °C dur- ular in the form of hot water tanks for both domestic hot water
ing the thermostat settings and never higher than 23 °C. this and space heating (in the case of hydronic systems). However,
operational flexibility enables the heater to preheat the building the system needs to be carefully designed to override the peak
ahead of an expected peak event or when excess wind is on the consumption on extreme cold days, where again diversity is
system. for the week in figure 9, the building indoor tempera- likely to decrease substantially (figure 4). If the thermal stor-
ture is preheated very notably at least three times (days 2, 5, and age system is adequately designed for extreme days, then it can
7) and reduces peak load compared to the standard/nonflexible yield a substantial level of flexibility under normal conditions
Hp operation. In practice, rolling out these systems will require that can be utilized for demand response or more in general
detailed building sensing equipment and may not be acceptable DsM, in this case to deal with both thermal and voltage issues.
for all customers. this can, for example, happen in the presence of electric heat-
ing paired to photovoltaic (pV) systems, where pV produc-
Distribution-Level Impact of Heat Electrification tion and self-consumption can be maximized by making use
a variety of drivers that impact heat electrification have been of the flexibility enabled by thermal storage. More generally,
discussed so far, in particular heat consumption, outdoor and this distributed storage could be used to support active network

76 ieee power & energy magazine july/august 2018

management in a lV network to
deal with voltage control issues due
to pV and thermal issues due to 300 300

Transformer Utilization (%)

Transformer Utilization (%)

electric heating. 250 250
Detailed distribution network
modeling is also ongoing as part of 200 200
the realValue project. While poten- 150 150
tial system benefits are outlined in
100 100
“Coupling small-scale electric
Heating and thermal storage” sec- 50 50
tion, it is also important to consider
0 0
the impact of these scenarios on the

0
10
20
30
40
50
60
70
80
90
100

0
10
20
30
40
50
60
70
80
90
100
distribution network. In addition, the
potential of using the resource for HP Penetration (%) HP Penetration (%)
congestion management and poten- (a) (b)
tial upgrade deferral is also explored. 800-kVA Transformer
Building models are included in a 500-kVA Transformer
three-phase unbalanced ac optimal One Standard Deviation
power flow that allows heat demand
to be treated as a variable. Differ-
ent demand-response strategies are
figure 10. The average transformer utilization level (with one-standard deviation
examined, considering both the im- range) for (a) ASHP and (b) GSHP. Note that he penetration level refers to percentage
pact on the network and the thermal of houses with HPs. (Source: Navarro-Espinosa and Mancarella, 2014.)
comfort of customers.

Business Case for Multi-Energy Systems

While, as seen above, heating electrification can affect electri-
cal infrastructure requirements, it can also bring major ben-
efits in terms of cost-effective decarbonization of the heating
sector. However, this must be done in the context of a multi-
energy system whereby heat, electricity, and fuel (gas, in par-
ticular, but also, for example, hydrogen and other potentially
low-carbon fuels such as biofuels) are optimally integrated
together in both operation and planning. In particular, in the
context of a multi-energy system, heating can play a key role
in terms of providing cost-effective sources of flexibility to a
low-carbon, renewable energy-based electricity sector.
thermal storage, intrinsic thermal inertia, or hybrid heat-
ing solutions can bring substantial flexibility in the form of
demand response to support system operation, particularly
when the balancing requirements become more stringent in the
presence of variable renewables. this can be implemented at
different aggregation levels, for instance, individual buildings
(via physical thermal storage and thermal inertia of the build-
ings). this demand response flexibility can also be deployed
at the aggregate level, e.g., in the form of community energy
systems and smart districts, and up to the city level. In these
cases, heat networks, where cost-effective, bring the oppor-
tunity to use the heating sector as a flexibility provider for
electricity arbitrage and other demand response services, pri-
marily to support the cost-efficient integration of renewable
electricity. When multiple advanced technologies are brought
together [e.g., combined heat and power (CHp), Hps, and
thermal storage as in the high-flexibility multi-energy scheme
shown in figure 11], the system can substantially increase
Electricity Grid and Markets
Electricity
EHP
CHP
Electrical
and
Fuel Heat
Thermal
Demand
Auxiliary TES
Boiler
figure 11. A general distributed multienergy scheme.
(Source: Martinez-Cesena et al., 2016.)
its flexibility and, subsequently, the relevant benefits from
the provision of multiple market services (e.g., price arbi-
trage and real-time demand response). furthermore, there
can also be substantial gains from cross-technology support
that avoids asset redundancy and decreases plant size, again
reducing the overall cost and improving altogether the busi-
ness case for high-efficiency technologies. for example, the
system in figure 11 can effectively arbitrage between CHp
and Hps, and with additional flexibility support from thermal
storage, based on the spark spread between electricity price
(that can be very volatile in a renewables-based system) and
gas. In addition, real-time demand response can be provided
by ramping up/down the Hp (and in a slower way the CHp
too) and relying on the support from auxiliary boiler and

july/august 2018 ieee power & energy magazine 77

 thermal storage. In this respect, policy makers and regulators
can play an important role in promoting the deployment of
low-carbon, multi-energy systems by facilitating participa-
tion of such systems in grid services markets.
as in most other investment options, but especially when
dealing with the complexity of a multi-energy system, extreme
care needs to be taken at the planning stage to deal with long-
term uncertainty, primarily energy prices. stochastic optimiza-
tion tools are emerging in the literature of multi-energy sys-
tems to enable planning flexibility (see Martinez-Cesena et al.,
2016, in the “for further reading” section) and will hopefully
be embraced soon by industry too. In fact, these tools allow
for the management of long-term uncertainties by identifying
an optimal set of investment strategies that can dynamically
respond to changes in the underlying variables, for example,
gas and electricity prices. the ultimate aim is to improve the
business case of high-flexibility, high-efficiency, multi-energy
systems that, besides support heating sector decarbonization,
can also support low-carbon power system operation.
Consumer and Societal Values
the role of the consumer or building owner in the future of
electric heating cannot be understated. Consumer investment
decisions (heat technology, heat distribution system, thermal
storage, and building insulation) and behavior largely deter-
mine the shape of the future load curve. Most engineering,
economic, and policy research represents consumers as eco-
nomic agents and do not consider that consumers’ decisions
are largely impacted by noneconomic factors such as con-
venience, habitual routines, cultural values, and feedback
from acquaintances.
a crucial area of research is understanding consumers’ will-
ingness and ability to provide system flexibility or, in other
words, participate in demand response. While technology is
an important enabler of demand response, user engagement
is paramount. to estimate the flexibility potential of demand
response, consumer willingness and the ability to change
behaviors and expectations must also be understood. for a given
heating technology, end users will have varying thermal comfort
expectations that will impact on the realizable flexibility poten-
tial from such a resource. It is important that these sociotechni-
cal aspects are considered within the modeling framework when
demand response potential is being explored.
Summary
the electrification of heat is a huge opportunity for europe
to increase energy security and reduce carbon emissions by
shifting away from fossil-fuel sources. However, heat demand
today represents roughly half of the final energy demand in
the eu, which means that efficient electricity system integra-
tion will hinge on increasing building energy efficiency and
harnessing heat system flexibility.
Heat can be stored considerably more economically and effi-
ciently than electricity, which also serves to decouple electricity
demand from heat demand by using thermal stores, the thermal
78 ieee power & energy magazine
storage of buildings, and other energy-system integration solu-
tions such as hybrid heaters. these solutions increase system
flexibility and provide investment and operational benefits
across different levels of the electricity system from wholesale
markets to lV feeders. Valuing and remunerating these benefits
will improve the business case for multi-energy systems.
next to efficient grid integration, the future of electric
heating will depend on its capability to present a compelling
business case while at the same time meeting the expecta-
tions of consumers for warm homes and hot water.
Acknowledgments
C. o’Dwyer acknowledges support from the european
union Horizon 2020 realValue project under grant agree-
ment no. 646116. p. Mancarella acknowledges the support
of the uK epsrC My-store project (ep/n001974/1).
s. Heinen acknowledges support prior to July 2017 from the
fonds national de la recherche, luxembourg (project ref-
erence 6018454) and the Centre for It-Intelligent energy
systems (CItIes) project, Denmark (project reference
1305-00027B/Dsf).
For Further Reading
a. navarro-espinosa and p. Mancarella, “probabilistic mod-
elling and assessment of the impact of electric heat pumps
on low voltage electrical distribution networks,” Appl. En-
ergy, vol. 127, pp. 249–266, aug. 2014.
e. a. Martinez-Cesena, t. Capuder, and p. Mancarella,
“flexible distributed multi-energy generation system expan-
sion planning under uncertainty,” IEEE Trans. Smart Grid,
vol. 7, no. 1, pp. 348–357, Jan. 2016.
M. Bakhtvar, C andrade-Cabrera, g Buttita, o neu, and
a. Keane, “a study of operation strategy of small scale heat
storage devices in residential distribution feeders,” in Proc.
IEEE Power and Energy Society Int. Conf. Innovative Smart
Grid Technologies, turin, Italy, 2017, pp. 1–6.
real value project. (2018). [online]. available: http://
www.realvalueproject.com/
s. Heinen, D. Burke, and M. J. o’Malley, “electricity,
gas, heat integration via residential hybrid heating technolo-
gies: an investment model assessment,” Energy, vol. 109,
pp. 906–919, May 2016.
s. Heinen, W. turner, l. Cradden, f. McDermott, and M.
o’Malley, “electrification of residential space heating con-
sidering coincidental weather events and building thermal
inertia: a system-wide planning analysis,” Energy, vol. 127,
pp. 136–154, 2017.
Biographies
Steve Heinen is with Vector ltd., new Zealand.
Pierluigi Mancarella is with university of Melbourne,
australia, and university of Manchester, united Kingdom.
Ciara O’Dwyer is with university College Dublin, Ireland.
Mark O’Malley is with national renewable energy lab-
oratory, united states. p&e
july/august 2018
Electrification
and the Future
of Electricity
Markets
Transitioning to a
Low-Carbon Energy System

E
End-usE ElEctrification is onE of thE thrEE pillars
of deep decarbonization, which are fundamental strategies necessary
to eliminate energy-system carbon emissions. as a low-carbon strategy,
electrification is only effective when paired with reductions in the emis-
sions intensity of electricity generation. thus, any discussion of electri-
fication’s role in reducing carbon emissions, either through the direct
electrification of end uses but also fuel switching to electricity derived
fuels, must include supply–demand interactions within the electricity
system. this article highlights what these interactions may look like in
the transition to a very low-carbon energy system in the united states
and some potential implications for the future of the grid and wholesale
electricity markets. the concepts are primarily illustrated using updated
scenarios from the 2014 u.s. deep decarbonization pathways project
(ddpp) technical report.
©istockphoto.com/gorica

By Ryan Jones, Ben Haley, Gabe Kwok,

Jeremy Hargreaves, and Jim Williams
Digital Object Identifier 10.1109/MPE.2018.2823479
Date of publication: 18 June 2018

july/august 2018 1540-7977/18©2018IEEE ieee power & energy magazine 79

 research in the united states indicates that reducing overall
greenhouse gas emissions to 80% below 1990 levels by 2050,
a level adopted by the united states and other governments as
compatible with international agreements to protect the climate,
will require a 90%+ reduction in the average emissions intensity
of electricity generation in the united states over the same time
period. three observations help explain the need for a dispro-
portionally large reduction in emissions from electricity.
✔ Electricity generation in a very low-carbon energy sys-
tem will need to grow significantly, from 4 pWh/year
today to around 7 pWh/year in 2050, an increase
of 1 pWh/decade. this growth is a necessary conse-
quence of widespread electrification, even assuming
rapid progress in energy efficiency.
✔ Because the carbon benefits of electrification are
contingent on emissions reductions in electricity, the
greater those reductions are, the more attractive and
effective electrification will become as a carbon strat-
egy. put another way, because their combined effect on
emissions is multiplicative, not linear, these strategies
are mutually reinforcing.
✔ reducing emissions intensity in electricity is of moder-
ate cost compared to alternatives in other sectors, and
thus it makes economic sense from society’s standpoint
to reserve any remaining emissions budget for high
marginal abatement cost applications in transportation
and industry. reducing emissions by less than 80% in
these areas implies larger reductions in other areas,
particularly electricity generation, if the overall emis-
sions goal for the united states is to be met.
accepting the premise that deep decarbonization requires
profound changes in the electricity mix, three generation
types are likely candidates: renewables, nuclear, and car-
bon capture and storage (ccs). We define renewables here
Month
Jan. Apr. July
600
Growth in Average GW
500
400
300
200
100
0
1 24 1 24 1
figure 1. The month-hour average load growth from 2018 to 2050 in the DDPP
high-renewables scenario by end use. The vehicle-charging profile shown before
flexible charging is applied.
80 ieee power & energy magazine
broadly, but the primary focus is on wind and solar given that
these two resources have the largest technical and economic
potential. Between the three generation types, renewables have
emerged as the front-runner and have made up most of the
capacity added in the united states over the last ten years. this
trend is expected to continue, in part due to strong policy support
on a regional basis. Even in prior deep decarbonization scenarios
emphasizing nuclear and ccs, results suggest that renewables
will still comprise a plurality of generation by midcentury.
Physical Characteristics
of Highly Electrified and
Highly Renewable Electricity Systems
Given two premises, that deep decarbonization requires both
high levels of electrification and a near complete decarbon-
ization of electricity and the most likely method of electricity
decarbonization is large deployments of renewable energy,
some significant changes in the physical supply-demand dy-
namics of the future u.s. electricity system can be antici-
pated. in studies, these characteristics start to emerge when
the penetrations of wind and solar production exceed about
50% of energy consumption, driven by the nature of both newly
variable generation and electrified loads.
Newly Electrified Loads Will Be Large
but Potentially Flexible
the largest loads with the potential to be electrified are heat-
ing, electric vehicles, and, in some cases, electric fuel produc-
tion (for example, electrolysis of water to produce hydrogen).
figure 1 shows demand-side load growth between 2020 and
2050 in a high-electrification scenario, broken down by end
use. a key feature of these loads is that heating, vehicles, and
electric fuels all have inherent storage potential (thermal or
chemical) that can be used to shift energy in time.
Electric fuel production is a
nonfirm load (meaning it can be
Oct. curtailed) with a high operating-
to-capital-cost ratio that can shift
Light-Duty Vehicles
Water Heating vast amounts of energy seasonally.
Space Heating heating and vehicle loads are fun-
Other gible over much shorter timescales
but can still solve shorter times-
cale energy imbalance and, criti-
cally, mitigate their own impact
on the need for new system infra-
structure build.
New Loads Can Have Very
High and Unpredictable
Instantaneous Demand
24 1 24
Electric vehicle rapid-charging has
become a selling point for new elec-
tric vehicles, and as battery charge
management continues to improve,
public dc charging of light-duty
july/august 2018
electric vehicles in the hundreds of kilowatts or up to 1 MW for
heavy-duty transportation will be common. in terms of instan-
taneous demand, these power demands are more concentrated
than all but large industrial loads and will have large locational
impacts on transmission and distribution systems. Without the
development of real-time vehicle-grid communications, the
exact timing of these loads will be difficult to predict.
New Loads Will Change
Seasonal Load Profiles of Many Utilities
for most of the united states today, summer electric loads
are strongly temperature dependent above 65 ˚f due to air-
conditioning. But below 65 ˚f, load today is only weakly cou-
pled to temperature, except where electric heating is already
common. if a much larger share of heating across the united
states is electrified, load will have a strong dependence on
temperature across the entire range, and many more utilities
will find themselves with both summer and winter peaks.
this is already evident in places like florida with large air-
conditioning load but also infrequent cold snaps and high
penetration of electric resistance heating.
Energy Imbalance Has a Long Timescale
in High Renewable Systems
that the production patterns for wind and solar generally do
not match load on a diurnal basis is a well-understood problem
that can be addressed in many ways, including generator ramp-
ing, load shifting, electricity trading, and energy storage. not
as well understood is that, in very high renewable systems, sys-
temic overgeneration will occur in some seasons and systemic
energy deficit in other seasons. for the united states, the for-
mer will occur in spring and fall and the latter in summer and
winter, as illustrated in figure 2. particularly in hydro-wind
dominated systems, even longer timescale imbalance, includ-
ing interannual, can also be observed.
solving seasonal energy imbalance is arguably the main
impediment to reaching aspirational 100% renewable energy
goals called for by localities across the united states. sea-
sonal energy imbalance is a multifaceted challenge and will
require some mix of dispatchable generators, long duration
storage, seasonal flexible load (e.g., electric fuel production),
and overbuilding and curtailing renewables. Many of the most
widely discussed solutions for integrating wind and solar, for
example, flexible loads in buildings, flexible vehicle charging,
and battery electric storage, will be ineffective for balancing
on seasonal timescales. in the scenario shown in figure 2,
electric fuel production in spring and fall plus limited thermal
generation in summer and winter are used to solve this imbal-
ance, in addition to seasonal renewable curtailment.
Seasonal Energy Imbalance Extends
Across a Large Geographic Area
regional coordination to increase the diversity of load and
generation can be one of the most cost-effective ways to inte-
grate renewables by reducing relative forecast errors, thermal
july/august 2018
ramping, and round-trip efficiency losses from storage. how-
ever, increased regional coordination will not greatly dimin-
ish the fundamental drivers of seasonal energy imbalance
common to most of the united states for example, across
north america, spring is characterized by low loads due to
moderate temperatures, while solar insolation grows with the
approach of the summer solstice, higher pressure differen-
tials in temperate regions drive high winds, and spring snow-
melt produces high hydro. therefore, regional coordination
(in the absence of a globally connected grid) can only address
seasonal imbalance to a degree.
Renewable Curtailment Occurs
Across Many Hours
overgeneration-driven renewable curtailment occurs when
nondispatchable and must-run generation exceeds load,
including storage. the quantity of curtailment that a system
experiences is highly nonlinear, and as the amount of solar
and wind on the system approaches the levels required in
deeply decarbonized systems, is projected to be significant,
as shown in figure 3.
resources that provide flexibility can derive value from
arbitraging between periods with and without curtailment;
however, diminishing returns and a long-tailed distribution of
hours with substantial curtailment make it certain that some
amount of curtailment will always be economical. ultimately,
the most cost-effective system design and the level of resulting
curtailment will depend on the relative costs between adding
additional renewable capacity versus adding resources that
provide flexibility. the lower the cost of renewables and the
higher the cost of flexible resources, such as energy storage,
the higher the optimal level of curtailment.
the u.s. ddpp high-renewables scenario, with 80%+
renewables by energy, results in curtailment in 20–40% of all
hours, depending on location. actual curtailment may be less
than these levels due to new types of loads currently unantici-
pated that can take advantage of electricity that is zero cost
Nondispatchable Generation
800
600 Load
TWh
400
200
0
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
figure 2. The firm load versus nondispatchable genera-
tion in 2050 by month for the U.S. DDPP high-renewables
scenario.
ieee power & energy magazine 81
 at the margin yet not eliminated
given the large amount of cur-
49% 130% 45% 44% 45% 45% 47% 49% 53% 58% 64% 70% tailed energy concentrated in short
46% 120% 42% 41% 41% 42% 44% 46% 50% 55% 62% 68% periods of time and the difficulty
Renewable Portfolio Standard (%)

43% 110% 39% 37% 37% 38% 40% 43% 47% 53% 59% 65% of using intermittently available
39% 100% 35% 32% 32% 33% 36% 39% 44% 50% 56% 62%
energy economically.
35% 90% 30% 27% 27% 28% 31% 35% 40% 46% 52% 59%
30% 80% 26% 22% 21% 23% 26% 31% 36% 42% 48% 55%
Wind and Solar Alone
24% 70% 20% 17% 16% 17% 21% 25% 31% 37% 43% 50%
18% 60% 14% 11% 10% 11% 15% 19% 25% 31% 37% 43%
Do Not Provide
11% 50% 9% 7% 6% 6% 9% 13% 18% 23% 29% 35% Resource Adequacy
6% 40% 4% 3% 2% 2% 3% 6% 10% 15% 20% 25% systems with high variable gen-
2% 30% 1% 1% 0% 0% 1% 1% 3% 5% 9% 13% eration have large needs for dis-
0% 20% 0% 0% 0% 0% 0% 0% 0% 0% 1% 2% patchable capacity, called residual
0% 10% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% capacity, to provide reliability, fre-
0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% quently at times of the year when
Solar Fraction 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% energy deficits make it more dif-
Wind Fraction 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% ficult for short duration storage to
fill this need. figure 4 shows the
residual capacity need in a 100%
figure 3. The percentage of renewables curtailed without storage or flexible load
renewable scenario for the u.s.
for a simulated system in New York State. Renewable delivery respects a must-run
requirement of nuclear, cogeneration, and hydro (16% by energy). A renewable port- Eastern interconnection using a
folio standard of greater than 100% implies that renewables are overbuilt to increase single year of data from 2011. in
the amount of delivered renewables, with excess energy curtailed. This type of chart an ideal scenario, the net load (load
can be generated for other geographies with the exact pattern depending on load minus renewables) would be zero,
and renewable profiles. meaning that no residual capacity

800
123
179 190
700 227
261 280 298 317 323 322 300

600

500
Capacity (GW)

400
691
635 624
300 587
553 534 516 497 491 492 514
200

100

-
100% Solar,
0% Wind

90% Solar,
10% Wind

80% Solar,
20% Wind

70% Solar,
30% Wind

60% Solar,
40% Wind

50% Solar,
50% Wind

40% Solar,
60% Wind

30% Solar,
70% Wind

20% Solar,
80% Wind

10% Solar,
90% Wind

0% Solar,
100% Wind

Residual Capacity Need Capacity Met with Renewables

figure 4. The capacity met with 100% renewables by energy in the U.S. Eastern Interconnection with different wind/
solar compositions.

