Enterprise Asset

Management

Improving asset availability, reliability and capacity

with next-generation digital asset performance
management solutions

View our ebook

Volume 17 • Number 1 • January/February 2019
 magazine

Volume 17 • Number 1 • January/February 2019

www.ieee.org/power

on the 73
cover

©istockphoto.com/sibeRianaRt
people: ©istockphtoto.com/Rawpixel
icons: footage fiRm, inc.

features

contents
20 European Union Electricity Markets 53 A Change Is Coming
By Tomás Gómez, Ignacio Herrero, Pablo Rodilla, By Gianluca Fulli, Marcelo Masera,
Rodrigo Escobar, Salvatore Lanza, Amanda Spisto, and Silvia Vitiello
Ignacio de la Fuente, Maria Luisa Llorens,
and Paula Junco 67 Wholesale Electricity Markets
in the United States
32 Electricity Markets in the United States By Emma Nicholson and Arnie Quinn
By Eugene Litvinov, Feng Zhao,
and Tongxin Zheng 73 Three Waves of U.S. Reforms
By Benjamin F. Hobbs and Shmuel S. Oren
43 The New Zealand Electricity Market
By Andy Philpott, Grant Read, 82 PV-Battery Systems for Critical Loads
Stephen Batstone, and Allan Miller During Emergencies
By Chanaka Keerthisinghe,
Mareldi Ahumada-Paras, Lilo D. Pozzo,
Daniel S. Kirschen, Hugo Pontes,
Wesley K. Tatum, and Marvi A. Matos

columns &
departments
4 From the Editor 94 Society News
10 Letters to the Editor 98 Calendar
14 Leader’s Corner 104 In My View
18 Guest Editorial

94
Digital Object Identifier 10.1109/MPE.2018.2879124

january/february 2019 ieee power & energy magazine 1


Editor-in-Chief
Michael I. Henderson, mih.psat@gmail.com
Associate Editors
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,
J. Paserba, C.E. Root, H. Rudnick,
magazine
IEEE Periodicals/Magazines Department
445 Hoes Lane, Piscataway, NJ 08854 USA
www.ieee.org/magazines
M. Shahidehpour, G.B. Sheblé, J.C. Smith, Geraldine Krolin-Taylor, Senior Managing Editor
M. Thomas, E. Uzunovic, S.S. Venkata, Janet Dudar, Senior Art Director
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 S. Ninalowo, Women in Power Publications

S. Rahman, President J. Benedict, Marketing Publications Board Chair, M. Crow
F. Lambert, President-Elect M. Seif, Membership Development Editors-in-Chief
M. Armstrong, Vice President, Chapters IEEE Electrification Magazine, I. Husain
F. Rahmatian, Vice President, Technical Activities Technical Council
F. Rahmatian, Chair, V. Vittal, Vice Chair IEEE Power Engineering Letters, M. Fotuhi-
E. Uzunovic, Vice President, Education Firuzabad
B. Pal, Vice President, Publications H. Chen, Secretary, M. Sanders, Past-Chair
IEEE Trans. on Energy Conversion, J. Jatskevich
W. Bishop, Vice President, Meetings Technical Committee Chairs IEEE Trans. on Power Delivery, W. Xu
J.C. Montero Q, Vice President, Membership K. Schneider, Analytical Methods for IEEE Trans. on Power Systems, N. Hatziargyriou
& Image Power Systems IEEE Trans. on Smart Grid, J. Wang
S. Bahramirad, Vice President, New Initiatives/ K. Chen, Electric Machinery IEEE Trans. on Sustainable Energy, B. Chowdhury
Outreach W.T. Jewell, Energy Development & Power IEEE Power & Energy Technology Systems Journal,
N. Hadjsaid, Treasurer Generation S. Sudhoff
J. Bian, Secretary C. Ashton, Energy Storage & Stationary Battery IEEE Power & Energy Magazine, M. Henderson
D. Novosel, Past-President E. Bascom, Insulated Conductors eNewsletter, S. Fattah
D. Diaz, Region Rep., U.S. & Canada T. Koshy, Nuclear Power Engineering PES Representative to IEEE Press, Open
J. Milanovic, Region Rep., Europe, Middle East, M. Dood, Power System Communications Website, Open
& Africa & Cybersecurity
S. Leon, Region Rep., Latin America C. Vournas, Power System Dynamic Performance Meetings
D. Sharafi, Region Rep., Asia & Pacific J. McBride, Power System Instrumentation Committee Chairs
B. Enayati, Member-at-Large & Measurements General Meeting Steering, D. Hall
R. Kappagantu, Member-at-Large L. Barroso, Power System Operation Planning Joint Technical Committee Meeting Steering,
H. Koch, Member-at-Large & Economics S. Ward
J. Yu, Member-at-Large R. Patterson, Power System Relaying & Control Technically Cosponsored Conferences Steering,
S. Chandler, Smart Building, Load & J. Mitra
IEEE Division VII Director Transmission & Distribution Conference &
B. Meyer Customer Systems
J. Gravelle, Substations Exposition Committee for North America
Steering, C. Segneri
IEEE Division VII Director-Elect S. Hensley, Surge Protective Devices
Innovative Smart Grid Technology Conference–
M. Sanders T. Irwin, Switchgear
S. McNelly, Transformers Europe, D. Van Hertem
PES Executive Director G. Chang, Transmission & Distribution Innovative Smart Grid Technology Conference–
Patrick Ryan, +1 732 465 6618, Asia, TBD
fax +1 732 562 3881, p.ryan@ieee.org Technical Council Coordinating Committees Website, Open
D. Houseman, Intelligent Grid & Emerging
Standing Committee Chairs Technology Education
H. Louie, Constitution & Bylaws D. Alexander, Marine Systems Committee Chairs
V. Kolluri, Fellows D. Lew, Wind & Solar Power Power and Energy Education, S. Brahma
N. Hadjsaid, Finance Scholarship Plus, J. Peer
A. Apostolov, History Technical Council Standing Committees Website, Open
F. Lambert, Long-Range Planning M. Sanders, Awards
D. Novosel, Nominations & Appointments V. Vittal, Technical Sessions New Initiatives and Outreach
H. Chen, Organization & Procedures Committee Chairs
Chapter Representatives T. Burse, Standards Coordination IEEE Smart Village, R. Larsen, Chair
D. Diaz, B. Djokic, B. Gwyn, M. Hunt, D. Toland, Webmaster R. Podmore, Vice-Chair
J. Khan, N. Logic, A. Nassif, P. Pabst, Open, Industry Education Industry Outreach, J. Giri, Chair
M. Papic, M. Patel, S. Pawale, Website, K. Anastasopoulos
M. Stryjewski, C. Wong
Chapter Committee Chairs
Z. Wang, Electronic Communications
C. Diamond, Awards & Resources IEEE Power & Energy Magazine
N. Nair, Distinguished Lecturer Program IEEE Power & Energy Magazine (ISSN 1540-7977) (IPEMCF) is published bimonthly by the Institute of Electrical and
Z. Wang, Chapters Website Electronics Engineers, Inc. Headquarters: 3 Park Avenue, 17th Floor, New York, NY 10016-5997 USA. Responsibility for  the
contents rests upon the authors and not upon the IEEE, the Society, or its members. IEEE Operations Center (for orders, sub-
Membership & Image Committee Chairs scriptions, address changes): 445 Hoes Lane, Piscataway, NJ 08854 USA. Telephone: +1 732 981 0060, +1 800 678 4333.
T. Doggett, VP Deputy Individual copies: IEEE members US$20.00 (first copy only), nonmembers US$84 per copy. Subscription Rates: Society members
A. Vaccaro, Awards & Recognition included with membership dues. Subscription rates available upon request. Copyright and reprint permissions: Abstracting is permitted
J. Hofman, Young Professionals 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)
L.F. Gaitán, Social Media those post-1977 articles that carry a code at the bottom of the first page, provided the per-copy fee indicated in the code is paid through
the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923 USA; 2) pre-1978 articles without fee. For other copying,
G. González, Web Site Development
reprint, or republication permission, write Copyrights and Permissions Department, IEEE Operations Center, 445 Hoes Lane, Piscat-
away, NJ 08854 USA. Copyright © 2019 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.2018.2879125 IEEE Operations Center, 445 Hoes Lane, Piscataway, NJ 08854 USA. Canadian GST #125634188 printed in U.s.a.

2 ieee power & energy magazine january/february 2019

Fully integrated power system analysis software

DIgSILENT announces

PowerFactory 2019
DIgSILENT has set standards and trends in power system modelling,
analysis and simulation for more than 25 years. The proven advan-
tages of the PowerFactory software are its overall functional integra-
tion, its extensive modelling and analysis capabilities for generation,
transmission, distribution and industrial networks, and its powerful
data management tools to ensure data consistency and traceability
while enhancing overall work efficiency.
The new version PowerFactory 2019 addresses two major goals for
the integrated analysis of power systems. A new Unit Commitment
and Dispatch Optimisation tool allows the network analysis to be
complemented by a market simulation, with a single entry point of
information for your simulation model, thereby maximising data
consistency and work efficiency. Secondly, the new co-simulation
for single and multiple time domains (RMS-RMS, RMS-EMT and
EMT-EMT) allows unlimited combination of RMS/EMT-type models of
different complexity requiring different integration steps. The mixed
co-simulation is particularly well suited for the analysis of systems
with high penetration of converter-based generation and power
electronics in general.
SELECTED NEW FEATURES:
G 08(1)
Type Gen 08
~
SG

Unit Commitment and Dispatch Optimisation tool

Bus 37(1) Line 26 - 29(1)
Line Type 26 - 29

2,00
Load 26(1) Load 29(1)
Trf Type 25 - 37 YNy0

General Load Type General Load Type

G 10(1)
Trf 25 - 37(1)

Type Gen 10 Line 26 - 28(1) Line 28 - 29(1)

Line Type 26 - 28 Line Type 28 - 29
~
SG

1,00
Bus 25(1) Bus 26(1) Bus 28(1) Bus 29(1)
Bus 30(1)
Line Type 26 - 27

Trf Type 29 - 38 YNy0

Line 26 - 27(1)

Line 25 - 26(1)
Trf 29 - 38(1)
Line Type 02 - 25
Trf Type 02 - 30 YNy0

Line 02 - 25(1)

Line Type 25 - 26
Trf 02 - 30(1)

0,00
Load 25(1) Load 28(1)
General Load Type General Load Type

to minimise generation and re-dispatch costs. Built-in

Bus 27(1)
Bus 38(1)
Bus 02(1) Line 17 - 18(1)
Line Type 17 - 27

Line Type 17 - 18
Line 17 - 27(1)

-1,00 Load 27(1) Bus 24(1) SG

Line Type 01 - 02

~
Line 01 - 02(1)

General Load Type

G 09(1)
Type Gen 09
Bus 18(1) Bus 17(1)
Line Type 02 - 03

-2,00
Line 02 - 03(1)

Load 24(1)
Line Type 03 - 18

Line Type 16 - 24
Line 03 - 18(1)

Line 16 - 24(1)

General Load Type

0,80 0,90 1,00 1,10 1,20 [s] 1,30

Bus 01(1)
Line Type 16 - 17
Line 16 - 17(1)

Load 18(1)
General Load Type

linear programming solver as well as support of the

G 06(1)
Type Gen 06
~

AC voltage (EMT)
SG
Bus 09(1): Phase Voltage A in p.u. Bus 03(1)

Bus 09(1): Phase Voltage B in p.u. Bus 16(1)

Bus 09(1): Phase Voltage C in p.u. Bus 35(1)
Line Type 01 - 39
Line 01 - 39(1)

Line Type 15 - 16
Line 15 - 16(1)

Load 03(1)
Line Type 16 - 21
Line Type 03 - 04

Trf Type 22 - 35 YNy0

Line 03 - 04(1)

Line 16 - 21(1)

General Load Type

Trf 22 - 35(1)

Load 16(1) Line 21 - 22(1)

Load 04(1) General Load Type Line Type 21 - 22
General Load Type
G 01(1)
Type Gen 01 Line 04 - 14(1) Bus 15(1)
Line Type 04 - 14
~
SG
Line Type 23 - 24
Line 23 - 24(1)

Bus 21(1) Bus 22(1)

external solvers IBM CPLEX and GUROBI

Line Type 14 - 15
Line 14 - 15(1)

EMT Simulation
Bus 04(1)
Load 15(1)
Line Type 16 - 19
Line 16 - 19(1)

General Load Type

Bus 39(1)
Line Type 04 - 05
Line 04 - 05(1)

Bus 14(1) Load 21(1)

General Load Type
Line Type 22 - 23
Line 22 - 23(1)

Load 39(1) Bus 05(1)

General Load Type Load 12(1)
General Load Type
Line Type 05 - 06
Line 05 - 06(1)

Load 23(1)
General Load Type
Line Type 13 - 14
Line 13 - 14(1)

1,30 Bus 06(1)

Bus 12(1)
Line Type 06 - 07

Trf Type 13 - 12 YNy0

1,10
Bus 19(1) Bus 23(1)
Line 06 - 07(1)

Trf 13 - 12(1)
Line Type 05 - 08

Trf Type 11 - 12 YNy0

Line 05 - 08(1)

Trf 11 - 12(1)
Line Type 06 - 11
Line Type 09 - 39
Line 09 - 39(1)

Line 06 - 11(1)

Trf Type 19 - 20 YNy0

Trf Type 23 - 36 YNy0

Fully integrated, easy to configure Co-Simulation for

Load 20(1)
General Load Type
Trf 19 - 20(1)

Trf 23 - 36(1)

Bus 07(1)

0,90
Trf Type 06 - 31 YNy0
Trf 06 - 31(1)

Trf Type 19 - 33 YNy0

Bus 13(1)
Trf 19 - 33(1)
Line Type 07 - 08
Line 07 - 08(1)

Bus 11(1) Bus 20(1) Bus 36(1)

0,70
Load 07(1)
General Load Type
Line Type 10 - 13
Line 10 - 13(1)

Trf Type 20 - 34 YNy0

Line Type 10 - 11
Line 10 - 11(1)

Trf 20 - 34(1)

SG
~

0,50
Bus 08(1) G 07(1)
Type Gen 07

0,80 0,90 1,00 1,10 1,20 [s] 1,30 0,40

Bus 31(1)
Line Type 08 - 09
Line 08 - 09(1)

Bus 10(1) Bus 34(1) Bus 33(1)

single and multiple time domain (RMS-RMS, RMS-EMT

AC voltage (RMS)
Load 08(1)
Trf Type 10 - 32 YNy0

Bus 29: Line-Ground Voltage, Magnitude A in p.u. General Load Type

SG
Trf 10 - 32(1)

~
Bus 29: Line-Ground Voltage, Magnitude B in p.u.

0,00
Bus 09(1) G 02(1) Load 31(1) SG SG
Type Gen 02 General Load Type ~ ~
Bus 29: Line-Ground Voltage, Magnitude C in p.u. G 05(1) G 04(1)
Type Gen 05 Type Gen 04

Bus 32(1)
G 08

-0,40
Type Gen 08
~
SG

Line 26 - 29(2) SG
Line Type 26 - 29 ~
G 03(1)
Type Gen 03
Bus 37

-0,80
Line 26 - 29
Line Type 26 - 29

and EMT-EMT). Co-Simulation with external solver via

Load 26 Load 29
Trf Type 25 - 37 YNy0

General Load Type General Load Type

G 10
Trf 25 - 37

Type Gen 10 Line 26 - 28 Line 28 - 29

Line Type 26 - 28 Line Type 28 - 29
~
SG

Bus 25 Bus 26 Bus 28 Bus 29

-1,20
0,80 0,90 1,00 1,10 1,20 [s] 1,30
Bus 30 Line 26 - 29(3)
Line Type 26 - 29

AC currents (EMT)
Line Type 26 - 27

Trf Type 29 - 38 YNy0

Line 26 - 27

Line 25 - 26
Line Type 02 - 25
Trf Type 02 - 30 YNy0

Trf 29 - 38

Line Type 25 - 26
Line 02 - 25
Trf 02 - 30

Load 25 Load 28
General Load Type General Load Type Line 26 - 29(2): Phase Current A/Terminal j in p.u.
Bus 27 Line 26 - 29(2): Phase Current B/Terminal j in p.u.

IEEE C37.118 communication interface

Bus 38
Line 26 - 29(2): Phase Current C/Terminal j in p.u.
Bus 02 Line 17 - 18
Line Type 17 - 27

Line Type 17 - 18

RMS Unbalanced
Line 17 - 27

Load 27 Bus 24 SG
Line Type 01 - 02

General Load Type ~

Line 01 - 02

G 09
Type Gen 09
Bus 18 Bus 17
Line Type 02 - 03
Line 02 - 03

Load 24
Line Type 03 - 18

Line Type 16 - 24

General Load Type

Line 03 - 18

Line 16 - 24

Bus 01
Line Type 16 - 17
Line 16 - 17

Load 18
General Load Type

G 06
Type Gen 06
~
SG
Bus 03

Simulation
Bus 16
Bus 35
Line Type 01 - 39

Line Type 15 - 16
Line 01 - 39

Line 15 - 16

Hosting-Capacity tool for the evaluation of spare gen-

Load 03
Line Type 16 - 21
Line Type 03 - 04

Trf Type 22 - 35 YNy0

General Load Type

Line 16 - 21
Line 03 - 04

Trf 22 - 35

Interconnection to Load 16 Line 21 - 22

General Load Type Line Type 21 - 22
Rest of U.S.A. / Canada Load 04
General Load Type
G 01
Type Gen 01 Line 04 - 14 Bus 15
Line Type 04 - 14
~
SG
Line Type 23 - 24
Line 23 - 24

Bus 21 Bus 22
Line Type 14 - 15
Line 14 - 15

Bus 04
Load 15
Line Type 16 - 19

General Load Type

Line 16 - 19

Bus 39
Line Type 04 - 05
Line 04 - 05

Bus 14 Load 21
General Load Type
Line Type 22 - 23

eration or load capacity in predefined network areas

Line 22 - 23

Load 39 Bus 05
General Load Type Load 12
General Load Type

0,80
Line Type 05 - 06
Line 05 - 06

Load 23
General Load Type
Line Type 13 - 14
Line 13 - 14

Bus 12
Bus 06

0,60
Line Type 06 - 07

Trf Type 13 - 12 YNy0

Bus 19 Bus 23
Line 06 - 07
Line Type 05 - 08

Trf Type 11 - 12 YNy0

Trf 13 - 12
Line 05 - 08

Line Type 06 - 11
Line Type 09 - 39

Trf 11 - 12
Line 06 - 11
Line 09 - 39

Trf Type 19 - 20 YNy0

Trf Type 23 - 36 YNy0

Load 20
General Load Type
Trf 19 - 20

Trf 23 - 36

0,40
Bus 07
Trf Type 06 - 31 YNy0

Trf Type 19 - 33 YNy0

Trf 06 - 31

Bus 13
Trf 19 - 33
Line Type 07 - 08
Line 07 - 08

Bus 11 Bus 20 Bus 36

Load 07

0,20
General Load Type
Line Type 10 - 13
Line 10 - 13

Trf Type 20 - 34 YNy0

Fast AC linearized contingency analysis using

Line Type 10 - 11
Line 10 - 11

Trf 20 - 34

SG
~
Bus 08 G 07
Type Gen 07

0,00
Bus 31
Line Type 08 - 09
Line 08 - 09

Bus 10 Bus 34 Bus 33

0,80 0,90 1,00 1,10 1,20 [s] 1,30

Swing Node
Load 08
Trf Type 10 - 32 YNy0

General Load Type

SG
~
Trf 10 - 32

Bus 09 G 02 Load 31 SG SG
Type Gen 02 General Load Type ~ ~

AC currents (RMS)
G 05 G 04
Type Gen 05 Type Gen 04

Line 26 - 29(2): Phase Current, Magnitude A/Terminal i in p.u.

Bus 32 Line 26 - 29(2): Phase Current, Magnitude B/Terminal i in p.u.

sensitivity based method, which can be combined with

CoSim_39 Bus New England System Project: Example
Line 26 - 29(2): Phase Current, Magnitude C/Terminal j in p.u. DIgSILENT
39 Bus 10 Machine New England Power System Graphic: Grid
GmbH
Date: 06.11.2018
PowerFactory 2019 Beta Annex:
SG
~
G 03
Type Gen 03

the full load contingency analysis for fast screening of

contingency cases
User-defined data objects (classes) extending the
existing data extensions concept for a fully customisable
data model
Additional data security features, such as password se-
curity rules, authentication modes and privacy settings,
to comply with general data protection regulations

For further information about PowerFactory 2019 please visit our

website www.digsilent.de and see the What’s New document.

DIgSILENT GmbH | Heinrich-Hertz-Straße 9 | 72810 Gomaringen | www.digsilent.de | mail@digsilent.de

 from the editor
I
In 2017, new england experI­
enced the second­lowest annual whole­
sale electric market prices in more than
a decade. well­designed markets
enhance the reliable and econom­
ic operation and planning of the
system by providing incentives
for improved performance and
efficient investments. new
and existing resources show
extraordinarily high avail­
ability, which generates re­
venue as well as energy. For
example, nuclear plants cur­
rently boast availabilities
above 95%, a considerable
improvement over the rates of
65% that were common during
the 1980s. The start­up times of
fossil units are much faster now
than they were before the wholesale
markets were introduced. Markets also
encourage capital improvements to the
system, such as the addition of very
efficient generating units that reduce op­
erating costs and lower emissions. Com­
petition has fostered the use of modern
combined­cycle generating units with heat
rates below 5,700 Btu/kwh.
The 1990s saw the advent of whole­
sale electric markets in europe and the
United States. Since then, the grid has
been transformed by the large­scale de­
velopment of variable energy, demand­
side, and distributed resources. Because
of their lack of fuel costs, new wind and
photovoltaic resources depress energy
Digital Object Identifier 10.1109/MPE.2018.2877897
Date of publication: 7 January 2019
4 ieee power & energy magazine
electric markets
effective designs & flexible responses
market prices, which is often the primary
source of revenue for existing resources
in the current wholesale electric markets.
Higher prices for other wholesale electric
markets, such as for capacity and ancil­
lary services, can make up for revenue
shortfalls and meet the system’s need for
flexibility. when used in concert, prop­
erly designed markets and the large­scale
addition of distributed resources can
greatly improve a system’s reliability and
its economic and environmental perfor­
mance. The role of demand resources and
storage in these markets is noteworthy,
as it can improve the overall operation of
Michael Henderson
the system by providing demand elastic­
ity and flexibility.
How should the markets be design­
ed, given the increased amounts of
variable energy resources, energy
efficiency, storage, and demand
resources that have transform­
ed our industry? It is our hope
that the contents of this issue
image licensed by ingram publishing
will trigger future discus­
sions about possible market
improvements that will en­
hance the reliable, econom­
ical operation and planning
of the system.
In This Issue
The articles in this issue provide
a global view of electric markets
and consider the large penetra­
tions of variable energy resources
and the ability of demand to provide
a flexible response. In their guest edi­
torial, antonio J. Conejo and ramteen
Sioshansi discuss the challenges and some
of the potential solutions for effectively
designing markets in europe, the United
States, and new Zealand. The issue also
examines the effect of public policies on
the markets and power systems and sum­
marizes the wide experience of authors
from europe, the United States, and new
Zealand. Six well­written articles discuss
the following topics:
✔ the current state of electric mar­
kets in europe, policies affecting
their integration across national
boundaries, and how coordination
can be achieved across different
national and regional markets
january/february 2019
 Ready
To
SwiTch?
Receive up to 60% off
when you switch
to SKM.
You’ll also receive
free service to convert
competing software
projects over to SKM!
“The main advantage with
Power*Tools for Windows is
its user-friendliness. PTW is
easy to use and the ready-
to-use standard models
enable you to speed up
your work. Changing data is
simple and makes evaluating
different cases very easy.”
47
th
IVERSARY
ANN
19 2 ~ 2019
7 Interactive Filter Design. Evaluate and correct harmonic distortions.
One Pearl Street
Redondo Beach, CA 90277
1-800-500-4SKM
www.skm.com • sales@skm.com
ARC FLASH EVALUATION
Design safer power systems while ensuring compliance with NEC,
OSHA, BFPA, IEEE, NESC, and CSA standards
Fully integrated with Short Circuit, Over-Current Coordination,
and Equipment Evaluation modules
Evaluate alternatives quickly and easily to establish an optimal design
Improve safety margins with user-definable arcing fault tolerance
Save time by automatically generating arc flash labels and work permits
Determine required PPE levels instantly
Increase flexibility by creating custom labels in any size with
user-defined logos, text, comments, and field placement.
Avoid potential fines, post productivity, and increased insurance
and litigation costs.
Compare results from multiple project scenarios in a single table
ADDITIONAL SOFTWARE FROM SKM
DAPPER® Integrated Electrical Analysis Software. Comprehensive AUTO DESIGN Automatically sizes Cables, Transformers,
Three Phase Short-Circuit Studies, Load Flow Study, Demand Load Buses, and Protective Devices to achieve proper size and
Study, Feeder and Transformer Sizing Study, Impact Motor Starting ratings that meets criteria limits set by the Equipment Evaluation
Study, and Load Schedules. (Unbalanced Studies also available) study. Utilize user-defined Design libraries. Best suited for new system
designs or to resolve undersized and/or overdutied equipment for
CAPTOR® Time-Overcurrent Coordination. Graphical Time-Overcurrent
Coordination. Integrated with one-lines, short-circuit modules, Equipment existing systems. (Requires Equipment Evaluation module)
Evaluation, and Arc Flash. Comprehensive protective device library. I*SIM Dynamic Simulation and Transient Stability. Model, Simulate,
and Analyze Dynamic Responses to Power System Electro-Mechanical
AUTO COORDINATION & EVALUATION Instantly identify
Disturbances. Includes Generator Sizing and Stability, Rotor Angle and
coordination and protection issues by evaluating each
Voltage Stability, Flux Level Machine Representation, Identify Fault
equipment for adherence to the National Electric Code (NEC).
Clearing Times. User Defined Graphical Models for Exciter, Turbine
Automatically make changes to resolve violations.
Governor, PSS, and other controllers.
CABLE AMPACITY Accurate calculation of ampacity and
temperature rise in various power cable installations. Based DISTRIBUTION RELIABILITY Calculates the reliability indices of
individual load points and the overall distribution systems with either
on and fully compliant with IEC 60287.
A_FAULT ANSI Short-Circuit Study. Three Phase Short-Circuit Study
radial or loop configuration.
based on the ANSI/IEEE C37 Standards. Separate solutions for low, DC SYSTEMS ANALYSIS Battery Sizing, Load Flow, & Short Circuit
medium and high voltage systems and for symmetrical, momentary Analysis. Evaluate all loading conditions for DC duty cycle loads and
and interrupting calculations. (Unbalanced Studies also available) AC emergency loads. Complies with IEEE std. 485, 1115, 399, 946,
and IEC std. 61660.
EQUIPMENT EVALUATION Equipment Evaluation Report. Automatically
compares short-circuit ratings, withstand ratings. Applies de-rating GROUND MAT Substation Ground Grid Design and Analysis. Optimizes
adjustments and user defined pass/marginal/failed criteria. Includes grid design using general purpose finite element algorithm for
error checking for input data and topology. potential analysis and graphical facilities to validate grounding
systems efficiency.
TMS Transient Motor Starting Simulation. Time-based motor starting
simulation with graphical output. Includes reduced voltage and Data Exchange Seamlessly share project data between
capacitor starting, graphical motor and load models. SKM projects and other third-party software. Exchange
format includes Tab Delimited, CSV, and SKM XML. Also includes
HI_WAVE Power Quality Investigation and Filter Design. Frequency
an interface to exchange data with Autodesk Revit and Intergraph
Scan, Harmonic Current, Voltage Distortion, Harmonic Load Flow and
SmartPlant Electrical.
SKM Power*Tools Training Classes
Visit skm.com to find a class near you today! 101/102 Introduction to SKM Power*Tools
Hurry, classes fill up fast! 103/104 Intermediate Topics in SKM Power*Tools
201/202 Power System Application – Coordination & Arc Flash
Training more than five engineers? 204/205 Motor Starting and Harmonics
Email training@skm.com for a 206 DC System Analysis & Arc Flash
personalized On-Site Training quote. 207 Grounding and Applications
Chosen By The Top 40 eleCTriCal engineering Firms in The World
 ✔ an overview of U.S. wholesale
electric markets, their implemen­
tation by independent system op­
erators, challenges being faced,
and potential future solutions
✔ a history of the new Zealand mar­
kets and how their restructuring is
meant to address inefficiencies in
this isolated electric power system
✔ a discussion of how policy deci­
sions affected the current euro­
pean Union electricity system and
markets, new policy and regula­
tory initiatives aimed at redesign­
ing these markets, and ideas about
the interactions among regulatory
and technical advances and their
effects on future power systems
✔ background on the role of the U.S.
Federal energy regulatory Com­
mission in competitive wholesale
electric markets, the challenges it
faces, and principles for market
reforms to address key issues
✔ the historical and future evolution
of electricity markets, which needs
to reflect lessons learned and in­
creased use of demand­side, re­
newable, and distributed resources.
The “In My View” column supports
an opinion that the sound market prin­
ciples used to introduce wholesale elec­
tric markets during the 1990s should
still apply to systems with large pen­
etrations of renewable and demand re­
sources. Care must be taken with market
implementation details.
This issue also features a standalone
article, “pV­Battery Systems for Critical
loads during emergencies.” The extent
of damage to infrastructure after natu­
ral disasters can be devastating. people
with severe medical conditions are par­
ticularly susceptible to long­term electri­
cal outages, which can prove fatal. The
article discusses the sizing and operat­
ing experience of photovoltaic­battery
systems deployed in puerto rico, which
provided reliable and economical elec­
tric supply to critical loads.
PES Updates
The Ieee power & energy Society (peS)
has almost 40,000 members located in
150 countries, with 258 local Chapters
and 358 Student Branch Chapters. The
“leader’s Corner” column by Saifur rah­
man, peS president, discusses many key
peS activities, including technical, stan­
dards, and education/training activities;
the peS Corporate engagement program;
remote working groups that promote the
peS’s worldwide vision and focus; and
publications activities. Our leadership
welcomes your thoughts on future direc­
tions. You and the peS benefit from your
participation in the wide array of activities
and outreach to new members.
Awards
Tight publication deadlines resulted in two
awards not being mentioned in previous

World Leader in Developing

BIGWOOD SYSTEMS, INC. Powerful and Field-Proven
www.bigwood-systems.com
+1.607.257.0915 On-line and Off-line Tools
sales@bigwood-systems.com for Power Grid Modeling, Analysis, Assessment,
NY, USA
Head Office Branch Offices Enhancement and Optimization
Ithaca, NY Taipei, TWN | Beijing, CHN
Innovation prevails!

Network Applications for the Control Center

“Bigwood Systems’ Online Voltage
Stability product is simply the best in the
field.” - CAISO Project Leader
“BSI On-line Transient Stability
Assessment is a superb product.”
- TEPCO Manager
“The BSI VAR Management System is a
best practice tool for all utilities.”
- SERC Audit Staff
• Real-time and Look-ahead Static and Dynamic Security
Assessment (patented screening, ranking, and detailed
analysis)
• Reactive Power Reserve Monitoring, Management &
Enhancement
• Actionable Control Recommendation engine to
eliminate violations and mitigate unstable contingencies
• New-Generation State Estimation for Transmission and
Distribution Networks with a low number of measurements
using SCADA & PMU data
• Global Optimal Power Flow (for over 250,000 control variables with
AC/DC constraints)
• Distribution Network Management Tools
• Renewable Energy Forecasting, Uncertainty Modeling,
Representation & Management
• Advanced Measurement-Based Power System Modeling
(Generation, Transmission & Load)
World-Wide Customer Base
Over 40 electrical utilities served world-wide, including long-standing
customers such as California ISO, PJM Interconnection, Tokyo Electric
Power Company, Tennessee Valley Authority, and UK Power Networks,
among others. BSI’s advanced technology, innovation and satisfied
customers are key differentiators.

6 ieee power & energy magazine january/february 2019

 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 coordination: 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.
issues of this magazine. please congratu­
late the following 2018 awardees:
✔ Bhuvaneswari gurumoorthy for
the Ieee peS wanda reder
pioneer in power award, “for
outstanding research in power
quality converters and for educat­
ing new engineers”
✔ Marina Mondello, for the Ieee
peS wanda reder pioneer in
power award, “for leadership and
for serving as an inspiring role
model and mentor to young wom­
en in the power industry.”
peS members are encouraged to submit
nominations for 2019 awards by 31
January 2019 at https://www.ieee­pes
.org/pes­communities/awards/pes­
society­level­awards and https://www
.ieee­pes.org/pes­communities/awards.
Changing of the Guard
we offer congratulations and wish all
the best to leaders assuming new Ieee
positions in 2019. Miriam p. Sanders
will hold the incoming division VII
delegate­elect/director­elect position. In
addition, the following peS board mem­
bers took office on 1 January 2019:
✔ vice president, publications: Bi­
kash pal, Imperial College, london
✔ vice president, Meetings: wayne
Bishop, Omicron energy, Boston
✔ member­at­large, global Out­
reach: Yu Jun, State grid Corp
of China, Beijing
✔ member­at­large, global work­
shops: ramakrishna Kappagan­
tu, Bangalore, India.
Our deepest gratitude goes to the
following outgoing peS Board Mem­
bers for successfully completing their
tenure and for their many significant
contributions to peS:
✔ vice president, p ublications:
Mariesa l. Crow, Missouri Uni­
versity of Science and Technol­
ogy, rolla
✔ vice president, Meetings: Thom­
as w. Mayne, Mayne Consultants,
lacombe, louisiana
✔ member­at­large, global Outreach:
C.Y. Chung, University of Sas­
katchewan, Saskatoon, Canada
✔ member­at­large, global work­
shops: luis (nando) Ochoa, Uni­
versity of Melbourne, australia.
Thanks
a special note of appreciation to Mel
Olken, who continues to provide guid­
ance and tutelage, and to the Ieee pub­
lications staff who make this publication
possible. Thanks to the many contribu­
tors to this issue, especially our guest
editors, antonio J. Conejo and ramteen
Sioshansi, and our authors. a particular
note of appreciation to associate editor
John paserba and robert C. Henderson,
who provides editorial assistance.
p&e
 letters to the editor
R
ReadeRs aRe encouRaged to
share their views on issues affecting the
electric power engineering profession.
send your letters to Michael Henderson,
editor-in-chief: mih.psat@gmail.com.
Letters may be edited for publication.
A View into U.K. Education
the article in the september/october
2018 issue, “electrical Power engineer-
ing education: cultivating the talent in
the united Kingdom and Italy to Build
the Low-carbon economy of the Fu-
ture” (IEEE Power & Energy Magazine,
vol. 16, no. 5, pp. 53–63, sept.-oct. 2018),
overlooked something fundamental in
the description of university education in
the united Kingdom, which consists of
england, Wales, northern Ireland, and
scotland. the article states that B.eng.
degrees in the united Kingdom are three
years long and M.eng. degrees are four
years long. this is not the case in scot-
land, where B.eng. programs require four
years of study and M.eng. degrees take
five years. the differences result from
the secondary school structures in scot-
land and the rest of the united Kingdom,
which also explains the differences in en-
trance requirements.
—Martin Queen
Author’s Response
the article should have described the
structure of B.eng. and M.eng. elec-
trical engineering courses in scottish
Digital Object Identifier 10.1109/MPE.2018.2877898
Date of publication: 7 January 2019
10 ieee power & energy magazine
share your thoughts
send comments to mih.psat@gmail.com
universities and the entrance grades
required for admission to the courses
as being different from the rest of the
united Kingdom. I agree with Martin
Queen’s comments and will now explain
them for our readership.
In england, Wales, and northern Ire-
land, most students joining a three-year
B.eng. or four-year M.eng. eee degree
program will have successfully passed
three or four advanced-Level general
certificate of education
qualifications (referred
to as A-levels) at an ap- In Scotland,
propriate grade. they
must include mathemat-
most students
ics and usually physics, study five so-
and an appropriate grade
is normally an a*, a, or called Highers
B for a Russell group
university, which repre-
at school.
sents the top universities
in the united Kingdom. these grades are
obtained from a percentage numerical
mark, where a* > 90%, a = 80–89%, B
= 70–79%, c = 60–69%, d = 50–59%,
and e = 40–49%. F < 39% indicates an
examination failure.
In scotland, most students study five
so-called Highers at school. these are
generally awarded in the fifth year of sec-
ondary school and considered halfway
between the general certificate of second-
ary education and the a-level. advanced
Highers are an optional qualification, nor-
mally studied in the sixth year of second-
ary education and generally considered
slightly more difficult than a-levels.
the entrance requirement for a typical
four-year B.eng. eee course in scotland
is four Highers at the a grade and in-
cludes math, physics, or engineering sci-
ence. students studying a-levels, mainly
in other parts of the united Kingdom, re-
quire three a-levels at grade B, including
math and physics. However, students with
very good a-levels may be allowed direct
entry into the second year of a scottish
B.eng. course if they achieved grade a in
math, physics, and computing. emphasis
is placed on mathematics, engineering
science, circuits, basic
electrical engineering,
and computing during
the first year of an eee
B.eng. course in scot-
land. It is possible for
some students—those
with high a-level grades
in appropr iate sub -
jects—to complete their
B.eng. degrees within
three years, i.e., similar to the rest of the
united Kingdom.
the entrance requirement for a typi-
cal five-year M.eng. eee course in
scotland is five Highers at grade a in-
cluding math, physics, or engineering
science or three a-levels at grade B in-
cluding math and physics.
the second year of a four-year B.eng.
or five-year M.eng. eee course in scotland
is similar to the first year of a three-year
B.eng. or four-year M.eng. eee course
in england, Wales, and northern Ireland.
although the term similar recognizes that
the exact contents of different B.eng. or
M.eng. eee courses in u.K. universities
vary significantly, all accredited B.eng. or
M.eng. eee courses have been assessed
january/february 2019
by one or more of the u.K. professional to assert that general electric (ge)
engineering institutions, and all satisfy the was proposing “three phase” when the
educational criteria for an incorporated or war of the currents was
a chartered engineer. still raging at that time
could have been further
There is evidence
—Peter Crossley explained. as we know that several
from Marc seife r’s
Alternating Developments research, Forbes was Westinghouse
My letter is in regard to the “History”
column in the september/october 2018
a university professor
from anderson’s uni-
Company
issue, “the niagara alternators: com- versity, glasgow, scot- employees were
peting alternatives and Frequencies” land who had tested a
(IEEE Power & Energy Magazine, vol. 16, number of inventions. jealous of Tesla.
no. 5, pp. 96–104, sept.-oct. 2018). to He concluded that tes-
read that Forbes did all of the design la’s invention of the induction motor and
work on the niagara alternators with- ac power system was by far the best one
out any mention of the several patents and recommended it to the niagara Falls
by tesla on those generators does not Power commission (see Wizard: The
tell the complete story. to suggest that Life & Times of Nikola Tesla, p. 111).
“tesla’s motors were still under develop- there is evidence that several West-
ment” begs the question of his alternator inghouse company employees were
development, which is not mentioned. jealous of tesla, particularly stilwell
and shallenberger. Benjamin Lamme,
however, realized that the tesla system,
as tesla described it,
would work properly.
Lamme traveled to
new York frequently
with albert schmid so
that they could get ad-
vice from the master
on how to perfect the
motor and the entire ac
system so that it would
work at niagara. tes-
la’s name is mentioned
nine times on the patent plaque. schmid
is mentioned twice, and so is stanley.
the system was clearly tesla’s, and
that’s why he spoke as the inventor at the
inauguration of the power system. ad-
ams also credited tesla as the inventor.
—Thomas Valone

Advanced modeling and analysis features include:

Dynamic Security Assessment Software
Offering a simple, user-friendly interface
with extensive analysis options, DSAToolsTM
is a suite of software tools for power system
analysis. The software is designed for
applications in both off-line studies and
on-line dynamic security assessment.
Experience and support:
www.dsatools.com
• Comprehensive capabilities for voltage, transient, small signal, and
frequency stability assessment
• Integration with EMS/WAMS for on-line DSA to assess security
of real-time system conditions
• Computation of secure operation regions (stability limits)
• Recommendation of remedial actions
• Prominent provider of the on-line DSA technology with over
40 installations worldwide including 7 out of 9 ISOs in North
America
• Consulting services for custom software and model development,
system studies, and training

VSAT TSAT SSAT

Contact us:
dsainfo@powertechlabs.com
81022-0005

january/february 2019 ieee power & energy magazine 11

 Author’s Response
I am not aware of tesla’s alternator de-
velopment. If the writer is referring to
tesla’s patents, these were not develop-
ments. It’s a long road from defining the
parameters of an ac polyphase system
to designing a working machine. In his
book The Power Makers, Klein says,
“Westinghouse engineers resented tesla
for his pomposity and the large sum he
received for an invention they thought
worthless until scott and oliver shallen-
berger improved it.” the “History” col-
umn tones this down a bit by using the
phrase “under development.” this is the
reference: “By december of 1892, West-
inghouse has submitted its two-phase
ac design …ge’s entry … employed
three-phase ac,” taken from Jill Jonnes,
Empires of Light, p. 291.
this “History” column doesn’t elabo-
rate on this because the piece does not
12 ieee power & energy magazine
focus on ge using three-phase ac. ge
was bidding on Forbes’s design, which
called for polyphase ac. In any case, by
this time, ge had realized that dc was a
limited market and by 1896 had signed a
cross-licensing agreement with Westing-
house. see Harold c. Passer, The Elec-
trical Manufacturers, p. 331. the war of
the currents may have been “still raging at
that time,” but ge was about to capitulate.
It is clear from the patents that tesla
invented the polyphase power system.
It’s my contention that he didn’t design
the alternator. [Editor’s note: coinci-
dently, the “History” column from the
november/december 2018 issue was
on the subject of the tesla lab at 89 Lib-
erty street, new York city, which was
credited as the birthplace of commer-
cial polyphase power.]
the column uses information from
a statement prepared by Benjamin
Lamme. that statement was made at the
request of edward dean adams for pre-
sentation at the close of the edison gold
Medal ceremonies where the medal was
presented to Mr. Lamme. It is repro-
duced on pp. 409–421 of adams’s book,
Niagara Power, Volume 2. In it, Lamme
states, “From the standpoint of arma-
ture reaction and reactance this machine
[Editor’s note: this refers to the niagara
alternator based on Forbes’s design] was
excessively bad … the full load magne-
tomotive force was about three times as
great as the no-load field magnetomo-
tive, an inoperative condition from the
modern standpoint.” He goes on to say
that the “polar construction would have
rendered the machine inoperative.”
—Robert D. Barnett
p&e
january/february 2019
 Aalborg, Denmark - June 4-7, 2019

Key Themes Location

KeyInternational
Themes CIGRE Symposium Location Aalborg has a rich and long history, going back to
The aim of the 2019 CIGRE International
Key organized by the Danish National Location
Committee the Vikings. Successful heavy industries have
aim ofThemes isAalborg has a rich and long history, going back to
The the 2019 CIGRE International Symposium to provide a forum for recent gradually transitioned into an environment of
the Vikings. Successful heavy industries have
International CIGRE Symposium
and issupported
Symposium to provide a forum
The aim of the 2019 CIGRE International
by eight
for recent
research results and system operations experience
Study Committees
research results and system
Aalborg has a richoperations experience
and long history, going back to
gradually transitioned into an environment of
the Vikings.
related to the technology-rich
rapid Successful heavy
transformation industries have
and challenges
technology-rich enterprises, a feat not least due to
relatedGoing offshore - Challenges of onthe future
Symposium is to provide a forum for recent enterprises, a feat not least due to the presence of Aalborg University, organizer of
organized to the rapid
research results and by the operations
transformation
system Danish National
and challenges imposed
experience Committee
gradually
networks
the presence
technology-rich
of and
transitioned
and markets
Aalborg supported
into
by
enterprises,
an environment of
increased
University,
a feat
by
organizer of eight
not least due to
Study Committees
the symposium. Aalborg Congress & Culture Centre
imposed
Going power
on
offshore
related
networksgrid and markets
to the rapid-transformation
Challenges
by increased
of the
and challenges amounts of
future powerrenewables
the in
grid
the symposium.
presence
the energy mix.
AalborgUniversity,
of Aalborg
Perhaps
Congress & Culture of
organizer Centre (uk.akkc.dk) will host all meetings and technical
amounts
imposedofon renewables
networks and in the energybymix.
markets Perhaps
increased exemplified more than anywhere,
(uk.akkc.dk) will host Northern
all meetings and technical
Aalborg, Aalborg,
Denmark
exemplified thanDenmark
- anywhere,
June 4-7, -mix.
June
2019 4-7, 2019 the symposium. Aalborg Congress & Culture Centre sessions, and is located centrally, only 6 km from
amounts ofmore
renewables in the energyNorthern Europe
Perhaps is about to increase
sessions, and
(uk.akkc.dk) isits
will
already
located
host
large share
centrally,
all meetings onlytechnical
and 6 km from Aalborg airport.
Europe is about
exemplified moreto increase its already
than anywhere, large share
Northern of wind powerAalborg
with further
airport.offshore plants and
ofEurope
wind power with sessions, and is located centrally, only 6 km from
is about tofurther
increase offshore plants
its already and
large share
offshore interconnectors.
Aalborg airport.The objective of the
offshore
of windinterconnectors.
power with further The objective
offshore of the
plants and symposium is to address recent technical advances
symposium is to address recent technicalofadvances
Key Themes and
offshore interconnectors.
policy that allow
The objective
transmission and
symposium is to address recent technical advances
the and policy that allow transmission and distribution
distribution
networks to enter their next era, in particular those
The aim of the 2019 CIGRE International
networks
and policy Symposium
to enter
that is toin provide
theirtransmission
allow next era, particular athose
forum
and distribution
related to offshore generation, transmission and
Key Themes
related
system
to offshore
networks
for recent research results anddistribution. to entergeneration,
operations Location
their next era,
experience
transmission
in particularand
related those
related to offshore generation, transmission anddistribution. The preferential topics reflect the
The preferential topics reflect the
to the rapid transformation andabove, challenges
and beyond,
distribution. Theimposedacross
preferential on networks
technology,
Aalborg topics and
planning,
hasreflect
a rich the above,
and long and beyond,
history, going back across
to technology, planning,
The aim of the 2019 CIGRE Internationaloperations
above, and and markets.
beyond, across technology, planning,operations and markets.
markets isbytoincreased
Symposium provide a forumamounts of
for recent renewables in the energy mix. Perhaps
operations and markets.
the Vikings. Successful heavy industries have
gradually transitioned into an environment of
exemplified
research more
results and thanoperations
system anywhere, Northern Europetechnology-rich
experience is about to increase enterprises, its a feat not least due to
large share of wind Programme
related to the rapid transformation and challenges
already power with further offshore plants of
the presence and Programme
offshore
Aalborg University, organizer of
imposed on networks and markets by increased Programme Abstracts
the symposium. Aalborg Congress & Culture Centre
interconnectors.
amounts of renewablesThe objective
in the of the
energy mix. symposium is take
Perhaps to address
The main symposium will(uk.akkc.dk) place on
recent
willWednesday
Abstracts
host all meetings and technical Abstracts
exemplified
technicalmore than anywhere,
advances Northern
and policy that
5thThe allow
main
June transmission
andsymposium
Thursday will
6th and
sessions,
take
June. placedistribution
and on
Tutorialsis Wednesday
located Abstracts
The main symposium will take place on Wednesday
will takecentrally, onlyParticipants
6 km from wishing to present papers are invited
Europe is about to increase its already large share 5th June and Thursday 6th June. Tutorials will takeare invited Participants wishing to present papers are invited
of networks
wind powerto enter
with their
further next plants
offshore era, 5th
in
place June and
particular Thursday
Tuesday 4ththose
and
Aalborg
6th June.
related
June and airport.
Tutorials
to
a technical visitwill
offshore to take Participants
Participants
to submit their wishing
wishing
synopsis toENGLISH
to present
in present
papers papers
before 12th are invited to submit their synopsis in
place Tuesday
offshore 4th June
wind power plantand a technical
ANHOLT to place TuesdayOctober
visitplace
will take 4th
to June
submit andto:
their
2018
a https:/
technical
synopsis visit
in ENGLISH to before
/cigreaalborg2019.dk/12th to submit their synopsis in ENGLISH before 12th
generation,
offshore transmission
interconnectors. and distribution.
The objective ofoffshore
the The
wind power
Friday 7th June. preferential
plant
Spouses’ tripANHOLT topics
will visitwill reflect
take place
Grenen, the
offshore
Art wind ENGLISH
power plantbefore
ANHOLT 12th
will October
take place 2018
October 2018 to: https://cigreaalborg2019.dk/ to: https://cigreaalborg2019.dk/
October 2018 to: https://cigreaalborg2019.dk/
symposium is to address recent technicalFriday advances
7th June. Spouses’ trip will visit Grenen, Friday
Art 7th June. Spouses’ trip will visit Grenen, Art
andabove, andallow
policy that beyond, acrossand
transmission technology,
Museum
distribution
Museum of
of planning,
Skagen
Skagen
and operations
Brøndums
andgroup
Brøndums
and
Hotel. markets.
Study
Hotel.areStudy Museum of Skagen and Brøndums Hotel. Study
committee and working meetings
networks to enter their next era, in particular those
Programme
related to offshore generation, transmission 4th
committee
organized and working
between group
Saturday 1stmeetings
June and are
and between Saturday 1st June and Tuesday
organized
June.
Tuesday committee and working group meetings are
organized between Saturday 1st June and Tuesday
The mainThe
distribution. symposium will take
preferential topics place
reflect on Wednesday 5th June and Thursday
theJune.
4th 4th June.
above, and beyond, across technology, planning,
6th June. Tutorials will take place Tuesday 4th June and a technical visit to
operations and markets.
offshore wind power plant ANHOLT will take place Friday 7th June. Spouses’ International CIGRE Symposium
trip will visit Grenen, Art Museum of Skagen and Brøndums Hotel. Study organized by the China National Committee
Programme
committee and working group meetings are organized between Saturday 1st and supported by six Study Committees
June
mainand Tuesdaywill4th
takeJune.
Abstracts
The symposium place on Wednesday
5th June and Thursday 6th June. Tutorials will take Participants wishing to present papers are invited
Towards active, sustainable digital networks
Location
place Tuesday 4th June and a technical visit to to submit their synopsis in ENGLISH before 12th that are resilient and integrated from UHV
offshore wind power plant ANHOLT will take place
Aalborg has a rich and long history, going back to October
Friday 7th June. Spouses’ trip will visit Grenen, Art
2018 to:
the Vikings. Successful More
More information
https://cigreaalborg2019.dk/
information to distribution
heavyofindustries
Skagen andhave gradually
Hotel.transitioned into an environment of
Museum
technology-rich
committee
Brøndums
enterprises,
and working group meetings
Study
are not least due to thePlease
a feat Pleasevisit
presence our
ourwebsite
of Aalborg
visit website for
for more information
more information
More information
Chengdu, China - September 20-25, 2019
organized between Saturday 1st June and Tuesday https:/ /cigreaalborg2019.dk/
University,
4th June. organizer of the symposium. Aalborg Congress & Culture
https:/Centre
/cigreaalborg2019.dk/ Please visit our website for more information
(uk.akkc.dk) will host all meetings and technical sessions, and is located
https://cigreaalborg2019.dk/
centrally, only 6 km from Aalborg airport.
Theme International CIGRE Symposium
Venue
organized are
Today, new global energy challenges by the China National Committee
emerging. Chengdu is one of the most famous historical
The application of the huge and supported
number of newby six Study Committees and cultural cites of China, and it is a city that
technologies and new equipment makes the
Towards active, sustainable digital you willnetworks
never want to leave once you come. It
International CIGRE Symposium
electric power systems more complex. The deep
that are resilient and integrated offers a great variety of attractions, entertainment
from UHV
organized by the Danish National integrationCommittee
of information technology and power
to distribution
and lifestyles. The Intercontinental Century
systems brings not only intelligence but also City Chengdu will be the venue for all technical
and supported by eight Study
More information Committees Chengdu,
International CIGRE Symposium
Going offshore - Challenges
challenges to the energy systems.
those new ofchallenges
the future
China
To deal with- September 20-25,
and achieve goals such as
sessions.2019

organized by the China National

Please visit our website for more information Committee and
the UN Sustainable Development Goals for 2030,supported by six Study Committees
Towards power grid
https:/active, sustainable digital
/cigreaalborg2019.dk/
Theme
Parisnetworks
Agreement and that air are resilient
pollution abatement,
Venue
and integrated from UHV to distribution
Chengdu,Aalborg, Denmark -20-25,
June we4-7,
need to 2019
reconsider and investigate the issue
China - September Today, new
2019
ofglobal
energy development
energy and provide effective
challenges are emerging. Chengdu is one of the most famous historical
The application of the huge number of new
systematic solutions for the upgrading of the and cultural cites of China, and it is a city that
technologies and new equipment makes the you will never want to leave once you come. It
power
electric power systems.
systems more complex. The deep offers a great variety of attractions, entertainment
Themes integration of information technology
The aim
systems brings of the
not only
attractions,
CIGRE
intelligence
and power
butsymposium
also
entertainment
Chengdu and
2019
City Chengdu
lifestyles.
and lifestyles. The Intercontinental
The Intercontinental Century
will be the venue for all technical
Century City Chengdu
Today, new global energy challenges are emerging. The application of the
challenges is totothe
huge provide willTobedeal
a forum
energy systems. forthe venueresearch
recent
with for all sessions.
technical
findings sessions.
those new challenges and achieve goals such as
and operation experiences related to sustainable,
number of new technologies and new equipment makesLocation the electricthe
power systems
UN Sustainable Development Goals for 2030,
Key Themes
more complex. The deep integration of information technology and power
active
Paris Agreement
systems
we need system(From
and
and airdigital
pollutiondevelopment
UHV
to reconsider and investigate
abatement,
toback
distribution),
the issue
of power
and in particular,
Aalborg has a of
rich anddevelopment
energy
long history, and
going
provide effective
to
brings
The aim ofnot
theonly
2019intelligence but also challenges to the energy
CIGRE International systems.
the Vikings. To deal
Successful
systematic withindustries
tosolutions
address
heavy recent have technical
for the upgrading of the
advances and policy
Symposium is to provide a forum for recent that
intoallow transmission of and distribution networks
those new
research challenges
results and system and
operations goals such as the UNgradually
achieveexperience transitioned
Sustainable Development
power systems.
tothe
an environment
enter
technology-richTheenterprises,
aim of CIGREa their
feat next
not
symposium era.due 2019
least
Chengdu to
Goalstofor
related the2030, Paris Agreement
rapid transformation andand air pollution abatement,
challenges we need
the presence is to
of to reconsider
provide
Aalborg aUniversity,
forum for recent researchof
organizer findings
imposed on networks and markets by increased and operation experiences related to sustainable,
and investigate the issue of energy development
amounts of renewables in the energy mix. Perhaps and provide effective
the symposium. systematic
Aalborg
active and
Congress
digital development
& Culture
of power
Centre
(uk.akkc.dk) will host all meetings and technical
solutions more
exemplified for the upgrading
than anywhere,ofNorthern
the power systems. system(From
sessions, and is
to located
UHV to distribution),
centrally,
address recent technical only
and in particular,
6 km and
advances from policy
Europe is about to increase its already large share
Programme Submission of synopses
that allow transmission and distribution networks
of The
windaim
powerof the
withCIGRE
further symposium
offshore plantsChengdu
and 2019 is Aalborg airport.
to provide a forum for recent
to enter their next era.

research
offshore findings andThe
interconnectors. operation experiences
objective of the related to sustainable, active and The technical programme sessions opens with Participants wishing to present and publish
symposium is to address recent technical advances
anddigital
policydevelopment of power and
that allow transmission system(From
distributionUHV to distribution), and in particular,
keynotes and welcome reception on September papers should submit synopses by the following
to address recent technical advances and policy that allow transmission
networks to enter their next era, in particular those and
Programme 23, 2019. Before Submission of synopses
that day, we will offer aSubmission of synopses
technical instructions on www.cigre2019-chengdu.org,
related to offshore generation, transmission and visit to±500kV convertor station or Zipingpu where the synopses and paper templates are also
distribution networks to enter their next era. Participants
The technical programme sessions opens with
wishing to present and publish papers should
Participants wishing to present and publish
submit synopses
papers, by
andthe
distribution. The preferential topics reflect the keynoteshydropower station.
and welcome reception Tour of Chengdu will
on September alsoshould
papers be submit synopses by available. All synopses,
the following presentations
Programme
above, and beyond, across technology, planning, 23, 2019. organized
Before that day, we following
for accompanying
will offer a
visit to±500kV convertor station or Zipingpu
instructions
persons. on www.cigre2019-chengdu.org,
technical instructions on must be in English.
www.cigre2019-chengdu.org,
where the synopses and paper templates are also
where the
Please synopses
note authors should be
operations and markets. Individual Study Committee meetings and Working available to present at the symposium.
The technical programme sessions opens with keynotes and welcome reception and paper
hydropower station. Tour of Chengdu will also templates
be are also available.
available. All synopses,
All synopses, papers, papers,
and presentations
organized Group meetings
for accompanying will take place from September
persons. 20 Please note authors should be
must be in English.
on September 23, 2019. Before that day, we will offer a technical visitIndividual
to±500kV
toStudy and presentations
21. Committee meetings and Working mustavailable
be in toEnglish.
present atPlease note authors should be available to
the symposium.
Programme Group meetings will take place from September 20
present at the symposium.
convertor station or Zipingpu hydropower station. Tour of Chengdu will
to 21.also be

organized
The for accompanying
main symposium persons.
will take place on Wednesday
Abstracts
Individual
5th June and Study Committee
Thursday meetings
6th June. Tutorials willand
takeWorking Group meetingswishing
Participants will take place papers are invited
to present
place Tuesday 4th June and a technical visit to to submit their synopsis in ENGLISH before 12th
from September
offshore wind power 20 toANHOLT
plant 21. will take place October 2018 to: https://cigreaalborg2019.dk/
More information More information
Friday 7th June. Spouses’ trip will visit Grenen, Art
Venue
Museum of Skagen and Brøndums Hotel. Study
Please visit our website for more information
Please visit our website for more information
www.cigre2019-chengdu.org
Chengduand
committee is one of the
working groupmost famousarehistorical and cultural cites of China, and it is
meetings
organized between Saturday 1st June and Tuesday
a June.
city that you will never want to leave once you come. It offers a great variety of
www.cigre2019-chengdu.org
4th