82 ieee power & energy magazine july/august 2018

is needed. instead, the residual capacity need at minimum
is 60% of the gross load peak, indicating that a significant
amount of flexible load or dispatchable generation is needed
to maintain current reliability levels.
an illustration of this basic idea is well demonstrated in
analysis of the effective load-carrying capability (Elcc) of
renewable generation, a concept that accounts for the outage
rates of thermal generation and the probability distributions
of renewable production, along with decades of historical
weather conditions, that impact peak load. large-scale elec-
trification is likely to materially alter Elcc calculations due
to heating load and other load-shape changes.
the size and persistent nature of residual capacity
requirements make it unlikely that imbalance in high-
renewables systems can be solved with load flexibility alone.
as already noted, short-duration storage is not able to effi-
ciently provide this capacity because it is energy limited
during peak seasons. these concepts are elaborated in the
section “capacity Market Must have Multiple products
that represent capacity of different time durations.”
While an ensemble of complementary resources will be
helpful in reducing residual peak load, significant amounts
of “persistent” balancing capacity, meaning those that aren’t
severely energy limited, will be required to match today’s
electric reliability.
Five Existential Questions for Future
Low-Carbon Electricity Markets
While they seem to be technically feasible, highly renewable
and high-electrification electricity systems will be radically
different from today’s systems dominated by dispatchable
thermal capacity and inflexible loads, not only in physical
characteristics but in market behavior. in competitive elec-
tricity markets today, unit commitment and dispatch are
handled by a security-constrained economic dispatch where
all generators are paid the nodal clearing price, may provide
a diverse set of ancillary services, and, depending on the
market, may receive additional capacity payments. limited
out-of-market payments are used to make whole those gen-
erators needed for reliability/stability but not competitive in
the market.
Market design is not agnostic to resource type. Each mar-
ket rule, from lead time for unit commitment decisions to
qualifying for regulation markets, inherently embeds the
technical characteristics of existing loads and resources.
these markets were designed under a paradigm in which
most generators are assumed to be dispatchable and have
nonzero marginal cost and where load is passive and far more
difficult and costly to control than supply. these assump-
tions are flipped under a high-renewable system paradigm in
which almost all costs are fixed, supply is variable, and new
technology enables unprecedented demand side flexibility.
this leads to five basic issues that future market designs
must address for high-renewables, high-electrification sys-
tems to succeed.
july/august 2018
How Will Conventional Power Plants Get Paid?
as noted earlier, a significant quantity of residual capacity
will be needed to maintain reliability in a high wind and
solar system, but those generators, if not equipped with car-
bon capture, must operate at low-capacity factors or else car-
bon budgets will be exceeded. Essentially, all fossil plants
that cannot sequester carbon must transition to being peaker
plants under deep decarbonization.
figure 5 shows the capacity and utilization for fossil
thermal power plants in the u.s. ddpp high-renewables sce-
nario contrasted against the 2017 annual Energy outlook.
thermal capacity declines by 31% in the high-renewables
scenario, due to the resource adequacy provided by wind and
solar (but offset somewhat by electrification). capacity fac-
tors, on the other hand, decrease by 69%, to just over 10%,
for the fossil thermal fleet.
What figure 5 indicates is that the revenues earned by gen-
erators in an energy market will decline over time, even if all
caps on market bids are removed. in jurisdictions with capac-
ity markets, the difference between the gross and net cost of
new entry will decrease. this will present special challenges
for energy-only markets such as the Electric reliability coun-
cil of texas, with greater volatility likely as fixed costs must
be recovered in fewer and fewer hours. out-of-market pay-
ments may increase and new ancillary service products may
be devised, trends that have already begun, but these are not
likely to completely replace lost energy market revenues.
the importance of capacity markets will grow; where they
do exist, these mechanisms will need to become more sophis-
ticated regarding the types of products and characteristics
of the resources that qualify. for instance, a 6-h battery that
may provide capacity today may not contribute to reliability
in a high wind and solar system where the binding constraint
(say, in august) is the availability of energy. here, concepts
that have been long used in hydro-dominated systems, such
AEO 2017
High Renewables
Reference
40
Thermal Power Plant
Capacity Factor (%)
600
Capacity (GW)
30
400
20
200 10
0 0
2015 2050 2015 2050
Thermal Power Plant Capacity
Capacity Factor
figure 5. The fossil thermal power plant capacity and
utilization for the 2017 U.S. Department of Energy’s annual
energy outlook reference scenario and the U.S. DDPP
high-renewables scenario.
ieee power & energy magazine 83
 Establishing and frequently updating a clear road map to
deep decarbonization in electricity is critical to ensure goals
are met on time and at reasonable cost.
as sustained peaking capability, may need to become an inte-
gral part of all markets. this is illustrated by the 21st century
screening curve in the “capacity Market Must have Multiple
products that represent capacity of different time dura-
tions” section.
How Will Fixed Costs Be Allocated
Among Electricity Consumers?
today, variable costs already make up a relatively small
share of total electricity system costs (encompassing genera-
tion, transmission, and distribution), and their share of the
total electricity revenue requirement is set to plummet in a
high-renewables scenario, as shown in figure 6 using data
from the u.s. ddpp scenarios. in the high-renewables sce-
nario, only 2.5% of costs are variable by 2050, thus 97.5%
of the revenue requirement must be collected regardless of
how, when, or whether electricity gets used.
it can be argued that many new electric loads, including
electric fuel production, will be able to balance the system
by consuming energy that is free at the margin. however,
this ignores a serious cost-allocation problem, namely that
other market participants must pay for the renewable gen-
erator providing that marginally free electricity at the price
of a long-term power purchase agreement (ppa). thus, if
20%
US$110B
of Revenue Requirement
US$85B
Variable Cost Share
15%
10%
5% Scenario
AEO 2017 Reference
High Renewables US$34B
0%
2020 2025 2030 2035 2040 2045 2050
Year
figure 6. Variable costs as a share of the total U.S. elec-
tricity revenue requirement. This is created using generator
fuel costs rather than market prices. In electricity markets
where a marginal price is paid to all generators, a greater
portion of fixed costs are recovered on a variable basis
than shown.
84 ieee power & energy magazine
direct access customers are not ppa off-takers, they become
a fixed-cost free rider and receive energy subsidized by other
market participants. this could be desirable from a public
policy perspective, for example, to encourage zero-carbon
transportation using hydrogen generated from marginally
free electricity, but it raises legitimate concerns about cross-
subsidies between customer classes.
How Can Flexible Load
Be Induced to Participate?
flexible load has a large potential to solve energy imbal-
ance at low cost compared to many other solutions. antici-
pated new electric loads have inherent flexibility, either as
a function of their thermal storage (electric space and water
heating) or chemical storage (electric vehicle batteries or
hydrogen electrolysis), which can be leveraged to manage
supply–demand imbalances.
the role that flexible load could play in a future system is
illustrated in figure 7, showing average load for three simu-
lated months in 2050 in a high-renewables scenario. the first
panel shows load net of all nonthermal generation technolo-
gies, including dispatchable hydro and nuclear. as seen, the
average net load is frequently negative, indicating a surplus of
energy on the system. Given the seasonal features shown in
figure 2, the month of april has the greatest surplus energy.
the second panel shows the average hourly operation of dif-
ferent balancing solutions in each month. Balancing solutions
above the x-axis effectively increase load and absorb surplus
energy, while those below the x-axis reduce load. Electric fuel
production plays a large role in reducing curtailment in the
spring, and flexible end-use demand, including vehicle charg-
ing flexibility and flexible thermal loads, plays a large role in
reducing daily net load peaks. the final panel shows the resid-
ual net load (summing panel one with panel two) that in this
scenario is met with thermal dispatchable generation. these
three panels demonstrate the importance of flexible loads
among the balancing solutions. in these scenarios, energy
storage plays a relatively small role because it does not have
the storage duration to replicate the role of electric fuel pro-
duction in seasonal balancing, and it is more expensive than
flexible load for solving diurnal imbalance.
conceptually, it is useful to divide flexible load into two
types: short duration, which is primarily traditional firm load
made flexible through the use of control technology, and long
duration, which includes new nonfirm loads with low marginal
energy-storage costs, for instance, grid-scale electrolysis. the
july/august 2018
 Jan. Apr. July Jan. Apr. July Jan. Apr. July
800
Nonthermal Net Load Residual Load
600

400

200
Average GW

–200

Balancing Resource
–400
Renewable Curtailment
Electric Fuel Production
–600 Flexible End-Use Demand
Energy Storage
–800
1 241 241 24 1 24 1 24 1 24 1 24 1 24 1 24

figure 7. An example of the magnitude of flexible demand that operates in a high-renewables system taken from the U.S.
DDPP high-renewables scenario.

first type of flexible load can solve diurnal energy imbalance
and reduce load at peak to avoid capital investments (i.e.,
substations and transmission lines). the second type of flex-
ible load provides a productive use for excess energy during
seasons with extensive overgeneration and allows for further
renewable build at lower marginal curtailment by reducing
energy deficits during other times of year. ultimately for both
types of flexible load, the goal is to flatten net load.
short- and long-duration flexible load each face differ-
ent barriers to adoption. interruptible load for peak shav-
ing is a mature market for large commercial and industrial
applications; recently, interruptible programs have expanded
into small commercial and residential applications, with
many small interruptible loads aggregated into large ones.
More recent still are load-shifting programs, which are an
important step past traditional demand response. for most
customers, though, even given the necessary smart controls
in demand equipment, the price signals sent do not have the
temporal or spatial granularity to unlock the full cost-of-ser-
vice reductions possible from utilizing flexible load.
the barriers to long-duration flexible load also include
its exclusion from full participation in energy markets.
today the impact that price-responsive flexible load has on
the shape of load must be accounted for in load forecasts.
this works fine when flexible load is marginal, but if very
large nonfirm flexible loads such as electric fuel production
develop, symmetry between supply and demand in electric-
ity markets will be critical for optimal outcomes, for two
reasons. first, wind and solar vary enough day to day that
a prearranged schedule will always be suboptimal. second,
the quantity of electric fuel production could be substantial
and thus will have a systemic impact on markets.
How Can Developers Be Induced
to Make Long-Term Investments
Given Uncertain Revenues?
today, most renewable generators are compensated through
bilateral ppas, which guarantee fixed volumetric rates over
many years and insulate project developers from variable mar-
ket prices. few merchant nondispatchable generators exist
today, and it seems unlikely that merchant exposure is a robust
model for future development given the long payback times
for renewable generators and the significant risk of curtail-
ment and market price depression over the project lifetime.
thus, bilateral ppas are likely to remain the dominant struc-
ture for paying renewable generators.
technologies that provide balancing, such as energy stor-
age or electric fuel production, face similar risks in energy
markets due to unknown future patterns of renewable devel-
opment and/or the deployment of competing balancing
resources. While risk is a normal part of the merchant elec-
tricity equation, the rapidity of changes on the supply side
may produce so much uncertainty in long-term revenues that
no project is bankable or the deployment of balancing sub-
optimally lags the build of variable generation.
How Will Future Electricity-System
Planning Be Conducted?
if the goal is reaching 2050 electricity emission rate targets
consistent with deep decarbonization, it is necessary for the

july/august 2018 ieee power & energy magazine 85

 Given the challenges posed by rapid electrification
and renewable build, markets will need to keep pace
on several different fronts simultaneously.
electricity system to go through several decades of unprec-
edented change. Widespread electrification will lead to load
growth not seen in decades. planning in a system dominated
by variable generation is more complex because it intro-
duces a temporal component into every planning problem.
the timing and complementarity of the many different but
necessary electricity investment decisions are critical, and
many 20th century planning paradigms and tools are likely
insufficient to address the challenges ahead.
compounding the planning challenges, deregulation has
decentralized planning by putting many planning activities in
the hands of market participants. in current competitive mar-
kets, no stakeholder has visibility into all parts of the system,
which will seriously hamper execution of an overall long-term
vision. if uncertainty about policy, market rules, or timing of
complementary resource deployment calls future revenue
streams into question, then carbon targets may not be met, the
cost of financing may increase, and sub-optimal development
strategies pursued.
Electrification exacerbates planning challenges by involv-
ing sectors that historically have no strong links to electric-
ity, most notably transportation. as mentioned earlier,
transportation loads represent significant energy but, per-
haps even more importantly, may have very high and unpre-
dictable instantaneous power demand. the new spatial
planning challenge, of developing transportation charg-
ing infrastructure in which the highest demand is located
where the electricity grid is best able to accommodate it,
is considerable.
Concepts for Future Market Design
Given the challenges posed by rapid electrification and
renewable build, markets will need to keep pace on sev-
eral different fronts simultaneously. it can be argued that
a vertically integrated, rather than deregulated, structure
is best suited for coordinating the scope of electricity
system changes brought about by deep decarboniza-
tion. But this argument neglects functions that markets
are better suited to accomplish, namely communicating
information between participants and driving efficiency
improvements. the challenge is thus how to get the best
of both: 1) the flow of information and efficient opera-
tions from markets and 2) effective long-term planning
from a central decision maker. Below are some concepts,
challenges, and principles that such an optimal approach
should consider.
86 ieee power & energy magazine
Energy Market Compensates Balancing
Services, with Full Symmetry Between
Supply- and Demand-Side Balancing
in a world with near-zero marginal cost for energy and uncon-
trollable supply, energy markets primarily become a place to
compensate and schedule balancing services. as seen in fig-
ure 7, these balancing services can take many forms, including
flexible end uses, new nonfirm loads, energy storage, thermal
resources, and renewable curtailment. such resource types
would provide the bulk of balancing in a highly electrified,
high-variable generation system, but today these resources are
not integrated into markets in a way that would allow them to
operate in the way pictured.
in such a system, large loads would bid a demand curve
in which different amounts of energy consumption would be
offered at different prices, in a way that is completely analo-
gous to how generators bid an energy supply curve today.
for small customers, this could be done by an aggregator on
their behalf that then sends control signals to building loads
after considering a customer’s expressed preferences.
one technical challenge when scheduling flexible load
or energy storage is state-of-charge constraints, which are
not yet integrated into markets. today, most resources
that track state of charge, such as hydro or energy stor-
age, self-schedule generation, which works fine when those
resource are marginal, when net-load is predictable and
well understood by the scheduler, or when few competing
balancing resources exist (there is no real need to exchange
information). as supply becomes less flexible and the
opportunity cost of not using newly electrified flexible
load increases, markets will play a critical role as a clear-
inghouse for balancing resources. it will be important for
regional transmission operators (rtos) to gain experience
scheduling flexible loads at a small scale before balancing
needs grow massive. the alternative is more reliance on
conventional resources, which in the context of deep decar-
bonization could add significantly to cost or make carbon
goals unattainable.
Capacity Market Must Have Multiple Products
That Represent Capacity of Different Time
Durations (“21st Century Screening Curve”)
capacity markets serve two key functions in electricity today.
the first is as a longer-term planning mechanism to ensure
that sufficient resources are built and maintained to meet
expected loads, months or years in advance. the second is
july/august 2018
as an operational tool to make sure sufficient resources are
available in energy markets to cover forecast errors or meet
expected and unexpected changes in net load or generation.
under this definition, most ancillary service products can be
categorized as near-term capacity markets.
in a highly electrified, highly variable electricity system,
all capacity markets will need to increase in sophistication.
ancillary service products must accommodate increased
forecast error, more extreme generator ramping, and certain
grid services provided today by conventional generation
but not replicated by renewables such as primary frequency
response. this is an area in which rtos have been active in
recent years, and much progress has been made.
the changes necessary for long-term capacity markets
are not as well understood, and debate still revolves around
exactly what capabilities are necessary for future capacity
and how much will be required. it is often assumed that a
combination of wind, solar and 6–24 h of energy storage
will be sufficient to create a reliable low-cost decarbon-
ized electricity system, but this does not stand up to seri-
ous scrutiny.
to illustrate these concepts, a new type of screening
curve is presented here for systems with large amounts of
inflexible supply that demonstrates the capacity required
from balancing resources of different durations. figure 8
shows the residual capacity over the modeled year 2050 for
two cases, the u.s. department of Energy’s 2017 annual
Scenario
600
Residual GWh
400
AEO 2017
Reference
200
0
600
Residual GWh
400
High
Renewables
200
figure 8. The residual capacity need across simulated 2050 scenarios. Residual capacity is equal to total load minus all
zero-carbon resources, including nuclear and hydroelectricity. In the modeled scenarios, this residual was met with ther-
mal resources.
july/august 2018
Energy outlook and the high-renewables scenario from
the u.s. ddpp. residual capacity is calculated as gross
load minus all zero-carbon resources (excluding bioen-
ergy) and includes the impact of all flexible loads and ener-
gy storage.
this residual is instructive because it is the energy that must
be met with either thermal resources or a greater deployment
of the zero carbon resources already in the system. the key
question is what the characteristics for other resources must be
to replace thermal and reduce emissions or cost. this question
is approached by analyzing the residual capacity requirement
to understand the number of hours a given capacity must be
maintained to ensure a reliable system. this result is shown
in figure 9. in the high-renewables case, capacity resources
had to be able to provide at least 422 GW for 1 h, at least
376 GW for 10 h, and at least 141 GW for 100 h. the flatter
these screening curves, the higher the likely capacity factors
for residual resources, and the longer the duration, the more
difficult it is for energy-constrained balancing resources, such
as battery storage, to meet the need.
the most striking aspect of the curve is how little the
shape of the curve changes between the two cases. despite
huge quantities of renewable generation, residual capacity
needs with significant duration remain a necessity for reliable
system operation. one important caveat is that this screen-
ing curve does not consider the need for long-duration stor-
age resources to charge as well as discharge. thus, a storage
Hour of the Year
ieee power & energy magazine 87
 resource with 100 h of duration may need 100 h of charging
after meeting a long-duration event and may be unavailable
for subsequent events that theoretically could have been met
if operated differently. the operation of such a resource to
maintain system reliability and without being overly conser-
vative is a large challenge for future system operators.
caveats aside, figure 9 demonstrates a critical challenge
for the provision of long-term capacity in future markets. for
markets today, the binding constraint is nearly always in the
first 6 h of peak duration because most generators are not
energy limited. however, in future systems with sufficient
load flexibility and short-duration storage, it is possible that
longer duration constraints will become binding, and capac-
ity markets will need to expand to specifically procure per-
sistent balancing resources.
A New Mechanism for Allocating Socialized
Fixed Costs (Renewable PPAs, Grid Electrolysis
Plants, and Energy Storage Systems) Among
Energy Consumers Must Be Devised
in the highly electrified, highly variable generation system
in which 97.5% of all electricity costs are fixed, the question
of how to allocate these costs among market participants
will be critical. the question is made more difficult because
cost allocation is as much a policy question as an economic
one. today, fixed costs are allocated using auctions and a
combination of volumetric, capacity, and fixed charges. fig-
uring out how to fairly allocate fixed costs to loads using
these mechanisms will be an important step in the overall
development of decarbonized electricity systems.
as an example of the challenge, consider the options for
allocating the cost of nondispatchable renewable generation.
Scenario
AEO 2017 Reference High Renewables
800
600 583 GW
Capacity (GW)
as discussed previously, energy market prices will be insuf-
ficient to cover the costs of renewable generation if all genera-
tors bid their marginal cost, which is near zero. to purchase
a renewable generator is essentially to prepay for the energy,
and the question is what mechanism will be used to allocate
the cost of prepaid energy between market participants. three
potential options include the following:
✔ no special cost allocation—renewable off-takers make
or lose the spread between market prices and ppas and
other loads benefit from merit order effects and poten-
tial curtailed energy.
✔ some or all loads must purchase prepaid energy cred-
its, equivalent to renewable energy credits (rEcs) today.
this is similar to how many renewable portfolio stan-
dard mechanisms work today. the necessity for different
types of loads to purchase rEcs would be policy driven,
and market participants could trade rEcs among them-
selves. any loads not required to purchase credits are
given an implicit subsidy paid for by other loads.
✔ renewable costs are pooled and administered in the
market as part of a charge analogous to a transmission
access charge.
Risk Must Be Pooled to Attract Long-Term
Investment with Reasonable Financing
Mechanisms will be needed to pool the risks associated
with many of the new technologies discussed, including
large flexible loads and balancing resources. the basic
issue comes back to uncertainty regarding future revenues,
which will drive up the cost of financing or cause underin-
vestment when measured against what is required for deep
decarbonization. Mechanisms to address this issue must
be developed, playing a role si -
milar to the one played by ppas
for renewables.
capital assets in energy have
long lifetimes, depend on a long-
term revenue stream, and aren’t
mobile between markets. utilities