Digital Object Identifier 10.1109/MPE.2018.2885099

page-aalborg-chengdu.indd 2 22/10/18 15:54

 leader’s corner
I
I would lIke to update you
on the progress and priorities of our
Society. Currently, we have more than
38,900 members, which is over 1,000
more than last year. we are truly a global
organization with members located in
150 countries, but we also value our con­
nections at the local level through our
Chapters, which provide us with our grass­
roots contacts. the Ieee power & energy
Society (peS) now has 258 such Chap­
ters (seven more than last year) and 358
Student Branch Chapters (60 more than
last year), in almost every country where
our members live.
In addition to our technical, stan­
dards, and education/training activities, I
am focusing on developing stronger and
sustaining relationships with these pro­
fessional Chapters and Student Branch
Chapters globally. Chapter members are
our foot soldiers, providing links to in­
dustry engineers, government agencies,
and academics from all over the world.
they can respond to local needs and
priorities and bring important programs
to meet those needs. to strengthen the
local Chapters and help them interact
more closely within their country or re­
gion, we are now forming peS Chapter
Councils. these councils will directly
work with their respective peS Region
representatives and, therefore, have ac­
cess to the peS governing board. they
Digital Object Identifier 10.1109/MPE.2018.2872428
Date of publication: 7 January 2019
14 ieee power & energy magazine
increasing engagement
an update on Chapters & councils
will have budgets to support travel within
their country or region, as appropriate.
these councils will facilitate peS activi­
ties in their areas, focusing on these top­
ics (and others as needed):
✔ membership
✔ conferences
✔ technical activities
✔ education, including certificate
programs
✔ standards.
Councils will interact with their nation­
al/regional electrical engineering organi­
zations to develop joint programs, build
awareness among engineers and academ­
ics, and promote peS membership. So far,
such councils have been formed in China,
India, and africa. plans are afoot to create
councils in latin america, North america,
and europe as well.
another signature effort of peS to raise
awareness and build a stronger member­
ship base is the peS Corporate engagement
program. participation in this program of­
fers corporations and other organizations
an opportunity to help their employees to
stay up to date on the latest advances, in­
cluding renewable energy system design
and integration, smart grid technologies,
electric vehicles, wholesale market design,
and operation and asset management. By
interacting with peS through conferences,
committee work, the distinguished lec­
turer program, and so on, employees of
an organization can have many opportu­
nities to engage professionally with their
peers all over the world. this opportunity is
Saifur Rahman
implemented through the Ieee Corporate
Membership program where a company
or an organization can become a corporate
member by paying the peS membership
fee for their employees so they can join
peS as professional individual members
(minimum 50 members).
to give our members outside North
america the opportunity to participate in
technical committee activities and stan­
dards development, we have formed vari­
ous remote working groups with the help
of the peS technical Council. For ex­
ample, the peS Smart Buildings, loads,
and Customer Systems (SBlC) technical
Committee has formed the asia­pacific
working Group (apwG), whose scope
includes meeting with stakeholders for
SBlC within the asia­pacific region to
regularly address issues within the scope
of the committee, generally described as
electrical systems behind the meter, and
integration of such systems with grid ser­
vices. the peS website, http://sites.ieee
.org/pes­sblc/, has more details about our
committees and their activities.
these regional working groups hold
meetings in various locations in person,
with electronic access. For example, sever­
al meetings of the apwG have been held
in Guangzhou, China, hosted by China
Southern Grid. the working group’s last
meeting was held at the 2018 peS Gen­
eral Meeting in portland, oregon. these
working group meetings have resulted in
the ongoing development of three project
authorization requests (paRs), with two
january/february 2019
 ADVERTISEMENT

•
etap 19
•
•
•
 ESMO 2019
NEW HORIZONS
BOUNDLESS
POSSIBILITIES
Early Bird Registration Now Open
June 24-27, 2019 | Columbus, Ohio | ieee-esmo.com

Where Engineering Principles

Meet Field Realities
ESMO 2019 is set to be the industry’s largest “how to” event in North America. ESMO serves electric utility
professionals, contractors, and consultants who are interested in hands-on solutions for the safe engineering,
construction, operation and maintenance of the world’s power delivery systems. The four day event features
two days of outdoor field demonstrations and exhibits and a two-day technical conference program with
indoor exhibits.
IEEE Power and Energy Society and American Electric Power have partnered together and pooled their
resources to bring you this unique event where manufacturers and service providers connect with those who
engineer, design, operate and maintain the electrical network. PES and AEP envision this conference to be
the premier Safety event in the industry and have engaged T&D World to help make that happen.

For Information & Program Updates, Visit: ieee-esmo.com.

For Exhibit & Supporter Information, Contact:

Sam Posa | Exhibit Sales | 800.525.9154 | sam.posa@informa.com
Steve Lach | Midwest, Canada | 708.460.5925 | steve.lach@informa.com
Matt Wintringham | Eastern US | 508.277.8397 | matthew.wintringham@informa.com

General Information, Contact:

Digital Object Identifier 10.1109/MPE.2018.2885105 Eriks Surmanis | e.surmanis@pdc-cables.com
 Most importantly, a particular skill. to formalize various
training and lifelong learning opportu­
I hope all these activities will go a long
way to make peS even more relevant to
please remain nities, peS is in the process of estab­ the global working professional. as your
lishing the peS university under the president for 2018–2019, I would like to
engaged with Ieee umbrella. this will be a one­stop hear your suggestions, questions, and ad­
your Society and shop for members or groups of engineers
as well as students and faculty to learn
vice. But most importantly, please remain
engaged with your Society and ask your
ask your friends new technologies, processes, deployment friends and colleagues to join.
p&e
experiences, and so on.
and colleagues
to join.
approved from SBlC­apwG efforts ad­
dressing matters of interest to the com­
mittee for Quick Response of Customer
loads (submitted to the Ieee Standards
association in September 2018) and
p2781—load Modeling and Simula­ IEEE 1584 Updated for 2018
tion. an additional paR concerning the
integration of loads from electric rail is The Second Edition of IEEE 1584 - IEEE Guide for
being developed in coordination with the Performing Arc Flash Hazard Calculations is finally
Ieee Vehicular technology Society. we here. New challenges are ahead and include:
encourage our members outside North
america to engage with the peS tech­ · New data collection requirements
nical Council (https://www.ieee­pes.org/ · New electrode configuration considerations
technical­activities/technical­council) if · New enclosure size considerations
they want to form such working groups
that can operate remotely.
· New calculations and procedures
peS publishes six technical journals · New recommendations and limitations
and two magazines. In addition to hard Simplify a transition to the new standard with
copies, all of these periodicals are available EasyPower software.
online. one of the six journals, IEEE
Power & Energy Technology and Sys-
tems Journal, is open access, and papers
are practice oriented. In this journal, in­
dustry engineers can publish their case
studies, best practices, and unique field · Updated software
deployment experiences for the benefit
of working engineers, who can access · Intro to IEEE 1584-2018
the material for free. More details can
be found at https://www.ieee­pes.org/ · Comparing 2002 vs. 2018
publications/transactions/power­and­
energy­technology­systems­journal. · Instructional videos
peS is deploying significant resourc­
es for training and lifelong learning
· Practical guides
activities for our practicing engineer
members. there are numerous short Explore more online and
courses, webinars, and tutorials— request a free demo copy at:
both online and in person—available
for members throughout the world. un­ www.EasyPower.com/1584
der the peS Corporate engagement
program, we can also offer customized
courses at the company level for groups ®
of engineers needing to be trained in Power made easy.
 guest editorial
T
The firsT efforTs To resTruc­
ture electricity markets in europe and the
united states were in the late 1990s. At
that time, electricity production based
on variable weather­dependent sources
was virtually nonexistent. As a result,
those early efforts did not take into ac­
count the properties of those resources.
Today, such generators are dominant
in some markets, such as in california,
Texas, the iberian Peninsula, and Den­
mark. The contribution of variable re­
newable resources in other countries and
regions is increasing at a rapid pace.
on the flip side, electricity demand
was largely fixed and inelastic in the
1990s, with the exception of some indus­
trial and large commercial customers.
These types of customers display more
price sensitivity in their loads, especially
if energy constitutes a major input to their
production processes. even though elec­
tricity demand remains largely fixed to­
day, new loads, such as those pertaining
to electromobility, and behind­the­meter
production facilities are slowly increas­
ing demand elasticity and flexibility.
The operational and planning conse­
quences of these major changes (such
as the dramatic increase in weather­de­
pendent electricity production and the
incipient increase in demand flexibility)
need to be reflected in the design of elec­
tricity markets of the future. it is crucial
to spur a conversation to scrutinize to
what extent (if any) these formative mar­
ket designs need to be updated or sim­
ply replaced by a new design that better
Digital Object Identifier 10.1109/MPE.2018.2871737
Date of publication: 7 January 2019
18 ieee power & energy magazine
Antonio J. Conejo and Ramteen Sioshansi
electricity market
a conversation on future designs
incorporates weather­dependent renew­
able producers and increasing demand
flexibility. This issue of IEEE Power &
Energy Magazine is intended to promote
such a conversation with the ultimate
goal of achieving increasingly efficient
and fair markets for electricity.
This issue contains six articles and
concludes with the “in My View” column.
The first three articles describe the evolu­
tion of market designs and forthcoming
challenges in europe, the united states,
and New Zealand (with the latter being a
highly renewable­dominated system). The
following two articles summarize market­
design challenges from the perspective of
regulators in the united states and europe.
The sixth article gives an in­depth analysis
of current market designs from a historical
perspective and suggests redesign criteria.
The “in My View” column, reviews and
clarifies the fundamental economic prin­
ciples on which markets rely.
in the first article, “european union
electricity Markets,” Gómez and his
coauthors provide an overview of the
current state of electricity markets in
europe. They emphasize policies that
have been pursued by european in­
stitutions to achieve greater market
integration of member states through
the electricity Target Model. The day­
ahead, intraday, and real­time markets
are revisited, detailing what has been
achieved in terms of integration and
what lies ahead. The capacity alloca­
tion and congestion management regu­
lation is particularly important as it
allows coordination between different
national and regional markets.
The second article, “electricity Mar­
kets in the united states,” by Litvinov,
Zhao, and Zheng, reviews the main fea­
tures of electricity markets in the united
states that resulted from federal energy
regulatory commission (ferc) order
888. The focus, in particular, is on pric­
ing and market­coordination issues and a
review of day­ahead and real­time energy,
forward reserve, and forward capac­
ity markets. A description of the market
management system components is given,
which is the internal architecture of the
system operator that allows it to effectively
function. They identify the main challeng­
es faced by markets in the united states
today, including those that pertain to the
grid, and provide ideas to tackle them.
The next article, “The New Zealand
electricity Market,” by Philpott et al.
describes the historical evolution, main
features, and governance of the New
Zealand market. New Zealand pro­
vides an interesting case study, as it is a
completely isolated island system with
seasonal interannual variability in the
availability of (hydroelectric) renewable
energy resources. The authors survey
the historical market and regulatory in­
efficiencies, which market restructuring
in New Zealand has tried to address.
“A change is coming,” the fourth
article, by fulli et al. provides a vision
of how the supply and use of electricity
within europe is expected to evolve over
the next 30 years. They show how regu­
latory changes and research, demonstra­
tion, and development projects supported
by european institutions will change
the nature of customer interactions with
january/february 2019
electricity. in the authors’ vision of the implementation details, not from the ing electricity markets increasingly effi­
future, customers are more engaged in underlying economic principles. cient and fair for all market agents. The
making electricity supply­and­demand our goal in putting together this issue ultimate goal, as stated by the electricity
decisions, as opposed to leaving these to is to spur a much­needed conversation on Authority of New Zealand, is promoting
utilities and supply companies. market redesign, which should involve “competition in, reliable supply by, and
in the fifth article, “Wholesale elec­ operators, regulators, market agents, and the efficient operation of, the electricity
tricity Markets in the united states,” the research community. such a conver­ industry for the long­term benefit
Nicholson and Quinn provide t hei r sation will certainly produce innovative of consumers.”
p&e
view of the ferc in regulating and ideas that will contribute to make exist­
reforming competitive wholesale elec­
tricity markets in the united st ates.
They discuss the means by which the
ferc has encouraged the development
of increasingly competitive wholesale making life visibly safer
markets. The authors also outline three
challenges that these markets will have
to contend with in the near future:
1) the effects of declining wholesale
prices on capacity investment, 2) the
need for a new fleet of flexible energy
resources and the most efficient means
by which to incentivize the develop­
ment of such assets, and 3) coordina­
SpanLite TM

tion between electric power and natural

Self-Illuminated
gas systems. Power Line Marker
in the final feature article, “The
Three Waves of u.s. electricity Market Meets FAA Advisory
reforms,” hobbs and oren provide a Circular 70/7460-1L
detailed analysis of the historical and (Dec. 2015)
future evolution of electricity markets.
Installs directly on
They identify this as taking place in
live lines up to 500 kV
three waves. The first, unbundling
and debunking the natural monopoly
myth, was the formative step in market
restructuring. The second, focus on
economic efficiency and incremental
improvements, was the natural genesis
of lessons learned from the california
market crisis of 2000 and 2001. The
third, making way for demand­side,
renewable, and distributed resources, is
the natural evolution that we are cur­
rently undergoing. They also introduce
us to some of the major figures of mar­
ket restructuring.
The issue concludes with an “in My
View” column by hogan. he insight­
fully notes that the basic premises of BirdMark® & FireFly® FAA Obstruction SpanMarkTM & QuikMarkTM
Bird Diverters Lights Low Line Markers
supply and demand, market clearing,
and pricing that applied when market
restructuring took place in the 1990s
will still apply to a future power sys­

making life visibly safer

tem with renewable energy supply and
demand flexibility. he concludes that 800-722-8078 • pr-tech.com
6851

any potential market flaws stem from

january/february 2019 ieee power & energy magazine 19

 European
Union Electricity
Markets

Current
Practice and
Future View

Digital Object Identifier 10.1109/MPE.2018.2871739 ©istockphoto.com/tommasourbinati,

Date of publication: 7 January 2019 flag—©istockphoto.com/artJazz

20 ieee power & energy magazine 1540-7977/19©2019IEEE january/february 2019

By Tomás Gómez, Ignacio Herrero, Pablo Rodilla,
Rodrigo Escobar, Salvatore Lanza, Ignacio de la Fuente,
Maria Luisa Llorens, and Paula Junco
T
The european union (eu) aims To develop an
integrated electricity market as an efficient instrument to
achieve the energy policy targets of security, affordability, and
sustainability. many milestones have already been reached,
including the implementation of a european platform for
the simultaneous clearing of day-ahead markets, the design
of several platforms for coordinating intraday and regional
balancing markets, and the enforcement of guidelines for the
design of capacity mechanisms (Cms). There are still many
challenges ahead for this market: the integration of large
amounts of renewable energy, the proliferation of distributed
resources, higher consumer engagement, and the sheer scale
and complexity of a market serving 500 million citizens. This
is a fascinating task that will keep us busy in the coming years.
The formal beginning of the european electricity mar-
ket took place in 1996, when the european parliament and
the european Council approved the first electricity direc-
tive (96/92/eC) aimed at liberalizing the energy sector by
establishing competitive electricity markets. The subsequent
energy packages continued the development and integra-
tion of european electricity markets, not only with an inter-
est in deregulation but as part of the overall european
energy policy.
one of the eu’s main objectives is to share a single market
that guarantees the free movement of goods, capital, servic -
es, and labor. in this context, a single market for electric-
ity allows greater integration of the member states, which is
also necessary for the eu to have a common and coordinated
external energy policy. Therefore, the electricity market has
been designed as an instrument to achieve the goals of supply
security, affordability, and sustainability of the eu.
The common electricity market has been (and is being)
built gradually, beginning with its institutional arrangement.
The first major transition is the unbundling of generation,
transmission, distribution, and retail activities. The eu leg-
islation allows some freedom to the member states to imple-
ment different schemes for unbundling transmission, most
commonly the full ownership unbundling through a transmis-
sion system operator (Tso), which owns and operates the grid
and must not have any interest in generation activities. The
other basic building block in this organization is the power
exchange (pX) or, more specifically, the nominated electricity
market operator (nemo) when it refers to the operation of the
integrated day-ahead and intraday markets. market operators
january/february 2019
provide a marketplace where wholesale buyers and sellers can
exchange electricity in an organized way, with a set of rules,
at public prices.
Therefore, the two basic functions necessary in a deregu-
lated power system, system operation and the organization
of competitive electricity markets, are split into two separate
entities: Tsos and nemos. This institutional framework is
clearly different from the most common structure found in
the united states, in which independent system operators
perform both market and system operation functions. on
top of these foundations, additional rules must be defined to
achieve the main goal of the eu electricity market, which is to
allow cross-border trading of electricity and optimize the use
of interconnections, while pursuing a level playing field for
all market participants. rules are set by defining framework
guidelines, which are then detailed in binding regulation as
network codes or guidelines. These are developed by the
european Commission (eC), the agency for the Cooperation
of energy regulators, the national regulatory authorities
(nras), and the european network of Transmission system
operators for electricity in a cooperative process.
This cooperative process has led, slowly but steadily, to
a great degree of integration among the eu electricity mar-
kets. however, the process is still far from being fully com-
pleted. This article reviews the major milestones that have
already been reached in the integration of the eu electricity
markets as well as the major challenges ahead.
EU Target Model
The eu Target model refers to the institutional arrangements
and conceptual market definitions that have been developed
to ensure efficient cross-border exchanges in the european
electricity market. The formal definition of those arrange-
ments has been established in the Capacity allocation and
Congestion management (CaCm) regulations, which cover
cross-zonal capacity allocation in the day-ahead and intraday
time frames as well as in the Forward Capacity allocation
(FCa) regulations for long-term capacity allocation. in addi-
tion, the Guideline on electricity Balancing (GleB) regula-
tion complements those regulations with common principles
and methodologies for the procurement and settlement of
operating reserves and their activation in real-time operations.
The overall market model uses sequential markets: day-ahead,
intraday, and balancing.
ieee power & energy magazine 21
 The formal beginning of the European electricity
market took place in 1996.

in day-ahead markets, the nemos match demand and spain and portugal zones; epeX spot operates in the
supply bids to set hourly energy prices in each bidding zone. Germany, austria, France, Great Britain, The netherlands,
The european day-ahead market is a zonal price market, in Belgium, and switzerland zones; and nord pool operates
contrast with nodal price markets (as in the united states). in the norway, sweden, Finland, denmark, Great Britain,
as shown in Figure 1, most member states feature a single estonia, lithuania, and latvia zones. other nemos man-
bidding zone, but some are divided into several zones to age single-country zones, such as Gestore mercati energet-
reflect structural internal congestions, i.e., italy and swe- ici (Gme) in italy.
den. in addition, the nemos may manage several bidding in the intraday time frame, the nemos are working to
zones belonging to different member states. For instance, set a european continuous implicit trading platform. in the
the operador del mercado ibérico (omie) manages the future, intraday cross-zonal capacities will be priced reflect-
ing congestion. in some peripheral
markets, such as in the iberian pen-
insula, italy, and ireland, and more
recently in the nordic region, the
continuous market could be com-
SW1 plemented with intraday auctions
that increase liquidity and price
transparency. Cross-zonal capac-
NO4 ity was previously determined
SW2 FIN by the Tsos and then implicitly
NO3
allocated in day-ahead and intra-
NO5 NO1 day markets by the corresponding
SW3 clearing algorithms.
NO2 Currently in europe, most bal-
ancing markets for the procure-
SW4
DK1
ment and settlement of reserves
NI DK2 and real-time operation are local
in scope and mainly managed
IE GB
independently by each Tso. how-
NL PL ever, several regional initiatives
GE have already started to coordinate
BE
LX CZ the activation and exchange of dif-
SK ferent types of balancing energy
AU among neighboring Tsos, in line
HU
FR SZ with the mandates of the GleB
SLO regulation. The final goal of this
IT1
regulation is to set a progressively
pan-european market with a com-
IT2 mon merit order for activation of
PR IT3 balancing energy. These main
SP IT4 regulations (FCa, CaCm, and
IT5 GleB) are complemented with a
set of other network codes on grid
IT6 connection and system operation
that complete the presented eu
figure 1. The bidding zones in European power markets. (Image courtesy of the Of- Target model to ensure an inte-
fice of Gas and Electricity Markets, “Bidding Zones Literature Review.” London, United grated, efficient, and secure euro-
Kingdom, 2014.) pean electricity market.

22 ieee power & energy magazine january/february 2019

NEMO Markets: Day Ahead and Intraday
Role of the NEMOs
The CaCm provides the legal definition, designation pro-
cess, and roles for the nemos. each member state has the
ability to assign the nemo role to one (legal monopoly) or
more (in competition) entities. market participants in a given
member state submit bids and receive market results through
their assigned (in case of a monopoly) or preferred nemo,
which gives them access to the european day-ahead- and/
or intraday-coupled markets. The most relevant task to be
performed by the nemos is the market coupling operator
(mCo) function, defined as the task of matching orders for
the day-ahead and intraday markets for different bidding
zones and simultaneously allocating cross-zonal capacities.
The mCo function is a fundamentally centralized task and,
therefore, requires tight cooperation and coordination among
nemos. in addition, given that day-ahead and intraday mar-
kets in the eu perform the implicit allocation of intercon-
nection capacity, close coordination with Tsos is required
(since they are responsible for computing available transmis-
sion capacities).
The CaCm assigns the mCo function jointly to all
nemos. This implies that the same entities required to
compete at the local level to perform the nemo role have
to cooperate at the central level to provide the mCo func-
tions. all designated nemos have a mandate to agree on a
plan that sets out how to jointly set up and perform the mCo
functions. This plan was approved by all nras in June 2017,
and according to the CaCm, it must be implemented within
12 months. Consequently, all nemos in coordination with
the Tsos achieved a number of
contractual milestones (i.e., a
cooperation agreement and sev-
eral operational agreements) and
technical milestones specific to
either the day-ahead or the intra-
day market.
Day-Ahead Markets:
PCR Initiative
Implementation Challenges
The price coupling of regions
(pCr) initiative developed the
methodology for the joint clear-
ing of the day-ahead market in
europe. as already introduced,
this requires a common process
to calculate electricity prices
in all european bidding zones
and a single market-clearing
a lgor ith m, which calls for a
high degree of harmonization
among local markets. however, figure 2. The PCR users. (Data courtesy of PCR with our own elaboration.)
january/february 2019
the CaCm does not trigger this process from scratch: many
countries already had well-established local markets, each
of them having specific characteristics (e.g., the time for
opening and closing of the market session).
The need to converge to harmonized rules entails that some
or all parties must modify some of their existing characteris-
tics (such as the price limits imposed on the market, the length
of the market session, and settlement rules). This, inevitably,
creates conflicts among parties, and it is fair to acknowledge
the significant effort made by the nras and market partici-
pants to bring the pCr project to fruition. Figure 2 shows the
high level of harmonization already achieved in europe with
many markets using the pCr.
most of the eu member states are already coupled under
the so-called multi-regional Coupling (mrC) project, which
is based on the use of the pCr algorithm. only one mrC
member (Bulgaria) is not physically coupled to the rest of the
mrC area. at this stage, there is another separate regional
coupling mechanism called 4mmC using the pCr algorithm
that gathers four eu member states. other countries have
plans to implement the pCr algorithm and couple with their
neighbors in the future. Finally, two other countries already
use the pCr algorithm but are not coupled to any neighbor-
ing country (independent users).
Benefits
The benefits of integrating electricity wholesale markets
are multiple: efficient day-ahead capacity allocation, higher
liquidity, wider relevant market, consistent energy price for-
mation with the same algorithm (known as euphemia), and
reduced information asymmetry across regions. Furthermore,
Markets Coupled
MRC
Markets in MRC
Not Fully Coupled
Markets Using PRC
Algorithm 4MMC
Independent
Users
Future Markets
Foreseen for
Coupling
ieee power & energy magazine 23
 short-term efficiency (capacity is always implicitly allo-
cated to those who value it the most) supports long-term effi-
ciency (investment in additional capacity is based on reliable
scarcity signals).
it is difficult to provide a quantitative estimate of these
benefits in a wholesale electricity market serving 500 million
citizens. one of the most noticeable benefits is an improve-
ment in the use of interconnections. Figure 3 shows the usage
of network capacity in the “right direction,” i.e., when day-
ahead market prices dictate that it is economical to do so.
Coordination among member states clearly improved as they
joined the pCr initiative, which included the interconnec-
tions in 30 of 42 eu borders by the end of 2016.
price convergence is also a desirable outcome of the single
market. in this regard, it should be recalled that the eu Target
model is based on a zonal system; transmission congestion
between bidding zones is reflected in zonal price differences.
in principle, the bidding zones are defined so no structural
congestion is expected within them, but the optimal configu-
ration of bidding zones is a controversial issue and difficult
to implement in practice because it would also evolve with
system evolution and changes. on the one hand, some argue
that if bidding zones are too small, market liquidity may be
negatively affected. on the other hand, if bidding zones are
too large, internal congestion is not solved through market-
based capacity allocation mechanisms, and, consequently,
an efficiency loss occurs. price convergence should not be
achieved artificially, i.e., by merging bidding zones, but it
should be attained through adequate interconnection capac-
ity among zones. additionally, it is worth noting that, in many
cases, bidding zones coincide with member states; therefore,
political arguments play a nonnegligible role in the defini-
tion of the bidding zone configuration. Finally, from a physi-
cal perspective, bidding zones are conventional aggregates
100
90
Net Transfer Capacity (%)
39% 33%
80
70
60
Available
50
40
61% 67%
30
20
10
0
2010 2011
Capacity Used in the Right Direction
figure 3. The percentage of available transfer capacity used in the “right direction” under a >1€/MWh price differential
on 37 European electricity interconnectors. (Chart courtesy of the ACER/CEER, Annual Report on the Results of Monitor-
ing the Internal Electricity and Gas Markets in 2016, October 2017.)
24 ieee power & energy magazine
of nodes. Because of this approximation, commercial and
physical flows hardly coincide, which creates unscheduled
flows and loop flows.
Current Discussions Around Day-Ahead Markets
To be approved by nras, the market-clearing algorithm
must comply with a number of requirements set out by the
CaCm. however, such requirements might turn out not to
be jointly attainable.
Improvements to the Algorithm
according to the CaCm, “the algorithm aims at maximizing
the economic surplus,” which is also known as the total wel-
fare and is equal to the sum of the consumer surplus and
the producer surplus. in practice, the time to run the algorithm
is limited, and the solution does not necessarily achieve the
maximization of the economic surplus. This entails a wel-
fare loss, whose measurement is critical to assess which
investments should be made by the nemos to improve the
efficiency of the algorithm. For this reason, metrics (indica-
tors) are needed to assess the magnitude of the welfare loss
and investment in algorithm improvements are required.
Scalability
The pCr project, based on a voluntary agreement among
some european pXs, had a limited geographical scope. The
CaCm requests the nemos to use the existing solution (i.e.,
the pCr) as a basis for extending day-ahead market coupling
to all eu member states and even to go beyond the eu (nor-
way, switzerland, and some energy community contracting
parties). since the project was not initially designed for such
a broad scope, scalability problems may arise. although the
pCr initiative developed its initial steps with ease (because
few actors were involved initially), shifting from the pilot
15% 16% 14%
25% 23%
85% 84% 86%
75% 77%
2012 2013 2014 2015 2016
Capacity Unused in the Right Direction
january/february 2019
version to the commercial one proved to be more challenging.
however, the scalability issue is not limited to the geo-
graphical scope; the performance of the current version of
the algorithm can be seriously degraded, not only because of
the number of selling and buying orders but also when some
challenging combinations of complex orders are used in the
same bidding zone. in the future, many limitations should
be removed by algorithmic and technological improvements,
but this process will likely require numerous incremental
improvements.
Complex Orders
as more member states joined the pCr, the market-clear-
ing algorithm grew in complexity by incorporating new
order types from each nemo. This diverse set of order
types allows market agents to reflect different operating
and economic constraints, some of them depending on
market results, complicating the computation of market
prices. Furthermore, the algorithm must apply the uni-
form pricing principle within each bidding zone. This
approach is, among other reasons, aimed at improving
transparency: an accepted market order is always profit-
able (in the money) at given market prices, facilitating
the verification of market results. however, some orders
might be paradoxically rejected even if apparently in the
money. Because of the complexity of the algorithm, the
reasons of the rejection may not always be clear or trans-
parent to the market participants, hindering confidence in
the clearing process.
Intraday Markets: XBID Initiative
Status
The Cross-Border intraday market project (also known as
XBid) is an initiative of several nemos (epeX, Gme,
nord pool, and omie) to implement the eu Target model
for the intraday time frame, which is a continuous energy
market with an implicit continuous allocation of cross-
zonal transmission capacity. The XBid solution is based
on a shared order book, linking the local trading systems
operated by the nemos, and on a centralized capacity
management module, which gathers available cross-zonal
transmission capacities provided by the Tsos. Therefore,
orders entered by market participants in one bidding zone
can be matched with orders submitted by market partici-
pants in any other bidding zone within the information
technology (iT) systems’ reach, provided there is cross-zonal
capacity available. according to the CaCm, negotiation in
the continuous intraday market should be possible until (at
least) 1 h before delivery time.
after several years of project development, the XBid
platform became operational. The implementation plan
comprises local implementation projects, which means
that the operations of XBid will not start at the same time
in all member states. There will be a stepwise geographical
january/february 2019
extension based on readiness of the concerned nemos
and Tsos.
XBID Implementation Challenges
one of the main challenges of the XBid project has been the
cooperation among project parties (some of them being com-
peting entities) with different positions and interests. For that
reason, one key issue was to ensure that all parties would be
granted equal treatment by the trading system.
From the technical point of view, the main challenge was
to build a reliable and scalable trading system, capable of
processing a considerable number of transactions per sec-
ond. also, the local nemo and Tso systems must comply
with demanding requirements to successfully interact with
the central platform.
at present, there are some intraday continuous trad-
ing platforms at a regional scale, but these can hardly be
considered a relevant benchmark for the XBid platform,
which has a pan-european scope and is designed to com-
ply with the requirements of the CaCm. The liquidity in
the XBid and the amount of offers that may be submitted
to the system in peak times are difficult to predict. There-
fore, even though the project parties designed realistic
scenarios to test system performance, it will only be pos-
sible to assess to what extent it serves market needs after
XBid goes live. There are some pending future evolutions
and requirements, such as processing flow-based param-
eters, incorporating losses on a high-voltage direct current
interconnector(s) between bidding zones during capacity
allocation, minimizing the number of bidding zone bor-
ders on the path between the matched orders, allowing for
route prioritization using interconnector-specific cost coef-
ficients, and so on.
Current Discussions Around Intraday Markets
in the intraday time frame, the target of the CaCm is to have
pan-european continuous energy trading, with reliable pric-
ing of intraday cross-zonal capacity reflecting congestion.
Where there is sufficient liquidity, regional auctions may com-
plement the implicit continuous allocation mechanism.
The rationale for promoting a pan-european continu-
ous energy market is based on the understanding that the
main purpose of the intraday market is to enable market
participants to trade energy as close to real time as pos-
sible to balance their position. indeed, the bulk of the
energy trading is supposed to take place in longer-term
time frames and be incorporated by the Tso in its pro-
gram. in practice, in many member states, the volume
negotiated in the intraday time frame is rather low (around
one order of magnitude lower than the day-ahead nego-
tiation). however, it is expected that an increasing share
of renewable generation and the expected development of
demand flexibility will result in increased negotiation in
the intraday market. Furthermore, in the day-ahead market,
market participants implicitly pay for cross-border capacities
ieee power & energy magazine 25
 In day-ahead markets, the NEMOs match
demand and supply bids to set hourly energy
prices in each bidding zone.
whereas they currently use it for free in continuous trad-
ing. Therefore, shifting trading volumes from day ahead to
intraday might be economically convenient for some mar-
ket participants.
The continuous market provides a platform where any
market participant may adapt its position at any time, until
1 h before the time of delivery, if allowed by market condi-
tions, i.e., enough availability of buy/sell orders and avail-
able transmission capacity. on the other hand, intraday
auctions allow participants to adjust their positions only at
specific times, which is less agile but helps to pool liquid-
ity and, most importantly, provides a single clearing price
for all of the associated transactions. Therefore, intraday
auctions may be preferred by smaller market participants
who cannot operate in a continuous market as effectively as
larger players.
regarding the pricing of cross-zonal capacity in a con-
tinuous trading framework, pursuant to the CaCm, all the
Tsos proposed to organize at least an intraday auction near
the end of the day ahead of delivery that would allocate the
available capacities in an implicit way, originating conges-
tion rents. Besides, the CaCm also allows for complemen-
tary regional auctions.
There is a general problem with price formation mecha-
nisms in the target model: the day-ahead market is based on
implicit auctions (uniform clearing price for energy reflect-
ing cross-zonal congestion), whereas the intraday market is
organized in a dual manner. For example, in the continuous
trading mode, energy is priced pay as bid with the associ-
ated cross-zonal capacity allocated for free until conges-
tion is reached, whereas, in the intraday auction, the pricing
is equivalent to the day-ahead mechanism. This dual intraday
market may produce some inconsistencies in prices since
the bidding behavior of market players is adapted to each
allocation mechanism. This can create confusion and be
detrimental to the less-sophisticated market players.
Balancing Markets
Status
european balancing markets put the “latest” price on energy,
reflecting real-time imbalances and the corrective actions
of the Tsos to ensure the balance between generation and
demand at every moment. The Tsos organize and are respon-
sible for these markets that involve real-time operational
security and the deployment of reserve capacities. There-
fore, they require close coordination so the Tsos can update
26 ieee power & energy magazine
the cross-border capacities available for balancing, incorpo-
rating nemos’ nominations for intraday schedules. Besides
this, balancing gate-closure timing also relies on previous
intraday timing.
Balancing markets have not yet reached the harmoniza-
tion of day-ahead and intraday markets. They differ in sev-
eral important aspects that need to be addressed before the
full integration of balancing markets across europe can be
reached. First, the definition of the balancing products, i.e.,
the operating reserve types to be used, should be established.
Three categories have been defined:
✔ Replacement reserve (RR): manually activated bal-
ancing reserve characterized by the slowest activation
time (up to a 30-min activation time).
✔ Manual frequency restoration reserve (mFRR): equiv-
alent to tertiary reserve (less than a 15-min activation
time, mostly provided by spinning reserve).
✔ Automatic frequency restoration reserve (aFRR): equiv-
alent to secondary reserve (regulation reserve in unit-
ed states).
The definition of these balancing products is currently
under harmonization. modifying the definition of balancing
products requires significant changes to the Tso operating
procedures and their corresponding iT infrastructure. mar-
ket participants also need to adapt, mainly balancing service
providers (Bsps) and balance responsible parties, as defined
in the GleB. For this purpose, the Tsos are closely cooper-
ating with european market stakeholders to define technical
rules and products.
settlement rules for balancing products also need to be
harmonized. Currently, several countries apply marginal pric-
ing while others apply pay as bid, or regulated prices. Figure 4
illustrates the different settlement systems for mFrr energy
in europe.
Benefits
an integrated european balancing market should unlock
synergies between Tsos, such as the netting of balancing
needs of different signs, upward and downward. netting
cross-border imbalances would bring important benefits by
reducing the amount of balancing energy activated, which
would not only reduce balancing costs but also imply higher
availability of reserves, which could further reduce operat-
ing reserve needs.
higher Tso coordination would facilitate renewable
production integration, and similarly to the day-ahead and
intraday markets, a wider balancing market is expected to
january/february 2019
 In the intraday time frame, the NEMOs are
working to set a European continuous
implicit trading platform.