486 GW
422 GW
and regulators have a large role to
400 376 GW play in deep decarbonization by
383 GW providing stability in future revenue
278 GW streams for projects deemed optimal
200
by long-term planning exercises.
141 GW Market mechanisms, such as reverse
12 GW
0
auctions, should be employed within
1

10

100

1,000

10,000
1

10

100

1,000

10,000

this framework to ensure the lowest

cost solutions are built; regulators
Duration (Hours) Duration (Hours) should maintain technology neutral-
ity, defining the technical character-
figure 9. A 21st-century screening curve showing the number of hours and different
amounts of capacity that must be maintained for a reliable system. This curve was istics for generic new resources and
created using residual net load (gross load minus nonthermal resources) for the an- allowing markets to respond.
nual energy outlook reference scenario and the U.S. DDPP high-renewables scenario the pace of change required
by systematically counting the number of hours that load stayed above a certain in electricity to reach deep decar-
capacity level. bonization by micentury may very

88 ieee power & energy magazine july/august 2018

 Due to the pace of change in technology and the
speed at which those technologies need to be deployed,
agility in planning and market design is paramount.
likely produce less-optimal planning decisions than would
otherwise happen under a slower transition. the main risk is
that too many poor decisions could derail the entire process,
driving up costs while missing the carbon goals. this high-
lights the need for expanded long-term planning.
Concluding Thoughts: An Elevated
Role for Planning Is Needed to Ensure
Prudency and Guide Sequencing
Toward Deep Decarbonization Goals
Establishing and frequently updating a clear road map to deep
decarbonization in electricity is critical to ensure goals are met
on time and at reasonable cost. in creating and administering
such road maps, utilities and their regulators play an indispens-
able role and must rise to the occasion for the enterprise to
succeed. We conclude by offering some guiding principles for
success over the coming decades.
✔ a key to future planning will be the efficient and trans-
parent sharing of information among stakeholders,
some of which have not coordinated historically.
✔ regulators should strive to understand which changes
will be best accomplished through markets and which
through regulations.
✔ it is better to delay reaching 2050 goals than to make
poor decisions because the goals may end up even
further delayed and have a damping effect on achiev-
ing long-term decarbonization. deliberation must be
properly balanced against urgency.
✔ all near-term decisions should always be viewed through
the lens of long-term goals. this will help minimize
carbon lock-in or stranded assets.
✔ regulators have an important role to play in democra-
tizing risk and in transforming markets by deploying
novel technologies.
this article raises many questions regarding markets
and regulatory structures at large associated with a highly
electrified electricity system with significant amounts of
renewables. due to the pace of change in technology and
the speed at which those technologies need to be deployed,
agility in planning and market design is paramount. this
is perhaps the single most important concept for regulators
and system planners in the coming decades: maintaining
the ability to correct course quickly and frequently will be
july/august 2018
more valuable than the ability to perfectly follow a blue-
print created years before.
For Further Reading
J. h. Williams, B. haley, f. Kahrl, J. Moore, a. d. Jones,
M. s. torn, and h. McJeon. (2014). pathways to deep decar-
bonization in the united states: the u.s. report of the deep
decarbonization pathways project of the sustainable devel-
opment solutions network and the institute for sustainable
development and international relations. revision with tech-
nical supplement, nov. 16, 2015. [online]. available: http://
usddpp.org/downloads/2014-technical-report.pdf
J. h. Williams, B. haley, and r. Jones. (2015, nov. 17).
policy implications of deep decarbonization in the united
states: a report of the deep decarbonization pathways proj-
ect of the sustainable development solutions network and the
institute for sustainable development and international rela-
tions. [online]. available: http://usddpp.org/downloads/2015-
report-on-policy-implications.pdf
J. Williams, r. Jones, G. Kwok, and B. haley, “deep de-
carbonization in the northeastern united states and expand-
ed coordination with hydro-Quebéc,” 2018.
G. Kwok and B. haley. (2018). portland general electric
decarbonization study. [online]. available: https://www
.evolved.energy/single-post/2018/02/23/portland-General-
Electric-decarbonization-study
r. Jones. (2017). conference presentation: realities
of balancing electricity systems with 100% renewables.
[online]. available: https://www.evolved.energy/single-
post/2017/07/11/realities-of-balancing-electricity-systems-
with-100-renewables
Biographies
Ryan Jones is with Evolved Energy research, denver,
colorado.
Ben Haley is with Evolved Energy research, denver,
colorado.
Gabe Kwok is with Evolved Energy research, portland,
oregon.
Jeremy Hargreaves is with Evolved Energy research,
san francisco, california.
Jim Williams is with the university of san francisco,
california. p&e
ieee power & energy magazine 89
 An Electrified
Nation By Carla Frisch,
Paul Donohoo-Vallett,
Caitlin Murphy, Elke Hodson,
and Nathaniel Horner

H
Historically, tHe term electri-
fication has meant providing electricity to
customers who lack access to electric infra­
structure. From the earliest urban electric
services in the 1890s to the electrification
era in rural america from the 1930s to
the 1950s, access to electricity drastically
improved quality of life, including thermal
comfort, hygiene, food availability, illumi­
nation, mobility, commerce, communica­
tion, and entertainment. as a result of these
early electrification efforts, almost all U.s.
homes are fully or partially electrified, with
notable exceptions in some Native ameri­
can and alaska Native communities.
an expanded definition of electrifica­
tion allows us to contemplate a new era of
electrification that is now in its early stages,
driven by economic opportunities, new tech­
nologies, consumer demands, environmental
issues, and changing societal and industrial
structures. in this context, electrification can
be broadly understood to include three means
of increasing electricity demand.
1) Expanded adoption of existing ser-
vices: increasing the extent to which
electricity is used to provide existing
energy services (e.g., growth in the use
of electric cooling technologies).
2) Fuel switching: substituting electric­
ity for another fuel to provide the same
energy service (e.g., using electric heat
pumps for space heating instead of fuel
oil or using electric vehicles for trans­
portation instead of gasoline vehicles).

Digital Object identifier 10.1109/MPe.2018.2823480

Date of publication: 18 June 2018

90 ieee power & energy magazine U.S. Government work not protected by U.S. Copyright. july/august 2018
 A Review of Electrification Study Scenarios
and Future Analysis Needs for the United States
3) New end uses: adopting electric technologies that
provide an entirely new service demand for electricity
(e.g., smartphones, data centers, and vertical farming).
a range of recent studies examining strategic, beneficial,
or efficient electrification show substantial growth potential
for electricity use in transportation, buildings, and industry.
july/august 2018
they include a focus on consumer value (choice, convenience,
productivity, and environmental impacts) and business value
(expanded opportunities for electric providers). electrifica­
tion’s economic potential has primarily been studied through
the lens of reducing energy­related greenhouse gas (GHG)
emissions. substantial electrification, however, is not solely
driven by decarbonization and merits addi­
tional analysis.
in the context of current and historical
electrification trends, this article summarizes
recent studies that model future pathways
that increase electricity’s share of energy
consumption in U.s. buildings, industry, and
transportation. We discuss a range of output
metrics mentioned in the studies, including
total electricity demand, electrification by
sector, environmental impacts, costs, eco­
nomic growth, and energy productivity, to
identify broad trends and possible impacts
of electrification. in addition, we compare
methodologies among studies, identify areas
for additional research, and discuss policy
implications of common outcomes of elec­
trification across studies.
Current and Past
Electrification Trends
Historically, electrification has meant gener­
ally gaining access to electricity and new elec­
tric technologies. a simple metric for evaluat­
ing electrification is electricity consumption
as a percent of total delivered energy. Based
on this metric, Figure 1 shows that, until 2010,
electricity’s share of delivered energy in the
United states grew steadily, reflecting the
increase in end uses and technologies that are
powered by electricity.
in the context of electrification, the flat­
tening of electricity’s share of delivered ener­
gy since 2010, shown in Figure 1, may seem
©istockphoto.com/Filograph
surprising. the change reflects a combina­
tion of economic and technology­related fac­
tors, but it also highlights this metric’s key
limitation. in contrast to historical trends,
the current drivers of electrification are not
limited to gaining access to electricity but
focus more on using electricity for a service
ieee power & energy magazine 91
 electricity demand for light­duty vehicles doubled between
20% 2015 and 2017, driven by decreasing battery costs, increases
18% in electric vehicle offerings, and policy incentives. in residen­
Transportation
of Delivered Energy

16%
Industrial tial heating, electricity’s share of energy demand grew in all
Electricity as %

14%
12%
10%
Commercial
8%
6%
4%
Residential
2%
0%
1950 1960 1970 1980 1990 2000 2010 2016
figure 1. Electricity as a percentage of delivered energy in
the United States. (Source: The EIA’s Monthly Energy Review.)
that uses a different fuel source (e.g., electric heat pumps
replacing heating oil for space heating). electricity’s share
of delivered energy shown in Figure 1 does not account for
increasing efficiency of electric technologies. For example,
substituting light­emitting diode and compact fluorescent
lightbulbs for older incandescent bulbs does not decrease
demand for the lighting service, but it does reduce energy
consumption from lighting on a per­unit basis.
currently, energy consumption by fuel varies greatly by sec­
tor, as shown in Figure 2. electricity constitutes approximately
half of energy consumed on site in buildings in the residential
and commercial sectors. By contrast, electricity makes up just
13% of industrial and less than 1% of transportation energy con­
sumption. subsectors, geographic areas, and time trends show
additional variation. For example, electricity’s share of building
energy use varies significantly by climate region.
Within these sectors, both new technologies and the increased
deployment of existing technologies contribute to trends po­
tentially driving electricity consumption growth. technol­
ogy deployment is driven by economics, technology maturity
and availability, consumer preferences, changing population pat­
terns, and policies. in the transportation sector, for example,
30
Direct Renewable
Delivered Energy (Quads)
regions of the United states between 2005 and 2013, partly due
to increased electric heat and partly due to population migra­
tion to milder climate regions. in the industrial sector, electricity
has fueled improved manufacturing services, such as induction
heating and the electric arc furnace. indoor or vertical farming
is a growing area of commercial­industrial electricity demand,
and the deployment of information and communication tech­
nologies is affecting electricity use across all sectors.
Comparison of Recent
Electrification Scenarios
the studies summarized here highlight potential pathways
that accelerate the current electrification trends and explore
what it might look like if the United states were to enter a
new era of electrification. Here we build on and update a 2015
review by Peter loftus and coauthors that surveyed low­car­
bon pathway studies, which all project increased demand for
electricity, particularly by passenger vehicles and heat.
the studies reviewed are recent ones that generally use
national energy­sector models to explore the electrification of
U.s. building, transportation, and industrial sectors. each study
includes multiple scenarios that model energy, economic, and
environmental impacts out to the year 2050 based on a set of
assumptions regarding fuel and technology costs, economic
growth, policy goals, and other factors. in this context, a sce­
nario is the set of assumptions and constraints modeled, while
a projection of a particular metric is the output of the modeled
scenario for a future year. scenarios and their resulting projec­
tions are meant to reflect economically and technically plau­
sible pathways, but they are not forecasts or predictions.
the exception is the 2017 annual energy outlook (aeo)
reference case from the energy information agency (eia);
this projects 2050 energy use under the assumption that cur­
rent policies remain this same and is used here as a bench­
mark to evaluate the impact of the other study scenarios.
many of these studies incorporate electrification as only one
of several critical elements, such as land use, that can shape

25 Biomass
Petroleum future energy pathways. For each study, we review the scenario
20 Natural Gas that specifically focuses on achieving high electrification if
Coal one is provided; otherwise, we examine results from the core
15 Electricity
or main scenario. all of the scenarios share certain attributes:
10 they assume rapid electrification of end uses, rely on a balanced
portfolio of low­ and zero­carbon electric generation technolo­
5
gies, achieve high­decarbonization goals, and assess the impli­
0 cations for the energy system to meet those goals.
Residential Commercial Industrial Transportation
While many of the scenarios are conceptually similar, the
Percent
Electricity: 44%
models used to generate the scenarios or examine the impacts
52% 13% <1%
of exogenously defined scenarios differ in terms of resolu­
figure 2. The total delivered energy by fuel type and sec- tion, scope, and operation. For example, the Global change
tor for the United States in 2016, not including electricity- assessment model (Gcam­Usa) and the regional energy
related losses. (Source: The EIA’s Monthly Energy Review.) Deployment system (reeDs) model are both optimization

92 ieee power & energy magazine july/august 2018

models, but Gcam­Usa includes both energy and land­
use sectors while reeDs only models the electricity sector
(see the article by mai et al. in this issue). a third model,
PatHWays, is completely user specified and a detailed
accounting tool (see the articles by Haley et al. and mahone
et al. in this issue).
the methodology for determining the extent to which end
uses are electrified is exogenous for each scenario; adoption
rates of electric devices are developed externally to the
model. in some cases, a literature review informed the level
of electrification, and in other cases the scenarios are based
on expert judgment. in general, these scenarios reflect rates
of electrification that are technically feasible.
to compare how each scenario examined opportunities
and challenges for electrification, we compare the follow­
ing commonly used metrics from the studies described in
table 1: total electricity demand, energy demand met by
electricity in each subsector, and GHG emissions.
Electricity Demand
Figure 3 compares the 2050 total electricity demand pro­
jected across scenarios in the various studies, within the context
of historical electricity demand. the data point marked with
a diamond in each study’s output corresponds to the scenario
in which electrification is most emphasized. Note that there
may be differences among the studies as to what is included
in demand.
these electrification scenarios project significant increases
in electricity demand over 2017 levels and over 2050 levels in
the aeo reference case. the scenarios range from approxi­
mately 1.5­fold to approaching three­fold increases over 2017
electricity demand by 2050. Given
different analysis approaches and
scopes, the range of values seen in 12,000
the figure can be viewed as a general
trend, roughly doubling electricity
10,000
demand. in addition to increased elec­
Total Electricity Demand (TWh)
tricity demand, electricity increases
as a proportion of overall energy con­ 8,000
sumption. Not all studies report this
metric, but those that do project a dou­
6,000
bling in electricity’s share of energy
consumed in 2050 (including fuels
produced from electrical processes), 4,000
Historic (EIA)
from around 25% in the reference
projections to approximately 50% in
the high­electrification scenarios. 2,000
the modeled electrification is
not uniformly distributed across the -
buildings, industrial, and transporta­ 1970
tion sectors in these scenarios. Fig­
ure 4 compares the studies’ projected figure 3. The total historic and projected electricity demand in recent studies.
electrification of each end­use sector (Note: Iyer results are for electricity generation.) Some scenarios show possible
in 2050, with aeo included as a ref­ 2050 electricity demand of more than double or nearly triple today’s value, a sig-
erence benchmark. the height of the nificant increase over AEO 2017 reference projections. MCS: Midcentury Strategy.
july/august 2018
bars indicates delivered energy in 2050, which varies across
the studies, in large part, due to energy­efficiency assumptions.
the buildings sector, including residential and commercial,
is almost fully electrified across the studies, ranging from 74 to
100%. industrial sector electrification varies more widely
based on assumptions about technology availability and cost,
ranging from 25 to 54%. Finally, in the transportation sector,
most studies assume that light­duty vehicles are almost com­
pletely electrified by 2050, whereas medium­ and heavy­duty
vehicle electrification varies more widely. Hydrogen produced
by electrolysis and used in fuel cells is largely considered elec­
trification. overall, electricity’s share of transportation energy
consumption ranges from 24 to 100%.
Emissions
in these scenarios, electrification was accompanied by
the decarbonization of electricity generation. overall, the
studies that completed new analysis reduced economy­wide
GHG emissions in 2050 by 71–80% below 2005 levels.
However, in the absence of significant decarbonization of
the electric sector, electrification is still expected to result
in GHG emissions reductions; the steinberg study found
that even if modest decarbonization of electricity genera­
tion is achieved (similar to rates with existing policies at
the time of analysis), electrification of end­use services
alone results in a 41% reduction in U.s. GHG emissions by
2050. these projected emissions reductions are due to both
the high efficiency of electric end­use technologies and
that new electric power supply is projected to be dominated
by low­carbon resources, both high­efficiency natural gas
combined cycle and renewable energy.
3 × 2017
2 × 2017
Steinberg et al. (2017)
Williams et al. (2015)
Weiss et al. (2017)
Iyer et al. (2017)
AEO 2017 Reference
MCS (2016)
AEO (2017)
1990 2010 2030 2050
(2050 Projections)
ieee power & energy magazine 93
94
table 1. The methodology and objectives of the studies reviewed in this analysis.
Methodology Methodology Methodology
Electrification for Estimating for Estimating for Estimating
Entity Responsible Scenario Name in Model Used for Building Industry Transportation Decarbonization
Study for Modeling Study Study Objective Scenario Analysis Electrification Electrification Electrification Target
EIA AEO EIA Reference case Business as usual NEMS, Business as usual Business as usual Business as usual None
(2017) reference case energy system plus technology plus technology plus technology
optimization model improvement over improvement over improvement over
time time time

ieee power & energy magazine

Iyer et al. Pacific Northwest High electrification Top-down energy GCAM-USA Literature review; Based on Steinberg Electric vehicles Exogenous;
(2017) National sector assessment integrated assumes significant et al. (2017) assumed cost economy-wide net
Laboratory to reach low GHG assessment and adoption of competitive GHG emissions
goal optimization model electrified with internal reductions of 80%
technologies for combustion engine below 2005 levels
end uses by 2030 for light- by 2050
duty vehicles and
2050 for medium-
and heavy-duty
vehicles
Steinberg National High electrification Bottom-up ReEDS, electric Expert judgment Based on market Adoption rates Exogenous; power
et al. Renewable Energy with CO2 emissions energy sector sector capacity- potential analysis exogenously sector 83% below
(2017) Laboratory goal assessment of high expansion (EPRI 2010) determined; 2005 levels by
electrification optimization model vehicle stock 2050
with and without model estimates
decarbonization electrification
potential given
adoption rates
Mid- Pacific Northwest Benchmark Top-down GCAM-USA, Exogenous; Exogenous; Exogenous; Exogenous;
century National vision for high integrated technology technology technology economy-wide net
Strategy Laboratory decarbonization, assessment model, assumptions assumptions assumptions GHG reductions of
(2016) including for the optimization model developed by the developed by the developed by the 80% below 2005
energy sector U.S. Department of U.S. Department of U.S. Department of levels by 2050
Energy Energy Energy
Weiss et The Brattle Group Full electrification Top-down No model; Exogenous; water Omitted Exogenous; light- Output: 72%
al. (2017) of heating and calculation of calculations made and space heating duty, commercial, below 2015 levels
transport electrified heating based on EIA AEO gradually fully and freight land- by 2050
and transport 2015 electrify based transport
fully electrify
Williams Energy and Mixed case Top-down; PATHWAYS: Specified by user; Specified by user; Specified by user; Exogenously
et al. Environmental (balanced feasibility of low energy sector, implemented in implemented in implemented in determined; 80%
(2014) Economics (E3) renewables, GHG emissions, accounting model model through model through model through below 1990
nuclear, fossil including for the adoption curves adoption curves adoption curves levels by 2050
fuel with carbon energy sector
capture and
storage)

july/august 2018
 Transportation Sector 2050 Buildings Sector 2050 Industrial Sector 2050
35
Delivered Energy in 2050 (EJ)

30
25
20
15
10
5
0
O

er

er

er

g
ss

ss
am

am

am
er

er

er
C

C
AE

AE

AE
Iy

Iy

Iy
M

ei

ei

M
nb

nb

nb
illi

illi

illi
W

W
ei

ei

ei
W

W
St

St

St
Electricity Other Energy

figure 4. Electrification in 2050 by sector as projected by various studies. The total height of the bar is delivered energy
in 2050; the darker-shaded portion is amount of delivered energy derived from electricity. Note: *Weiss et al. assump-
tions of full electrification for residential and commercial heating and light-duty vehicles and trucks applied here to the
2017 AEO. Weiss et al. used the 2015 AEO and do not address industrial uses.

Other Metrics a significant impact. the Weiss study discusses the potential
in addition to the commonly used metrics described previously, for utility sales growth.
there are many additional metrics that could inform decision
makers and define the potential opportunities and challenges Societal and Consumer Costs
that could arise from a new era of electrification. table 2 builds the electrification of end­use services will certainly impact
on the loftus approach to show a variety of such metrics that consumer energy expenditures, but the magnitude and direc­
are either quantitatively or qualitatively discussed in at least tion of the impact will depend on regional electricity rates,
one study. a broader consideration of these metrics in future retail rate structures, the relative efficiencies and prices of
analyses will improve the community’s understanding of the end­use technologies, and the relative price of displaced fuels
economic and environmental impacts of electrification across (oil, natural gas, and propane). some studies quantify expen­
multiple geographic and temporal scales. ditures but vary in their treatment of costs; some include
price impacts, others include bill impacts, and still others
Pace of Change and Economic Growth include societal costs. For example, steinberg et al. explore
the availability and rate of technology replacement are key the wholesale price of electricity and Williams et. al include
metrics for determining the degree of electrification in a both household and overall costs.
scenario. a few studies in this review explore the pace of current results emphasize the importance of conventional
change required to meet a specific policy target, assessing technologies’ capital and fuel­cost assumptions and energy­
how quickly the electricity system and supporting techno­ efficiency technology and policy assumptions, in addition to
logies would need to evolve
across regions to avoid put­ table 2. Treatment of alternative metrics across studies.
ting investments at risk. For
AEO
example, the Williams study 2017 Iyer Steinberg MCS Williams Weiss
notes that the rate­of­tech­
Pace of change and
nology replacement limits how economic growth
quickly and to what extent sig­ Societal and consumer
nificant electrification could energy costs
be achieved in certain end Energy productivity
uses with long­lived assets
Resilience and reliability
(e.g., industrial boilers). How­
ever, even without forcing an Energy security
early retirement of long­lived Other environmental
assets in modeled scenarios, impacts
the increased penetration of Key: No treatment; Qualitative treatment ; Quantitative treatment.
electric technologies can have

july/august 2018 ieee power & energy magazine 95

 the projections of electric technology cost and performance, in
assessing the net costs of electrification scenarios in future stud­
ies. For example, layering on additional aggressive assump­
tions about improved energy efficiency of end­use technologies
lowers prices as less electric infrastructure is required to meet
the same level of service.
Energy Productivity
energy productivity (output, often gross domestic product, per
unit of energy) is an important consideration for electrification
and electrification studies since many electric technologies (e.g.,
electric vehicles) are more efficient than their fuel­powered ana­
logs, even when considering the energy consumed at power plants
to produce the electricity under certain generation assumptions.
energy­productivity improvements are generally discussed quan­
titatively. For example, Williams et al. identify significant oppor­
tunities for industrial energy productivity including electrifying
industrial boilers and heating and accelerating the use of electric
arc furnaces. the midcentury strategy (mcs) notes that energy
productivity improvements in the industrial sector are an impor­
tant strategy for reducing emissions and would benefit greatly
from improved energy efficiency of heating and motors.
While not an explicit focus of many studies, the underlying
assumptions for how efficient electric technologies can become
affects overall energy productivity and impacts the opportuni­
ties for electrification. in particular, these effects are synergistic:
more efficient electric technologies allow greater electrification
through the more efficient use of electric infrastructure, and
electrification enables greater overall energy efficiency relative
to combustion technology alternatives. the steinberg analysis
provides an example of the former, showing how more efficient
technologies can provide a buffering effect for electric infra­
structure build out, allowing greater electrification with a lesser
need for new infrastructure. Without the underlying efficiency
assumptions in each of the studies, total electricity demand in
the electrification scenarios would be even higher than the dou­
bling in 2050, highlighting the importance of technology effi­
ciency assumptions in scenario outcomes.
Resilience and Reliability
electrification’s impact on energy­system resilience and
reliability depends on a number of disparate system­level
characteristics, including the extent to which flexibility is
built in to electricity supply and demand, the electricity gen­
eration mix, relative threats among various fuel types and
end­use technologies, consumer choice, and supporting poli­
cies, among other factors. as such, studies do not conclude
whether electrification has a positive or negative impact on
these system attributes but do provide some insight into infra­
structure and operational challenges, along with the potential
benefits of flexible electric load. the steinberg study finds
that electrification could reduce peakiness, by increasing
system utilization and decreasing need for capacity assets
that are only used a fraction of the time, which may improve
energy system reliability.
96 ieee power & energy magazine
Energy Security
some studies address infrastructure needs or resource poten­
tial for given resources such as biofuel production, but none
have systematically addressed the impacts of electrification
on energy production and trade. U.s. electricity is predomi­
nantly a domestic resource (98% domestic versus 55% for
crude oil and 73% for petroleum as reported by the eia),
and further electrification could impact policy goals includ­
ing energy independence, enhanced manufacturing, and the
balance of international trade. the Nems model used to
produce the aeo explores trade flows and some manufac­
turing issues. the Williams study shows how transportation
electrification and the increased reliance on domestic fuel
sources could reduce exposure of the economy to volatile
global commodity markets.
Environmental Impacts
Finally, the electrification of end­use services that currently
rely on direct fuel use would impact air quality and other
environmental issues, but the magnitude and direction of the
impact will depend on the extent of electrification, the type of
fuel displaced, and the evolution of the electricity generation
mix. Zero tailpipe emissions from battery and fuel­cell elec­
tric vehicles, for example, can reduce local air pollution and
health impacts in major cities. Depending on the source and
type of the electricity production, however, reduced local air
pollution could lead to changes in air pollution at the site of
the electricity generation. energy­economic models currently
track criteria pollutants at a regional scale but not in further
detail, particularly for end­use services. in this review, the stud­
ies did not explore local air pollution. With the exception of
the mcs, no studies mentioned the impact on water quality or
resources. Water impacts, which could vary dramatically based
on the future generation mix, merit further exploration. Future
electrification studies could assess temporal and geographic
distribution of criteria pollutant emissions (at both a local and
statewide level), local human health and medical expenditure
benefits, and impacts on water quality or resources.
Summary and Recommendations
for Future Research
the results reviewed here show the high potential for electri­
fication to alter the future U.s. energy pathway. such a large
transition away from the current trajectory, for example, in
aeo, warrants continued research, and we conclude by high­
lighting several areas ripe for further exploration.
Examine Sensitivity of Electrification
to Key Parameters Across Models
While these scenarios are either explicitly or implicitly
designed to achieve increased electrification, they differ in
the manner in which they achieve it. to understand the dif­
ferent pathways through which each model achieves electri­
fication, it is essential to understand the inner workings of
energy models used to explore electrification.
july/august 2018
 While these scenarios are either explicitly or implicitly
designed to achieve increased electrification, they differ
in the manner in which they achieve it.
Beyond the economic parameters that impact a consum­
er’s decision to electrify (e.g., up­front technology costs and
operating costs), other parameters within the model architec­
ture impact the extent to which a given model pursues elec­
trification, including consumer choice parameters, which can
be oversimplified or based on historical preferences for elec­
trification that may not reflect future preferences; barriers to
switching to electricity from direct fossil fuel use; support­
ing infrastructure requirements, such as the ratio of electric
vehicles to charging stations; restrictions on the pace of addi­
tional generating capacity; and limited ability to incorporate
research­, development­, and demonstration­stage technolo­
gies such as electrified heavy­duty vehicles. Understanding
the sensitivities of results to these parameters and how dif­
ferent models treat them is important for contextualizing and
evaluating model outputs.
We therefore recommend that studies prioritize and report
sensitivity analysis on key parameters and a new model in­
tercomparison study be carried out that explores the repre­
sentation and treatment of key parameters for electrification
across models.
Assess Electrification with
Additional Objectives and Metrics
additional objectives, side effects, and perspectives merit
further exploration to successfully investigate the system­
and society­wide implications of electrification. For exam­
ple, studies will need to incorporate and report a broader
range of metrics, including economic growth; energy costs
for consumers and society; energy productivity; and the
resilience, reliability, and security of the U.s. energy sys­
tem. analyses of a broader scope may be required to fully
explore the potential impact of electrification on globalized
commodities such as oil.
similarly, there are potential challenges associated with
significant growth in electricity demand that are not fully
explored in the studies assessed: the need for additional grid
infrastructure, such as transmission expansion; the stability
of grid operations; and the viability of other fuel supplies,
such as heating oil. electrification could also propagate side
effects throughout the economy and society, including effects
on technology leadership, manufacturing, jobs, cybersecu­
rity, health, and income and wealth distribution.
Given a range of potential effects, one key question for
electrification scenarios is value to consumers, including cur­
rent and future cost­effectiveness compared to other options.
july/august 2018
moving beyond energy price to incorporate a richer assess­
ment of consumer value will be an important part of assessing
the benefits of electrification. First, the value of electrification
should be modeled to include the location and time­dependent
benefits and costs of increased electricity consumption. sec­
ond, the value of electrification should be expanded to include
noneconomic energy service characteristics that drive choice
(e.g., noise pollution, convenience, safety, and performance).
in addition to system­level costs, the economic interests
of different stakeholders need to be more fully examined.
For example, stress to existing utility business models, caused
in part by lack of electricity demand growth, could be alle­
viated by increased demand associated with electrification.
changing job opportunities and availability could be a con­
sequence. Jobs have not been addressed frequently in the
current literature, perhaps because most studies look across
the energy system and would not capture movements among
subsectors. Future electrification studies could explore pace
of change, economic growth, and associated stakeholder
impacts through investment needed and avoided along with
other job and economic activity related metrics.
Increase Resolution of Analyses
on Technology and Energy System
the studies reviewed here necessarily rely on fairly gross
assumptions about the extent and mechanisms of electrifi­
cation within each sector. a higher­fidelity analysis of the
economic and technical potential of electrification in each
sector, subsector, process, and technology would help refine
the inputs to large system models and, importantly, also help
move discussions on electrification from broad visions to
concrete implementation plans. such sector­ or technology­
specific explorations should address feasibility, cost, con­
sumer preference, cobenefits, system interfaces and enabling
technologies, and barriers to deployment.
in support of these studies, more expansive monitoring of
existing electric technologies is needed to provide additional
and higher­quality data on uptake, performance, and life­
cycles of existing electric technologies, such as heat pumps
and electric vehicles. as an example, consider that, in eia
data, consumption for electric vehicles charged at residential
or commercial locations is included in the residential and
commercial sector data and not in the transportation cat­
egory. While this amount is still relatively small compared
to total electricity consumption, it will become more signifi­
cant as electric vehicle deployment grows. improving data
ieee power & energy magazine 97
 collection would allow updating past work, inform the line
of exploration recommended above, aid in the development
of guides and procedures for implementing electrification,
and support future technology development by guiding pilot
projects and larger­scale investment decisions.
the studies also focus on a system­level view of the
electricity sector, which includes generation facilities
and transmission needs in scope, but they generally did
not explored the impact of electrification on the distribu­
tion system. this inclusion could be especially critical if
buildings with significant amounts of electrified technolo­
gies (such as electric vehicles) are clustered on the same
distribution system. the impact of electrification on the
distribution system could be substantial and needs to be
better understood.
Finally, electricity demand growth in the models re­
viewed here derives largely from the first two modes of
electrification: expansion of existing services and fuel
switching. modeling the third mode—new uses—is neces­
sarily more difficult. Nonetheless, emerging technologies,
like autonomous vehicles and information and communi­
cation technology services, merit additional consideration.
While estimating the effects of new technology is plagued
by high uncertainty, expanding the common group of sce­
narios used here to envision a range of new sociotechnical
deployment paradigms could highlight feasible and desir­
able implementation pathways.
Assess the Impact of Policy on Electrification
adopting these proposals would provide insight into the
costs, benefits, and feasibility of electrification. additional
research should also be conducted on how policy enables or
inhibits electrification pathways. there are several federal,
state, and local policy factors to consider regarding electri­
fication, both directly and indirectly. some states and cities
(rhode island, for example) have already made an overarch­
ing policy decision to support increased electrification, in
part through fuel switching. other entities may have existing
policies, designed for reasons unrelated to electrification,
that impact electrification.
Historically, electricity and gas have been regulated
separately, limiting the opportunities for electric­gas coor­
dination or strategic fuel switching. Fuel specific policies
and incentives have developed over time and may inadver­
tently limit electrification. For example, fuel­specific effi­
ciency incentives could limit additional efficiencies from
fuel switching. rate design, energy efficiency investment,
decoupling policies, and the role of utility infrastructure
investment are all major considerations that have yet to be
fully examined in the light of greater electrification.
the scenarios reviewed demonstrate that the United states
could be embarking on a significant shift in energy system
dynamics driven by increased electrification. this new era of
electrification could perhaps be similar in relative scale and
pace to the original U.s. electrification. the recommendations
98 ieee power & energy magazine
of this article aim to promote clear and comprehensive analy­
sis of the costs and benefits of these potential pathways.
Acknowledgment
the views of the authors do not necessarily reflect those of
the United states Government or any agency thereof.
For Further Reading
G. iyer, l. clarke, J. edmonds, P. Kyle, c. ledna, H. mcJeon,
and m. Wise, “Gcam­Usa analysis of U.s. electric power
sector transitions,” Pacific Northwest National laboratory,
tech. rep. PNNl­26174, 2017.
P. J. loftus, a. m. cohen, J. c. s. long, and J. D. Jenkins,
“a critical review of global decarbonization scenarios: What
do they tell us about feasibility?” Wires clim. change, vol.
6, pp. 93–112, 2015. doi: 10.1002/wcc.324.
D. steinberg, D. Bielen, J. eichman, K. eurek, J. logan,
t. mai, c. mcmillan, a. Parker, l. Vimmerstedt, and e.
Wilson, “electrification & decarbonization: exploring U.s.
energy use and greenhouse gas emissions in scenarios with
widespread electrification and power sector decarboniza­
tion,” National renewable energy laboratory, tech. rep.
Nrel/tP­6a20­68214, 2017.
J. Weiss, r. Hledik, m. Hagery, and W. Gorman, “elec­
trification: emerging opportunities for utility growth,” the
Brattle Group White Paper, cambridge, ma, tech. rep.
7376, 2017.
J. H. Williams, B. Haley, F. Kahrl, J. moore, a. D.
Jones, m. s. torn, and H. mcJeon, “Pathways to deep de­
carbonization in the United states,” the U.s. report of the
Deep Decarbonization Pathways Project of the sustainable
Development solutions Network and the institute for sus­
tainable Development and international relations. revision
with technical supplement, san Francisco, ca, Nov. 16,
2015, 2014.
(2016). United states mid­century strategy for deep
decarbonization, the height of the bars indicates deliv­
ered energy in 2050, which varies across the studies in large
part due to energy efficiency assumptions. [online]. avail­
able: https://unfccc.int/files/focus/long­term_strategies/
application /pdf/m id _centur y_ strategy_ repor t­f ina l
_red.pdf
Biographies
Carla Frisch is with the U.s. Department of energy, Wash­
ington, D.c.
Paul Donohoo-Vallett is with the U.s. Department of
energy, Washington, D.c.
Caitlin Murphy is with the National renewable energy
laboratory, Golden, colorado.
Elke Hodson is with the U.s. Department of energy,
Washington, D.c.
Nathaniel Horner is with the U.s. Department of en­
ergy, Washington, D.c.
p&e
july/august 2018
 By J.D. Willson and M.E. Long

©istockphoto.com/theerapong28
System
T
TwenTy-five years ago, J.D.
willson was the control center manager
with one of the United states’s largest
utilities—the Pennsylvania, new Jersey,

Restoration
Maryland interconnection (PJM)—
and also served as chair of the ieee
Power system restoration Training
Task force. M.e. Long was a training
manager at Pepco and chair of the ieee
working group on operator Training

Readiness
(wgoT). PJM was a pioneer in study-
ing ways in which utilities could inter-
connect their power systems to increase
reliability while reducing overall costs
and also led efforts to improve system
operator training.

Interconnected Power
The Evolution in North America Systems Drove Need
for Restoration Plans
interconnected power systems origi-
Digital Object Identifier 10.1109/MPE.2018.2824101
nally operated so that they could man-
Date of publication: 18 June 2018 age the loss of the single largest unit,

july/august 2018 1540-7977/18©2018IEEE ieee power & energy magazine 99

 Millions with seeming increases in major storms and a growing
Year (89–15) Cause of Outage interest in responding to terrorism, we wondered, consider-
Impacted
3/13/89 Geomagnetic Storm 7 ing all the nerC requirements, how prepared our industry is
10/17/89 Earthquake 1.4 for a restoration of any major portion of the north american
7/7/91 Windstorm 1 power system and found that we’ve come a long way. How-
8/24/92 Hurricane Andrew 1.4 ever, there is no room for complacency, as figure 1 suggests.
The first large blackout in north america occurred on
10/4/95 Hurricane Opal 2
9 november 1965 and impacted more than 30 million people.
7/2–3/96 Overheated Line 2
Largely as a result of that blackout, the north american electric
8/10/96 Heat 4
reliability Council was formed in 1968 and began work
1/98 Ice 3.5
as a nonprofit organization with the goal of developing and
5/31/98 Windstorm 2 sharing a set of reliability-related principles to support
8/14/03 Multiple–Not Weather 55 bulk power system operations and so minimize the risk of
9/19/03 Hurricane Isabel 4.3 future blackouts.
8/31/05 Hurricane Katrina 2.6 working under the north american electric reliability
12/8–12/07 Ice 1 Council umbrella, industry experts developed and approved
2/26/08 Equipment Failure-Fire 4 reliability-related system operating principles published as
9/13/08 Hurricane Ike 2 the NERC Operating Manual (with compliance being volun-
12/12/08 Ice 1.5 tary). specific policies and supporting documents (“Policy 5:
8/27–28/11 Hurricane Irene 5 emergency operations,” “Policy 8: operating Personnel and
9/8–9/11 Human Error 5
Training,” “appendix 8B1: suggested items for system opera-
tor Training Courses,” and the “system restoration reference
10/11/11 Snowstorm 2
Document”) dealt with emergency operations and training.
6/29/12 Thunderstorms 3.8
at the same time that the north american electric reli-
10/29–30/12 Hurricane Sandy 8
ability Council was working on developing its operating pol-
figure 1. Most major outages are caused by weather. icies, the ieee formed a system restoration working group
(Psrwg) that began to study technical issues associated
or critical transmission element, without causing a more wide- with system restoration and report its findings at conferences
spread system failure or blackout. as more and more systems and workshops. almost 100 of those reports are included
were interconnected, it became obvious that, in some instances, in the 2000 book Power System Restoration: Methodologies
weather or other events would likely result in the simultaneous and Implementation Strategies by Psrwg Chair Mike M.
loss of more than one facility. This caused power system engi- adibi and published by wiley-ieee Press.
neers to build system models to study the ways in which inter- Despite these efforts, on 14–15 august 2003, another major
connected systems could separate and ways to safely reconnect blackout impacted approximately 50 million people. following
those separated systems in the shortest possible time without that blackout, experts from Canada and the United states con-
damaging expensive equipment. This was an exciting time for ducted a very detailed analysis that showed, among other things,
power system engineers and resulted in the formation of both a need to provide greater assurance that utilities were follow-
the ieee system restoration working group and the ieee ing the policies established in the NERC Operating Manual,
system restoration Training Task force. including policies associated with developing restoration plans
in this article, we look at where the industry was 25 years and providing system restoration training to operators. in 2005,
ago when power system engineers were just beginning to U.s. President george w. Bush expanded the authority of the
develop company-specific system restoration plans and train federal energy regulatory Commission (ferC), giving it the
their system operators in the principles behind those plans. we power to select an entity for developing and enforcing manda-
compare that to the situation today, when virtually every entity tory reliability standards and providing ferC the authority to
registered with the north american electric reliability Cor- review and approve those standards. The north american elec-
poration (nerC) as a reliability coordinator (rC), balancing tric reliability Council reformed itself as the north american
authority (Ba), or transmission operator (ToP) is required to electric reliability Corporation, and ferC selected nerC
✔ develop a technically sound restoration plan to, among other things, establish a set of operational standards
✔ staff its control room with system operators who have with mandatory compliance. entities found noncompliant are
attained a nerC certification to attest to a minimum subject to fines of up to Us$1 million a day per infraction.
set of operating knowledge, including system restora- These standards were intended to identify which func-
tion principles tional division within an electric organization would be
✔ train its system operators on the principles and steps responsible for compliance. Hence, a classic electric util-
supporting its specific restoration plan ity might have one or more nerC-defined roles in meeting
✔ conduct restoration drills. these standards.