increase liquidity and improve competition. The implementa-
tion of cross-border balancing energy exchanges will also
help to optimize the use of interconnection capacities and
reinforce operational security in europe. Therefore, building
the european integrated balancing market is fully aligned
with the achievement of european energy policy objectives.
a cost-benefit analysis for 2013 estimated that the net ben-
efits that would be gained if six neighboring european coun-
tries [involved in the Trans-european replacement reserves
exchange (Terre) project described later] cooperated
in the activation of replacement reserves would amount
to €150 million per year. in addition, it was shown that those
benefits would increase significantly if some of the cross-
border interconnection capacities were upgraded.
on the other hand, the described harmonization effort
among european balancing markets is focused on the acti-
vation of balancing energy. Currently, balancing capacity is
mainly procured by european Tsos in local reserve markets of
different types. These markets present even more differences
from one member state to another. some Tsos organize daily
auctions, whereas, in other cases, the Tsos rely on weekly,
monthly, or even longer commitments for reserve procurement.
Future Balancing Markets
The GleB regulation became effective in december 2017,
and its main goal is to create european balancing platforms at
different time scales, while harmonizing different aspects of
local balancing markets. Besides this, it promotes the pro-
vision of balancing services from renewable units, flexible
demand and energy storage units under the same conditions as
conventional generation units, and the application of marginal
pricing (rather than pay as bid) for the settlement of balancing
energy. The GleB regulation also aims at an absence of price
caps and floors.

Frequency Restoration Reserve (Manual)–Energy–Settlement Rule

Definition of Question
Settlement Rule The pricing rules for settlement.

Definition of Answer
Hybrid Combination
Marginal Pricing Marginal pricing is the change in total cost arises
when the quantity produced changes by one unit.
Pay as Bid Contracted parties who provides a service are
paid based on their offer price.

Regulated Price The price for this service is based on a price that
is set by the relevant regulatory authority.

Key
Missing Data
N/A
Pay as Bid
Marginal Pricing
Regulated Price
Hybrid

figure 4. The different settlement rules of the mFRR energy in Europe. “Missing data” refers to the TSOs that did not re-
spond the questionnaire, and “N/A” refers to the TSOs that did not answer to this specific question. (Image courtesy of the
ENTSO-E, Survey on ancillary services procurement, balancing market design, March 2017.)

january/february 2019 ieee power & energy magazine 27

 The CACM provides the legal definition,
designation process,
and roles for the NEMOs.
another relevant point is the harmonization of gate-
closure times for Bsps to submit bids to the corresponding
balancing platforms, for which the prior harmonization of
intraday gate-closure times (as described above) is crucial. in
addition, a harmonization of imbalance settlement periods to
15 min is encouraged before January 2025. The GleB also
encourages the financial settlement of unintended deviations,
in contrast with the current scheme in continental europe
based on the pay-in-kind approach (which nets energy deliv-
ery at equivalent peak/off-peak periods). Finally, the GleB
considers the possibility of cross-border capacity reservation
for balancing purposes when justified through a previous
cost-benefit analysis.
Implementation Steps
The gradual implementation of the GleB would be facilitated
by regional cooperation through the development of different
platforms to allow information exchange among the Tsos
and market participants. each balancing process will require
a separate platform, and each platform will involve a different
set of the Tsos and entail different technical complexities,
so each will face a different deadline, as shown in Figure 5.
in Figure 5, two different deadlines are set for each platform.
The first deadline establishes when the platform should be
implemented according the GleB requirements. The second
deadline is when all concerned Tsos should be integrated in
the platform. Both deadlines are the same for the rr activa-
tion platform and the imbalance netting (in) platform.
RR Activation and Platform
The Terre project has been launched and was designated by
enTso-e (european network of Transmission system oper-
ators for electricity) as a reference to implement the future rr
Platforms 2017 2018
Replacement Reserve
Manual Frequency Restoration Reserve
Automatic Frequency Restoration Reserve
Imbalance Netting
GLEB Entry into Force
Deadline for Implementation of the Platform According to the GLEB
Deadline for Integration of all Concerned TSOs
figure 5. The road map for the implementation of European/regional balancing energy activation platforms.
28 ieee power & energy magazine
platform. The iT platform liBra is composed of different
modules: 1) an optimization algorithm for a common balanc-
ing market merit order and 2) a settlement and capacity man-
agement module. The Tsos from France, italy, portugal,
spain, switzerland, and Great Britain are actively participat-
ing. in the future, additional countries may be incorporated
into the rr platform, including romania, Czech repub-
lic, poland, hungary, and Bulgaria. The energy exchanged
in this platform will make use of the remaining cross-border
capacity after the last intraday market.
The mFRR Activation and Platform
unlike the rr activation that is only mandatory for coun-
tries using the rr, the mFrr activation, under the GleB
regulation, is mandatory for all european interconnected
Tsos. The current reference project for the future mFrr
activation platform is the manually activated reserves ini-
tiative. This project is currently in a design phase involv-
ing 19 european Tsos, and six Tsos are in the process of
becoming observers.
The aFRR Process and Platform
The aFrr activation will be voluntary for european inter-
connected Tsos. For instance, Great Britain and ireland do
not currently use the aFrr. The reference project is called
the platform for the international Coordination of the auto-
matic Frequency restoration process and stable system
operation. This project is under design, and eight european
Tsos are full members of this initiative, including Tsos
from Belgium, austria, The netherlands, and France as well
as the four Tsos from Germany. The remaining european
interconnected Tsos using the aFrr are currently in the
process of joining this initiative.
2019 2020 2021 2022 2023 2024
january/february 2019
 In the future, intraday cross-zonal capacities will be
priced reflecting congestion.
IN Platform
The goal of the in platform is to minimize the activation of the
aFrr energy by counteracting the systems’ aFrr real-time
imbalance requirements of opposite sign (upward and down-
ward). The in platform, called the international Grid Coop-
eration Control (iGCC), is already in a go-live stage, and is
operated by the German Tso Transnet BW. For nordic Tsos,
joining the platform is optional, since they belong to a differ-
ent synchronous area. The united Kingdom and ireland do not
currently use the aFrr, so they cannot be involved either.
The in platform is envisaged, in principle, as a tempo-
rary process, since the aFrr platform, once established,
will also perform implicitly imbalance netting as part of the
economic optimization. although the in platform does not
use economic information, it minimizes the activation of
the aFrr by taking advantage of synergies among systems
with opposing aFrr needs subject to available cross-border
capacities and additional exchange security limits. This plat-
form has been proven useful because reducing upward and
downward aFrr activations, through the netting process,
can significantly reduce balancing costs, since aFrr energy
is, by nature, the most expensive balancing energy.
Additional Features of GLEB
The GleB road map will face the following inherent hetero-
geneity aspects of the interconnected european systems that
will likely remain in the future.
Proactive Versus Reactive Balancing
proactive balancing actions take advantage of forecast-
ing tools, for demand and the production of wind and solar
resources, to activate slow balancing products (rr) that com-
pensate for forecast deviations in advance. The aim is to solve
imbalances as soon as they are identified, therefore using bal-
ancing energy that is expected to be less costly. The accuracy
of forecasting tools relies on the real-time observability of the
system. Therefore, Tsos using proactive balancing actions
make use of all balancing products (rr, mFrr, and aFrr)
in sequence. This approach is typically adopted in systems
with a high level of system observability (high accuracy of
forecasts) and where intraday markets’ gate closure takes
place up to 1 h before real time. This is the case of Tsos in
France, spain, portugal, the united Kingdom, italy, poland,
and hungary, among others. in these contexts, proactive use
of manual reserves avoids making excessive use of more
expensive balancing energies (mFrr and aFrr).
reactive balancing actions wait until deviations are
almost detected very close to real time. There is a lack of
january/february 2019
observability, and, therefore, forecasting imbalances is more
difficult. in this case, aFrr, sometimes complemented with
mFrr for resetting the usage of aFrr, is the main resource
used by the Tsos to counteract imbalances. reactive bal-
ancing is used in markets with intraday gate closures very
close to real time, e.g., 30 min before delivery in the German
case, and also a high level of interconnection capacity with
the rest of the neighboring systems.
Central Versus Self-Dispatch
in centrally dispatched systems, such as those in poland and
italy, Tsos assume the responsibility of physically rescheduling
resources immediately after the day-ahead market gate closure,
so intraday markets only play a financial role at centrally dis-
patched systems. alternatively, self-dispatched systems (most
european systems) still allow market actors to change their
physical schedules until the gate closure of intraday markets.
Future View
Clean Energy for All Europeans
The eu energy policy sets decarbonization commitments for
member states for 2030 and beyond. different instruments,
such as renewable and energy-efficient mandatory targets and
carbon dioxide emission pricing, are promoted and imple-
mented at the national or european level. in particular, the
electricity sector is called to play a central role in this transi-
tion. The increasing shares of renewable electricity genera-
tion (mainly from wind and solar), electrification of vehicles
and buildings, and adoption of distributed energy resources
by electricity prosumers create new challenges for the design
and operation of the integrated european electricity market.
in this context, in november 2016, the eC presented the
Clean energy package, a set of legislative proposals for nearly
every energy-related sector. These proposals allow glimpses
of the upcoming challenges in the implementation of the elec-
tricity market.
regarding renewable energy integration, the challenges are
not exactly new. as reviewed in this article, integrating all
european power markets requires major market reforms in the
directions signaled by the CaCm and GleB regulations. Well-
functioning and coordinated intraday and balancing markets are
an increasingly important way to achieving the much-needed
flexibility necessary to complement intermittent and variable
wind and solar generation. To ensure the best possible outcome,
these markets should be open to the participation not only of
conventional generators but also of new flexible resources, such
as demand response and storage. at the same time, all market
ieee power & energy magazine 29
30
Capacity Auction Strategic Reserve
(Since 2014, First (Since 2007)
Delivery in 2017/2018)

Strategic Reserve
Capacity Payments (Since 2004), Gradual
(Since 2007) Considering Phase-Out Postponed
Reliability Options to 2025

ieee power & energy magazine

Capacity Requirements Network Reserves
(Certification Started
1 April 2015,
First Delivery in 2017) Strategic Reserves

Capacity Payments Strategic Reserves

(Since 2008, Part of Them (Envisaged End 2018)
Do Not Apply to New
Capacity Installed as of
1 January 2016)–Tendering Strategic Reserve
for Capacity Considered (from 2016 on, Extended
But No Plans Until the End of 2019)

Capacity Payments Strategic Reserve

(Since 2010 Partially (Since 1 November 2014)
Suspended
Between May 2011
and December 2014)
Tender
(Since November 2013)
Reliability Options (The
Date for the First Auction
Has Not Been Set. First (Capacity Payments
Delivery of Contracted Existed from 2006 to 2014)
Capacity Is Expected New Capacity Payments
in 2020. Ongoing from 1 May 2016 to 30
Discussions with the EC) April 2017 Approved by
the EC

No CM (Energy-Only Market) CM Proposed/Under Consideration CM Operational

january/february 2019
figure 6. The CMs in Europe. The main changes compared with 2015 are highlighted in red. (Image courtesy of the ACER/CEER, Annual Report on the Results of Moni-
toring the Internal Electricity and Gas Markets in 2016, October 2017.)
participants must bear the same responsibilities, including balanc-
ing production schedules (from which renewable generation
has traditionally been exempt, as a way of promotion).
harmonized market rules across europe enable the real-
ization of synergies between energy resources in different
member states, guaranteeing an efficient operation of the
power system. however, an integrated market also requires a
higher level of cross-border interconnections among member
states, and the eC has identified increasing interconnection
capacity as a strategic objective. Building consensus around
the development of new transmission projects, and especially
about the cost allocation of such projects, is an ongoing pro-
cess led by the Tsos and regulators.
New Resources, Complexity, and Scalability
The Clean energy package envisions a power market in which
consumers can engage in demand response, self-consump-
tion, or self-generation and participate in markets. distributed
energy resources, such as decentralized generation, storage, and
demand response would be allowed to participate in wholesale
electricity markets directly or through aggregators meeting the
required technical capabilities, in a level playing field with con-
ventional generators. That would involve changes not only in the
design of current retail and wholesale markets but also in the
interactions and coordination between the Tsos and the distri-
bution system operators.
in general, the presence of distributed energy resources
will increase the number of market participants, challenging
the implementation of a power market that is already under
great strain due to its computational complexities. Future
developments in day-ahead and intraday markets will be
characterized by the unavoidable conflict between the need
to simplify implementation due to technical constraints and
to allow for the more complex participation models required
by new resources such as demand flexibility or storage.
Scope of Markets
The Clean energy package makes a clear statement in favor
of markets as the best tool to enable efficiency and transpar-
ency for participants. Well-functioning “energy-only” markets,
as presented in this article, have been considered the most
efficient design for an integrated european electricity market.
however, this paradigm has been, and is frequently, challenged
by adequacy concerns that lead to different types of Cms.
Figure 6 shows that a significant number of member states have
already implemented or plan the implementation of different
national Cms.
The eC has provided guidelines for member states to jus-
tify the need for Cms and, if that is the case, to implement
Cm designs that minimize distortions in the functioning of
short-term markets and support neighboring areas in case of
adequacy problems. in addition, it recommends that mem-
ber states identify the market flaws that may constitute an
obstacle to the provision of adequacy and only if these flaws
cannot be removed to resort to a Cm. among those obstacles
january/february 2019
are caps on wholesale market prices, retail price regulation,
support schemes for fossil or nuclear generators that do not
incentivize a flexible operation, and the previously discussed
poor designs of intraday and balancing markets.
For Further Reading
european Commission. (2016, nov.). proposal for a directive
of the european parliament and of the Council on common
rules for the internal market in electricity november 2016.
europa. [online] available: http://ec.europa.eu/
Commission regulation (eu) 2015/1222 of 24 July 2015
establishing a guideline on capacity allocation and con-
gestion management. entsoe. [online]. available: https://
electricity.network-codes.eu/network_codes/cacm/
Commission regulation (eu) 2017/2195 of 23 november
2017 establishing a guideline on electricity balancing. entsoe.
[online]. available: https://electricity.network-codes.eu/
network_codes/eb/
nemo Committee. (2017, July 24). all nemos proposal
for the mCo plan approved by all nras of the eu on 26 June
2017. europex. [online]. available: http://www.europex.org/
all-nemos/all-nemos-mco-plan/
agency for the Cooperation of energy regulators. (2017,
oct. 6). annual report on the results of monitoring the in-
ternal electricity and gas markets in 2016. aCer. ljublja-
na, slovenia. [online]. available: https://acer.europa.eu/
official_documents/publications/aCer%20market%20
monitoring%20report%202016%20%20document%20
histo/aCer%20market%20monitoring%20repor t%
202016%20-%20eleCTriCiTY%20-%20original.pdf
i. J. perez-arriaga, C. Knittel, a. Bharatkumar, m. Birk, s.
Burger, J. p. Chaves, p. duenas-martinez, i. herrero, s. hun-
tington, J. Jenkins, m. luke, r. miller, p. rodilla, r. Tabors,
K. Tapia-ahumada, C. vergara, and n. Xu. (2016, dec.). utility
of the future: an miT energy initiative response to an industry
in transition 2016. miT energy initiative. [online]. available:
http://energy.mit.edu/research/utility-future-study/
Biographies
Tomás Gómez is with the universidad pontificia Comillas,
spain.
Ignacio Herrero is with the universidad pontificia
Comillas, spain.
Pablo Rodilla is with the universidad pontificia Comillas,
spain.
Rodrigo Escobar is with the operador del mercado
ibérico, spain.
Salvatore Lanza is with the autorità per l´energia, italy.
Ignacio de la Fuente is with the red eléctrica de españa,
spain.
Maria Luisa Llorens is with the red eléctrica de
españa, spain.
Paula Junco is with the red eléctrica de españa, spain.
p&e
ieee power & energy magazine 31
 Electricity
Markets in the
United States

F
Following order 888 From the Federal regulatory energy
Commission (FerC) in 1996, the u.S. power industry began its restructuring pro-
cess. this undertaking involved a transition from a vertically integrated utility
structure where generation, transmission, and distribution are combined to serve
consumers. through advancing to a market environment, generation companies
(genCos) can compete with each other to provide energy through open transmis-
sion and distribution systems managed by transmission (transCos) and distribution
companies (disCos). the goals of the restructuring are the efficient produc-
tion of electricity and efficient generation investment through competition. under the

Digital Object Identifier 10.1109/MPE.2018.2872300

Date of publication: 7 January 2019

32 ieee power & energy magazine 1540-7977/19©2019IEEE january/february 2019

 ©istockphoto.com/vector, store window items—image licensed by ingram publishing

restructured environment shown in Figure 1, wholesale electricity markets are managed by independent system
operators (iSos) or regional transmission organizations (rtos), whose responsibility is reliable system opera-
tion, market administration, and system planning. genCos, large consumers, load aggregators, marketers, and
load serving entities (lSes) buy and sell electricity through FerC-regulated wholesale electricity markets.
Small consumers like individual households, some large consumers, lSes, marketers, and load aggregators
can transact energy either at a regulated retail rate overseen by the public utility commission (PuC) or through
retail competition. the restructuring process has been complex, especially in trying to harmonize the whole-
sale and retail sides, and involves many stakeholders from both public and private sectors (see Figure 1).
u.S. electricity markets have been considered largely successful,
and the rto footprint continues to grow. today there are seven rtos
Power Industry operating in the united States (see Figure 2), bringing substantial
benefits to end users. the benefits achieved by iSo new england’s
Restructuring (iSo-ne’s) wholesale electricity markets are summarized in table 1,
as one example.
Processes for the wholesale electricity markets evolved with changes in regulatory
policy, technological innovations, and economic conditions since the
Present and Future start of electric restructuring. during the past two decades, the u.S.
power industry has experienced a transformation with the retirement
of coal, oil, and nuclear resources; increased electricity production
and investment in natural gas facilities due to environmental and sit-
ing factors as well as relatively low gas prices; increased penetration
of renewable resources as a result of state environmental policies; and an increase in demand resources and
distributed generation technologies. these changes have created challenges to power system operation and
planning and electricity market design. we will review the main features of existing wholesale electricity
markets and several key components. as pricing is particularly important for an effective market design,
we will discuss ongoing pricing issues. Finally, we will take a brief look at the evolution of the electric grid
and future challenges and then identify and discuss several key market design initiatives stemming from
these challenges.

By Eugene Litvinov, Feng Zhao, and Tongxin Zheng

january/february 2019 ieee power & energy magazine 33

 coupled with the inefficiency of the uplift payment-based con-
Small gestion management model, led the wholesale electricity mar-
GenCo LSE Aggregator
Customer kets to gradually converge to a pool-based model similar to
those of PJm and nyiSo. today this model involves a set of
ISO Retail common features, including the two-settlement system [day-
TransCo
Wholesale DisCo Regulation/ ahead market (dam) and real-time market (rtm)], lmP, co-
Market Competition optimization of energy and ancillary services, zonal reserve
products, financial transmission rights, and a forward capac-
Large Large ity market (FCm).
Marketer
Customer Customer
the goal of wholesale electricity markets is to increase the
power system’s economic efficiency without compromising
FERC PUC
its reliability. therefore, market products traded in most u.S.
electricity markets are designed according to the needs of bulk
figure 1. The deregulated power industry in the United States. power system operation and planning. as a result, electricity
markets always interact with the physical power system plan-
ning, scheduling, and operation. Figure 3 presents markets with
Main Features of different time frames and their corresponding physical systems.
U.S. Wholesale Electricity Markets we will discuss the main market products and major features
wholesale electricity markets emerged in the united States associated with each of the markets in more detail.
in the late 1990s with different models. PJm and new york
iSo (nyiSo) started locational marginal pricing (lmP)- Energy
based energy markets with a two-settlement system in 1998 energy is the main product traded in u.S. electricity markets,
and 1999, respectively. California iSo (CaiSo) adopted a which typically include the dam and the rtm. although
day-ahead power exchange with the real-time energy market in market designs vary in different iSo regions, all have struc-
1998. iSo-ne opened its real-time energy market with a single tured the two markets in a two-settlement system with the for-
market clearing price and an uplift payment-based congestion ward dam and the spot rtm. all markets also have a similar
management model in 1999. after the California energy crisis objective of maximizing the total social surplus, and lmP is
in the early 2000s, the power exchange model was considered used to manage the transmission congestion.
to be prone to market power abuse and gradually abandoned electricity producers and consumers may be located at
in the united States. FerC’s standard market design efforts, different areas but are connected through the grid. delivering

ISO New
England
Midcontient ISO
NYISO

PJM Interconnection

Southwest
Power Pool
CAISO

Electric Reliability
Council of Texas

figure 2. The U.S. ISOs/RTOs (Online: https://www.ferc.gov/industries/electric/indus-act/rto.asp).

34 ieee power & energy magazine january/february 2019

energy from a generator to a consumer at a different loca-
tion not only incurs energy losses but also may overload
the transmission lines. this, in turn, prevents low-cost
generation from being delivered to the customer, a phe-
nomenon often called congestion. lmP recognizes the
locational differences in energy production and consump-
tion and has been the main approach to pricing electric-
ity and managing congestion in the transmission system.
lmP reflects the cost of serving an increment of load at
a specific location in the transmission system. electric-
ity producers and consumers are paid or charged by the
corresponding lmP at their locations. as a type of uni-
form pricing, lmP provides strong incentives for truth-
ful bidding in energy markets and long-term investment
signals for building new generators at the locations where
they are needed (indicated by higher lmPs), including
the effect of marginal losses. mathematically, lmP can
be decomposed into energy, congestion, and loss com-
ponents, where the congestion component in the dam is
used to settle Ftrs.
lmP calculations solve an economic dispatch (ed) prob-
lem. due to the nonconvex features such as unit commitment
(uC) decisions that are not modeled in the ed, out-of-mar-
ket side payments to the generators are often used by iSos
in addition to the uniform lmP-based market payment. the
total payments then allow generators to recover their start-up
and no-load costs. the out-of-market payments distort the
market and are not transparent, therefore creating problems
in market participants’ risk hedging. reducing these out-of-
market payments has been a difficult problem that will be
discussed in later sections.
table 1. ISO-NE wholesale market benefits.
Generation Increased from 75% in 1997 to
Availability 88% in 2016
Emissions reductions 29% lower carbon dioxide emissions
between 2001 and 73% lower nitrogen oxide emissions
2016 98% reduction in sulfur dioxide
emissions
Wholesale energy 2016 average annual energy market
prices prices were lowest since 2003
Private investment More than US$8 billion
New resources More than 15 GW of new generation
since 1999 and about 2.7 GW of new
demand
Because iSo systems are connected by interties to facili-
tate the power exchange among neighboring iSos, specific
locations are used for pricing of imports or exports of energy.
Currently, most iSos assign a proxy bus for pricing energy
imports and exports. the selection of these proxy buses is
often arbitrary and may not conform to the actual power
flow. as a result, the interchange schedule between iSos is
often inefficient.
the dam is often considered a forward financial market
where participants submit bids to buy or offers to sell elec-
tricity at each market location. the bids and offers can be
physical and associated with generating units or demands
or virtual and not associated with any physical assets. Bids
and offers come in the form of price–quantity pairs in a
descending or increasing price order at any market location.
in addition, each physical asset submits its physical operating

Real-Time Operation Short-Term Planning Midterm Planning Long-Term Planning

System Monitoring and Alarming Outage Coordination Generator Maintenance Resource Adequacy
Automatic Generation Control Load Forecasting Scheduling Transmission Planning
Load and Wind Forecasting Reliability Assessment Transmission Outage Strategic Planning
Security Analysis Operations Planning Scheduling
Economic Dispatch
Emergency Operation
Look-Ahead Unit Commitment Physical System Operation and Planning
Transaction Scheduling
Intraday Reliability Assessment

Real-Time Day-Ahead Forward Reserve Market Forward

Market Energy Market FTR Market Capacity Market

Market Operation

figure 3. The markets and systems. FTR: financial transmission rights.

january/february 2019 ieee power & energy magazine 35

 parameters, such as economic minimum and maximum,
ramp rate, start-up and no-load costs, and minimum up and
down times. the market clears at the point where supply
meets demand at a maximal social surplus subject to net-
work security constraints. the market clearing problem is
called the security-constrained UC (SCUC) problem. dam
results are financially binding, which leaves only the real-
time deviations from the day-ahead schedules to be exposed
to the real-time price (two-settlement system). as a result, the
iSo’s dam design provides a way for market participants to
hedge against rtm price volatility or arbitrage between the
two markets.
the dam closes at a particular time of the day prior to
the delivery date. market participants submit their bids into the
market before the closing time. once the market closes, the
market administrator runs software solving a 24-h (or longer)
SCuC problem. the problem is typically a mixed integer lin-
ear program (milP) because it involves both hourly binary
on/off decisions for physical generating units and continuous
unit output level decisions. dam clearing prices, or locational
market prices, are determined from the constraint shadow
prices calculated by the ed process for each hour.
the rtm is a spot market where electricity is physically
delivered to the consumers in a least-cost and reliable fash-
ion. it is closely coupled with the power system operations
by including all of the system and security constraints into
the market clearing problem, i.e., security-constrained ed
(SCed). different from the dam, the rtm usually does
not require demand bids due to loads’ lack of response to
rtm prices.
in practice, after the dam is cleared and prior to the
operating day, iSos often run a 24-h or longer reliability
uC (ruC) process committing additional units that have not
been selected in the dam to meet the forecasted load (which
may exceed the dam cleared as-bid load). the reliability
commitment problem is similar to the dam uC problem,
(US$/MWh)
rve
Cu
er
n Off
r atio
G ene B (i.e., areas a and B) and the lmP payment received by the
LMP
C
A A + B 2 A + C), the generator has a revenue shortage of
(MW)
Generator Output
Generator Cost = A + B
LMP Payment = A + C
Uplift Payment = B – C
Uplift = max {0, Cost-Payment}
figure 4. An illustration of uplift payment.
36 ieee power & energy magazine
except that only physical units are considered and no market
clearing prices are produced. the ruC process may be run
again throughout the operating day to look ahead a few hours
and determine possible needs for additional resources.
with the committed units from the dam and ruC, iSos
run the rtm, which determines the generation schedule and
rtm clearing prices. market participants are allowed to sub-
mit rtm bids that differ from their dam bids. iSos solve an
SCed problem that minimizes the total dispatch cost subject
to network and individual unit constraints. different iSos
may model SCed differently. iSo-ne implements a sin-
gle-interval dispatch for its rtm, whereas nyiSo solves a
multi-interval dispatch problem. Special pricing rules for the
fast-start resources are often considered in the rtm to reflect
their deployment costs in addition to the incremental costs in
the lmP calculation. these rules often differ among rtos
but share the similar principle of eliminating the nonconvex-
ity of the resources by relaxing certain characteristics, such
as relaxing the minimum output level to be zero or amortiz-
ing the commitment costs into the dispatch costs.
energy trading in the iSo-managed wholesale markets
often creates two financial positions for each market partici-
pant: dam and rtm (cleared transaction quantity). a two-
settlement system is adopted to settle each financial position.
the dam settles on the dam position with dam lmPs, and
the rtm settles on the energy deviations from the cleared
dam position with rtm lmPs. under such design, physi-
cal players, such as generators and loads, are able to hedge
their risk by locking in the price of their transactions with
dam locational market prices. Financial players who place
virtual transactions in the dam without delivering in rtm
can arbitrage the price difference between dam and rtm
for their cleared dam quantity. the two-settlement system
has worked well in u.S. electricity markets, closing the gap
between dam and rtm prices.
although lmP reflects the marginal cost of producing
an increment of energy, it does not capture the entire cost
of generation, such as the start-up cost of an online unit. as
a result, the generation as-bid cost of a resource may not be
fully recovered by the lmP payment. an illustration of cost
and payment is shown in Figure 4, where the as-bid cost is
represented by the area under the generation offer curve
generator is lmP multiplied by the generation output (i.e.,
areas a and C). when the cost exceeds the payment (i.e.,
B-C. Compensation for such revenue shortage, often termed
a make-whole payment or an uplift payment, is needed and
has been widely used in existing iSo markets to incentivize
generators to follow iSo dispatch instructions.
uplift payments are caused by the nonconvexity of the bids,
e.g., the integer commitment costs and the presence of mini-
mum generation level. they are unit specific as opposed to the
uniform market clearing price lmP and are often opaque to
the market participants as being very difficult to hedge. iSos
january/february 2019
have been striving to reduce such out-of-market payments, but,
due to generators’ nonconvex physical characteristics, they are
unlikely to be eliminated. an alternative approach is to force
one-part bidding and thereby eliminate the physical param-
eters that cause nonconvexity in their bids. this effectively
removes the uplift payment at the cost of increased risk for
participants that have to embed the nonconvexity-related costs
into their convex bids, and it presents increased difficulty for
market power monitoring. Currently, all u.S. markets have
adopted the three-part bidding (i.e., incremental cost, start-
up, and no-load cost) and uplift payments approach instead of
the convex bidding seen in some other countries.
Ancillary Services
in addition to meeting the system load, power system operators
have to prepare for generation contingencies and load fluctua-
tions in real time. these additional reliability needs can be
satisfied by ancillary service products besides energy, e.g., reg-
ulation, reserves, and ramping. resources with these additional
capabilities are compensated through the ancillary service
markets. ancillary service products could be quite different
among iSos because the iSo systems may have different reli-
ability needs. Consequently, the ancillary service markets are
diverse. CaiSo, e.g., considers a 5-min ramping product, mid-
continent iSo projects a 10-min ramping product, and iSo-ne
does not currently have a ramping product. with the increasing
penetration of renewables, new system features arise and may
create a need for new ancillary services.
Because energy and ancillary services share the capacity
of generation units and transmission network, they are often
co-optimized to achieve the most efficient capacity alloca-
tion of resources, leading to an SCed optimization problem
that simultaneously determines the energy and ancillary ser-
vices designations for resources. Similar to energy lmPs,
the clearing prices for ancillary services are obtained from
the dual variables of the SCed problem.
Financial Transmission Rights
an Ftr is a financial instrument for hedging against trans-
mission congestion; it has a source and a sink location as well
as a megawatt quantity and entitles its holder to collect con-
gestion revenue based on the difference of lmP congestion
components between the sink and source locations. Ftrs
have been implemented in all u.S. iSos. existing Ftrs
are point-to-point annual or multiyear products and can be
obtained through corresponding Ftr auctions or allocated to
market participants based on grandfathered rights.
iSos sell the Ftrs and collect the auction revenue in the
annual Ftr auction. depending on the region, the Ftr auction
revenue is allocated to the lSes, transmission owners, or other
entities in accordance with the market rules. Ftrs are usually
settled on the dam prices. due to the discrepancy between
the transmission model used in the Ftr and the model used
in the dam, the congestion revenue in the dam and rtm
may not cover 100% of the entitled payments to Ftr holders,
january/february 2019
i.e., Ftr underfunding. to remedy the problem, iSos often
auction a reduced transmission capacity in the Ftr auctions.
Forward Capacity Product
generation investment is a key step to secure the long-term
supply of electricity to meet demand growth in the future.
in some regions, responsibility for resource adequacy has
remained with vertically integrated utilities. in other regions,
such as iSo-ne and PJm, where utilities divested their gen-
eration assets, resource adequacy is generally met through the
wholesale markets. due to the presence of price caps in the
energy markets and the manual actions of system operators
that may depress the energy price, revenues from the energy
market alone may not be sufficient to cover a unit’s capital
investment cost. this phenomenon is termed the missing
money problem. Facing the revenue shortage risk, generating
units may not be available when needed, potentially causing
a shortage of capacity and, therefore, reliability issues. one
solution to the problem that has been implemented in iSo-
ne is the FCm, where units’ generating capacity is obtained
in advance and compensated through the capacity product
and the corresponding market clearing price. an important
enhancement to the FCm, the pay-for-performance mecha-
nism, reduces revenues for resources that do not perform and
rewards resources that perform during critical system events.
the FCm provides a steady revenue stream in advance for
existing resources to recover the missing money and for new
resources to reduce the investment risk.
as a forward market, FCm auctions run several years
to several weeks ahead prior to the real-time operation of the
capacity commitment period, depending on different market
designs. Qualified resources, existing or new, are required to
submit their capacity offers to the iSo, which, in turn, clears the
FCm auction amounts and determines the clearing price. local
capacity zones may be modeled in the FCm to reflect local reli-
ability needs. Capacity requirements are administratively set
by the iSo through reliability simulations that are similar to
other loss of load expectation resource adequacy studies.
with fixed capacity requirements, the FCm auction clear-
ing prices may be volatile, i.e., the market clearing price
tends to substantially increase when there is any capacity
shortage from the fixed requirement and plummet when there
is a capacity surplus. Such price volatility increases the risk
for capacity investment and load payment and impedes the
formation of adequate price signals for long-term investment.
Furthermore, such a market may be prone to market power.
to remedy this issue, iSo-ne has replaced the fixed require-
ments with demand curves based on the tradeoff between
reliability and cost of capacity.
Interchange Scheduling
and Market Coordination
the evolution of the power grid has led to today’s large inter-
connected power systems, each operated by an iSo. an iSo
has complete information of its own control area but may not
ieee power & energy magazine 37
 have information from other iSos, such as network model,
generation offers, and system condition. Furthermore, iSos
may run electricity markets with different designs. as a
result, the dispatch solutions of individual areas may not be
optimal for the entire interconnected system, i.e., economic
efficiency could be impaired. Such a problem caused by the
heterogeneous market designs and insufficient information
sharing is termed the seams issue. to achieve higher eco-
nomic efficiency, coordination among neighboring iSos is
needed. the purpose of such coordination is to drive each
iSo’s area dispatch solution toward the optimal solution of
the entire region’s dispatch problem.
the coordination between two markets can be performed
through interchange transaction scheduling and generation
redispatch. there are two market-based interchange schedul-
ing approaches. one is scheduled by market participants by
submitting bids to buy from one market and offers to sell to
another market. Both buy bids and sell offers must be cleared
in both markets to schedule the physical interchange flow.
this scheduling process is generally conducted on an hourly
basis. another is the coordinated transaction scheduling (CtS)
implemented by iSo-ne, PJm, and nyiSo. in this approach,
market participants submit CtS bids that represent their will-
ingness to buy or sell energy based on the price difference
between two markets. the CtS market commonly runs on
15-min intervals to increase the scheduling efficiency. in addi-
tion to interchange scheduling, a market-to-market coordina-
tion scheme is also available for coordinated congestion man-
agement. under this scheme, two markets exchange the cost of
redispatch in their markets and relieve the congestion through
dispatching more efficient generation.
Market Management System Components
the market management system (mmS) components are
central to the operation of wholesale electricity markets in
iSo control centers. the aforementioned market features
are all implemented in this system. different from financial
market operation, electricity market operation, especially
real time, requires constant interaction with the underlying
physical power system, making the mmS an essential part of
system control. a typical mmS is shown in Figure 5, where it
interacts with the physical control system, such as the energy
management system (emS), and the financial settlement
system. the physical system provides essential operating
parameters, such as load forecast, system topology, and trans-
fer limits to mmS, which determines generation schedules
(a type of control signal) for the emS and market clearing
prices for the financial settlement system. mmS is complex
but made up of several basic building blocks, e.g., uC, ed,
security analysis, and pricing.
Architecture
typical electricity market systems produce generation sched-
ules and clearing prices by solving an optimization problem.
the common components in the optimization problem are
38 ieee power & energy magazine
uC, ed, security analysis, and pricing. the uC determines
the integer on/off decisions for generation units. the preced-
ing components can be combined to solve different problems.
the security analysis, e.g., is iterated with ed until there
are no security violations. the components can be organized
in different ways depending on the market requirements. Fig-
ure 6 shows an implementation of the SCuC with these compo-
nents. notice that the architecture has two loops. the inner
loop is often termed SCED. the individual components are
explained as follows.
UC
the uC component is an optimization problem that determines
each unit’s on/off status for each time interval by respect-
ing the unit-level constraints, such as minimum up/down
time, ramping limit, and system constraints, such as energy
balance and security. typically, the dam involves a 24-h or
longer uC module for hourly on or off decisions. the uC prob-
lem is generally hard to solve because its integer solution space
grows exponentially as the number of generation units increases.
the most commonly used solution techniques for the uC are
lagrange relaxation and milP. enabled by the increasing com-
puting power in recent years, the milP solution that allows
more modeling flexibility has become a dominant approach to
uC. the milP performance can be much affected by the prob-
lem formulation. therefore, an efficient milP formulation is
very important. Studies on tightening uC formulations can be
found in the literature. however, it is generally considered that
there is no universally good formulation for all uC problems
due to the various factors that affect the path of a milP solver.
ED
the ed component is an optimization problem that deter-
mines each unit’s generation output, subject to unit and system
constraints. the objective of the optimization problem is to
minimize the total dispatch cost based on units’ submitted cost
curves. the uC statuses determined by the uC problem are
fixed, so there are generally no integer decisions for the ed
component. depending on the offer curve characteristics, the
ed problem can be formulated as a linear or quadratic program.
Security Analysis
the security analysis component examines the dispatch solution
against n-1 contingencies. For each contingency, a power flow
problem is solved to check and identify any network constraint
violations. the violated constraints are linearized and fed back
to the next iteration of the ed problem, forming the inner loop
of SCed. if the SCed cannot resolve all security constraint vio-
lations, then the remaining violated constraints are fed into the
next iteration of uC until all violations are resolved, forming the
SCuC problem.
Security analysis could be time-consuming due to the
large number of contingencies. Because the simulation of
different contingencies is independent, parallel computing
techniques are often adopted.
january/february 2019
Pricing and Settlement
whereas all of the preceding components are intended to
obtain a secure commitment and dispatch solution for all
units, the clearing price is critical for a good market design
because it acts as the signal to the market participants. if prop-
erly priced, a unit will have no incentive to deviate from the
iSo’s schedule, which is a desired market design property that
allows market participants to bid truthfully. as we will discuss
next, pricing could be particularly challenging in the presence
of market nonconvexity or multiple-interval dispatch.
with the market clearing prices, settlement is defined by a
set of market rules of who pays or gets paid and by how much.
the settlement component collects information from the mmS
as well as emS to calculate uplift payments and so-called set-
tlement positions for each market product and time frame.
Electricity Pricing Challenges
Pricing is a critical part of a market because proper market
prices provide the right signals to market participants for
achieving an efficient market outcome. Pricing in electric-
ity markets could be complex in the presence of transmis-
sion congestion and losses. Furthermore, nonconvexity in a
market is known to cause problems for conventional linear
prices. unfortunately, nonconvexities are abundant in elec-
tricity markets as a result of physical unit characteristics.
Pricing with nonconvexity has therefore become one of the

System Planning
Market Operation Financial Settlement
and Operation

Day-Ahead Scheduling DAM Energy Trading Two-Settlement System

Mitigated
Outage Operations
Offer, Unit Energy
Coordination Planning
Topology, Schedule Awards
Limits LMP
DAM Energy
Load Market UC ED SFT and Uplift
Forecasting Mitigation
Awarded
Reserve
Security UC Schedule
Transmission Reservation RMR
Transaction Constraints
FTR Payment

CRF
Real-Time RTM Energy
Operation Trading
EMS System
Cost Allocation

Conditions
SCADA
Load and RUC
Renewable RTM Energy
State Estimation
and Uplift
UC
Market
Contingency Short-Term Schedule
Clearing
Analysis Outage
RTUC
Short-Term RTM Ancillary
Load Limit Service
Calculation Fast Start Payment
UC Schedule
System
Monitoring Tie-Line
Scheduling
Constraint RTM
Performance
Management AS&
AGC Units Payment
Capacity Tie-Line
AGC Analysis Schedule
Unit Dispatch

figure 5. The MMS architecture. RTUC: real-time UC; AGC: automatic generation control; SCADA: supervisory control
and data acquisition; CRF: congestion revenue fund; RMR: reliability must run units; SFT: simultaneous feasibility test.

january/february 2019 ieee power & energy magazine 39

 Pricing could be particularly challenging in
the presence of market nonconvexity or
multiple-interval dispatch.

central topics in electricity markets. a less discussed pric- mitment decisions. as a result of the fundamental mathematical
ing problem, but attracting more interest recently due to the properties of nonconvex problems, there’s no such set of per-
emerging ramping needs in several iSos, is the appropri- fectly consistent prices for the uC problem. Consequently, for
ate pricing and settlement design with multiple scheduling any given market clearing prices, some units may have incen-
intervals. this problem is further complicated by the practi- tives to deviate from their optimal schedule. recognizing this
cal implementation of multi-interval scheduling with rolling nonexistence of perfect prices, in reality, markets are designed
horizons. we will discuss the solution of a multisettlement to pursue a second-best set of prices with certain criteria.
system as an extension to the existing two-settlement system. most of the existing markets use the uplift payment as the
additional compensation alongside the market clearing prices
Nonconvexity for units to follow the iSo’s optimal commitment and dispatch
wholesale electricity prices, or lmPs, are typically produced instructions. the total uplift payments are then allocated to
from the shadow prices of the SCed problem. Because the demands in the market. different markets may have differ-
SCed problem is often convex, i.e., being a linear program or ent uplift calculation and allocation rules. most iSos pay the
a quadratic objective function with linear constraints, the con- make-whole payment for online units to cover the revenue
straint shadow prices exist and can be conveniently used to cal- shortage, although some iSos (e.g., iSo-ne) also pay the so-
culate market clearing prices. these prices are consistent with called lost opportunity cost (loC) for certain situations even
the optimal dispatch solutions in the sense that, given the prices, when the unit has no revenue shortage. nevertheless, the uplift
each unit will not have financial incentives to deviate from the payments cannot be efficiently allocated through a market
dispatch results. in reality, however, a unit’s decisions are also mechanism and are often considered nontransparent. many
based on the commitment costs in addition to the incremental iSos have explored ways to reduce the uplift payments by
costs. the clearing prices from the convex dispatch problem adopting alternative market clearing prices in place of exist-
reflect the marginal cost of dispatch but do not capture commit- ing lmPs. one approach, termed convex-hull pricing, uses the
ment costs and thus may not be consistent with the unit’s com- dual variable solutions of the nonconvex uC problem to arrive
at the minimal level of total uplift
and loC (although the uplift pay-
ment may actually increase). other
SCUC
approaches typically relax the non-
Supply Offers UC convex features of the problem, such
Demand Bids as removing the minimal genera-
tion output parameter or allowing
Commitment
Schedules
Constraints

partial commitments of units for

Unit
Schedules LMPs obtaining a price that at least relates
Publishing to the commitment costs. a variety
Binding
Constraints of different relaxations have been
implemented by the iSos; how-
Outage ED ever, in theory, there is no perfect
Scheduler price that can eliminate the need for
Unit Schedules
Constraints

uplift payments due to the presence

Schedules

SCED

External
Transaction Schedules of the duality gap for nonconvex
problems. the practical question
Constraints remains: what are the objectives and
Security desired properties of the pricing?
Analysis different iSos may have different
External answers. the decision on the trad-
Transactions eoff between the level of noncon-
vexity and appropriate set of prices
figure 6. The SCUC architecture. is central to a good market design.

40 ieee power & energy magazine january/february 2019

 The power industry is facing yet another
revolutionary change.
Pricing Multi-Interval Solutions
given that uC and dispatch decisions are temporally coupled
through ramp limits and minimum up/down time constraints,
the market clearing prices for different time intervals are
naturally related. although the multi-interval dispatch problem
can be convex, and the prices for each interval conveniently
produced out of the dual variables, the challenge of pricing
with multiple intervals lies in its real-time applications that
run on the rolling horizon. nyiSo and CaiSo, for exam-
ple, solve a multi-interval dispatch problem but only use the
first interval prices for settlement; prices for other intervals
are only advisory and will be determined when the interval
becomes first in later runs. with the rolling horizon, each
multi-interval dispatch run includes new time intervals that
are not modeled in the previous intervals. as a result, the
prices produced from different runs for the same interval
could be different. the iSo’s settlement typically uses the
prices from the latest run where the interval being priced
is the first. these prices for different intervals are from
different runs, and so they may not be consistent with the
dispatch solutions for different intervals. this may cause the
incentive compatibility problem, i.e., deviation from the iSo
instructions, as the units chase the settlement prices.
a natural way to remedy this problem is to make all
prices from different runs financially binding, i.e., one inter-
val would be settled multiple times instead of just for the
last time. this effectively requires a multisettlement design
where multiple levels of markets in addition to the existing
dam and rtm are involved. iSo-ne is developing a new
approach to appropriately price ramping constraints.
New Market Initiatives
the existing market designs are shaped by the historical,
physical, and regulatory features of the system as manifested
by the difference in the designs of different iSo markets.
Currently, the power industry is undergoing a major transfor-
mation toward a system with a lower carbon footprint, more
decentralized controls, and an increasingly flexible load.
these changes bring fundamental challenges to power sys-
tem operations, planning, and market design. in the follow-
ing, we highlight the new challenges associated with the new
electric grid, wrapping up with a discussion on future market
design to accommodate the challenges.
A New Electric Grid
modern power systems went through different stages of evo-
lution driven by technical, economic, and regulatory events.
january/february 2019
they went from a decentralized, loosely coupled grid to
large, highly interconnected, and centrally controlled sys-
tems. the sheer size, large number of components, and pro-
liferation of power electronics have made the electric grid a
complex cyberphysical system.
Currently, the power industry is facing yet another revo-
lutionary change. government directives to lower the carbon
footprint and, as a consequence, high penetration of renewable
energy resources and smart grid technologies are transform-
ing power system planning and operational paradigms. dis-
tributed energy resources (ders) are being built deeply into
the distribution networks, and the boundary between transmis-
sion and distribution is blurring. a significant part of ders is
unobservable to the transmission system operators, introduc-
ing an unprecedented level of uncertainty not only spatially
but also temporally due to the intermittent nature of renewable
resources. even the concept of contingency is changing from
being binary (e.g., on and off events) to continuous, e.g., the
system load or generation can swing by several gigawatts in a
comparatively short period of time. this behavior, previously
considered as abnormal or emergency, becomes part of normal
operation. as a result, system operators will be challenged to
deal with completely different patterns in the new grid archi-
tecture, such as the one shown in Figure 7. another property
of large cyberphysical systems is high interdependence with
different infrastructures. not only do we have to monitor the
electric grid contingencies but also the failures in communica-
tion, information technology, and fuel delivery systems. today’s
preventive approach to power systems planning and operation,
µGrid µGrid µGrid
Transmission Backbone
Virtual Power Demand PHEV
Plants Aggregators Aggregators
figure 7. A new power grid. PHEV: plug-in hybrid electric
vehicle.
ieee power & energy magazine 41
 i.e., every contingency is treated with equal probability and no
contingency is allowed to cause violation of the n-1 reliabil-
ity criteria, becomes economically prohibitive in the new
grid era. more corrective actions must be introduced to make
power system planning and operation less expensive. this all
requires a significantly different control paradigm, new mod-
els, and new algorithms. in response, the market design needs
to be changed as well to accommodate these new changes in
physical systems.
Future Market Design
although the changes in the electric grid impose significant
challenges to power system operation, planning, and control,
adapting existing electricity market designs and developing
new design solutions to address these challenges is an even
more daunting task, given the relatively short history of elec-
tricity markets. in this section, we discuss several key market
design problems that are likely to occur in the near future and
suggest possible solutions.
the unprecedented level of uncertainty due to the high
penetration of renewable resources and ders requires more
flexibility to deal with uncertainty. this leads to the need
for a new flexibility market mechanism to create incentives
for investment in flexible technologies. Some iSos, such as
CaiSo, have introduced a 5-min ramping product to accom-
modate the steep change of the net load curve due to significant
swings in photovoltaic production. the design of such a new
product should consider its interaction with existing products,
such as reserves. it is important that the new products provide
unique and different services from the existing products so as
to attract targeted flexible resources.
an alternative way of addressing these uncertainty challenges
is to manage them through more sophisticated commitment and
dispatch processes. instead of the existing deterministic com-
mitment and dispatch problems currently solved by the iSos,
stochastic programming or robust optimization that explicitly
models the uncertainty within their formulation through scenario
tree and uncertainty set seem to be promising. however, a bigger
issue is how to appropriately price the market clearing decision
and compensate the selected resources based on multiple possible
realizations of uncertain parameters. From the market design
perspective, transparent compensation schemes, proper informa-
tion policy, and incentive compatibility issues should be resolved
before we adopt these advanced risk management techniques.
direct demand participation (where demand bids are sub-
mitted to the market) in the rtms can bring significant ben-
efits to system operation. it enhances system reliability and
also improves the price formation in the electricity market.
however, an increasing level of indirect demand participation
(demand responds to the price on its own) can create chal-
lenges in load forecasting, system control, and, thus, real-time
pricing. to overcome these issues, more and more iSos start
incorporating iSo-developed demand curves into the market.
the electric reliability Council of texas, for example, has
applied the operating reserve demand curve to its real-time
42 ieee power & energy magazine
reserve market. iSo-ne has recently developed system and
zonal capacity demand curves for its FCm based on tradeoffs
between the costs of reliability and capacity. the development
of these demand curves based on different reliability needs is
a promising direction of the new market design.
the electric industry is moving into a new era with new car-
bon emission limits, new resilience needs, and the zero mar-
ginal cost economy where the incremental cost of producing a
good approaches zero. all of these new features bring signifi-
cant challenges for electricity pricing. with the carbon limits
and resilience needs, prices for these features need to be defined
for the market. the existing marginal cost pricing may lead to
zero or negative market clearing prices, when more and more
renewable resources with zero marginal costs and state/federal
subsidies participate in the market. therefore, the marginal
pricing principle may need to be augmented with some new
constructs, including new scarcity pricing rules.
Acknowledgment
the views expressed in this article do not represent those of
iSo new england, inc.
For Further Reading
e. litvinov, t. Zheng, g. rosenwald, and P. Shamsollahi,
“marginal loss modeling in lmP calculation,” IEEE Trans.
Power Syst., vol. 19, no. 2, pp. 880–888, 2004.
e. litvinov, “design and operation of the locational mar-
ginal prices-based electricity markets,” IET Generation,
Transmission, Distribution, vol. 4, no. 2, pp. 315–323, 2010.
F. Zhao, t. Zheng, and e. litvinov, “Constructing demand
curves in forward capacity market,” IEEE Trans. Power Syst.,
vol. 33, no. 1, pp. 525–535, 2017.
F. Zhao, e. litvinov, and t. Zheng, “a marginal equiva-
lent decomposition method and its application to multi-area
optimal power flow problems,” IEEE Trans. Power Syst.,
vol. 29, no. 1, pp. 53–61, 2014.
d. Schiro. (2017). Flexibility procurement and reimbursement:
a multi-period pricing approach. iSo new england. [online].
available: https://www.ferc.gov/CalendarFiles/20170623123635-
Schiro_FerC2017_Final.pdf
w. w. hogan. (2013). electricity scarcity pricing through op-
erating reserves. harvard univ. Cambridge, ma. [online]. avail-
able: https://sites.hks.harvard.edu/fs/whogan/#working papers
r. li, Q. wu, and S. S. oren, “distribution locational mar-
ginal pricing for optimal electric vehicle charging management,”
IEEE Trans. Power Syst., vol. 29, no. 1, pp. 203–211, 2014.
Biographies
Eugene Litvinov is with iSo new england, holyoke, mas-
sachusetts.
Feng Zhao is with iSo new england, holyoke, massa-
chusetts.
Tongxin Zheng is with iSo new england, holyoke,
massachusetts.
p&e
january/february 2019
The New Zealand
Electricity Market