100 ieee power & energy magazine july/august 2018

 Question 1992 2016

Do you have a system restoration plan? 84% 100%

Most used Technical

What is the basis for your system restoration plan?
“experience” analyses

100% for
Does your plan include actions of other companies? 50%
RC and TOP

100% for
Do you have a training program to support your restoration plan? Fewer than 50%
RC and TOP

Do you provide annual system restoration training to your system operators? About 50% 100%

100% if identified in
Do you involve plant operators in your system restoration training? About 40%
the RC or TOP plan
100% if identified in
Do you involve neighboring personnel in your system restoration training? About 25%
the RC or TOP plan

Do you use drills in your training? About 60% 100%

100% for
Do you perform system restoration drills to test your restoration plan? About 33%
RCs or TOPs

Do you use a system-specific simulator for your system restoration training/drills? About 25% 100%

What is the average length of your system restoration drills? About 4–8 h About 8 h

figure 2. Changes in drills over the past 25 years.

Many in the industry did not wait for mandatory stan-
dards to develop detailed system restoration plans or system
restoration training for system operators: they began devel-
oping/testing restoration plans and devising/implementing
restoration training long before the nerC reliability stan-
dards became enforceable. These entities recognized that,
with the complexity of interconnected power systems, real-
time operations are not linear, making it is essential that system
operators understand the principles associated with isolated
power systems and then apply those principles to a variety of
possible restoration scenarios.
Restoration Plans Circa 1992
in 1991, the ieee published the paper “Power system res-
toration issues,” emphasizing the need to perform detailed
analytical studies to ensure a sound technical basis for res-
toration plans. in 1992, the ieee wgoT and the Psrwg
surveyed the industry to assess the state of system restoration
training provided to system operators. almost 100 entities
representing different-sized entities from various regions of
both the United states and Canada returned this survey (see
the column headed “1992” in figure 2). Most respondents
(84%) indicated that they did have a restoration plan; however,
“lessons learned through the experience” of system operators,
system engineers, and system managers provided the foun-
dation for these plans rather than detailed technical studies.
Many respondents (about 60%) indicated that they revised
their plans based on lessons learned from system restoration
exercises or drills. while nerC had published its first version
of the guidance document (“electric system restoration ref-
erence Document”) listing several technical issues to address
in system restoration plans, only 67% of the respondents with
restoration plans indicated they used the information in that
document as a basis for their restoration plan.
early restoration plans focused primarily on the actions
of the entity’s own system operators. while approximately
25% of the plans included some actions of system operators
in other companies, only about 40% included the actions of
personnel such as plant operators in the same company. and,
although only 20% involved corporate communications, most
(76%) did include mobilization plans.
early restoration training programs used a variety of train-
ing methods, with almost half using some self-study as well
as lectures led by system operators. approximately 25%
used a company-specific simulator, and fewer than 10% used
a generic simulator. a little over 30% incorporated drills into
the training (figure 3 shows uses of drill evaluations). while
Plan Tools
Evaluate
Processes Staff
Training
figure 3. Drill evaluations identify needed changes.

july/august 2018 ieee power & energy magazine 101

 over 60% included evaluations in their training programs,
the results of these evaluations were varied: over 50% used
the results to modify the training program, fewer than 50%
used the results to identify the need for more training, and
about 35% used the results to modify the restoration plan.
almost 60% of respondents to the 1992 survey indicated
that they performed annual system restoration drills, with
most drills lasting 4–8 h. The objective of most of these drills
was to test the restoration plans, not to meet training objec-
tives. early simulators used in these restoration drills were
expensive and required labor-intensive maintenance, and
most didn’t model electric systems with as much fidelity as
system operators desired. Because these simulators didn’t
act in the same way the actual power system did, system
operators found them to be unreliable for predicting how the
electric system would perform in real time. system operators
felt that generic simulators were good tools for learning the
principles associated with restorations but didn’t provide the
desired real-life experience.
Further Refinements to
Restoration Plans and Training
in 1994, wgoT held a workshop for system operator trainers
that focused primarily on system restoration drills. following
the workshop, wgoT developed the paper “system res-
toration guidelines: How to set Up, Conduct, and evaluate a
Drill,” which was presented at the 1996 ieee Power & energy
society winter Power Meeting. at that time, many utilities
were just beginning to develop more sophisticated system res-
toration plans and associated training programs. The informa-
tion in the paper provided guidance based on lessons learned
from utilities willing to share their experiences. Most utilities
still did not have high-fidelity simulation capability and relied
on seminars and a variety of exercises or drills both to test their
restoration plans and to prepare their system operators.
✔ seminars provided system operators with the principles
of restoration and familiarized them with their respon-
sibilities relative to their company’s restoration plan.
✔ Tabletop exercises simulated an emergency situation
and then used group discussion to determine the best
resolution. These varied in complexity, with some ex-
ercises testing the capability of a single function or
a single complex activity within a function and other
full-scale exercises evaluating the interactive opera-
tional capability of emergency management systems
over a substantial period of time.
✔ Computer-based simulations were rare.
The 1996 paper highlighted the importance of involving
various personnel in the planning and execution of drills and
noted that these early drills served multiple purposes.
✔ Plan improvements reveal planning weaknesses and
resource gaps, clarify roles and responsibilities, and
improve interdepartmental coordination.
✔ system operator training improves individual perfor-
mance and develops proficiency and confidence.
102 ieee power & energy magazine
✔ Company obligations satisfy specific requirements of
certain agencies or organizations and enhance inter-
agency relationships.
The 1996 paper also made several recommendations, includ-
ing the following:
✔ Have clearly defined objectives.
✔ ensure that communication leads all involved to an
understanding of what is going on throughout a drill.
✔ assign a drill coordinator who has time, over several
months, to dedicate to drill planning.
✔ Develop realistic scenario narratives.
✔ avoid accelerating time.
✔ Provide system operators with the same tools as those
used for normal operations.
✔ involve as many departments as are expected to par-
ticipate in an actual emergency.
✔ evaluate to determine whether the drill achieved its
objectives, and identify any improvements needed in
• the plan and supporting processes
• training and staffing deficiencies
• operations tools and equipment needed to better sup-
port a restoration.
Current NERC Requirements
nerC now requires that system operators obtain and
maintain certification to show, among other things, they un-
derstand the principles associated with restoring a power
system that has separated from the rest of the power grid.
nerC reliability standards also require entities to have sys-
tem restoration plans and provide system restoration training
to various classes of nerC-certified system operators based
on those plans.
each rC is required to
✔ have a technically verified restoration plan for the fa-
cilities under its control
✔ provide its system operators with annual system res-
toration training
✔ conduct at least two system restoration drills, exercises,
or simulations each year and invite ToPs and generator
operators (goPs) mentioned in its restoration plan to
participate in its drills, exercises, or simulations.
each ToP is required to
✔ have a technically verified restoration plan for the fa-
cilities it operates
✔ provide annual system restoration training for its nerC-
certified system operators
✔ participate in one of its rC’s drills, exercises, or simu-
lations at least once every two years
✔ develop a physical security plan for critical facilities
it controls.
if a goP has a blackstart resource, it is required to
✔ provide a minimum of 2 h of training every two years
to its operating personnel
✔ participate in one of its rC’s drills, exercises, or simu-
lations at least once every two years.
july/august 2018
 with the increased threat of terrorism, nerC has a set present real staffing challenges. Most entities are hard-
of cyber and physical security standards that include a wide pressed to staff their control rooms fully while also staff-
array of requirements (more than 30) aimed at ensuring that ing for an 8-h drill.
entities implement security processes and controls and train Drills have become an intricate part of a restoration train-
their personnel (including system operators) to follow those ing program. The scope and complexity associated with sys-
processes. These security standards are ever-evolving: some tem restoration drills and the objectives of those drills involve
have been through six revisions as the industry tries to keep huge variations: some drills involve entities from throughout
pace with changes in cybersecurity. north america, while others involve only a region, subregion,
Patrick everly has been chair of wgoT for more than or single entity. some focus on a series of events with multiple
15 years and holds a workshop for system operator trainers objectives; others involve only a single objective.
every year. everly is in a unique position to reflect on changes The most comprehensive drills are the grid security exer-
he has seen in system restoration training and drills because, cise (gridex) drills that nerC has been sponsoring every
in addition to serving as the chair of wgoT, he heads a other year since 2011 (see figure 4). The drill conducted in
company, oesna, that provides system-specific restoration 2015 involved over 4,000 people from more than 300 entities
training and drills for system operators. throughout north america and included the following:
During the 2016 wgoT workshop for system operator ✔ specific objectives
trainers, he collected data from the participants on the cur- ✔ a dedicated coordinator with the support of numerous
rent status of system restoration training throughout north technical volunteers from throughout north america
america. The participants represented the following nerC having sufficient time and resources to develop de-
functional entities: tailed plans
✔ rC ✔ involvement not only of numerous departments within
✔ ToP utilities but also of numerous government entities rep-
✔ Ba resenting different local, state, and federal agencies
✔ transmission owner ✔ realistic, complex scenarios to test response to wide-
✔ distribution provider spread, coordinated physical and cyberthreats
✔ goP ✔ tools for system operators that closely represent those
✔ generator owner. used in real time
The results of the 2016 survey show that system resto-
ration training and the plans supporting that training have
improved significantly over the past 25 years (see figure 2).
in 2016, all responding entities indicated they had a restora-
tion plan based on technical analyses. for entities serving as
an rC or ToP, all restoration plans included actions of sys-
tem operators in other companies, supported by a formal res-
toration training program. all entities had a system-specific
simulator and used the simulator both to test their restoration
plans and to train system operators. all represented system
operators participated in restoration training every year, and
drills were an important part of this training. nerC requires
all entities operating as a ToP or rC to “verify through anal-
ysis of actual events, steady-state and dynamic simulations,
or testing that its restoration plan accomplishes its intended
function.” not surprisingly, all respondents used drill results
to modify their restoration plans. Most drills take about
8 h; however, some entities have conducted drills that last
for several days.

Implementation of Restoration Drills

entities have implemented most of the recommendations
made in the 1996 ieee paper on system restoration drills.
The only recommendation they have been unlikely to im-
plement is the one suggesting that accelerating time be
avoided. This recommendation would be particularly chal-
lenging to implement because an actual widespread res-
toration may take several days to achieve, and this would figure 4. The 2016 GridEx III Report.

july/august 2018 ieee power & energy magazine 103

 ✔ a focus on crisis response and timely information sharing
✔ numerous observers and evaluators.
one of the recommendations for drills in 2006 was to
ensure clear lines of communication during restoration, and
one from the 2015 gridex drill offered a similar recommenda-
tion that the industry take a closer look at its communication
tools because some were redundant and failed to provide feed-
back to those who most needed the information. as the follow-
ing examples suggest, operating entities provide emergency-
related information through multiple channels:
✔ nerC emergency operations standard (eoP)-004
requires entities to file a report with nerC following
any of almost 20 different events, and entities must
report some of the same events to the U.s. Department
of energy (Doe).
✔ nerC eoP-011 requires Bas to report energy defi-
ciencies to their rCs and further requires the rCs to
disseminate the information through the rC informa-
tion system (rCis).
✔ in addition to reporting events to nerC and the Doe,
entities report some of the same events to additional
local and federal authorities.
✔ Many entities voluntarily report events to nerC’s
electricity information sharing and analysis Center;
nerC then shares that information with other entities
and governmental authorities.
✔ in addition to all the required external communica-
tions, companies have numerous internal report-
ing requirements.
while the gridex drills are a good test of the industry’s
response to serious terrorist threats and provide useful infor-
mation to target continuous improvements in crisis manage-
ment, these drills involve only a fraction of the pool of sys-
tem operators. Most system operators gain their experience
in restoration response through participation in drills planned
and conducted on a regional, subregional, or company-spe-
cific basis that focus on the types of scenarios most likely
to occur in that particular region or company (such as wide-
spread equipment damage due to major storms, wildfires, or
extreme temperatures). as shown in figure 1, weather and
other natural phenomena have caused the greatest number
of north american power outages.
Patrick everly reports that the drills oesna has been
developing and coordinating for the southeast reliabil-
ity Corporation (serC) region (and other entities) have
grown more sophisticated over time. The earliest drills were
tabletop and focused on following the steps in the region’s
restoration plan. with each successive year, the drills have
included more simulations, involved more utilities, and,
more recently, incorporated both nerC and ferC. one
such drill involved 25 entities within the serC region, with
several entities representing more than one nerC func-
tional entity (such as an rC that is also a ToP) and included
observers from two regional reliability organizations in
addition to nerC and ferC.
104 ieee power & energy magazine
with the expansion of drill complexity, oesna has found
ways of mimicking tools that can’t be taken offline for use
during drills and exercises. The Chatzy internet applica-
tion (figure 5) allows entities to join a private chat room
and exchange information solely among chat room mem-
bers; it is also used in drills to mimic the rCis used in real
time for communication solely among rCs. Using skype
with handheld phones serves as a backup for communica-
tions, and the finist simulator is used to develop high-fidel-
ity system displays. according to everly, the drills cover a
wide range of both likely and less likely scenarios, such
as response to
✔ widespread equipment damage from storms
✔ islands disconnected from the grid
✔ gas supply disruptions
✔ loss of the primary control facility
✔ total restoration using blackstart resources.
PJM participates in the gridex drills, state- and feder-
ally sponsored exercises, and drills involving all the ToPs in
its area. it also hosts individual ToP restoration drills every
year. in these drills, two or three ToPs perform drills con-
currently, with the focus on following each ToP’s individual
restoration plan using either PJM’s simulator (see figure 6)
or the ToP’s simulator. These smaller-scope drills allow for
more focus on specific ToP restoration plans and coordina-
tion between ToPs and the PJM rC.
interest in north america’s readiness to prepare for and
respond to a widespread power outage has been growing. in
late 2014, ferC, nerC, and the nerC regional entities ini-
tiated a study to determine, in large part, whether the power
industry was ready for such an event. in early 2016, their
findings were published in the Report on the FERC-NERC-
Regional Entity Joint Review of Restoration and Recovery
Plans. That report declared that the existing set of reliability
standards provides good support for both restoration plans
and restoration training. The study did, however, make some
recommendations regarding possible improvements to sys-
tem restoration plans and associated training, including more
focus on
✔ when to hand off control from a ToP to a Ba
✔ scenarios involving loss of computer systems used for
real-time operations
✔ sharing lessons learned, including job aids developed
for drills that could be useful in actual emergencies.
Perhaps the most satisfying assessment of a restoration
drill’s success is whether participation in the drill helped
in a similar real-life situation. for example, Paul Davis,
electric system operations specialist for the Jacksonville
electric authority, has explained that its drills focus on,
among other things, storm restoration. in fact, the author-
ity was in the midst of planning a drill focused on response
to a hurricane just days before Hurricane Matthew struck
in 2016. according to Davis, the preparation for that drill
focused on coordination and communication between oper-
ating entities, closely mimicked the actual restoration effort
july/august 2018
 Welcome to the 2015 SERC Drill–RCIS

PJM RC: MISO, what’s your direct number out there? TVA and ourselves are having October 2015
trouble finding a good one.
serc joined the chat October 2015
serc joined the chat October 2015
PJM RC: As of 14:15 EPT PJM’s EKPC zone has lost all of its generation and appears October 2015
to be fully blacked out. EKPC is performing a thorough assessment, updates to follow.
MISO RC: joined the chat October 2015
peve joined the chat October 2015
PJM RC joined the chat October 2015
serc joined the chat October 2015
TVA_RC joined the chat October 2015
TVA_RC: MISO RC are you on the line October 2015
TVA_RC joined the chat October 2015
TVA_RC: TVA reported approx. 2500 MWs lost in the SE region of TVA. Crews in route October 2015
TVA_RC: LGEE has loss all generation and load in their area. (Total system blackout.) October 2015
MISO RC joined the chat October 2015
MISO RC joined the chat October 2015
NERC joined the chat October 2015
MISO RC joined the chat October 2015
MISO RC: MEPA has currently Moselle #3, #4 & Sylvarena #3 online serving 101 MWs October 2015
total. 161 kV in service. Moselle to Hintonville Moselle to Columbia Columbia
to Plant Morrow Morrow to Purvis Bulk Purvis Bulk to Lumberton Hintonville
to Plant Morrow Hintonville to Bennedale Hitonville to Waynesborrow.

figure 5. The Chatzy Internet application used to replicate the RCIS.

following Hurricane Matthew, and helped minimize resto-

ration time.

Recommendations

Improve Sharing of Lessons Learned

The results of the 2016 wgoT survey, as well as the results
of the lessons learned from gridex iii, show that there is
some reluctance to share lessons learned from drills. almost
all respondents to the 2016 wgoT survey indicated they do
not want either their regional reliability organization or nerC
to monitor or evaluate their drills. similarly, only 24% of
gridex iii participating utilities elected to share their lessons
learned with nerC staff. There is a great deal to learn from
sharing lessons: nerC and the regional reliability organiza-
tions should seek ways of facilitating the sharing of lessons
learned without fear of compliance findings.

Increase the Number of Exercises or Drills

while most all respondents to the 2016 wgoT survey indi-
cated they felt their system restoration training was suf-
ficient to prepare their system operators for a restoration
event, participation in one or two drills a year isn’t likely to
build the type of “auto-response” that system operators and
others need to recognize and respond to an unusual system figure 6. The PJM interconnection training simulator.
event. with only one or two restoration drills conducted each (Used with permission from PJM.)

july/august 2018 ieee power & energy magazine 105

 The results of the 2016 survey show that system restoration
training and the plans supporting that training have improved
significantly over the past 25 years.
year, fewer than half the system operators in an organiza-
tion participate in the exercises. Therefore, it is important to
incorporate lessons learned from each drill in the organiza-
tion’s ongoing training programs.
Commercial airline pilots practice a wide variety of
scenarios on high-fidelity simulators before sitting in a
cockpit ready to address real-life emergencies. By com-
parison, most system operators have limited opportuni-
ties to experience a wide variety of emergency scenarios
before possibly facing a real restoration. More frequent
and more varied training exercises with high-fidelity simu-
lation would support a greater degree of system operator
readiness. The fact that most entities found a need to modify
their system restoration plans based on drill results is a clear
indication that the industry hasn’t yet explored all possible
ways of efficiently separating and restoring portions of the
power grid.
Improve the Efficiency of Internal
and External Communication
with the proliferation of cyber and physical attacks, it is
essential that system operators recognize and report these
events in a timely manner. There is an opportunity for
more information sharing with respect to what cyberattacks
look like to a system operator. entities are required to have
policies in place to ensure as-soon-as possible reporting of
cyberattacks to various authorities. The intent is to have the
authorities analyze and share that information so other enti-
ties will become aware of the event and to increase aware-
ness that an attack has taken place in one entity and could
expand or be duplicated elsewhere. such quick exchange of
information is possible only if a system operator can deter-
mine that an unusual operation or event is reportable, so
sharing information about what these events might look like
is critical.
Benefits of System Restoration
Training and Drills
Many benefits accrue from system restoration training and
drills. Perhaps most importantly, entities have used the lessons
learned to refine their company-specific restoration plans, and
many have shared these lessons learned with others in the
industry. a second outcome has been the development of job
aids for supporting restoration drills that would also be helpful
during a real restoration. a third benefit has been providing
106 ieee power & energy magazine
system operators with near real-time experience in restoring
systems from a state in which their own system is disconnected
from the grid; they are thus better prepared in the event they
face the need to restore their systems.
Conclusions
Many in the industry are doing much more than nerC and
ferC require: nerC reports and ieee technical papers
serve as a sound foundation for both generic and company-
specific system restoration training and drills. Because power
systems are complex and power system operations are not
linear, it is essential that system operators understand the
principles associated with a restoration. Completely accu-
rate prediction of exactly how, when, or where a system
will disconnect from the rest of the grid is unlikely. we must
depend on system operators to apply their knowledge of sys-
tem restoration principles to a wide variety of scenarios—
many of which are unanticipated.
Multiple advances have been made in the development of
system restoration training and drills over the past 25 years.
Today’s system operators are better prepared than ever before,
but there is still an opportunity for improvement. for more
information about wgoT, contact Patrick everly at peverly@
oesna.com.
For Further Reading
M. M. adibi, Power System Restoration: Methodologies and
Implementation Strategies. new york: wiley, 2000.
J. D. willson, “system restoration guidelines: How to set
up, conduct, and evaluate a drill: a report by the operator
training working group,” IEEE Trans. Power Syst., vol. 11,
no. 3, pp. 1619–1629, aug. 1996.
“grid security exercise gridex iii report,” nerC, at-
lanta, ga, Mar. 2016.
ferC, nerC, and regional entities, “report on the
ferC-nerC-regional entity joint review of restoration and
recovery plans,” Jan. 2016.
Biographies
J.D. Willson, retired, was with the Pennsylvania, new Jer-
sey, Maryland interconnection.
M.E. Long, retired, was with the north american elec-
tric reliability Corporation.
p&e
july/august 2018
Join your colleagues at the 2018 IEEE International Forum
on Smart Grids for Smart Cities (SG4SC) that will be held
November 26 through November 28 this year at Thor
Central, EnergyVille, Genk, Belgium.