T
Trading in The new Zealand wholesale
electricity market (nZeM) pool began on 1 october 1996. The
new Zealand pool market was the first one to use two ele­
ments of what has since become the north american standard
market design, with locational marginal pricing and ancillary
service cooptimization, but it was implemented in a physical
and regulatory environment that differs markedly from that in
north america. This article reflects on 20 years of experience
with new Zealand’s 1996 market design, assesses its points of
difference from other jurisdictions, and speculates on some
possible future directions.
new Zealand is subject to the vagaries of a south Pacific
climate, is geographically prevented from importing or ex­
porting power, and has a high penetration of renewable
electricity. Key considerations include promoting the most
economical use of stored water and ensuring security of sup­
ply. Unlike most north american markets, the new Zealand
design currently does not include any form of capacity market

By Andy Philpott, Grant Read,

Stephen Batstone, and Allan Miller

Challenges of a
Renewable Energy System

Digital Object Identifier 10.1109/MPE.2018.2871705

©istockphoto.com/NZsteve
Date of publication: 7 January 2019

january/february 2019 1540-7977/19©2019IEEE ieee power & energy magazine 43

 or payment or day­ahead market, and, at least initially, the
market did not provide any financial transmission rights and
was not overseen by any sector­specific regulator.
Historical and Physical
Context of the NZEM
geographically, new Zealand is comparable in size and shape
to the islands of Britain and ireland or Japan. But it is much less
densely populated and too remote to allow interconnection
with any other country. Moreover, it is uneconomical to build
facilities to import relatively small volumes of liquefied natural
gas just for occasional electricity generation in years with low
hydro inflows. Thus, new Zealand must be self­sufficient in
electricity, gas, and generally also coal, with occasional recent
imports of the latter. The layout of the country’s transmission
grid is shown in Figure 1, and the evolution of its generation
mix is shown in Figure 2.
Culy et al. (1996) describe the history of electricity in new
Zealand, starting with developments by private companies in
the 1880s. From the early 20th century, however, all signifi­
cant power stations were government built and owned. From
1945 to 1978, electricity was supplied to local distribution
authorities by a government department, which was able to
develop the system in ways the private sector probably could
not have. large­scale hydro schemes were developed, with
associated new canals, roads, and towns, despite gradually
increasing environmental opposition. new Zealand pioneered
the development of large­scale geothermal generation during
the 1960s. The backbone of the national transmission system
was built in the same period, with new Zealand pioneering
the use of long­distance underwater high­voltage dc transmis­
sion to link the two islands in 1965.
Until the construction of a small coal­fired firming sta­
tion in 1958, almost all of the developments were hydro. Culy
et al. question whether this was actually the most economi­
cal strategy, but it reflected a philosophical commitment to
exploiting free renewable resources to further national devel­
opment in ways that were seen to be environmentally benign,
if not beneficial, at the time. in the late 1960s and 1970s, the
direction of development changed, as public attention focused
more on the negative environmental impacts of hydro devel­
opment, and new discoveries made gas­fired electricity gen­
eration economical. new coal­ and oil­fired stations were
also built to cope with wet/dry­year output variations in an
isolated hydro system where reservoirs large enough to store
water from year to year were (and still are) considered to be
environmentally unacceptable.
in 1979, the electricity department was integrated into a
Ministry of energy, which controlled demand­side manage­
ment programs and the planning, production, and pricing of not

50,000 1
45,000 0.9
North Island
1,700 MW Hydro 40,000 0.8
1,000 MW Geothermal 35,000 0.7
600 MW Wind
1,700 MW Gas and Coal 30,000 0.6
(GWh)

48 MW Solar 25,000 0.5

4,400 MW Maximum Demand
20,000 0.4
15,000 0.3
10,000 0.2
1,240-MW
High-Voltage dc Link 5,000 0.1
South Island 0 0
1975
1977
1979
1981
1983
1985
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2011
2013
2015
2017

3,500 MW Hydro
60 MW Wind
14 MW Solar
2,200 MW Maximum Demand
Major Generation Center
Major Load Center
High-Voltage ac Lines
High-Voltage dc Lines
figure 1. A map of New Zealand showing the main genera-
tors, largest lines in the transmission network, and maximum
loads over the year ended 31 December 2017. Over this
period, the North Island grid consumption was 24,735 GWh
and the South Island grid consumption 14,557 GWh.
Gas (12.8%/15.3%) Coal (2.3%/2.6%)
Oil (0%/0%) Other Renewable (1.4%/1.4%)
Solar (0.1%/0.2%) Geothermal (17.5%/17.3%)
Wind (5.4%/5.1%) Hydro (60.4%/58%)
% Renewable (84.8%/82%)
figure 2. The generation by type in New Zealand, with re-
newable generation. In the legend, the numbers listed beside
each generation type give the proportion of generation in
2016/proportion of generation in 2017. Hydro was reduced
in 2017 because of low inflows up to July 2017. (Data from
New Zealand Energy Quarterly, Ministry of Business, Innova-
tion, and Employment, June 2017 quarter, 5 October 2017.)

44 ieee power & energy magazine january/february 2019

only electricity at the wholesale level but also of fuels supply­
ing and competing with electricity. The goal was to facilitate a
more consistent and economical approach, and significant prog­
ress was made in trimming back the most extreme development
proposals, rebalancing hydro and thermal investment, using con­
sistent shadow pricing of thermal fuels across sectoral develop­
ment plans, and optimizing the reservoir management strategy.
oil­fired stations were built and then mothballed while work
continued on increasingly expensive and controversial hydro
developments. The government of the day also preferred to boost
demand by pursuing a controversial “Think Big” program pro­
posing discounted sales to new smelters, rather than scaling back
development to match lower demand growth.
end­user pricing remained inconsistent throughout that
period of government control. Prices to the public were held
artificially low, while even lower prices were occasionally
offered to a few major industries to boost national develop­
ment. But prices for most commercial and industrial users were
much higher. as long as the system remained 100% renewable,
the bulk supply tariff (BsT) charged to distribution authorities
was based solely on peak load, reflecting the economics of a
system most often constrained by transmission capacity limits.
as thermal power became more important, an equal weighting
was introduced on energy and peak demand, but no attempt
was made to reflect the wide variation in marginal costs from
year to year due to hydrological variations or even changing
economic conditions.
starting in 1984, though, a new government instituted sweep­
ing reforms, liberalizing all aspects of the economy and corpora­
tizing most productive activities previously undertaken by govern­
ment departments. The electricity Corporation of new Zealand
(eCnZ) was formed and was expected to act in a normal com­
mercial manner, with a view to privatization, if possible.
The BsT was replaced by a new regime under which all
wholesale electricity was bought and sold using hedge con­
tracts written against a half­hourly pseudospot price calcu­
lated a week in advance using eCnZ’s operational optimi­
zation model. internal generation groups were formed and
given incentives to improve efficiency, and experiments were
conducted with a form of internal market coordination.
The transmission system was also separated out into a sep­
arate government­owned corporation (Transpower), which
was intended to recover its costs under a pricing regime com­
bining nodal spot pricing and what was called beneficiary
pays funding of new projects, with costs already sunk at the
establishment date being allocated using a flow­tracing algo­
rithm. This controversial regime was not implemented as
planned, and transmission pricing remains contentious.
The institutional arrangements just described were all
intended to lay the foundations for development of an elec­
tricity market, the design of which was debated at great
length in the early 1990s, with eventual implementation in
1996. at that time, a collection of eCnZ assets was formed
into a new company and privatized, competing with the rest
of the eCnZ as a duopoly at the wholesale level. distributors
january/february 2019
still acted as the primary retailers in each local network but
were required to accommodate competitive entry. Full retail
competition, including the forced separation of distribution
and retail, followed in 1999, at which point the remainder
of eCnZ’s assets were formed into three separate compa­
nies, with all still under government ownership, but partially
privatized in 2012. From that point on these four major enter­
prises, along with a fifth controlling a collection of smaller
hydro plants, have accounted for most of the market.
Market Developments and Features
The new Zealand market pioneered the implementation of
nodal pricing concepts developed in north america, moti­
vated by the fact that, in a relatively long and sparsely devel­
oped transmission system, even marginal losses can create
significant marginal cost variations across the network. The
details of the pricing and dispatch system are described by
alvey et al. (1998).
The new Zealand dispatch model included the novel con­
cept of ancillary service cooptimization, which was seen to
be important because, in this small system, even the loss of
a 250­Mw unit requires significant contingency response, in
the 6­ to 60­s time frame. This ancillary service requirement
is also a major driver of locational price differences because
interisland transfers are often more limited by the need to
provide reserve support against link failure than they are by
power flow congestion. at times, there has been more capacity
on contingency reserve duty in the receiving island than has
been actually generated there. Thus, much of the congestion
rent produced by locational price differences is actually due to
these constraints and also to losses rather than to line capacity
limits, as may be typical elsewhere. Therefore, the market set­
tlement surplus is partly needed to pay actual loss and ancil­
lary service costs, and the remaining risk profile differs from
that assumed in classical financial transmission right (FTr)
theory. This partly explains why U.s.­style congestion­based
FTrs, which had originally been part of the market design
concept, were only recently introduced into the new Zealand
market. But the unavailability of such rights encouraged each
generator to seek customers close to its generation base and
then to see little need for FTr support.
Regulation
given its long history of direct government control over
electrical infrastructure, new Zealand had no institutional
history of sectoral regulation similar to that in, e.g., the
United states. initially, the market was implemented with­
out any sectoral regulator at all. This was a deliberate deci­
sion, partly motivated by a perhaps excessively pure economic
view that the electricity sector should not be treated differ­
ently from others and partly by a more pragmatic assessment
of the overheads involved in maintaining a full­blown regu­
latory function for a relatively small sector. This approach
has since been abandoned, with the formation of the electricity
Commission in 2003 and then the 2010 establishment of the
ieee power & energy magazine 45
 electricity authority (ea), which performs the role of an
independent market regulator.
regulatory market oversight of the new Zealand whole­
sale market by the ea is relatively loose by international
standards. Pricing above short­run marginal cost is widely
regarded in the electricity industry as a tolerated mecha­
nism for recovering the long­run marginal costs of gen­
eration. This view is not inconsistent with the doctrine of
workable competition espoused by the ea, in which the
electricity sector is expected to behave like a normal mar­
ket as much as possible and competition is encouraged by
focusing on low barriers to entry, so that market power is
curbed by potential new entrants rather than explicit market
oversight. with no formal price cap and capacity market,
the new Zealand system relies on prices (either spot prices
or those of derivatives) rising enough from scarce situations
to yield the required rents to cover long­run costs.
The regulatory oversight of hydro­dominated systems is
also more challenging than for purely thermal markets. in
systems dominated by hydro, electricity prices reflect the
marginal value that agents place on stored water, a number
that depends on their prediction of future inflows and assess­
ment of shortage risks that could be caused by dry winters,
outages, and the limitations of small, isolated fuel networks
with little flexibility to trade or adjust supply. when a poten­
tial water shortage looms, price increases are typically a
risk­averse, opportunity­cost response to a perceived risk of
future energy limitations and interactions in an almost com­
pletely closed system, rather than to clear capacity limits in
the electricity sector. Therefore, perfectly competitive out­
comes become difficult to determine with any certainty, let
alone to regulate. The ea has constructed a historical market
database and is developing an understanding of the effects of
uncertainty on wholesale prices.

Shortage Risk
although new Zealand’s electricity system is dominated by
4,500 renewable energy, storage­based hydro can easily accom­
4,000 modate short­term variations in demand and wind. But it
3,500
remains susceptible to periods of sustained low inflows, so
the challenges for system operation are more about varia­
3,000
tions in fuel over seasons rather than minutes and hours, as is
2,500
(GWh)

the case for most other countries with high­renewable ambi­

2,000
tions. This fuel risk is compounded by the fact that inflows
1,500 tend to be at their lowest during midwinter, when national
1,000 demand is highest.
500 There are no mandatory contractual arrangements for
0 hedging the risk of shortage. as originally conceived in 1992,
the new Zealand market was to have included a firm energy
y

ch

ay

ly

y
be

be
ar

ar
Ju
ar

M
nu

nu
em

em

mechanism similar to that subsequently implemented in

M
Ja

Ja
pt

ov
Se

Colombia, under which retailers would have been required to

N

(a) hold call options with a strike price set around the marginal
600 operating cost of an open­cycle gas turbine (oCgT) unit. Thus,
while the market price was to have been uncapped, consum­
500
ers would effectively have been protected against exposure to
400 price spikes above the oCgT level. The scheme was rejected,
NZ$/MWh

though, partly because it proved difficult to define exactly

300 what capacity measure should be used in each region. also,
200 there was a capacity surplus at the market start, and consumers
would have been paying a kind of insurance premium account­
100 ing for a substantial proportion of the market value.
That leaves new Zealand with no market price cap, with
0
participants potentially exposed to prolonged periods of very
y

ch

ay

ly

high prices under dry­year conditions that might be expected

be

be
ar

ar
Ju
ar

M
nu

nu
em

em
M

approximately once every 20 years. in theory, participants in

Ja

Ja
pt

ov
Se

(b) the wholesale market should expect to encounter high prices

when energy shortages loom. one can see this behavior in
Controlled Storage recent years in Figure 3(a) and (b), where the national reservoir
2017 2016 2015 2014 2013 storage and generation­weighted electricity price are plotted.
The risks of energy shortage are monitored by Transpower
figure 3. (a) The historical controlled reservoir storage in New in its role as the system operator. The shortage risks are rep­
Zealand and (b) the generation-weighted electricity price. resented by the curves in Figure 4, which displays the actual

46 ieee power & energy magazine january/february 2019

storage trajectory and a set of storage levels that indicate
increasing levels of risk at that time of year.
Vertical Integration and Contract Markets
originally, a strict vertical separation was planned, with
generators banned from retailing and retailers from gener­
ation. Thus, they would all be required to interact by arms­
length trading in contracts. it was soon realized that retail­
ers with little capital and no secure customer base would be
unable to enter into the kind of long­term power purchase
contracts needed to support new capacity investment. nor
could investors in 1996 rely upon what was then considered
a dormant fledgling contract market.
while traditional vertical integration (combining genera­
tion, transmission, and distribution) was banned, generators
were then allowed to purchase retailers. The risks of illi­
quidity in contracts were highlighted in 2001, when a major
retailer competing solely on the basis of bilateral contracting
was forced, by substantial financial losses, to exit the market
after failing to purchase enough forward cover in a dry year.
Consequently, the market is now dominated by five main so­
called “gentailers,” plus several small stand­alone retailers.
This degree of vertical integration is a disincentive to trad­
ing contracts, so establishing a liquid contract market became
a major goal for the ea. spot market participants now also
trade various contracts to reduce exposure to risk, both bilat­
erally (over the counter) and via a futures exchange. The most
popular are fixed­volume contracts for differences, with a pay­
out per megawatthour to the purchaser equal to the difference
of the observed spot price and the contract price.
Capacity Market
having rejected the original capacity market proposal, new
Zealand has operated, to date, as a one­part or energy­only
market. australia adopted a similar design but implemented a
market price cap. That cap is set at a very high level, adjusted to
generate enough revenue to support oCgT entry if the market
runs out of capacity and prices reach the cap for 4.5 h per year,
which corresponds to a specified reliability standard.
new Zealand has no such standard and no market price cap.
instead, the ea relies on the prospect of high prices, reinforced
through a regulated stress­testing regime that highlights the
financial risks associated with dry years, to provide both gen­
tailers and stand­alone retailers the incentives to invest in or con­
tract for dry­year capacity to meet their wholesale purchase obli­
gations for customers. a penalty regime also applies in the event
that the 10% risk storage levels occur (see Figure 4), whereby all
retailers are required to pay customers nZ$10 per week during
the savings campaign that is triggered by this event.
Day-Ahead Markets
nearly every spot market pool in north america and europe
consists of two markets, with a two­settlement system. Com­
mitted generation is cleared in a day­ahead market on the
day prior to dispatch, and then flexible plant production is
january/february 2019
delivered in a balancing market on the day of operation. This
process enables operators of large thermal plants to plan their
unit commitment, subject to start­up and ramping constraints.
The day­ahead market is typically dispatched using a mixed­
integer program (MiP) that accounts for these indivisibilities.
in contrast, the australian and new Zealand wholesale
markets are single­settlement systems. The fixed costs and
constraints of thermal plants are internalized in the prices of
their energy offers that are then dispatched in real time in merit
order. in new Zealand, with very few large thermal plants, the
advantages of using a MiP for unit commitment are minimal,
so a single­settlement market seems sensible.
on the other hand, new Zealand is dominated by hydro­
electricity, and the dispatch of hydrostations down a river chain
within a day involves a dynamic element that could benefit from
the coordination provided by a day­ahead dispatch. indeed, a
day­ahead market was established at the market start, but no
trading ever occurred. To improve river­chain coordination, a
design was also considered in which intraday river­chain man­
agement was directly incorporated into the real­time market.
But technical difficulties led to this being abandoned in favor
of a regime under which river­chain managers are dispatched to
aggregate station offers to match their planned flow regime but
are then allowed to rebalance generation among their units—
e.g., to operate units at peak efficiency—as required between
market clearing and actual dispatch time.
4,500
4,000
3,500
3,000
2,500
(GWh)
2,000
1,500
1,000
500
0
July January July January July January
2016 2017 2017 2018 2018 2019
Nominal Full Controlled Storage
Mean–All Records Mean–Last 20 Years
1% Risk 2% Risk 4% Risk
6% Risk 8% Risk 10% Risk
HRC Received
figure 4. The historical hydro risk curves. The available
controlled storage is plotted in blue. Dotted curves indicate
storage levels that would incur the displayed probability of
the controlled storage falling to zero if all thermal plants
were fully dispatched. If the controlled storage meets the
10% risk curve (red), the system operator will launch an
official conservation campaign, in which consumers will
be asked to conserve electricity.
ieee power & energy magazine 47
 Since 2010, the number of retail brands in the market has increased
from ten to over 30, with the market share of the five largest
gentailers declining from 97% to approximately 90% at present.

Retail Market Competition
since its establishment in 2010, the ea has been especially
active in promoting retail competition. Possibly the most sig­
nificant development here was the establishment of an electric­
ity futures exchange, which enabled small independent retailers
to hedge the significant medium­term price volatility charac­
teristic of new Zealand’s hydro­driven market. since 2010, the
number of retail brands in the market has increased from ten
to over 30, with the market share of the five largest gentailers
declining from 97% to approximately 90% at present.
when full retail competition was introduced in 1999,
smart meters were required only for larger typically industrial/
commercial consumers. studies suggested that the costs of
mandatory smart meter installation exceeded the benefits for
most households. This was left up to the market, except for a
regulation requiring recertification of all meters by 2015. The
smart metering rollout is relatively high now, at around 70%,
but this seems to have been driven more by the opportunity for
retailers to reduce meter­reading costs than by a strong desire
by consumers to actively manage their consumption patterns.
while recent growth has been strong, only a small proportion
of households actually face half­hourly spot prices, and, while
they receive real­time alerts of looming price spikes, they actu­
ally see their own data only about two days after real time.
There has been some consolidation in the smart meter market
to two main service providers and two main meter types.
8,000
Change Since 2007 (GWh)
Until recently, though, the tariff structure for a domestic
electricity consumer (as well as many mass­market business
customers) was almost exclusively a fixed­price, variable­vol­
ume tariff. new tariffs are emerging, however, most notably a
spot­based tariff offered by two retailers. one of these, Flick
energy, acquired over 20,000 customers (~1% of the market)
in a little more than three years. But 2017 was the first high
wholesale price year since Flick’s entry, and it remains to be
seen how durable the company’s customer base is in the face
of temporary high prices at the meter or whether the insula­
tion from wholesale signals (via a fixed­price, variable­vol­
ume contract) is ultimately preferred by customers, albeit at a
higher average price.
Investment and Entry
as discussed previously, investment in generation capacity is
at the discretion of market participants and solely incentivized
by participants who form views on future electricity prices in
an energy­only market, with revenue partially underwritten
by contracts with other market participants or retail custom­
ers. while investment is largely price driven, price projec­
tions depend on electricity demand growth and the investment
intentions of all market participants. new investment was low,
in the years immediately following deregulation, but accel­
erated after 2004. This investment was dominated by wind
and geothermal but also the construction of a large combined­
cycle gas turbine (CCgT) in 2006, bringing the total fleet of
CCgTs to four. although several hydro proposals have been
extensively investigated, environmental opposition has been

significant, and no substantial hydro project has actually pro­

6,000
ceeded since the market commenced in 1996.
4,000 Figure 5 shows the relative changes in both net supply
2,000 and demand (in terms of annual expected production) since
0 2007. over this period, electricity demand has remained
–2,000 largely flat, an unprecedented ten­year trend. while there
–4,000 have been some industrial closures during that period, we
–6,000 believe this trend is largely driven by energy efficiency. over
07
08
09
10
11
12
13
14
15
16
17

this time, participants continued to invest in renewables,

20
20
20
20
20
20
20
20
20
20
20

perhaps expecting demand growth to eventually recover. as

New Generation (Gigawatthours Expected)
Decommissioned Generation (Average Output)
Change in Demand Versus 2017
Net Supply Balance
figure 5. The new generation and decommissioned plant
generation in New Zealand, 2008–2017. (Image courtesy of
Whiteboard Energy.)
a result, thermal generation was increasingly displaced by
renewables, leading market participants to decommission
over 1,000 Mw of thermal capacity since 2013 (over 10% of
total capacity). a bilateral contracting arrangement between
market participants has delayed the mothballing of a further
480 Mw (two units at a 30­year­old coal/gas rankine plant)
until the end of 2022, amid concerns about the security of

48 ieee power & energy magazine january/february 2019

supply. it remains to be seen whether these ad hoc contracting
arrangements can continue to keep sufficient lightly utilized
thermal capacity viable to provide supply security in an era
of increasing renewable investment.
looking forward, the major disincentive to further invest­
ment continues to be lack of demand growth. This is com­
pounded by some degree of uncertainty over the future of a
large aluminum smelter, which makes up 12% of the country’s
total demand. on the other hand, as fossil fuels are phased out,
the electrification of transport and heat will put significant
upward pressure on demand. given new Zealand’s present
and potential renewable resources, it is in an enviable position
to fuel this electrification while contributing to its international
commitments regarding carbon emissions. however, the rapid
expansion of renewables implied by some future scenarios will
have to deal with considerable public concern about the visual
and environmental effects of both wind and hydro. dry­year
security will be an increasing challenge.
New Developments
in the Distribution Sector
during the reform process, the number of distribution
authorities was substantially reduced, but new Zealand still
has nearly 30 electricity distribution businesses (edBs) serv­
ing a total population of 4.5 million people. There is a range
of ownership models for edBs, including local government,
private investor, community­owned trust, and a combination
of private investor and community trust. Being considered
monopolies, edBs are subject to price­quality and informa­
tion­disclosure regulation, with strict limits on generation
or retail activities. The Commerce Commission determines
a default price­quality path, but edBs determine their own
pricing structures and access arrangements, within limits.
national retailers must create package offerings that deal
with many different cost structures, which poses a barrier to
retail competition and innovation.
Photovoltaic Solar
There has been no development of grid­connected photovol­
taic (PV) solar in new Zealand, but Figure 6 shows the his­
torical growth of rooftop PV solar, which is relatively low in
new Zealand, at about 13.7 wp per person. watt peak (wp) is
the peak power that can be produced by a solar energy system
under ideal sunlight and temperature conditions. The actual
power produced will vary substantially from the peak power
as sunlight (irradiation) and panel efficiency change with
temperature, shading, and surface cleanliness. about 90% of
the PV solar capacity is small­scale residential rooftop.
Unlike many other countries, the new Zealand govern­
ment has never provided explicit subsidies for rooftop PV

80

70

60
Installed Capacity (MW Peak)

50

40

30

20

10

0
2009-1
2009-3
2009-5
2009-7
2009-9
2009-11
2010-1
2010-3
2010-5
2010-7
2010-9
2010-11
2011-1
2011-3
2011-5
2011-7
2011-9
2011-11
2012-1
2012-3
2012-5
2012-7
2012-9
2012-11
2013-1
2013-3
2013-5
2013-7
2013-9
2014-11
2014-1
2014-3
2014-5
2014-7
2014-9
2014-11
2015-1
2015-3
2015-5
2015-7
2015-9
2015-11
2016-1
2016-3
2016-5
2016-7
2016-9
2016-11
2017-1
2017-3
2017-5
2017-7
2017-9
2017-11
2018-1
2018-3

Year and Month

figure 6. The PV solar installed capacity in New Zealand. The capacity reached 74 MWp (13.7 Wp per person) in April
2018. (Data through to September 2013 courtesy of the GREEN Grid project and data post-2013 courtesy of the EA.)

january/february 2019 ieee power & energy magazine 49

 solar. although some gentailers offered buyback rates of
around nZ$0.17/kwh prior to november 2014, these were
then reduced to approximately the average wholesale price of
nZ$0.08/kwh. growth in solar is still indirectly encouraged
by a low­use tariff regulation, which requires edBs to charge
residential consumers with low consumption a low daily fixed
rate, forcing transmission and distribution costs to be recovered
through a per­unit electricity price well above the wholesale
energy price, which is supposed to reflect the marginal energy
cost. This encourages energy efficiency but also incentivizes
potentially uneconomic use of nonelectric heating as well as
installation of rooftop solar water heating and PV.
studies, such as that by Miller et al. (2016), show that PV
solar is already commercially attractive for some households
with low discount rates and for some commercial customers,
even though the latter typically have a much lower variable
price and higher fixed price than residential consumers. in
fact, with current distribution pricing arrangements and fall­
ing PV prices, new Zealand seems close to a rapid adop­
tion of PV solar by residential customers, especially in those
regions with high irradiation.
But it remains to be seen whether PV solar can make a sig­
nificant positive economic contribution, given new Zealand’s
relatively cloudy climate and demand patterns dominated by
heating rather than cooling requirements. PV solar capacity
factors can be as high as 18% in the north of both islands,
where solar growth is highest, but only around 12% in the
south of the south island, and that is mainly during sum­
mer, whereas load peaks on long, dark, cold winter nights.
This contrasts with capacity factors well above 20% in parts
of australia and north america, in regions where demand
also tends to peak on hot, sunny days. nor does solar hedge
low hydro reservoir inflows, which fall to their lowest level
around June and July, when PV solar output is also lowest.
Electric Vehicles
new Zealand has no automotive industry and accounts for only
a minuscule proportion of global demand. in fact, the majority
of new Zealand households buy used vehicles from Japan and
thus choose from whatever was sold there several years ear­
lier. Thus, new Zealand has no discernible impact on global
research and development priorities or even manufacturing
strategies, so there is little economic incentive to foster experi­
mentation with new automotive technologies. The potential
value of electric vehicles in the new Zealand context is also a
complex question, depending heavily on correlations between
charging requirements, other loads, and renewable generation
sources across daily, weekly, and annual cycles, over a wide
variety of meteorological combinations.
still, the government has provided a fund for innovative
marketing and support of electric vehicles. electric vehicles
currently account for 0.16% of all light vehicles in the coun­
try, but Figure 7 shows that adoption has grown strongly in
the last two years, as more models become available interna­
tionally, range increases, and marketing ramps up.
Battery Storage
Batteries seem unlikely to contribute much to new Zealand’s
main supply security issue: storing enough excess production
in wet years to cover dry conditions occurring perhaps every
20 years or fewer. Batteries face stiff competition with exist­
ing hydro storage as a way of dealing with supply/demand
fluctuations over minutes, hours, or days. at a system level,
they will not make much contribution until solar and/or wind
output becomes so high that hydro cannot be backed off any
further and energy must be spilled.
at a household level, storage batteries can make PV solar
more attractive by maximizing self­consumption. But the cost
for this use alone is prohibitive, adding at least nZ$0.4/kwh

7,000
Number of Light Electric Vehicles

6,000 New Light Pure Electric

Used Light Pure Electric
5,000
New Light Plug-In Hybrid
4,000 Used Light Plug-In Hybrid

3,000

2,000

1,000

0
2014-1
2014-2
2014-3
2014-4
2014-5
2014-6
2014-7
2014-8
2014-9
2014-10
2014-11
2014-12
2015-1
2015-2
2015-3
2015-4
2015-5
2015-6
2015-7
2015-8
2015-9
2015-10
2015-11
2015-12
2016-1
2016-2
2016-3
2016-4
2016-5
2016-6
2016-7
2016-8
2016-9
2016-10
2016-11
2016-12
2017-1
2017-2
2017-3
2017-4
2017-5
2017-6
2017-7
2017-8
2017-9
2017-10
2017-11
2017-12
2018-1

Year and Month

figure 7. Light electric vehicles in New Zealand. Up to January 2018, the total number was 6,526 vehicles, a small
fraction of the approximately 3.6 million light vehicles in the country. (Data courtesy of the New Zealand Vehicle Fleet
Statistics, New Zealand Ministry of Transport.)

50 ieee power & energy magazine january/february 2019

 While New Zealand’s greenhouse gas emissions
are small on a global scale (about 0.15%), its per capita
emissions are comparatively high.
to the average cost of energy in a PV solar system. greater
benefit can be gained by situating batteries behind the meter,
thereby enabling one battery to deliver a bundle of network,
PV solar, demand response, and ancillary service potential ben­
efits. even so, studies suggest that it will be some time before
batteries become economical in new Zealand. nevertheless,
some edBs are experimenting with in­home and larger­scale
batteries to manage specific constraints in their networks. as
with interruptible­load systems, however, the regulators may
find it difficult to ensure equal access to the many benefits that
batteries may bring, while also allowing them to be bundled
effectively. household installation may be inhibited if new reg­
ulations require lithium­ion batteries to be installed in separate
enclosures outside homes for fire safety reasons.
The Future
The ea’s statutory objective is “to promote competition in,
reliable supply by, and the efficient operation of, the elec­
tricity industry for the long­term benefit of consumers.”
Consistent with this, the agency’s current priorities include
encouraging more retail competition in the market, improv­
ing information flows, finalizing prices closer to real time,
and enabling access to retail data, such as smart metering
data. it is also reviewing distribution pricing and exploring
options to encourage greater efficiency and competition by
opening access to distribution networks, developing alter­
natives to traditional electricity distribution, diversifying
supply sources, and better utilizing demand response.
some electricity distribution businesses are evolving from
low­voltage network owners into mini system operators, as
they seek to integrate the numerous new devices connecting
to the network and make economic tradeoffs with the new ser­
vices to manage system peaks and reduce the need to continu­
ally upgrade their networks. Challenges abound for low­volt­
age networks in optimizing the use of storage batteries, smart
electric vehicles, and increased demand­response, especially
where this requires integration with price signals emanating
from the wholesale market. Presently, the trend toward distri­
bution system operators is occurring within existing network
owners; it remains to be seen whether the separation between
asset ownership and system operation, which has been largely
adopted for the high­voltage grid market, will also eventuate
for distribution­level markets.
at a macro level, while the renewable generation share is
currently around 85%, a government goal to reach 100% renew­
able generation in a normal hydrological year by 2035 means
that the market now faces a very significant developmental
january/february 2019
and operating challenge. Moreover, new Zealand has ratified
the Paris agreement, with its nationally determined contribu­
tion (ndC) requiring a reduction in equivalent carbon dioxide
emissions by approximately 40% of current levels by 2030 and
net zero emissions by 2050. while new Zealand’s greenhouse
gas emissions are small on a global scale (about 0.15%), its per
capita emissions are comparatively high, mainly due to the
high level of methane and nitrous oxide emissions from agri­
culture. studies have shown that the electrification of private
vehicle transport and most stationary heating, and converting
of the remainder of electricity generation to renewable sources,
would almost reach new Zealand’s ndC. But the increase in
renewable generation required to achieve this would be about
130%. Thus, the question is what sources could be used to
deliver that level of renewable generation.
The sheer quantum leap of new generation required will
inevitably imply a major conflict between global environ­
mental concerns and localized environmental antagonism to
hydro or wind development, at least. But there are also three
specific problems to be addressed.
First, the short­term variability of renewable generation
sources will be a particular problem in the new Zealand
system, because it is relatively small and unable to import or
export energy or achieve the diversity typical of continental
power systems. new Zealand has so far incorporated 800 Mw
of wind generation (about 9% of installed capacity) with no
adverse effects on power system reliability, primarily because
flexible (predominantly hydro) generation has thus far been
able to manage the variations in wind generation supply. But
significant additions of flexible hydro capacity seem unlikely,
and the flexibility of existing operations is gradually being
reduced by environmental opposition to varying both flow and
storage levels.
There has also been significant pushback over the im­
pact of wind farms on landscape values. still, with about
3,000 Mw already approved for development, wind can be
expected to play a major role in achieving the 100% renew­
able target, along with geothermal, of which there is about
300 Mw endorsed for installation. studies by the green
grid project have investigated the technical impact on the
power system of adding up to 4,000 Mw more of wind gen­
eration (about 50% of 2018’s installed capacity). The results
show an increased need for frequency regulation, instanta­
neous reserves to deal with the reduction in inertia, and droop
response. ancillary service prices will eventually have to
rise to a level at which it becomes economical for investors
to provide these services, either from partially dispatched
ieee power & energy magazine 51
 geothermal generation, demand­side involvement, or battery­
based ancillary services.
second, new Zealand’s highest demand occurs during win­
ter evenings, when there is no solar and the prospect of cold,
windless conditions. Thus, for the near future, flexible hydro
will need to be supplemented by thermal peaking plants, albeit
operating at low load factors. in the longer term, though, the
storage capacity of hydro and demand­side heating/cooling
systems will probably need to be supplemented by storage bat­
teries, including distributed battery capacity in PV solar instal­
lations and electric vehicles.
Third, new Zealand already faces a significant mismatch
between seasonal demand and supply patterns that will not
be reduced by the renewable development options discussed
here. it must also cope with major wet/dry­year fluctuations,
which may well be exacerbated by climate change. indeed,
changing precipitation patterns may well change the effec­
tive storage capacity of existing hydro. arguably, these
factors could be mitigated by building much more hydro res­
ervoir storage capacity, but that seems unlikely to happen. if
burning stored fossil fuels is unacceptable, the only alter­
natives would seem to be burning stored biofuels, extreme
overbuilding of other renewable energy systems, or demand­
side management. in particular, electricity­intensive indus­
tries, such as smelting, could be developed further, provided
those industries maintained the flexibility to substantially
reduce consumption in dry years. effectively, varying the
net export of, say, smelted aluminum would be equivalent
to varying the net export of electricity, as would occur if
new Zealand were connected to a larger international power
grid. There would obviously be an economic cost to building
demand­side capacity to be idle in the occasional dry year.
But that cost may be significantly less than that of build­
ing substantial excess hydro/wind/solar capacity to be idle
in every other year. The same is likely to be true of many
islanded power systems.
Finally, increasing renewable contributions will also
increase the proportion of time that plants with near­zero
marginal cost are on the margin. offering more genera­
tion from these plants to the wholesale market will depress
the market clearing price, providing lower market returns
for most generation most of the time. To recover generators’
fixed costs in an energy­only market will then require more
periods with higher scarcity prices. arguably, this trend
may be so extreme as to pose an existential challenge to the
whole market design and its governance arrangements. it
is possible to calculate the average annual level of scarcity
prices required to incentivize a reasonable level of capac­
ity investment. in theory, we may argue that there is little
difference between suppliers collecting this revenue during
occasional price spikes and the government collecting it via
capacity charges, as was done when new Zealand last had a
100% renewable power system before 1958. But there may
be very significant differences in the way those two regimes
are perceived by risk­averse investors, the regulator, and the
52 ieee power & energy magazine
general public. Those perceptions may ultimately determine
the durability of the regime.
Perhaps the ea’s current approach will endure if the sec­
tor continues to be dominated by substantial vertically inte­
grated firms, with strong incentives to stay in business for the
long term. But that may change if the sector evolves toward
a purer version of the competitive market paradigm, with a
large number of less­substantial firms relying on supply from
and to a highly volatile spot market.
The situation in new Zealand is quite different from that
in, e.g., australia, where both investors and the general pub­
lic expect a fairly predictable number of short, sharp spikes
on hot days every summer. in new Zealand, a future genera­
tion of consumers. having experienced maybe 20 years of
low renewable energy prices, may find it quite hard to accept
that the accumulated cost of backup generation must be paid
for in a single dry winter of extremely high prices. Presented
with such an event, a future government may also find it hard
not to respond by intervening, either changing the market
design, overruling the regulator, or perhaps even reversing
privatization. risk­averse investors are likely to commit to
significantly less than the optimal capacity requirement
on the basis of such an irregular and unpredictable income
stream and even less if they believe the regime itself may
be changed so as to specifically prevent them from collect­
ing the revenue required to pay for that capacity when a
really dry year finally occurs. Thus, arguments about the
need for an explicit capacity market mechanism seem only
likely to increase.
For Further Reading
T. alvey, d. goodwin, X. Ma, d. streiffert, and d. sun, “a
security­constrained bid­clearing system for the new Zea­
land wholesale electricity market,” IEEE Trans. Power Syst.,
vol. 13, no. 2, pp. 340–346, 1998.
J. g. Culy, e. g. read, and B. wright, “structure and
regulation of the new Zealand electricity sector,” in Inter-
national Comparison of Electricity Regulation, r. gilbert
and e. Kahn, eds. Cambridge, U.K.: Cambridge Univ. Press,
1996, p. 312–365.
a. Miller, s. Mcnab, s. lemon, and a. wood, “The eco­
nomics and potential uptake of PV solar power by region and
PV system cost,” in Proc. Electricity Engineers Association
Conf. Exhibition, wellington, new Zealand, 2016.
Biographies
Andy Philpott is with the University of auckland, new
Zealand.
Grant Read is with the University of Canterbury, Christ­
church, new Zealand.
Stephen Batstone is with whiteboard energy and sapere
research group, Queenstown, new Zealand.
Allan Miller is with the University of Canterbury, Christ­
church, new Zealand.
p&e
january/february 2019
 A Change
Is Coming

©istockphoto.com/kamisoka

B
How Regulation and Based on current policy targets, projec-
tions, and expectations, electricity is set to play a cen-
Innovation Are Reshaping tral role in the european union’s (eu’s) economy. the
ambitious goals of decarbonization and energy-efficient
the European Union’s actions include decreasing greenhouse gas emissions
by 40% in 2030 and 95% in 2050 down to below 1990
Electricity Markets levels, increasing the renewable energy share to at least
27% of final energy consumption in 2030, and attain-
ing 27% energy savings by 2030 as compared to busi-
ness-as-usual rates of growth. correlated electrification,
decentralization, and digitalization trends suggest that the
By Gianluca Fulli, share of electricity in final energy consumption can grow
Marcelo Masera, from above 30% in 2030 to nearly 40% in 2050. decen-
tralized generation capacity could account for more than
Amanda Spisto, and 30% of all generation capacity in 2030 and could eas-
ily exceed half of installed generation capacity by 2050.
Silvia Vitiello intelligent inverters can double the distribution grid’s
capacity to the host photovoltaic systems. smart meters
Digital Object Identifier 10.1109/MPE.2018.2872303
are scheduled to reach nearly half of the eu’s citizens
Date of publication: 7 January 2019 by 2020, whereas internet-of-things (iot) devices (with

january/february 2019 1540-7977/19©2019IEEE ieee power & energy magazine 53

 The role of wholesale, day-ahead electricity markets in steering not
only as short-term, efficient resource allocations, but also medium-
to-long-term investments, is currently being debated.
embedded functions that interact with the grid) could serve
the majority of the eu’s population by 2030.
in such a rapidly shifting context, electricity system stake-
holders are striving to keep up with the pace of innovation and
anticipate the infrastructure and market arrangements neces-
sary for future electricity delivery and services. legislators
and regulators are confronted with the challenging task of cre-
ating the most effective actions to steer the electricity sector
transition for the customers’ benefit. the eu’s energy union
strategy issued in 2015 and the ensuing clean energy package
proposal (cep) put forward by the european commission in
2016 are examples of the efforts of eu institutions to under-
take those policy and regulatory issues.
against this backdrop, we describe some of the main
regulatory challenges underpinning the restructuring of elec-
tricity markets in the eu. We first analyze how policy deci-
sions have contributed to developing the present eu electric-
ity system and markets. then, we illustrate new policy and
regulatory initiatives that aim to redesign the eu markets.
in conclusion, we share some thoughts on the interplay of
regulatory and technical advances and their impact on future
power systems.
Current EU Electricity
Markets’ Regulations
Background and General Features
the demand for electricity within europe’s internal market
prompted the introduction of a radical project that triggered
significant challenges following the passage of the single
european act of 1986, a piece of legislation that advocated an
eu internal electricity market by 1992. three subsequent eu
energy packages, introduced in 1996, 2003, and 2009, have
created a single, competitive european market. national verti-
cally integrated undertakings, which, in the past, controlled the
whole electricity value chain, are now divided into four blocks:
generation and retail companies that compete in their respec-
tive domains and (natural) transmission and (local) distribution
monopolies. at the same time, the power system has progres-
sively moved from a scheme with centralized and predict-
able generation, which met a rather inelastic demand, toward
one with higher shares of renewable and distributed energy
resources that meets a more flexible expectation.
the european regulatory approach concerning monopolies
was implemented by the third energy package that, among
other things, called for the unbundling of transmission system
operators (tsos) according to one of the following options: 1)
54 ieee power & energy magazine
a full ownership-unbundled tso; 2) an independent system
operator managing, but not owning, the transmission network
assets; or 3) an independent tso legally separated but still
embedded in a larger energy company (Figure 1). in europe,
43 tsos and roughly 2,400 distribution system operators
(dsos), acting as regulated independent actors, comprise the
link between the system physics and the market transactions
by interacting with several electricity system players.
another main regulatory intervention addresses electric-
ity markets. today, electricity is traded in the eu either via
long-term bilateral contracts (which represent two-thirds of
the electricity transactions concluded in the major markets
in 2015) or via the four segments of the wholesale power
exchanges, i.e., the forward and derivatives markets (weeks
to years in advance), the day-ahead markets (daMs), the
intraday markets (iMs) (operating for an hour or fractions of
an hour), and the balancing markets (operating in real time).
the 2015 yearly trading volumes reached a few hundred ter-
awatthours in the major european spot markets, including
the italian power market exchange (ipeX), the centre-north
european power market exchange epeX, and the scandina-
vian nordpool. at the retail market level, roughly 3,100 local
suppliers offer electricity to consumers that is acquired from
either generators or the wholesale market.
the role of wholesale, day-ahead electricity markets in
steering not only as short-term, efficient resource allocations,
but also medium-to-long-term investments, is currently being
debated. as low-carbon technologies penetrate the system,
average spot-market energy prices tend to drop while price
volatility (if not constrained by rules) grows. the anticipated
decrease in wholesale, day-ahead energy prices does not seem
consistent with guiding long-term investments in electric-
ity generation and infrastructure. the widespread incentives
available for specific technologies triggered by supply secu-
rity, sustainability, or social inclusion policies may discourage
market-driven investments in some power system areas. in
particular, a large portion of public support in the electricity
sector, which increased from €36 billion to €63 billion during
the period 2008–2012, went to low-carbon solutions. the costs
related to redispatching and balancing capacity procurement
and capacity mechanisms have grown in recent years. the
tendency to increase the nonenergy component of the electric-
ity retail prices acts to reduce the amount of competition in
end-user markets. the interactions among the different poli-
cies affecting the market price signals are being scrutinized
to help understand the consequences on the level of competi-
tion and business opportunities for the investors. the goal of
january/february 2019
such revisited policies is to remodulate public-support schemes
while making low-carbon technologies cost-effective.
a third, significant eu regulatory effort focuses on research
and innovation (r&i) (also financed by private investors)
to determine how smart and digital energy systems, which
embed sizable amounts of distributed energy resources, could
be deployed on a large scale. in the past 15 years, €5 billion
has been invested in roughly 1,000 r&i projects, testing novel
smarter electricity systems on the ground (Figure 2). invest-
ment in smart energy system r&i have steadily increased in the
eu, growing from approximately one-fifth in 2010 to approxi-
mately one-quarter of the total expenditure (€27  billion) in the
energy union-related r&i priorities in 2014.
Unbundling Model
Full Ownership Unbundling
Independent System Operator
Independent Transmission Operator
Mix
Not Applicable
The
figure 1. A transmission operation unbundling status.
january/february 2019
Electricity End-User and Retail Market Status
the distribution system is the electricity value chain’s crucial
sector, incubating most of the innovation. the dsos facili-
tate the retail market’s functions by connecting parties to
the distribution grid, enabling supplier switching, provid-
ing metering data from customers, and delivering allocation
and reconciliation services. Fewer than 10% of the dsos are
partially (i.e., functionally or legally) unbundled, while the
smaller distributors, serving fewer than 100,000 customers
or isolated systems, are exempt from any unbundling. new
and emerging actors, often power system business outsiders
(including municipalities, house corporations, telecommuni-
cations, high-tech/security companies, or social media), are
Ownership
FS Fully State Owned
MS Mostly State Owned
MP Mostly Privately Owned
FP Fully Privately Owned
ieee power & energy magazine 55
 moving into the distribution, retail, and end-user spheres as
a result of the various socioeconomic drivers.
the rising penetration of distributed generation brings
more voltage stability and congestion-management issues to
dso operations. today, flexible distribution resources rarely
participate in the tso-governed ancillary services markets
(asMs) and rarely intervene to solve local network prob-
lems; fragmented products and rules are used across mem-
ber states (Mss).
in europe, the customers’ price is considered “the sum of
the following three main components: the energy and supply
component, the network component (transmission and distri-
bution), and the component comprising taxes, levies, fees, and
charges” [cf. art. 2 of annex to the proposal directive of the
european parliament and of the council on common rules
for the internal Market in electricity (recast)]. the wholesale
market’s dynamics are transferred to the retail level mostly
through the energy component of the retail prices, and they
make up just one-third of average eu retail electricity prices.
the nonenergy components, i.e., network costs on one side
and taxes and levies on the other, make up the remaining two-
thirds. the recent wholesale price drop only slightly reduced
the retail price energy component, which, through 2015,
resulted in escalating prices for consumers due to mounting
nonenergy components (2016 was the first year in a decade in
which retail prices declined). recently, liberalized countries
(generally in southeastern europe) have exhibited more non-
price related features (i.e., linked to renewable/environmental
aspects) in their electricity retail offers and lower consumer
switch rates to alternative suppliers compared to that of coun-
tries with more mature wholesale market competition (gener-
ally in northwestern europe, as shown in Figure 3).
the eu pursues deregulation actions on final electricity
prices to foster market entry, innovation, and competition.
Household retail prices consistently followed the wholesale
dynamics in the few cases where more competitive/dynamic
retail market schemes were in place, whereas fixed, regulated
prices are still used in nearly half of the Mss. specific con-
sumer categories indicate a preference for fixed-price contracts
over a financial reward for modulating consumption.

20° 10° 0° 20°

Budget Applications
European Finland 10
Commission
100
Joint Research Center 2017,
Directorate C Energy, Transport Detail for
and Climate, Gangale, F., Vasijevska
J., Covrig, F., Mengolini, A., Fulli, Norway Estonia € Million <€10 Million
G., Smart grid projects outlook 2017:
facts, figures and trends in Europe.
Latvia 0
EUR 28614 EN, doi: 10276015583 Other
Sweden Integration of Large-Scale
Denmark Renewable Energy Sources
United Kingdom Lithuania E-Mobility
Integration of DG and Storage
Ireland
50° DSM
The Netherlands Smart Network Management
Belgium Germany Poland

Czech Republic
Luxembourg
Slovakia
Austria
Switzerland Hungary
Slovenia
Croatia Romania

France

40° Bulgaria
Italy

Portugal
Greece

Spain
Cyprus
0 250 500
km Malta 20° 20°
figure 2. The smart grid research and innovation projects—a geographical distribution of investment by domain. DG:
distributed generation. [From the Joint Research Centre (authors’ elaboration).] E-Mobility: electric mobility; DSM: demand
side management.

56 ieee power & energy magazine january/february 2019

 approximately 25% of european household consumers
are already equipped with electricity smart meters. only
roughly half of those devices, however, embed the ten func-
tionalities (including more frequent readings and support
for advanced tariff schemes) recommended by the ec. this
makes dynamic pricing mechanisms, demand response, and
energy-saving strategies difficult to implement on a large
scale. demand response and dynamic pricing promise to
bring new flexibility resources to power system balancing;
however, the debate continues regarding the real value as
well as the costs and the benefits for all of the perspective
actors in the current market settings.
consumer protection and empowerment rights are incor-
porated into national and regional legislation. With the goal of
protecting and engaging consumers, some eu Mss have begun
to 1) provide more information on price changes, bills, actual
consumption, and costs; 2) introduce single contact points and
energy consumer checklists; and 3) simplify the supplier switch-
ing processes. the benefits are not clear regarding the accumu-
lated information offers provided and process improvements,
and their impact on consumer responsiveness remain unclear.
consumer rights vary greatly across europe, with some
consumers having little or no choice of supplier. energy
poverty and vulnerability are unresolved problems; it is

Percentage of
Total Consumers
0/No Data
0.1–5
5–10

10–15
15–20
20–25
> 25

The

0 250 500 km

figure 3. The percentage of final electricity consumers by EU country changing suppliers. (Data courtesy of the Agency for
the Cooperation of European Energy Regulators, 2016.)

january/february 2019 ieee power & energy magazine 57

 The first example of a regionally integrated generation adequacy
assessment carried out in collaboration with the seven European
TSOs was published in 2015 by the Pentalateral Energy Forum.

estimated that one in five european household energy consum-
ers is classified as “vulnerable.” poorer households, as com-
pared to wealthier ones, have energy expenditures that claim
a higher share of their income. in some Mss, suppliers of last-
resort mechanisms, who seek to replace failing suppliers and
protect inactive or poor consumers, cover more expenditures
than is necessary for consumers, thus hampering their proactive
involvement in the retail market’s dynamics.
Electricity Wholesale Market Status
the ultimate goals of the eu energy legislation are electricity
market integration (which features convergent prices across the
eu), improved investments in low-carbon technologies and
flexible solutions, and more focused market-based governmen-
tal support. despite the progress of energy-sector regulation in
the past 20 years, many improvements are necessary to address
several issues. these include low liquidity and low prices for
transmission rights that forward markets continue to offer as
well as the iMs and balancing markets that are underdeveloped
and only offer a portion of the short-term adjustments needed
for greater penetration of variable renewable energy.
electricity wholesale prices in europe have been subject
to considerable volatility over the last decade, although not
as significantly as in other world economies (Figure 4). nev-
ertheless, current levels of electricity prices in the wholesale
markets in europe vary significantly across Mss (Figure 5),
indicating that the goal of long-term price convergence for
electricity remains distant.
part of the difficulties with the implementation of the fully
integrated european electricity market comes from the two
waves of “disruptive innovations” that occurred in the power
system in the last decades, e.g., the widespread use of renew-
able energy sources (ress) (in 2015, 27.5% of electricity in
the eu was generated from renewables, a 12.6% increase from
2005) as well as the use of smart energy system interactions to
decentralize the consumption–production loop. some exam-
ples of the role of regulation in adaptating to unforeseen devel-
opments and coordinating between european and national
policies can be seen at both the european and national levels.
one of the main obstacles to an integrated european whole-
sale market is the lack of coordination in addressing power sys-
tem adequacy (or reliability), i.e., the ability of the power sys-
tem to meet electricity demand 99% of the time, as mandated
by the european system operation network code. currently,
the official reports on generation adequacy assessment are pro-
duced by each Ms (with complete autonomy) with respect to

200
Japan EU-Platts PEP
180
United States Australia
160

140

120
(€/MWh)

100

80

60

40

20

1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7
2010 2011 2012 2013 2014 2015 2016
Source: PJM and ERCOT Markets in the United States, Platts in the EU, Japanese Power Market Exchange (Japan),
and AEMO (Australia)

figure 4. The wholesale electricity prices in Europe, United States, Japan, and Australia. [Adapted from the European
Commission staff working document “Energy Prices and Costs in Europe,” (COM 2016) 769 final.]

58 ieee power & energy magazine january/february 2019

the methodology, data sets, scenarios, and time frames for the
analysis. the first example of a regionally integrated genera-
tion adequacy assessment carried out in collaboration with the
seven european tsos was published in 2015 by the pentalateral
energy Forum (pleF). the study compares the levels of loss of
load expectation (lole) from 1 october 2015 to 30 september
2016 for two different cases (table 1). in the first case, the reli-
ability assessment is undertaken in a noncoordinated way by
each individual country, i.e., reducing the net transfer capacity
along all of the borders (light-shaded area). in the other case,
countries carry out the assessment in a coordinated manner
(dark-shaded area). the last column on the right reports the dif-
ference between the two cases.
the costs (in millions of euros) associated with the lole
level (h) for each Ms are assessed using the methodology pro-
posed by Wu et al., who run a stochastic long-term (ten years),
security-constrained, unit commitment model and assess the
costs associated with the four levels of reliability that are set
as constraints in the model, i.e., 24 h, 120 h, 240 h, and 360 h.
the cost of lole in each european country is computed by
multiplying the number of hours of lole resulting from the
pleF’s adequacy study by the associated cost range calculated

Euro/MWh
0/No Data

0.1–35
35–40

40–45
45–50
> 50

The

0 250 500 km

figure 5. The average annual wholesale electricity prices per country in 2017. (Data courtesy of the European Commis-
sion, “Directorate General for Energy, Data, and Methodology” adapted from the “Quarterly Report on European Electricity
Markets” report.)

january/february 2019 ieee power & energy magazine 59

 according to the methodology in Wu et al., divided by ten (i.e., european national regulatory authorities approved the proposal
the number of years of optimization). this cost is then adjusted under the commission’s capacity allocation and congestion
to the 2016 dollar-to-euro exchange rate. Management (cacM) regulation, (eu) 2015/1222, that was
there are several benefits of a coordinated approach. introduced by the 17 nominated electricity Market operators
the first is that the results of the reliability indicators can (neMos) committee. the agreement seeks to establish a
be compared across countries. Moreover, the reliability european market coupling operator to further integrate daM
indicators for each Ms are more accurate because they are and iM functions across europe, a strategy referred to as the
calculated using data from neighboring countries, i.e., the MCO plan. the plan’s goals are to increase liquidity by orga-
available capacity for generation and transmission as well as nizing trade between zones and establish clear and detailed
hydrological data, which is provided homogeneously by the rules among the neMos.
tsos or other collaborating national bodies. the coordinated national regulation, on the other hand, serves as a bridge for
approach produces a more realistic estimation of the costs the implementation of european policies and plays a funda-
of each level of reliability. these costs are generally lower mental role as a bottom-up driver the transition of national
than for the uncoordinated case because, when countries can power markets toward an integrated system. power exchanges
rely on the available generation capacity of their neighbors, are the operational bodies of the national regulators and tasked
investments in such capacity can be more efficient. with implementing technical market rules. an example of a
the widespread implementation of market coupling-solu- structured market is the italian power market exchange (ipeX),
tions for cross-border congestion management in europe is which is managed by gestore dei Mercati energetici and orga-
another example of regional coordination contributing to the nized in a multisession structure from the day ahead to the day
realization of the internal electricity market. at the moment, of delivery (Figure 7), where it simultaneously (i.e., implicitly)
22 european countries (including the united Kingdom) are manages both energy and transmission capacity. during each
coupled in different combinations for the daM (Figure 6) market session, operators can change their bids/offers to new
with the goal of improving market liquidity, investment allo- and more refined demand and renewable generation forecasts.
cation management, and price signals. coupled markets use this structure adds flexibility to the system and creates business
a common algorithm for settling market transactions, taking opportunities for market operators to provide system services.
into account the constraints of the interconnectors linking the
countries. therefore, we can conclude that the coordination of Forthcoming Policy and
national energy systems across the eu is beneficial from the Regulatory Challenges
standpoints of coordinating generation capacity investment and
coupling energy markets. Overarching Challenges and Trends
there is ongoing discussion on the benefits of a euro- regulators must properly allocate risks associated with low-
pean market coupling for cross-border iMs. in june 2017, all carbon innovation investments and guarantee cost recovery
through market prices. the invest-
table 1. A comparison of reliability performances ment volume from now until 2030
in the isolated and coordinated cases. is estimated at €90 billion yearly,
which comprises power generation
Single-Country Coordinated Benefits of
Assessment Assessment Coordinated (55%) and grids (45%). policy deci-
Isolated Case Interconnected Case Assessment sion makers must intervene to revise
Average Cost of LOLE Average Cost of LOLE
prices and regulate charges for elec-
Countries LOLE (h) (€ Million) LOLE (h) (€ Million) (€ Million) tricity delivery and related services
so that there is a level playing field
Belgium 177 48.06 0 0 48.06
for all resources. at the same time,
France 217 58.21 14 4 54.21 they must promote new public–pri-
Austria 0 0 0 vate coordination mechanisms and
generation- or demand-side public
Switzerland 1,251 330.11 0 330.11
support for low-carbon solutions that
Germany 0 0 0 are not competitive.
The Netherlands 0 0 0 to improve individual storage
and electric vehicle (eV) solutions
Luxembourg 8,760 2,311.53 0 2,311.53
(possibly combined with generation),
Total 2,747.9 4 2,743.9 tailored remuneration schemes and
[Adapted from the Joint Research Centre, “The Pentalateral Energy Forum Support Group ownership structures may be needed
2 Generation Adequacy Assessment,” (authors’ elaboration) and on the basis of the for these technologies. r&i proj-
methodology developed in Wu et al., “Costs of reliability analysis based on stochastic unit ects are analyzing the participation
commitment,” IEEE Transactions on Power Systems, vol. 23, no. 3, 2008.]
of storage in the balancing market

60 ieee power & energy magazine january/february 2019

(e.g., netfficient) and proposing a cloud-based flexibility man-
agement system with eVs and batteries (e.g., inVade).
the cep proposal for integrating electricity grids provides
the dso–tso cooperation rules that manage supply- and
demand-side flexibilities during different time scales. clarifi-
cation is needed on the dsos’ tasks and challenges related
to procuring services from resources (not necessarily owned
by them), such as distributed generation, demand response, or
storage and implementation of market-based energy-effi-
cient measures.
the cep also promotes changes in the markets along
the following:
✔ the local spatial dimension, which allows the participa-
tion of innovative service providers on a cross-national
scale to extend market coupling and spatial optimiza-
tion of investments

Trilateral Market Market Coupling of Europe

Coupling Market Coupling of Spain
and Portugal

(a) (b)

Market Coupling of Europe Market Coupling of Europe

4M Market Coupling 4M Market Coupling
Market Coupling “Italian Borders”

(c) (d)

figure 6. Maps of market coupling in European power markets for (a) 2006, (b) 2010, (c) 2014, and (d) 2015. The “Italian-
borders” coupling considers the whole energy market of bordering countries to define the actual allocation of energy and
capacity for each border. 4M: four markets (Czech, Slovak, Hungarian, and Romanian).

january/february 2019 ieee power & energy magazine 61

 ✔ the time dimensions (from long

injections and withdrawals from the grid in real time. The ASM consists of the ex-ante ASM and the balancing market. B: balancing market; IM: intraday market; R: results of
previous ex-ante ASM sessions. In the balancing market, the TSO accepts energy demand bids and supply offers to provide secondary control services and balance energy
00:00 a.m.
term to real time) for the renewable in-
tegration and optimization of market
11:00 p.m.

figure 7. The IPEX market sessions’ 2016 scheme. In B1 (Terna), the Italian TSO takes into consideration the valid bids/asks that the participants have submitted in the
10:00 p.m.
transactions and network operations.