Learn from experts in energy, telecommunications, and

computing disciplines as we address some of challenges
and opportunities that arise from convergence of data and
energy infrastructures to enable a better life for the citizens
in smart cities.

Panel topics will include:

• Solar and renewables in the city
• Multi-carrier energy systems and storage
• Machine learning for smart grids
• Electromobility
• Cybersecurity
• New roles in the grid, customer engagements
• Smart cities in developing countries
• Smart cities: living labs, training and education

Register before 15 October 2018 to receive early bird pricing

Support opportunities are still available

Visit ieeesg4sc.org for more information
Digital Object Identifier 10.1109/MPE.2018.2835847
 history
G
Gold Street in lower Man­
hattan is a narrow passage laid out
when horse traffic was the predomi­
nant mode of travel. Just south of Ful­
ton Street, a large converted residential
structure at #33–43 bears a plaque with
the words Excelsior Power Co. and
the date A.D. 1888 (Figure 1). while
it may be a curiosity to the casual ob­
server, it is the city’s last remaining pub­
lic sign from the pioneer days of electric
power. two short blocks west of thom­
as edison’s famous Pearl Street station
of 1882, the excelsior Company was
a pioneer in providing power for me­
chanical tasks, as opposed to simple
lighting. it began with the provision of
steam power and then adopted electric
power as it was more efficient. the de­
velopment of the excelsior Company
paralleled the career of inventor leo
daft until mutual business interests
brought them together.
Leo Daft and Company
leo daft was born in Birmingham,
england. the son of a civil engineer,
daft showed an early aptitude for me­
chanical devices. at 15, he joined his
father’s firm as a draftsman and then
entered university studies and subse­
quently worked with his father’s firm
as an electrician. intrigued by develop­
ments in the United States, daft left
england for new York City in 1866 and
worked in a variety of jobs and fields
that included a steam locomotive en­
Digital Object Identifier 10.1109/MPE.2018.2819038
Date of publication: 18 June 2018
108 ieee power & energy magazine
Joseph J. Cunningham
forgotten pioneer
leo daft and the excelsior power company
Early electrification has a rich history in a wide range of diverse geographic loca-
tions, many of which have been highlighted in IEEE Power & Energy Magazine in the
pages of this “History” column. One of the most active areas is lower Manhattan,
New York, in the latter part of the 19th century, when it was a crossroads of inven-
tors, technology, entrepreneurs, and finance and with a large population of people
and businesses to be served.
Within this history, we don’t often hear the stories of the pioneers and com-
panies that experienced a rapid rise and an equally swift fall, as technology was
eclipsed due to rapid progress in the new fields of the day, as was the case with
electrification in the late 1800s. This column presents us with such a story on the
intersection of electrical pioneer Leo Daft and the Excelsior Power Company.
Joseph J. Cunningham returns for his eighth visit to these pages. He has re-
searched and authored numerous booklets, articles, and books on topics such as
industrial electrification, electric utility power systems, and electric rail transporta-
tion. His latest book, New York Power, was published in 2013 by IEEE History Cen-
ter Press. The author and the book were significant references stated in the N.Y.C.
Landmarks Preservation Commission decision to approve the Excelsior building for
landmark status. We welcome back Joe as our guest history author for this issue of
IEEE Power & Energy Magazine.
John Paserba
Associate Editor, History
gineer, drafting, photography, and, fi­ to Greenville, new Jersey, in 1881. His
nally, electric lighting and propulsion. initial successes included elevators in
in electric lighting and propulsion, Boston and industrial motors and light­
daft opened a small lighting company ing; he then tackled street railways and
on Centre Street in lower Manhattan, experimented with electric power on
which expanded to become the daft the ninth avenue elevated railway
electric light Company headquartered in Manhattan.
on lower Broadway near wall Street. By the late 1880s, daft’s firm was
as arc lighting became popular for at its peak with a wide variety of prod­
large spaces, a number of companies ucts, many that were a very sophis­
flourished amid considerable competi­ ticated design for their time. those
tion. daft also pursued motors, which included motors, generators, and con­
required more space, and he relocated trols (Figures 2 and 3). the range of
1540-7977/18©2018IEEE july/august 2018
products and the number of workers
required to produce them necessitated
yet another move to a large factory in
the Marion area of Jersey City. about
that time, the interests of daft and ex­
celsior joined (Figure 4).

Ensuring
The Excelsior
Power Company
the excelsior Power Company began
as the excelsior Steam Power Com­
pany and, in 1873, acquired the assets the Reliable
Performance
of previous steam­plant companies in
lower Manhattan. through a system

of Your
that would be considered bizarre and
arcane by modern standards, excelsior
transmitted steam power to industrial
customers by spinning shafts located
in tunnels beneath the city’s streets.
the company flourished, and, by 1887,
Assets
it operated three large steam engines
with a combined 800 hp in a plant at
• High Voltage Testing
13 Spruce Street. those engines supplied
power over a four­block area to busi­
nesses and newspaper printing plants • High Current Testing
on Frankfort, Spruce, william, and
Beekman Streets. Given the cumber­
• Mobile Cable Testing &
some nature of such a system, the con­
cept of power transmission by electric Commissioning
wire was alluring. See Figure 5 for a
geographic perspective of the area. • Asset Management &
the excelsior electric Company
Life Extension
was formed to enter the electric light
business through the patents of wil­
liam Hochhausen, an inventor and lo­ • Rotating Machines
cal businessman. Hochhausen had de­
veloped incandescent light systems to a
• Substation Automation
degree of perfection that exceeded edi­
son’s contemporary system. the com­ & Equipment Analysis
pany offices at 66–68 duane Street of­
fered a display of his lighting systems. • Advanced Transformer
By 1884 the company was billing itself
Testing
as the sole owner of the Hochhausen
patents, “the most perfect system of
electric lighting in the world.” Such in­ • Overhead Line
flated rhetoric was common in adver­ Inspection
tising of the day but, in this case, was
not unrealistic.
the system provided for the automat­
ic switching of faulted lamps and could
be combined with arc lights for a mix of
small­ and large­scale illumination, an www.kinectrics.com
approach not practical with the edison
system. as was the case with most of the /KINECTRICS /KINECTRICS @KINECTRICS
early pioneers, lighting was the primary
 focus of excelsior, but the provision of
power was not ignored as lighting pro­
duced insufficient return to amortize the
capital investment required.
daft installed a combined power
and light system in Boston in 1884, an­
other in rhode island, electric railways
in Baltimore and upstate new York,
and the first electric elevator in Man­
hattan. a demonstration of his motors
at the 1884 international electrical
exposition in Philadelphia was used to
print an entire edition of the prestigious
trade journal The Electrical World. ex­
celsior’s prime steam­power customers
were printing and publishing houses,
thus excelsior had a market for con­
version to electric power. daft products
figure 1. An intriguing and mysterious plaque in 2018. (Photo used by permis- and excelsior’s system made a per­
sion of Robert Lobenstein.) fect combination.

Save Energy.
Rent Test Equipment.

Advanced Test Equipment Rentals makes renting easy so you can focus your energy elsewhere.
Rent: • Regenerative Grid Simulators • AC / DC Loads
• Programmable Power Supplies • Cable Fault Locators
• Relay Testers • and more...

Advanced Test Start Saving Today!

www.atecorp.com
Equipment Rentals
888-404-ATEC (2832)

The Knowledge. The Equipment. The Solution.

110 ieee power & energy magazine july/august 2018

 OV E R 1400 I N S TA L L AT I O N S E N E R GY E X E M P L A R .CO M
ACROSS 50 COUNTRIES

How much does a

sub-optimal decision cost?
Just like the systems you attempt to optimize,
your energy, time and stakeholder’s tolerance
for sub-optimal decisions is finite.

Obtaining desired answers efficiently

(a)
and accurately is paramount.

Energy Exemplar now offers you two software

solutions that meet these objectives.

PLEXOS is a power, water and gas simulation

tool from which optimal scenarios emerge.

AURORA is an easy to use data and simulation

platform for forecasting and analysis.

Because Energy Exemplar believe optimally

derived knowledge yields power.

(b)

(c)

figure 2. A Daft motor applied to

a (a) fan, (b) pump, and (c) blower.
(Source: Daft Electric Light Company,
“Electrical Power Plants, Daft System,”
New York City, 1888.)
 Flexible. Accurate. Smart.

(a)

(b)

figure 3. (a) and (b) Two sizes of Daft electric motors.

(Source: Daft Electric Light Company, “Electrical
Power Plants, Daft System,” New York City, 1888.)

ART Series
Current Sensors
ART Series sensors are the first
to combine easy-to-use flexible
Rogowski coil technology with
Class 0.5s accuracy. When it comes figure 4. The Daft factory in Marion, New Jersey.
to measuring energy production, (Source: Daft Electric Light Company, “Electrical
Power Plants, Daft System,” New York City, 1888.)
monitoring substation transformer
health or real-time allocation of energy
costs and usage in smart buildings,
the daft Company (through what appears to have
ART Series sensors set the standard.
been a utility company named the electric Power
Company of new York) began operating an electric
At the heart of power electronics. light plant on the excelsior Company’s Spruce Street
property on 1 January 1884, nine months before the
Philadelphia exposition. two months later, excelsior
www.lem.com introduced electric motors for printers and other in­
dustries. demand was strong, and in the fall of 1887,
excelsior began the construction of an addition to the

LEMA12618_PrintAd_ART_SmartIndustry_IEEE.indd 1 5/9/18 2:38 PM

Gold Street steam­plant property it had being made to increase the capacity of an additional plant on Park Place
recently acquired (Figure 6). By July the system to 2,500 hp in the future, near west Broadway contained a 250­hp
1888, six basement boilers provided but no future change in transmission generator for a separate “west Side
steam to a 350­hp steam engine that voltage was noted. to what extent that circuit” that covered the territory from
drove seven 50­hp generators to supply expansion was completed has not come leonard Street to Cortlandt Street west
electric power to its customers. electric to light. of Broadway. that supplied customers
motors were rented at a monthly rate
that included maintenance and repair.
Soon the demand for electric mo­
tive power from printers, jewelers,
leather goods plants, and optical and

XGSLab - the Easy to Use,

metal working firms was such that two
separate systems evolved. when the
excelsior Building supplanted the old
Spruce Street property in July 1888,
the company supplied power at 110 V Full-Featured Grounding Solution
over the “east Side circuit” to more
than 100 industrial customers in the XGSLab is the full featured grounding solution that can
area of dense commercial development take you from a basic single-soil-layer step and touch
between Broadway and the east river
analysis to the most advanced multilayer/zone soil
south of the Brooklyn Bridge, which
models. All this comes at a fraction of the base price
was the prime territory of edison’s 1882
Pearl Street plant.
and cost of ownership of comparable packages.
excelsior’s customers employed
· Ground grid design and grounding system analysis
motors of 0.5–15 hp in the fall of 1887;
· Uniform, multilayer and multi-zone soil models
the most common sizes were 1.5, 3,
and 7 hp. the majority of those cus­
· Step and touch potential analysis
tomers were located on Spruce, Frank­
· Below and above ground systems
fort, Beekman, Fulton, John, william,
· Lightning effects and electro-
Pearl, and nassau Streets and also on
magnetic interferences
Maiden lane. Still more were located · Fault current distribution
on side streets and the waterfront, at · Time and frequency domain
Burling Slip, James Slip, and old Slip.
a few were located west of Broadway.
Most streets, lanes, or alleyways in
the area had at least one excelsior cus­ Global Solution
tomer. the maximum distance from
the power station appeared to have The only software in the
been 0.5 mi. market that takes into
additional motor loads included
machine tools, hoists, a few elevators,
account International (IEC),
book binding, tobacco processing, res­ European (EN) and USA
taurant ventilation, and a saw mill. in (IEEE) standards
some instances, one motor appeared
to have performed two functions, even
those as widely different in character
as an elevator and a paper cutter. often, Explore more online and
one customer was the primary lessee request a free demo copy at:
while businesses on other floors used
the power as provided by agreement. www.EasyPower.com/grounding
a total of 310 hp was listed—equal
to 231 kw, which would have required
® EasyPower is the exclusive representative
over 2,100 a given the pressure of Power made easy. of XGSLab software in the USA and Canada.
110 V. an excelsior publication dated
1 July 1888 indicated that plans were

july/august 2018 ieee power & energy magazine 113


A: Edison Duane Street Station
B: Excelsior Office
C: Second Excelsior Station
D: First Excelsior Plant
E: Excelsior Building
F: Edison Pearl Street Station
G: Edison Annex Station
figure 5. An 1899 map of the territory supplied by the Excelsior/Daft system.
(The author would like to thank Robert Colburn of the IEEE History Center for
locating the map of the area, source: Wikimedia Commons.)
114 ieee power & energy magazine
at a new standard of 220 V. as of July,
1888, 25 customers imposed an 82­hp
load or approximately 60 kw or 278 a
at 220 V. the motors were rated 1.5, 3,
and 7 hp, most in printing or lithogra­
phy with a few in other fields such as
tobacco working. the primary streets
were Broadway, Barclay, Greenwich,
warren, and leonard, with a few on
west Broadway.
Customers on both systems were
enthusiastic and declared their satis­
faction with the system that replaced
steam (Figure 7). excelsior also mar­
keted both arc and incandescent daft
lighting systems. this system peaked
about 1890 with some 200 customers,
to which it supplied a total of at least
400 hp or 3,000 kw, but then began a
rapid decline over the next few years.
Competition
while the daft/excelsior motors were
a “quantum” improvement over the use
july/august 2018
 Soon the located at each machine operated only
as needed without the continuous op­
tools by shafts and belts soon seemed
as archaic as the steam power it had re­
demand eration necessitated when a central placed. the Sprague motor freed sub­
daft motor was installed. Plant opera­ stantial space for other use, and space
for electric tion was simplified and more economi­ has always been a premium in Man­
motive power cal, and it was preferred by unions for
the enhanced safety of the elimination
hattan. that inherent efficiency and
flexibility, plus the compatibility with
from printers, of belts and shafts. By comparison, a edison lighting circuits, placed it well
large central daft motor connected to ahead of the competition.
jewelers, leather
goods plants,
and optical making life visibly safer
and metal
working firms
was such that
two separate
systems SpanLite TM

evolved. Self-Illuminated
Power Line Marker

of spinning shafts under city streets Meets FAA Advisory

for the delivery of steam power, they Circular 70/7460-1L
were not without drawbacks. Belts and (Dec. 2015)
shafts were still required on the cus­
tomer premises to transmit power to Installs directly on
live lines up to 500 kV
the machine being driven; essentially,
the electric motor simply replaced the
steam­power shaft. the motor had to
be kept running continuously during
working hours so that any user could
call upon power as needed (Figure 8).
another pioneer had paralleled the
efforts of daft. Frank Julian Sprague,
a graduate of the U.S. navy electrical
training program, the first in the na­
tion, had installed edison systems. in
so doing, he developed mathematical
formulas for system engineering and
perfected an efficient three­wire distri­
bution system that improved the edison
system to triple the area of distribution
and simultaneously reduce the copper
Bird LED Obstruction Low Line
requirement by 62%. Sprague sought a
Diverters Lights Flags & Markers
practical, efficient motor that would be
adequate to a wide variety of indus­
trial applications.
t he Sprague motor operated at
a constant speed, was compact, and
could be located at the workbench
making life visibly safer
800-722-8078 • pr-tech.com
6709

without shafts or belts. Small motors

july/august 2018 ieee power & energy magazine 115


figure 6. The Excelsior Power Building, 1888.
(Source: Daft Electric Light Company, “Electrical
Power Plants, Daft System,” New York City, 1888.)
DF and Electrifill Dielectric Fluids • Transformer Fluids • Solguard ACSR Anti-Corrosion Greases
Acetylene Black for Semi-Conductive Compounds • Naptel and Soltexfill Telecom Filling/Flooding Compounds
SOL Cable Fluids Ad - HALF PG - 011018.indd 2
116 ieee power & energy magazine
figure 7. The Excelsior catalog comparison of a compact Daft
electric motor with the space requirements for the single cylinder
steam engine that it replaced in the plant of Messrs. Snyder &
Black at 92 William Street. (Source: Daft Electric Light Company,
“Electrical Power Plants, Daft System,” New York City, 1888.)
Peak
SOLTEX
COMPOUNDS
AND FLUIDS:
Reliability
With your reputation on the line, get the compounds and
fluids trusted by major cable manufacturers, utilities and
electrical contractors. Soltex gives you high-performance
products, technical expertise, responsive and agile service,
plus the consistent quality you can only get from a world-
class manufacturing facility.
USA +1 281 587 0900 CANADA +1 613 966 8881
orderentry@soltexinc.com | soltexinc.com
1/10/18 2:29 PM
july/august 2018
 the Sprague motor earned the ap­
proval of edison as it did not produce
excessive flicker on lighting lines and
so was the only motor permitted on the
edison system. it debuted at the 1884
Philadelphia exposition and within
months had brought profitability to the
edison Pearl Street plant for the first
time as it provided a daytime load when
lights were used minimally, if at all.
daft had an experience similar to
that of the excelsior Company—his
manufacturing company fell rapidly
after its meteoric rise, was sold in 1890,
and was in receivership by 1896. daft
had moved to Seattle in the early 1890s
where he consulted on power projects.
He later returned to england and then
came back to the United States and
worked in the electrochemical field
prior to his death in 1922.
the short, yet intense, success of
daft and excelsior is a classic example
of the perils faced by pioneers in any

Daft/Excelsior Eclipsed
the daft/excelsior Company held its
position for a time. then edison’s sys­
tem expanded rapidly as the Sprague
motor made it fully competitive with
excelsior. in 1886, edison opened an
annex station on liberty Street and,
five years later, a mammoth (for the
time) 50,000­light (about 1,810­kw)
station at 55 duane Street. By 1895,
excelsior was absorbed into a succes­
sion of companies in a combination cre­
ated by utility magnate anthony Brady
and his system expert and inventor
thomas e. Murray. that combination
evolved into the new York edison Com­
pany in 1901, by which time 120/240­
Vdc utility systems had become firmly
established. those were supplied by
neighborhood substations that conver­
ted high­voltage ac from large­scale
central stations to dc for local distribu­
tion systems in an era when ac motors
needed much development and ac dis­
tribution in dense urban areas was still
problematic. dC motors had undergone
some 15 years of development, and dc
distribution allowed the use of battery
assemblies for peak loads and protec­
tion against outages.
the excelsior plant became a sub­
station in april 1899, although it re­
tained standby arc­lamp generators for
another 23 years. industrial customers
that had occupied space on the upper
floors of the Gold Street building re­

quality forged daily

mained there into the 1920s. there­
after, it operated for more than three
decades as a n.Y. edison Company
substation to supply dc in the area be­ www.hughesbros.com
tween Brooklyn Bridge on the north,
Park row and Broadway on the west, info@hughesbros.com
Maiden lane on the south and the east
river to the east.

july/august 2018 ieee power & energy magazine 117

 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
Call now 573-796-3812
ext. 2001
Fax 573-796-3770
www.tanawiremarker.com
TANA WIRE MARKERS
P.O. Box 370, California, MO 65018
118 ieee power
634276_Tana.indd 1 & energy magazine23/03/13 10:38 AM
figure 8. The Excelsior catalog depiction of a single, large, 15-hp Daft motor
that was in the plant of Jos. B. Stilwell & Co. printers at 20 Cliff Street. The single
motor operated a total of 23 machines of three companies. (Source: Daft Electric
Light Company, “Electrical Power Plants, Daft System,” New York City, 1888.)
field. Meteoric rise is often accompa­ steam­power customers able, willing,
nied by an equally rapid decline as a and eager to explore a better approach.
product or technique is surpassed by even more remarkable is that the
technical progress that tends to be very very center of the excelsior system was
rapid in a new field. Just how long did within some 200 ft of the world­re­
the eight­track­tape cartridges of the nowned edison Pearl Street station but
1970s dominate the market? How long is totally forgotten beyond that curious
did the (often overlooked) four­track wall plaque (Figure 1). ironically, the
version that preceded it last? the daft/ daft/excelsior system was predictive
excelsior combine lasted just under a of the future significance of electric
decade, with the true peak being two power for motors rather than light; edi­
years at most. its success was attribut­ son’s focus was electric incandescent
able to a ready market, a ready product, light until Sprague brought his motor
and an area of dense concentration of to the system.
july/august 2018
 Customers
on both
systems were
enthusiastic and
declared their
satisfaction
with the system
that replaced
steam.

another forgotten pioneer, the Ball

electric light Company of toronto,
produced a line of electric motors, gen­
erators, and arc and incandescent lights.
Ball had installations across north
america and as far away as plants in
figure 9. The contemporary appearance of 33–43 Gold Street as a residence in
Venezuela and extensively in the city
2002. (Photo used by permission of Robert Lobenstein.)
of Quetzaltenango, Guatemala, but it
never succeeded in penetrating the new
York market beyond a few dozen iso­
lated installations and a street lighting
contract. the reason for that failure is
a matter of conjecture. Perhaps the in­ THE TRANSFORMER
stallations of daft systems by excelsior INDUSTRY’S FULL LINE
and the Sprague motors on the edison MAGNET WIRE SUPPLIER
system left no territory of concentrated
industrial customers ready for electrifi­ With five experienced
cation by the Ball Company.
in later years, the excelsior station
manufacturing facilities, ENAMEL
INSULATED
was the site of various experimental Rea Magnet Wire is the
installations that included mercury leading wire supplier to
arc rectifiers in 1925 and a single stage the transformer industry PAPER AND
NOMEX® DRY TYPE
6,600­V, three­phase ac, 120/240­V in North America. Offering
“dyna­Motor” rotary converter in the
the most robust product
same period. For many years, it was the
location of the edison Company’s dc line in the industry and FORMVAR
training center and the storage site for a quality system that is
OIL FILLED
dc distribution materials. retired by second to none, Rea is DISTRIBUTION
Consolidated edison as a dc sub­ your ideal partner within MAGNEFLEX
st ation by 1961, it was sold in 1978 the transformer market.
for redevelopment into a residen­
tial property.
as has been noted in countless his­ TRANSFORMER POWER
tor ical accounts, novels, and even LEAD CABLE
For more
in science fiction tales, new York information,
City, particularly Manhattan, never
completely loses touch with its past. visit ReaWire.com/IEEE
Skyscrapers rise, are demolished, and or call 800-732-9473 CTC
 replaced by even larger structures. en­
tire neighborhoods evolve and change
as areas once known for a particular
occupation, be it meat packing, freight
yards, finance, the jewelry trade, or
ship building, evolve to find a new pur­
pose. recognition of the historic value
of its structures came to new York City
relatively late compared to other world
cities. Formal legislation for the pres­
ervation of landmarks did not come
about until the late 1960s, and many of
those efforts were fought by finan­
cial interests.
while the magnificence of Grand
Central Station and the Public library
have been preserved and enhanced,
Penn Station and the great downtown
Singer Building are among those lost.
Still, history lurks in the nooks and
crannies of side streets. an old house,
a church, or a commercial building of
120 ieee power & energy magazine
some significance, they are links to a
long ago and now dim, yet fascinat­
ing, past. the electrical industry has
its own past, located appropriately on
a narrow and dark street in the old­
est part of the city, the Gold Street sta­
tion of the excelsior Power Company.
a lone remnant of the earliest pioneer
days of electric­power distribution,
probably the oldest such in the region,
it was appropriately designated a land­
mark by the City landmarks Preserva­
tion Commission in december 2016.
Unfortunately, no samples of daft’s
pioneering technology remain, at least
not within the structure so far as is
known (Figure 9).
For Further Reading
M. Caratzas, excelsior Steam Power
Company Building, 33­43 Gold Street,
Manhattan, landmarks Preservation
Commission, december 13, 2016,
designation list 492, lP­0962. [on­
line]. available: http://www.neighbor­
h o o d p r e s e r va t io n c e n t e r.o r g /d b /
bb_files/2016—excelsior­Steam­Power­
Company­Building.pdf
J. l. Sprague, “Frank J. Sprague
invents: the constant speed dc electric
motor,” IEEE Power Energy Mag., vol.
14, no. 2, pp. 80–96, Mar.–apr. 2016.
J. J. Cunningham, New York Power.
new Brunswick, nJ: ieee History
Center Press, 2013
Electrical Power Plants, Daft Sys-
tem. new York: daft electric light
Company, 1888.
F. rowsome, Jr., The Birth of Elec-
tric Traction: The Extraordinary Life
of Inventor Frank Julian Sprague. new
Brunswick, nJ: ieee History Center
Press, 2014
p&e
july/august 2018
 2018

Europe

21-25 OCTOBER 2018

Sarajevo, Bosnia and Herzegovina
Meet your professional colleagues, today’s brightest students,
and industry experts at the 2018 IEEE PES Innovative Smart Grid
Technologies Conference Europe (ISGT Europe)

This year's conference will bring together researchers and industry experts from Europe and
all over the world to provide an international forum to share and discuss issues and
developments in the field of smart grid technologies and applications.

You won't want to miss the keynotes, plenary sessions, panels, industry exhibits, paper and
poster presentation, and also tutorials by experts on smart grid applications and systems
integration.

Support opportunities still available | Visit sites.ieee.org/isgt-europe-2018 for more information

Digital Object Identifier 10.1109/MPE.2018.2835848
 society news
meet the candidates
Division VII delegate-elect/director-elect position
B
Beginning in August, elec­ member of the Brook Byers institute
tions will be held for the position for sustainable systems and universi­
of ieee Division Vii delegate­elect/ ty center of excellence in Photovoltaic
director­elect 2019 and ieee Divi­ Research. His research
sion Vii delegate/director 2020–2021 activities are in the moni­
(the elected individual will serve both toring, analysis, and con­ There are two
as both delegate and director). Divi­ trol of power systems as
sion Vii consists of the ieee Power well as the development candidates, both
and energy society (Pes) only. there
are two candidates, both nominated
and applications of re­
newable and sustain­
nominated by
by Division Vii, vying for this office: able energy systems. He Division VII,
Miroslav M. Begovic and Miriam P. was editor of the sec­
sanders. the successful candidate will tion on transmission vying for this
serve as the ieee Division Vii direc­
tor in 2020–2021. to learn more about
systems and smart grids
in Springer Encyclope-
office: Miroslav
the candidates before casting your bal­ dia on Sustainability M. Begovic
lot, read the biographies and candidate (2012), guest editor of
statements that follow. the IET Generation, and Miriam
The Candidates
Transmission and Dis-
tribution special issue
P. Sanders.
on wide­area monitor­
Miroslav M. Begovic ing and control. He has authored near­
Miroslav M. Be­ ly 200 journal and conference papers
govic is a Fellow and ieee standards and technical
of the ieee. He is reports and has contributed over 100
t he d e p a r t m e nt keynote and invited presentations in all
head of the elec­ ieee Regions.
tr ical and com­
puter engineering IEEE Activities and
(ece) Department and carolyn e. and Accomplishments
tommie s. lohman ‘59 Professor at ✔ nominated in 2018 for ieee
texas A&M university. Prior to that, Division V i i d i re ctor­ ele ct
he was a professor and chair of the (2019) by the ieee Pes nomi­
electric energy Research group in the n a t ion s a nd Ap p oi nt m ent s
school of ece, georgia institute of committee.
technology, and an affiliated faculty ✔ Member, ieee nom i nat ions
and Appointments committee
Digital Object Identifier 10.1109/MPE.2018.2836838
(2017–2018) and ieee Fellows
Date of publication: 18 June 2018 committee (2017).
122 ieee power & energy magazine
✔ Pes president and chair, Pes
governing Board and Pes ex­
ecutive committee (2014–2015),
and ieee Pes Past
President (2016–
2017). As president,
he strongly supported
• membership and
chapter growth by
supporting chapter
chair orientation
trainings in all Re­
gions every year
(student chapters
grew 40% yearly
during the period)
• global outreach, es­
pecially in Regions
8 and 10, strength­
ening relations with
sister associations,
such as cigRÉ and the chinese
society of electrical engineers
• coordinating the first major reor­
ganization of the Pes technical
committees and strongly sup­
porting close coordination of
strategic relations of Pes with
ieee sister societies, the ieee
standards Association, and isO.
✔ Member, ieee Publication ser­
vices and Products Board com­
mittee (2015).
✔ Pes president­elect, member
of the governing Board and its
executive and Finance commit­
tees, and chair of the ieee Pes
long­Range Planning committee
(2012–2013), which developed
july/august 2018
 ®

2018
IEEE PES
GENERAL MEETING

5-9 AUGUST
PORTLAND, OR USA

OREGON CONVENTION CENTER

HEAR FROM INDUSTRY

Re-imagining the Electric Grid

LEADERS, INCLUDING:

Join us for the 2018 PES General Meeting in Portland, Oregon

REGISTRATION IS NOW OPEN!

The 2018 IEEE Power & Energy Society General Meeting will be held Mark Ahlstrom

from August 5-9, 2018 at the Convention Center and Crown Plaza.

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 with paper presentations, poster
and panel sessions, a number of technical tours, a student program Larry Bekkedahl
and companion activities.

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

with Super Sessions addressing such topics as:
• Extreme Events and Grid Resiliency
• Energy Storage
• Cyber and Physical Security Peter Jørgensen

• Distributed Generation (DG) Regulation, Engineering,

Modeling, and Impacts

Don’t miss this spectacular event- make plans now to attend!

Maria Pope

We look forward to seeing you in Portland!

IEEE PES – More Power to the Future TM

For more information visit: pes-gm.org/2018

Heather Rosentrater

Digital Object Identifier 10.1109/MPE.2018.2835851

 the ieee Pes strategic Plan
for 2014–2019.
✔ ieee Pes governing Board,
society treasurer, member of the
executive committee, and chair,
Financial committee (2010–
2011).
✔ chair, constitution and Bylaws
committee, ieee Pes govern­
ing Board (2006–2010).
✔ Vice­chair and chair, student
Activities subcommittee, ieee
Pes Power engineering education
committee (2004–2006); orga­
nized student travel to Pes general
Meeting for hundreds of students.
✔ secretary (2002–2003), vice­
chair (2004–2005), and chair
(2005–2009), ieee Pes coor­
dinating committee for emerg­
ing technologies; coordinated
development of the emerging
technologies Review, and intro­
duced a popular series of late­
Breaking news sessions at the
Pes general Meeting about in­
novations, standards, and emerg­
ing technologies.
✔ Member of the ieee Pes Relay­
ing committee for over 20 years,
chairing a number of its work­
ing groups that developed stan­
dards and reports.
✔ ieee Pes Atlanta chapter and
section, various leadership posi­
tions (1989–1999).
Statement
i have served the ieee for over 30 years,
as a member of Pes, and computer, and
circuits and systems societies, leading
working groups in its technical com­
mittees, and engaging in leadership
positions within Pes and the ieee
technical Activities Board as well as a
number of ieee committees.
During that time, and especially
within the past seven years, i have been
actively involved in advancing ieee
Pes and its solid revenue­supported
business model, building close relation­
ships with sister societies, developing
external relations with other profession­
124 ieee power & energy magazine
al associations, and building global out­
reach for the benefit of ieee members.
As an academic researcher closely
aligned with industry (neetRAc con­
sortium at georgia tech) and agile uni­
versity administrator (texas A&M uni­
versity), i am attuned to attracting and
promoting talent, identifying and nur­
turing innovation, and judiciously manag­
ing budgets, all useful experiences for
the ieee Board of Directors position for
which i am nominated.
it would be my honor to support ieee
in its stable progression of influence,
outreach, and continued growth to a
half­million global members.
Miriam P. Sanders
Miriam P. sanders
is a senior Mem­
ber of the ieee.
she received her
bachelor’s degree
in electrical engi­
neering from the
university of north carolina at charlotte
in 1980. she is a registered Professional
engineer in Florida and north caroli­
na. Her career has concentrated on pro­
tective relay with an emphasis in tele­
protection and communications. she
is currently with schweitzer engineer­
ing laboratories, inc. as senior com­
munications applications engineer. Her
previous experience includes five years
as a senior instructor with sel uni­
versity teaching protection and com­
munication fundamentals and 30+
years with Westinghouse, ABB, Pulsar
technologies, and Ametek Power in­
struments as product manager for
the Power line carrier products. she
has also done consulting work with
Quanta technology.
IEEE Accomplishments
and Activities
sanders is a 40­year member of the
ieee and a senior Member. she has held
many positions in the technical activi­
ties of the Pes. From 2013 to 2018, she
served as an officer of the Pes techni­
cal council, progressing from secretary
to vice chair and then to chair, concur­
rent with her position as vice president
of technical Activities. As chair of the
technical council, she facilitated the
reorganization of the technical commit­
tees to be aligned with today’s technol­
ogy. she also organized efforts to have
tutorials created on popular ieee stan­
dards for presentation around the world
to promote the great works the technical
committees accomplish. she has served
as committee chair, vice chair and secre­
tary for the Power systems Relaying and
control committee (PsRc) (2005–
2010), as well as standards liaison for
both the PsRc and the Power systems
communications and cybersecurity
committee. she has chaired several
working groups in these committees as
well. she became a member of Pes as
a student in 1978 and chaired the local
student chapter. she also received the
local Pes chapter’s scholarship during
her time as a student.
Statement
As a volunteer with Pes, i have served
the society for 35+ years participat­
ing in technical committee work. i
will endeavor to bring value to our
younger engineers as well as those ex­
perienced engineers on technical and
educational levels. Pes is one of the
largest societies in the ieee, and as
such, we bring great value to the
society. i will strive to ensure that
our desires are known and validated
at the ieee level to strengthen both
Pes and ieee. to continue to grow,
we need to continue to attract young­
er and diverse members with mean­
ingful projects and opportunities for
volunteers at all levels to the bet­
terment of society. the power and
energy industry is going through a
dynamic evolution with renewables,
energy storage, and digital technol­
ogies, and Pes needs to stay at the
forefront to help guide this evolution.
i will also continue my predeces­
sors’ great work to curtail cost and
keep the budget in check to benefit
the volunteers.
july/august 2018
EMBRACING CHALLENGES
for Power Engineering Sustainability

The 10th IEEE PES Asia-Pacific Power and Energy Engineering Conference 2018
(APPEEC) will be held at Hilton, Kota Kinabalu, Sabah, MALAYSIA on 7-10 October 2018.
The aim of the conference is to provide a premier platform for electrical
engineers and researchers to present their works and to share experiences
and ideas in power and energy engineering with experts and scholars
from around the world. Started in Wuhan in 2009, APPEEC is now an
annual power engineering conference organized in Asia-Pacific Region.

Plan to attend the premier conference for electrical engineers and

researchers to present their works and to share experiences and
ideas in power and energy engineering with experts and
scholars from around the world.

Attendees will enjoy:

• International speakers
• Tutorial sessions
• Technical exhibition
• Professional networking
• Technical paper presentations

REGISTER BEFORE 15 AUGUST TO

RECEIVE THE EARLY BIRD DISCOUNT

For more information, visit sites.ieee.org/appeec-2018 or email appeec2018@gmail.com

Digital Object Identifier 10.1109/MPE.2018.2835850

 In Memoriam: Mahmood (Mike) Adibi (1924–2018)
Earlier this year, we lost a wonderful engineer, true mentor,
and dear friend. Mahmood (Mike) Adibi passed away peace-
fully of natural causes on 28 January 2018 at the age of 93.
Mike received a bachelor of science degree in electrical en-
gineering with honors in 1950 from the University of Birming-
ham, United Kingdom. He earned a master’s degree in electri-
cal engineering in 1960 from Polytechnic
Institute of Brooklyn and a degree in
nuclear engineering in 1970 from the
University of Santa Clara, California.
Mike spent decades in the electric
power industry. He worked for General
Electric Company, British Petroleum,
EBASCO Services, and IBM Corp., where
he developed computer applications for
the engineering and operation of electric
utilities. In 1967, as a program manager at
IBM, he investigated the 1967 Northeast
blackout for the Department of Public
Service for the State of New York. In
1969, following the PJM blackout of 1967, Mike Adibi
he investigated bulk power security as-
sessment for Edison Electric Institute. Since 1979, and in the
aftermath of the New York Blackout of 1977, he developed
restoration plans for over a dozen international utilities.
Mike chaired the IEEE Power & Energy Society’s (PES)
Power Systems Restorations Working Group. These work-
ing group meetings remain one of the best learning experi-
Digital Object Identifier 10.1109/MPE.2018.2824138
Date of publication: 18 June 2018
ences and times of the heartiest laughs at any IEEE meeting.
Mike always made newcomers and old friends feel welcome
at all technical meetings. It was great fun learning together
and sharing funny stories.
Mike wrote over 50 IEEE transactions papers as a principal
author and edited many more. He had four books published on
power systems restoration and helped edit
the work of colleagues. Mike received
multiple IEEE PES awards. He started his
own consulting business, Industrial Re-
search and Development Corporation, in
the late 1970s. He was a consultant to PEP-
CO for many years, designing and testing
equipment and technology. He was an IEEE
Life Fellow, Chartered Engineer IET (United
Kingdom), and a registered professional en-
gineer in Maryland until the day he passed.
Mike was preceded in death by his
brother Ali and sisters Fatimah and Parri.
He is survived by his wife, Hallie; daugh-
ter Roxanne Smart (Ted) of Rockville,
Maryland; Clyde Adibi (Sherry) of An-
napolis, Maryland; grandchildren Riley, Aidan, Grady, and
Andrew; brother Iraj; sisters Batool and Fereshteh; and
many nephews, nieces, and cousins. Those wishing to honor
Mike’s memory may make a donation to the IEEE Foundation
in his name This can be done online at www.ieeefoundation
.org/tribute or by sending a check to IEEE Foundation, 445
Hoes Lane, Piscataway, NJ 08854. Donations in Mike’s mem-
ory will support the programs and collections of the IEEE
History Center.

p&e

Are You
Don’t miss an issue of this magazine—
update your contact information now!

Moving?
Update your information by:
E-MAIL: address-change@ieee.org
PHONE: +1 800 678 4333 in the United States
or +1 732 981 0060 outside
the United States
If you require additional assistance
regarding your IEEE mailings,
visit the IEEE Support Center
at supportcenter.ieee.org.
© ISTOCKPHOTO.COM/BRIANAJACKSON

126 ieee power & energy magazine july/august 2018

calendar
T
T H E I E E E P ow E r & E n E rgy
Society’s (PES’s) website (http://www
.ieee-pes.org) features a meetings sec-
tion, which includes calls for papers
and additional information about each
of the PES-sponsored meetings.
August 2018
IEEE PES General Meeting (GM
2018), 5–9 August, Portland, oregon,
United States, contact richard god-
dard, richard.goddard@pgn.com, or
Shane Freepons, Shane.Freepons@pgn
.com, http://pes-gm.org/2018/
September 2018
IEEE PES Transmission and Distri-
bution Conference and Exposition –
Latin America (T&D LA 2018), 18–
21 September, Lima, Peru, contact
Alex Arquinego Paz, alex.arquinego@
ieee.org, http://ieee-tdla2018.org/
October 2018
IEEE PES Asia-Pacific Power & En-
ergy Engineering Conference (AP-
PEEC 2018), 7–10 october, Saba h,
Malaysia, contact Zuhaina Zakaria,
zuhaina@ieee.org, http://sites.ieee.org/
appeec-2018/
IEEE PES Innovative Smart Grid
Technologies Europe (ISGT Europe
2018), 21–25 october, Sarajevo, Bos-
nia and Herzegovina, contact Senad
Huseinbegovic, shuseinbegovic@etf
.unsa.ba, http://sites.ieee.org/isgt-
europe-2018/
Digital Object Identifier 10.1109/MPE.2018.2822878
Date of publication: 18 June 2018
july/august 2018
PES meetings
for more information, www.ieee-pes.org
November 2018
IEEE Electronic Power Grid (eGRID
2018), 12–14 november, Charleston,
South Carolina, United States, contact
Johan Enslin, jenslin@clemson.edu,
http://ieee-et-d.org/
IEEE International Forum on Smart
Grids for Smart Cities (SG4SC
2018), 26–28 november, genk, Bel-
gium, contact geert Deconinck, geert.
deconinck@kuleuven.be, http://www
.ieeesg4sc.org/
December 2018
IEEE International Conference on
Power Electronics, Drives and Ener-
gy System (PEDES 2018), 18–21 De-
cember, Chennai, India, contact Arun
Karuppaswamy, akp@ee.iitm.ac.in,
www.ee.iitm.ac.in/PEDES2018/
January 2019
IEEE PES 201 Joint Technical Com-
mittee Meeting (JTCM 2019), 13–17
January, orange County, California,
United States, contact Solveig ward,
sward@quanta-technology.com, www
.pestechnical.org
March 2019
IEEE PES GTD Grand Internation-
al Conference and Exposition Asia
2019 (GTD Asia 2019), 20–23 March,
Bangkok, Thailand, contact nopbhorn
Leeprechanon, nopbhorn@engr.tu.ac
.th, http://www.ieeegt-d.org/
May 2019
IEEE International Con ference
on Electrical Machines and Drives
(IEMDC 2019), 11–15 May, San Di-
ego, California, United States, contact
Avoki omekanda, avoki.omekanda@
ieee.org
June 2019
14th International Conference on
Transmission & Distribution Con-
struction, Operation & Live-Line
Maintenance (ESMO 2019), 24–27
June, Columbus, ohio, United States,
contact Eriks Surmanis, e.surmanis@
pdc-cables.com, http://ieee-esmo.com/
IEEE PowerTech Milan (PowerTech
2019), 23–27 June, Milan, Italy, contact
Federica Foiadelli, federica.foiadelli@
polimi.it, http://ieee-powertech.org/
August 2019
IEEE PES General Meeting (GM
2019), 4–9 August, Atlanta, georgia,
United States, contact Matt Stryjewski,
matthew.stryjewski@ieee.org
p&e
ieee power & energy magazine 127
 in my view (continued from p. 132)
and industry, and there is some historical
animosity toward it that needs to be ac-
knowledged and laid to rest.
In the U.S. utility context, electri-
fication has a close historical associa-
tion with load building. Since the early
days, utilities have had strong incen-
tives to grow load to increase profits,
expand rate bases, and reduce average
rates. During the first half of the 20th
century, when access to inexpensive
electricity transformed first urban and
then rural life in the United States, elec-
trification served a powerful social pur-
pose. But since the 1970s, in the face
of energy security and environmental
concerns, electric load growth has often
been viewed negatively.
In the energy efficiency paradigm
pioneered by people like Art Rosenfeld
and Amory Lovins decades ago, using
less primary energy to provide the same
energy services made compelling sense
on many levels: lower fuel demand for
thermal generation, less pollution from
burning that fuel, and less need for gen-
erating capacity and, with it, lower cap-
ital requirements and fewer conflicts
over licensing and land use, as was (and
is) often the case for large hydro dams
and nuclear power plants.
In thermal-dominated systems where
the variable cost of fuel was the lion’s
share of generation costs, saving BTUs
meant saving money, and in the 1970s
when much of the generation fleet was oil
powered, it meant less economic depen-
dence on that volatile and conflict-fraught
commodity. Using electricity to heat water
and space in thermal-dominated systems
was about three times as energy intensive,
counting thermodynamic and line losses,
as bypassing the conversion to electricity
and using natural gas directly in furnaces
and water heaters.
While much of the logic of primary
energy efficiency remains valid today,
it has limitations when seen through the
lens of a low-carbon transformation. In
a power system dominated by renew-
able energy, with near-zero variable cost,
saving primary energy does not translate
directly into lower marginal or average
128 ieee power & energy magazine
cost. More important, even when the di-
rect combustion of fossil fuels is more ef-
ficient in a given end use from a primary
energy standpoint, it is generally a worse
option from an emissions standpoint if
the electricity is low carbon. Carbon, not
energy per se, dictates the logic of energy
systems in a climate-friendly future.
It is critical for industry, regulators,
and policy makers to recognize that deep
decarbonization cannot be achieved
through energy efficiency alone or even
a combination of energy efficiency plus
renewable electricity. Electrification is
absolutely required, and in many ap-
plications, it may complement or even
displace a focus on energy efficiency.
Beyond a certain level, conventional en-
ergy-efficiency investments can produce
diminishing returns for carbon reduc-
tion compared to a similar investment in
electrification, as long as the electricity
is low carbon. Electrifying end uses is
not counter to the fundamental purposes
that motivate investments in energy effi-
ciency. Indeed, modeling shows that, in
a U.S. low-carbon transition, the largest
source of energy efficiency will be elec-
trification itself, due to the thermody-
namic superiority of electric drive trains
and heat pumps over their combustion-
based alternatives.
Getting energy efficiency and elec-
trification to play nice with each other
in the regulatory and policy a renas
may be challenging. Clean-energy ad-
vocates have fought hard to incorpo-
rate renewable generation and building
energy efficiency in utility plans, and
many will be skeptical about electri-
fication if they see it as threatening
decades of hard-won gains. That may
change as the primary energy-efficien-
cy paradigm is reconsidered in the light
of deep decarbonization.
But there is also no clear mandate
for promoting electrification in cur-
rent policy. Indeed, if anything, there
are formidable barriers to fuel switch-
ing, not least of which are the inter-
ests of oil companies and gas utilities.
Current energy-efficiency programs
are designed to stay within their fuel-
type lanes and not change the game by
switching to new energy supplies.
There’s no question that energy effi-
ciency remains essential for decarbon-
ization. Some say that natural gas is the
bridge to a low-carbon future, but the
real bridge is energy efficiency. It will
play an outsized role in sectors with lim-
ited fuel-switching potential, for exam-
ple, freight trucking and industrial pro-
cess heat. In parts of the United States
with no history of efficiency programs,
grossly inefficient building shells, over-
sized HVAC systems, and antiquated in-
frastructure, efficiency will still be the
first tool out of the clean-energy tool-
box. Efficiency provides a brake on irre-
sponsible, high-carbon load building in
coal-based power systems with no tran-
sition plan. Even in a decarbonized sys-
tem, energy efficiency can help reduce
the scale, cost, and land use impacts of a
low-carbon infrastructure buildout.
Long-Term Policy in
the Electric Economy
Article 4.19 of the Paris Agreement
calls on all countries to develop mid-
century strategies for decarbonizing
their economies. The commitments
made at Paris are near term, out to 2025
or 2030, and only promised modest
emission reductions. But the Paris em-
phasis on the long-term future provides
a platform for strategizing and enacting
transformational changes. Given the
multidecade lifetimes of the most criti-
cal infrastructure on both the supply and
demand sides of the energy system (e.g.
power plants, buildings, industrial boil-
ers, and cars and trucks) and the poten-
tial for emissions lock-in and stranded
assets, the long-term perspective must
be a factor in near-term decision making
and investment. A key revelation that
emerges from long-term planning—and
is seldom visible in shorter-term analy-
sis—is the need for electrification as the
third pillar accompanying energy effi-
ciency and low-carbon generation.
Deep decarbonization requires a
suite of policies to transform infrastruc-
ture and markets, covering all three pillars.
july/august 2018
The usual suspects of policy approach-
es are unlikely to be sufficient, par-
ticularly for electrification. Carbon
pricing alone is inadequate for sending
long-term investment signals, especially
where emissions reductions are a func-
tion of interactions across sectors and
measures and over time (as in the case of
EVs, where net emissions depend on the
vehicles being replaced, the changing
carbon intensity of electricity, and their
uptake rate). Policy mechanisms must
be tailored to situations, but some gen-
eral areas of need are clear
A Regulatory Strategy for
Electricity Demand Under
Decarbonization
The necessity of electrification re-
quires rethinking regulatory strategy
for electricity demand. This includes
the mechanisms and relative priorities
of energy-efficiency and electrifica-
tion programs and investments and
how both of these relate to decarbon-
ized generation. The recent Illinois
decision to retain Exelon’s existing
nuclear plants in return for reducing
demand is a step behind in this think-
ing: the incentives for the utility to re-
duce demand work against low-carbon
electrification. Instead, a new para-
digm is needed, similar to the decou-
pling of demand from utility revenue
requirements to encourage energy ef-
ficiency: electrification of current fos-
sil-fuel end uses, when accompanied
by generation decarbonization, should
be considered “good load” for purposes
of incentivizing utilities.
A Planning Vehicle for
Cross-Sectoral Planning
and Coordination
Industrial, transportation, and building
electrification all need to be coordi-
nated with electricity sector planning,
across geographic boundaries and ju-
risdictional levels, to be effective and
economical. Yet neither market nor
regulatory mechanisms exist for such
coordination. Policy must drive joint
planning that aligns investment deci-
sions and timing in sectors that cur-
rently have little interaction but must
july/august 2018
interact much more in the future to
reach climate goals.
A Grand Bargain for Utilities
Deep decarbonization won’t be achieved
without a massive expansion of low-carbon
generation and electrification. Distrib-
uted generation cannot meet the first
of these requirements affordably and
does not address the second. While dis-
tributed generation has a role to play, it
cannot be the tail that wags the dog in
electricity policy. A grand bargain with
utilities is needed, in which they com-
mit to a timely schedule for meeting
these goals (providing supporting in-
frastructure, such as vehicle charging,
in the case of electrification) in return
for the long-term assurance of a viable
business model.
Wholesale Electricity
Market Design
Deep decarbonization means an en-
tirely new market environment for
electricity from what it has been his-
torically. The economics of a decar-
bonized electricity sector need clear
articulation in scholarship and expres-
sion in regulation and wholesale mar-
ket design. Some of the areas requiring
new theory, observation, and experi-
mentation include the following:
✔ A very high renewables system
will be dominated by fixed costs,
with variable energy costs near
zero. Current wholesale electric-
ity markets were not designed to
efficiently allocate fixed costs.
✔ In a system with very high lev-
els of inflexible supply, such as
wind and solar, the demand side
may contribute as much to sys-
tem balancing as the supply side.
Efficient wholesale ma rket s
must provide symmetric rewards
for flexible capacity without re-
gard to supply or demand.
✔ Large-scale flexible loads such
as hydrogen electrolysis could
potentially play a major role in
addressing seasonal imbalance
while also providing low-carbon
fuel, but the economics of elec-
tricity-produced fuels is under-
theorized, including areas such
as long-term market signals,
ownership, and the relationship
to supply-side planning.
✔ The definition of asset utilization
from the utility and regulatory
perspective will need to change
when net load factors (i.e., net of
generation and load, especially
when generation is inflexible) re-
place ordinary load factors as the
utilization paradigm.
The wholesale market design prob-
lem is complicated by a transition pe-
riod of two or three decades, in which
what works in the decarbonized future
also needs to work in the carbonized
present. One potential feature of the
transition that could provide bench-
marks for the timing of future changes
is the changing time signal of energy
imbalances with increased penetration
of intermittent renewables, from hours
to days in low renewables systems to
weeks to months in high renewables
systems. The economics of decarbon-
ized electricity will reflect this chang-
ing imbalance signature, as markets must
respond to it to keep the lights on.
The ea rl ier t hese problems are
thought th rough, the better future
obstacles can be anticipated and ad-
dressed before they sabotage the low-
carbon transition. This is already a
domain in which practice has moved
beyond theory, due, in part, to the
limitations of current analytical ap-
proaches. The 2050 time horizon has
little utility or regulatory motivation on
the power-system planning side, so it
is far outside of the usual scope; while
the limited granularity of integrated
assessments models, still the dominant
tool used to inform climate policy dis-
cussions at the national and interna-
tional levels, undermines their ability
to engage in such discussions. Recent
progress in the long-term, multisec-
tor analysis of deep decarbonization
that also includes granular, bottom-up
treatment of both the supply and de-
mand sides of the electricity sector is
an important first step toward bridging
these disparate worlds.
p&e
ieee power & energy magazine 129
ESMO 2019 • June 24-27, 2019 l Columbus, OH
New Horizons, Boundless Possibilities
The 2019 IEEE PES 14th International Conference on Transmission & Distribution
Construction, Operation & Live-Line Maintenance (ESMO 2019) will take place from
June 24-27, 2019, at the Greater Columbus Convention Center in Columbus, OH and
at host utility American Electric Power’s nearby Outdoor Demonstration area.

ESMO is the premier event for electric utility professionals, contracting, construction and
consulting companies who are interested in hands-on solutions for the safe engineering,
construction, operation and maintenance of the world's power delivery systems.

EXHIBITORS: ESMO 2019 will feature two days of outdoor field demonstrations and
exhibits and a two-day technical program combined with an indoor exhibit area.
For more information, please visit ieee-esmo.com/exhibitors.

SAVE THE DATES: June 24-27, 2019! ESMO IS ON!

For Information and Program Updates visit ieee-esmo.com

IEEE prohibits discrimination, harassment and bullying. For more information, visit www.ieee.org/web/aboutus/whatis/policies/p9-26.html.

Digital Object Identifier 10.1109/MPE.2018.2835849

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

Company page# URL Phone

Advanced Test Equipment Rentals 110 www.atecorp.com +1 888 404 ATEC

AFL Global 13 AFLglobal.com/IEEE +1 800 235 3423

ASPEN, Inc. CVR4 www.aspeninc.com +1 650 347 3997

Bigwood Systems, Inc. 8 www.bigwood-systems.com +1 607 257 0915

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

Delta Star, Inc. 118 deltastar.com

DIgSILENT GmbH 3 www.digsilent.com +49 7072 9168 0

EasyPower LLC 113 www.EasyPower.com/grounding

Electrical Studies, Corp. 17 www.electrical-studies.com/photon

Energy Exemplar Ltd. 111 energyexemplar.com

ETAP 7 etap.com/18 +1 949 900 1000

Flex-Core 114 www.flex-core.com +1 614 889 6152

Hughes Brothers, Inc. 117 www.hughesbros.com

Kinectrics, Inc. 109 www.kinectrics.com

LEM USA, Inc. 112 www.lem.com

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

Mitsubishi Electric 21 WWW.MEPPI.COM/HVDCDIAMOND

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

NR Electric Co., Ltd 127 www.nrec.com +86 25 8717 8888

OMICRON 16 www.omicronenergy.com

P & R Technologies 115 www.pr-tech.com +1 800 722 8078

Powertech Labs, Inc. 6 www.dsatools.com

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

Raytech USA, Inc. 15 www.RaytechUSA.com +1 888 484 3779

Rea Magnet Wire Company 119 ReaWire.com/IEEE +1 800 732 9473

Reef Industries, Inc. 23 reefindustries.com +1 800 231 6074

RTDS Technologies, Inc. 18 www.rtds.com

Siemens Power Technologies International CVR 3 siemens.com/power-technologies

SKM Systems Analysis, Inc. 9 www.skm.com +1 800 500 4SKM

Soltex, Inc. 116 soltexinc.com +1 281 587 0900

Tana Wire Markers 118 www.tanawiremarker.com +1 573 796 3812

Trakka Systems 23 trakkasystems.com

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.2017.2770702

july/august 2018 ieee power & energy magazine 131

 in my view
R
REDUCInG CARBOn DIOxIDE
(CO2) emissions to the level needed to
stabilize the climate will require very
deep reductions in emissions from en-
ergy suppliers and end users. In Europe,
the United States, and other industrial-
ized countries, this means reducing CO2
from fossil-fuel combustion at least 80%
below present levels by mid-century and
eliminating the emissions altogether a
decade or two after that. The technology
pathways for achieving such reductions,
while maintaining an energy system that
supports a modern economy, have re-
cently begun to be spelled out in consid-
erable detail in many countries, across
sectors, and over time.
Electricity and
Deep Decarbonization
Everywhere, the basic formula is the
same. There are three pillars required to
support a decarbonized energy system:
highly efficient energy use, electric-
ity produced with virtually no carbon
emissions (much lower than possible
with a natural-gas-based generation
fleet), and end uses that mostly run on
low-carbon electricity. These pillars
stand up across geographies and stages
of economic development because they
embody basic physics and chemistry.
There is a role for other types of
low-carbon fuels; for example, biofuels
may prove essential in aviation and
shipping. But biofuels are fundamen-
Digital Object Identifier 10.1109/MPE.2018.2820446
Date of publication: 18 June 2018
132 ieee power & energy magazine
the electric economy
moving toward a low-carbon future
tally constrained in quantity, given ex-
isting technology and land-use practic-
es. Carbon capture and storage (CCS),
which could allow the continued use
of fossil fuels in power generation and
industry, are constrained by the slow
pace of technology development and,
frankly, a lack of interest among policy
makers and the fossil-fuel industry.
The upshot is that currently there is
no feasible alternative to electricity in
most end-use applications—automo-
biles, buildings, industry—for supply-
ing the low-carbon energy required.
A mostly fossil-fuel-free energy sys-
tem means a major increase in elec-
tricity generation in the United States,
roughly a doubling from current levels
by mid-century, even accounting for
the offset of some load by behind-the-
meter solar photovoltaics (PVs). This
realization has not yet been absorbed in
some quarters. Many utilities see low
or negative load growth at present and
project that into their long-term load
forecasts. Yet an economy-wide low-
carbon transformation by mid-century
will require light-duty vehicles and
buildings to be more than 90% electri-
fied and for industry to double its cur-
rent electrification rate. This means a
lot of new load to be met by a lot of new
low-carbon generation.
Without a revival of nuclear power,
or an unexpected emergence of CCS,
most of the low-carbon electricity
needed will be provided by renewable
energy, especially wind and solar. Great
progress has already been made in actu-
James H. Williams
al practice in integrating renewables at
levels of around one-third of total gener-
ation, and there is now widespread con-
fidence in the ability of many systems
to reach well above this share. However,
above about a two-thirds share of inter-
mittent power, especially with a limited
carbon budget for gas generation, a new
mix of solutions for addressing energy
imbalances will be required, including
thermal, curtailment, storage, and in-
creased load and resource diversity from
expanded regional integration. Flexible
electric loads, including the production
of fuels like hydrogen and synthetic nat-
ural gas, may become key parts of the
balancing mix by mid-century.
Electrification
and Energy Efficiency
Integrating high levels of renewable gen-
eration poses some engineering-econom-
ic challenges, but it has a market driver
in falling wind turbine and PV module
prices, and it is widely understood to be
the path forward for power systems in a
climate-friendly world. The next frontier
of decarbonization is widespread elec-
trification. This will be driven, in part,
by consumer demand and technological
improvements, as we are beginning to
witness with electric vehicles (EVs) and
battery prices. But markets often need to
be kick-started by policy, and that will
require broad-based support. The need
for electrification hasn’t yet been univer-
sally embraced, especially in buildings
(continued on p. 128)
july/august 2018
 Join us at
the IEEE PES
General
Meeting

Envision a digital
future
Partner with Siemens PTI to
optimize your grid

From strategic advisory to technical consulting and state-

of-the-art software solutions, Siemens Power Technologies
International (PTI) offers a holistic approach to mastering
the technical and economical challenges of today’s and
future energy systems.
• Grid design, planning, optimization and operational support
• Energy infrastructure strategy and process consulting
• Industry leading tools for grid planning and simulation
With 60 years of industry experience, Siemens PTI can
provide the quality expertise needed to successfully
optimize the grid.

siemens.com/power-technologies
ASPEN
OneLiner
Short circuit and relay coordination program
trusted by utilities worldwide

ASPEN OneLiner ™
Powerful, fast, intuitive and easy to use

Applications Advanced features

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

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

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