In the balancing market, the TSO accepts energy demand bids and supply offers to provide secondary control services and balance energy injections and withdrawals
9:45 p.m.

8:00 p.m. With respect to digitization, cyberse-

7:00 p.m.
curity, and privacy (including elements

C*
such as distributed ledgers and artificial

R13-6
6:15 p.m.

5:00 p.m. intelligence), the cep advocates that all

4:45 p.m.
R7
consumers (as deemed necessary) should
3:00 p.m.
be equipped with modern smart meters

C*
that diligently protect their rights con-
R12-5
Day of Delivery

2:15 p.m.

1:00 p.m. cerning the processing and free movement

of personal data. this requires a careful
R6

11:45 a.m.

assessment of the costs and benefits.

In B1, the TSO takes into consideration the valid bids/asks that the participants have submitted in the previous ex-ante ASM session.
11:00 a.m.
C*

data protection and security, which in-

R11-4

10:15 a.m.

9:00 a.m. volve the technological, procedural, archi-

tectural, and behavioral aspects of custom-
R5

8:15 a.m.

ers’ personal and consumption information,

C*

7:00 a.m.

are receiving increased attention courtesy of

R10-3

6:15 a.m.

5:00 a.m. the eu’s general data protection regulation.

in this context, a few Mss linked their smart
R4

4:15 a.m.

metering roll-out plans to data protection

C*

3:00 a.m.

compliance. to secure smart grid opera-

R9-2

2:15 a.m.

1:00 a.m.
each market session. [From the GME (authors’ elaboration).] GME: Gestore dei Mercati Energetici. tions, several r&i projects have proposed
novel decision-support systems (sesaMe)
R3

00:15 a.m.

11:00 p.m. and innovative risk assessment and manage-

ment solutions (sparKs and segrid).
B2
B3
B4
B5
B6

10:30 p.m.

Various data-management models with

10:00 p.m.
R8-1

numerous ratios of traditional actors (par-

B1

9:45 p.m.

8:00 p.m.
ticularly tsos and dsos) and covering
different nonenergy systems and services
7:00 p.m.

are emerging across the eu. neutral data

6:00 p.m.
IM3
IM4
IM5
IM6
IM7

5:30 p.m.
brokers might arise as independent parties
R2

5:00 p.m.
Ex-ante ASM (Six Sessions)

4:30 p.m. feeding all players with the data required to

pursue their business. the cep clarifies the
One Day Before the Delivery

R1

3:30 p.m.

role of the parties responsible for data man-

The ASM consists of ex-ante ASM and balancing market.

3:00 p.m.

2:00 p.m. agement by introducing a common euro-

1:00 p.m. pean data format (some stakeholders suggest
making the existing national data models
IM1
IM2

12:55 p.m.
R1

12:00 Noon simply compatible rather than harmonized).

11:00 a.m. these new trends call for measures that tar-
10:00 a.m. get the data-management solutions’ scal-
9:00 a.m.
ability issues as they relate to iot and big
data applications.
8:00 a.m.
into/from the grid in real time.

several r&i projects, which often

7:00 a.m.

rely on the expanded common infor-

6:00 a.m.

5:00 a.m.

4:00 a.m.
mation model, study new digitization
3:00 a.m. solutions, including improved distribu-
2:00 a.m. tion grid observability and control solu-
1:00 a.m. tions (upgrid); a nondiscriminatory,
00:00 data-sharing platform for cross-eu data
exchange (FleXiciency); and a neutral
Nine Days

11:59 p.m.
Delivery

Note:
Before

market hub and common data platform for

DAM

8:00 a.m.
islanded energy management (upgrid
and netFFicient).

62 ieee power & energy magazine january/february 2019

 Flexibility may become the glue between the wholesale a significant influence on the design and management of the
and retail markets, and the overall benefits may grow when the electrical system. For this reason, strategies should be devel-
balancing and congestion-management tasks (for the most part oped between the different sectors to stimulate joint invest-
currently under the tsos’ control) are shared with the dsos ments and coherence in regulatory frameworks.
or other regulated and nonregulated market parties. demand several r&i projects include multiple electricity, gas,
response is one of these flexibility resources, and attention to water, heating and cooling systems, and different end-use
this technology is increasing in europe, although its deploy- sectors. some projects try to optimize heterogeneous energy
ment is scattered across the Mss. systems, including storage and renewable energy (smiles),
several r&i projects (see Figure 8 for the demand-side man- while others test integrated management platforms acting on
agement investments within the smart grid projects) investigate different energy carriers (pentagon).
new market schemes, including a third-party market operator
(sMartnet), the interplay of demand response with other Electricity End-User
flexibility resources (ecogrid and interFleX) and novel and Retail Market Challenges
prosumer’s flexibility concepts (FleX4grid). dynamic pricing entails passing a portion of the wholesale
the cep streamlines the procedure for expanding net- volatility to retail consumers. the cep entitles every cus-
work codes (and guidelines) from the transmission/wholesale tomer to a dynamic electricity price contract, which reflects
level down to the transmission/distribution interfaces and the “wholesale prices at intervals at least equal to the market
distribution/retail system. the standards for the future power settlement frequency.” dynamic pricing will only work if
grids are being developed based on the smart grids archi- wholesale markets become more efficient, detailed metered
tecture Model framework. data (ideally referring to each individual customer) are avail-
standardization and interoperability requirements are able, pricing intervals are small enough, and pricing options
tested in r&i projects on virtual power plant participation are carefully designed.
with wholesale/intraday markets (smartereMc2) and neigh- some players suggest not only intervening on the energy
borhood involvement in energy/flexibility markets (cooper- and supply component but also on the other two components
ate), whereas other projects (internet ppp and Finesce) (the network component and the component comprising taxes,
are demonstrating the value of open-source cloud platforms for levies, fees, and charges), which are not energy related, to make
the development of smart applications in multiple sectors. them more variable with the wholesale market dynamics.
necessary interactions between wider energy systems (such several r&i projects are studying how time-varying prices
as gas, heating, and cooling subsystems) and sectors (such as can shift consumption patterns in intelligently automated
information and communications technologies, transportation, households (linear) or how dynamic pricing can reduce
home automation, water, health, finances, and food) will have congestion problems and curtailment needs (goFleX).

250
Total
200 Private
206.9
Investment (€ Million)

150
158.1

144.3
140.4

137.2

133.2

100
106.7

37.3
85

82.2

31.1
78.9

26.8

25.5
73.8

73.2

50
70.6

15.1

14.2

14.8
13.8
24

12.4
54.2
42.1

48.8
43.5

6.8

7.1
28

0
Distribution System
Operator
ICT Company and
Telecom

University

Technology
Manufacturer

Consultancy

Research Center

Retail Company

Public Institution

Utility

Emerging Stakeholder

Other

Generation Company

Industry Association

Transmission System
Operator
Engineering Services

figure 8. The demand-side management investment in smart grid projects, per organizational type. ICT: information and
communications technology. (From the European Commission Joint Research Centre.)

january/february 2019 ieee power & energy magazine 63

 During times of constrained economic development, vulnerable
consumers—as opposed to the average household consumers—
end up being the most affected by price spikes.
the cep fosters a fair and cost-reflective consumer interac-
tion with the energy markets via improved billing information,
certified price-comparison tools, an easier choice of suppliers/
aggregators, the possibility to self-generate, the entitlement to
a dynamic price contract, and the installation of a smart meter.
consumer protection also entails defining clear conditions of ter-
mination, duration, price, and the conditions of switching suppli-
ers. decision makers will need to strike a balance between avoid-
ing situations in which simple consumers pay for the benefits of
the prosumers and that of rewarding active consumers/inves-
tors who deliver local/systemic benefits to the energy system
(i.e., via their flexibility potentials).
social acceptance issues as well as the perception of lim-
ited savings out of demand response practices may persist.
consumers must be protected from both current actors
and emerging aggregators, and the benefits/risks linked to
price volatility with dynamic pricing schemes need to
be properly explained. Fair and cost-reflective network
charges shall also be applied to users who do not support/
belong to the local energy community but are connected
to the grid operated by such a community. aspects to be
clarified include the definition of cost reflectiveness as well
as the calculation methodologies used for consumer net-
work charges and the supplier’s economic loss caused by
customers switching suppliers, the reward models used for
the active participation of citizens/communities in retail
markets, the consumption and generation of measurement
modalities, and the merits used for revisiting net-metering/
billing practices.
the cep reviews the “duly justified exceptions,” which are
typically linked to energy, poverty, and vulnerability, and allows
regulators to deviate from market-based retail prices. it also
makes the public service obligations that Mss may impose on
energy undertakings more transparent and nondiscriminatory.
during times of constrained economic development, vul-
nerable consumers—as opposed to the average household
consumers—end up being the most affected by price spikes.
consistent definitions of energy poverty and vulnerability are
still absent. eu Mss identify vulnerable consumers by several
characteristics that may overlap or extend well beyond their
relationships with the power system, including income, age,
health, social protection, and security. at times, this renders
consumer eligibility difficulty to ascertain.
solutions shall be implemented that enable consumers who
are vulnerable or experience energy poverty to participate in local
energy communities. Moreover, the effects of dynamic pricing
on certain household categories (i.e., large families or vulnerable
64 ieee power & energy magazine
consumers) need to be carefully assessed. several r&i proj-
ects address socioregulatory aspects by testing the consumer
response to information on renewable energy availability (e.g.,
nicegrid) and adopt social comparison strategies for individual
user engagement (e.g., intrepid and 3e-Houses).
local energy communities that manage distribution
resources, infrastructure, and services are defined by the
cep and open the door to many services and business oppor-
tunities. they differ from profit-driven energy companies in
their value-creation motivation and are entitled to own, estab-
lish, or lease community networks and autonomously man-
age them. they are expected to access all organized markets,
either directly or through aggregators or suppliers.
the regulatory challenge is to reconcile the development of
large energy communities with the reliable and secure opera-
tion of the power system. to this end, roles and responsibilities
of existing and emerging actors need to be clarified and recon-
ciled in a number of areas, including micro-grid management,
service quality, and flexibility services. How does one ensure
that all of the parties will bear a proportionate cost compared
to the benefits? an issue for debate involves the payment of
imbalance-induced compensation by the aggregators to the
suppliers, thus balancing all responsible parties or generators.
How small should those payments be to encourage the market
participation of demand response also through the aggrega-
tors? How big should they be to ensure that balancing costs
and benefits induced by the aggregators are fairly assigned to
market participants?
examples of r&i pilot studies include market-based energy
services and technological solutions for local energy commu-
nities and smart houses (eMpoWer and smartHouse/smart-
grid), schemes to help dsos operate their grid with higher
penetration of renewables, and interaction with commer-
cial aggregators (Wisegrid and ide4l).
Electricity Wholesale Market Challenges
the challenges that lie ahead for the european power sys-
tem, as addressed in the cep, have the wholesale electricity
market at the core of all the main policy objectives, such as
decarbonizing the eu energy system, taking stock of tech-
nology innovations, and further integrating the eu national/
regional energy markets while ensuring competition and
cooperation across borders at every relevant timescale.
While much progress has been made on power transmission
and the role of the tsos courtesy of the third energy pack-
age, completing the internal energy market remains one of
the top political priorities of the ec’s current agenda.
january/february 2019
 a nondistortive electricity market will be crucial not only
to guarantee the fair treatment of european consumers, pro-
ducers and prosumers but also to provide clear investment
signals, i.e., to install adequate generation capacity to meet
future demand and enough network capacity to ensure there
is sufficient transmission capacity to reduce price discrepan-
cies across all regions. in fact, the eu has chosen a “regional”
approach to european energy policies, i.e., most of the elec-
tricity system and market aspects are addressed via aggrega-
tions of bordering countries, named regions. an appropriate
and coherent governance of the interplay between regions,
control areas under the tsos’ responsibility, and scheduling
areas is needed to ensure coordination.
as mentioned previously, the impact of policy challenges
on the reliability of power systems is not yet clear since the
policy that encourages typically intermittent and decentral-
ized ress also increases the complexity of guaranteeing
system stability. additionally, specific markets that remuner-
ate reliability services (e.g., asMs, balancing, and reserves)
remain fairly underdeveloped in europe.
to guarantee adequacy throughout europe, the cep has
introduced new tools such as scarcity pricing and capacity-
remuneration mechanisms. scarcity pricing refers to the mecha-
nism used for price formation in electricity markets, especially
those closest to real time (e.g., balancing, reserves, and ancil-
lary services provisions). indeed, it might seem like a redundant
definition, since, in perfectly competitive markets, price-forma-
tion mechanisms are exclusively driven by demand and supply;
therefore, prices reflect conditions of scarcity of supply.
Forward markets and daMs are those in which the bulk
of electricity is traded, so the remuneration of investments for
many generators depends on the prices determined in these
market segments. as non-res generators will be less remu-
nerated in the daMs due to the market price decline (i.e., the
missing money problem), and as the share of ress in genera-
tion increase, the adequacy issue is expected to worsen.
the cep marks a mind-set shift in regarding temporary
price spikes in the future that may not be due to mere specu-
lation but, rather, due to scarcity pricing, i.e., a way to remu-
nerate the provision of electricity when the balance between
demand and supply is tight so that the investors who have
built plants that play a key role in “keeping the lights on” are
compensated enough to stay in the market.
as mentioned by the scholars who have addressed the
topic, continuous market monitoring must be ensured and is
provided for in the eu courtesy of the regulation on whole-
sale energy market integrity and transparency. scarcity pric-
ing is not enough to address the adequacy problem, however,
which is why the cep also provides for capacity remunera-
tion mechanisms, i.e., state subsidies to generators who play
a key role in keeping the power system running.
state subsidies tackle the longer-term perspective of ade-
quacy, but scarcity pricing in the short/medium term might
help in correcting unwanted distortions arising from capac-
ity remuneration mechanisms. in this sense, the eu has taken
january/february 2019
an open-minded approach in its cep package, which allows
the combination of both tools. nevertheless, given the unclear
organization of the future internal energy market, especially
for near real-time markets (i.e., intraday, balancing, reserves,
and ancillary services), in which a large variety of models cur-
rently exist across the eu, it is important to assess the interplay
among various tools and schemes in each specific situation.
the transition toward a single, integrated european mar-
ket (ieM) will surely help as well by allowing more and
more cross-border exchanges of balancing services. to guar-
antee nondiscrimination, all of these mechanisms should be
applied regardless of the country in which a generator is
located by considering its contribution to regional or conti-
nental power system stability.
an appropriate market design is likely to be the most
important factor in achieving a single european power mar-
ket in the future. the first logical steps are improving the
existing framework guidelines of the cacM regulation. the
cacM regulation represents a key moment for europe’s integra-
tion of electricity markets; however, the capacity calculation
and allocation principles described within it represent only the
first step and do not provide a univocal solution to europe’s
capacity-calculation problem.
one of the most important features of capacity calcula-
tions in europe is the choice of a zonal model. Zones can
be effective nodes (similar to those in italy) or countries.
redispatching/countertrading is carried out by tsos at the
national level at a considerable cost, especially for eu coun-
tries that chose the one-zone/one-country solution.
capacity calculation is strongly intertwined with energy
trading. the alternative locational marginal pricing for
energy, which also takes into account the available transmis-
sion capacity from/to each node, has been put on hold in favor
of an intermediate solution, such as zones. there are three
main reasons for using zones.
1) nodal pricing is computationally burdensome; there-
fore, it would be impossible to calculate the market
equilibrium for all of europe at the same time. this ar-
gument is refuted by the existence of nodal pricing in
large markets, such as the pjM in the united states. so,
in the future, it will certainly be possible to run a nodal
european market that manages all of the high-voltage
nodes in the grid.
2) nodal pricing and zonal pricing with small zones hinder
competition; therefore, generators have access to all of
the nodes/zones, but their electricity is “weighted” by the
network costs necessary to deliver it to each node. nodal
pricing will likely increase competitiveness by providing
optimal investment signals in the medium/long term.
3) optimal bidding zones are substantially equivalent to
nodal pricing. unfortunately, optimal bidding zones
do not exist and cannot be drawn beforehand. efficient
market zones must be time variant and rely only on
the pure physics of the transmission grid because the
power flow varies over time.
ieee power & energy magazine 65
 an appropriate capacity calculation that is consistent with
the features of the european transmission grid will allow the
expected benefits of transnational infrastructure projects—
e.g., the projects of common interest—to be assessed in a truly
impartial way.
Market liquidity is crucial to delivering electricity efficiently;
if generators, consumers, and prosumers do not take part in the
market, the price formation in that market will not reflect the real
conditions of the energy system. Because liquidity in eu power
markets is still relatively low, regulatory initiatives that ensure suf-
ficient participation in organized markets (versus over-the-counter
contracts) are fundamental to allowing power markets to deliver.
another missing piece of regulatory action that would help
achieve a fully efficient ieM concerns coordinating differ-
ent markets during different time frames. For the allocation
(market-based auctions for forward and daMs, continuous
trading for iMs) and entitlement methods to occur in markets
closer to real time, an effort toward substantial harmoniza-
tion may prove useful. such a lack of harmonization may give
rise to market manipulation by the most important operators,
which would optimize their profit arbitraging between mar-
kets, i.e., the relatively important generators may decide to
withdraw capacity from the daM to bid only on the more
lucrative, closer-to-real-time balancing and iM markets.
Regulation and Innovation
Interplay in Future Markets
if energy markets are to remain the instruments that enable
long-term policy and innovation initiatives, new regulatory
arrangements are needed to reconcile the security and com-
petitiveness of sustainable solutions. promising reforms are
outlined (e.g., in the electricity market time and spatial scales),
with the objective of properly distributing costs and benefits
among current and emerging actors. However, several aspects
and interfaces must be properly understood to govern the evo-
lution of the electricity delivery options and services. legisla-
tors and regulators must address a series of pressing questions.
1) How are business cases across interlinked energy and
nonenergy system configurations identified?
2) How is the balance between technology push and mar-
ket pull toward a fair energy transition designed?
3) How can private investors convert publicly supported
innovation into marketable products?
4) to what extent can emerging sociopolitical and tech-
nological trends subvert the wholesale transmission
and retail distribution boundaries and equilibria?
With this in mind, the eu and national decision makers are
making efforts to combine energy and climate change policy
actions with other proposals linked to, among others, digital
markets, circular economy innovation agendas, and capital
market/investment plans. large-scale and multidisci-
plinary pilots that target integrated architectures, interoper-
able applications, and harmonized standards are still needed
to test the merits and challenges of those potentially converg-
ing initiatives.
66 ieee power & energy magazine
a major regulatory challenge is to reconcile the stability
of the legal framework with the rapidity to react to the
pace of innovation. in future scenarios, predictably flexible
regulations should enable market players to assess the prof-
itability of investments through meaningful prices. Flexibil-
ity and resilience, as much as in the power system, surface
as must-have attributes in legislators’ quest for social wel-
fare maximization.
Acknowledgments
the authors are thankful to joint research centre colleagues
alexis Meletiou and Mircea ardelean for their comments
and map production.
For Further Reading
european commission. Market legislation. accessed on nov.
13, 2018. [online]. available: https://ec.europa.eu/energy/en/
topics/markets-and-consumers/market-legislation
agency for the cooperation of energy regulators. Mar-
ket monitoring. accessed on nov. 13, 2018. [online]. avail-
able: http://www.acer.europa.eu/en/electricity/market%20
monitoring/pages/default.aspx
european commission. (2016). clean energy for all eu-
ropeans proposal. [online]. available: https://ec.europa.eu/
energy/en/news/commission-proposes-new-rules-consumer-
centred-clean-energy-transition
g. Fulli, M. Masera, c. F. covrig, F. profumo, e. Bom-
pard, and t. Huang, “the eu electricity security decision-
analytic framework: status and perspective developments,
energies,” Energies, vol. 10, no. 4, p. 425, Mar. 2017. doi:
10.3390/en10040425.
e. Bompard, g. Fulli, M. ardelean, M. Masera, “it’s a bird,
it’s a plane, it’s a…supergrid!” IEEE Power Energy Mag., vol.
12, no. 2, pp. 40–50, Mar.–apr. 2014.
F. gangale, j. Vasiljevska, c. F. covrig, a. Mengolini, g.
Fulli, “smart grid projects outlook,” european commission,
luxembourg: publications office of the european union, 2017.
doi: 10.2760/701587.
j.-M. glachant and s. ruester, “the eu internal electric-
ity market: done forever?” Utilities Policy, vol. 30, pp. 1–7,
sept. 2014. doi: 10.1016/j.jup.2014.05.003.
l. Wu, M. shahidehpour, and t. li, “cost of reliability
analysis based on stochastic unit commitment,” IEEE Trans.
Power Syst., vol. 23, no. 8, pp. 1364–1374, 2008.
Biographies
Gianluca Fulli is with the european commission joint re-
search centre, ispra, italy.
Marcelo Masera is with the european commission joint
research centre, petten, the netherlands.
Amanda Spisto is with the european commission joint
research centre, petten, the netherlands.
Silvia Vitiello is with the european commission joint
research centre, ispra, italy.
p&e
january/february 2019
By Emma Nicholson and Arnie Quinn

I
In thIs artIcle, we gIve
our view of how the U.s. Federal
energy regulatory commission
(Ferc) typically approaches mar-
ket design issues in jurisdictional
wholesale electricity markets in
the United states and identify
three challenges that lie ahead for
these markets. the first challenge
is declining prices in wholesale
electric energy markets, which
can lead to revenue adequacy
issues for some resources. the
second challenge is to ensure that
wholesale electricity markets give
investors the proper incentives to
invest in resources that have the
operational flexibility needed to

©istockphoto.com/nazarkru

Wholesale Identifying Future

Challenges Facing

Electricity
Commercial Energy

Markets in the
United States
Digital Object Identifier 10.1109/MPE.2018.2872304
Date of publication: 7 January 2019

january/february 2019 1540-7977/19©2019IEEE ieee power & energy magazine 67

 One key advantage of RTO/ISO markets is the transparency
of market outcomes and the operational decisions that
drive those outcomes.
serve more variable net loads in the future. the third chal-
lenge is the interdependence between the natural gas and elec-
tric industries, which has implications for reliability.
FERC Background
Ferc has many responsibilities as a federal regulator, but the
most pertinent to this article is its role as the regulator of the
interstate transmission and sale of wholesale electricity for
resale. this role involves reviewing and approving the market
rules of six jurisdictional wholesale electricity markets in the
United states and ensuring that the resulting wholesale rates
in those markets are just and reasonable. Many areas related
to electricity service in the United states lie outside of Ferc’s
jurisdiction and within the purview of state public utility com-
missions; this includes, but is not limited to, the regulation of
retail electric and natural gas rates, the approval of the con-
struction of electric generation facilities, and overseeing reli-
ability issues related to electric distribution systems.
the six Ferc jurisdictional wholesale electricity mar-
kets in the United states are
✔ california Independent system Operator corporation
(caIsO)
✔ Midconti nent Independent system Operator, Inc.
(MIsO)
✔ IsO new england, Inc. (IsO-ne)
✔ new York Independent system Operator, Inc. (nYIsO)
✔ PJM Interconnection, llc (PJM)
✔ southwest Power Pool, Inc. (sPP).
these markets are operated by regional transmission orga-
nizations (rtOs) and independent system operators (IsOs).
Ferc approves the rules of the markets operated by the six
jurisdictional rtOs/IsOs, which include markets for energy,
ancillary services, capacity, and financial transmission rights.
all six operate energy, ancillary services, and financial trans-
mission rights markets, but only four (IsO-ne, MIsO, nYIsO,
and PJM) operate centralized capacity markets.
Ferc has historically promoted market-based outcomes
for wholesale electricity on the grounds that competition is
a means to create just and reasonable rates. For example, in
1996, Ferc issued Order 888, which promoted competition
through open-access nondiscriminatory transmission ser-
vices. In 1999, Ferc issued Order 2000, which advanced the
formation of rtOs/IsOs and codified their minimum char-
acteristics and functions. Ferc is also in a unique position
to identify market design issues across rtOs/IsOs, such as
barriers and inefficiencies that may arise with new resource
or technology types. For example, in February 2018, Ferc
68 ieee power & energy magazine
issued a rule to more effectively integrate electric storage
resources into rtO/IsO markets.
as former federal regulators, we are cognizant that
actions by Ferc to require market design changes will result
in changes to rtO/IsO priorities. a Ferc-mandated change
can crowd out or delay other market design changes. that
said, some market design changes require Ferc leadership
to be appropriately prioritized. For instance, market design
changes that have the potential to create clear winners and
losers (e.g., cost allocation issues) may not be amenable to
a stakeholder-driven process typically used to advise rtO/
IsO-sponsored market design changes. Ferc can play a role
in ensuring that such market design changes are adopted if
the status quo results in unjust or unreasonable rates.
a variety of factors are relevant when deciding where and
how to act. at the highest level, a market design ideally creates
incentives for resources to meet system needs during the oper-
ating horizon and invest in enhanced performance during the
planning horizon. Furthermore, one key advantage of rtO/
IsO markets is the transparency of market outcomes and the
operational decisions that drive those outcomes. Ideally, mar-
ket design changes enhance rather than introduce regulatory
certainty and remain robust under a variety of future condi-
tions. this can mean initiating action when a trend is emerg-
ing but before it has a material impact on market outcomes in
rtOs/IsOs. Finally, one needs to consider the complexity of
any proposed market design. rtO/IsO markets are inherently
complex, which has the potential to make it more difficult for
market participants to understand market outcomes and asso-
ciated value propositions. It can also make it more difficult for
rtOs/IsOs, regulators, and stakeholders to evaluate how well
a particular market design is functioning.
On the basis of our experience, generic actions directed
by Ferc are based on a methodical and evidence-based
approach to address rtO/IsO market design issues. we view
this as a three-step approach:
1) define a problem
2) develop and analyze record evidence
3) identify a just and reasonable solution.
Ferc often uses this approach when it proposes new rule-
making, which may result in a final rule, and when it investi-
gates market design issues in other proceedings.
the problem definition step is critical because it lays the
foundation for the inquiry that occurs in step two and the sub-
sequent solution in step three. the second step is to develop
a record of evidence that the commission will rely upon in
its decision making. rtO/IsO stakeholders (e.g., rtO/IsO
january/february 2019
management and staff, state governments, load-serving enti-
ties, resource owners, transmission owners, public power
systems, power marketers, consumer advocates, advocacy
groups, and trade associations) contribute to this record, often
based on targeted questions from Ferc and/or through par-
ticipation in technical conferences. the initial stages of the
record development process are designed to keep it focused
on the identified problem and identify potential solutions and
their associated tradeoffs. the analysis of the record devel-
oped in the initial stages of a proceeding is intended to deter-
mine whether the problem results in unjust and unreasonable
rates and, if so, to weigh the merits of the potential solutions
and select a proposed just and reasonable solution based on
the record evidence. It is important to note that Ferc must
follow an administrative process after identifying a proposed
solution to provide notice of a potential change. this admin-
istrative process serves to gather additional record evidence
that will support any final action to adopt the proposed solu-
tion. we believe this approach helps leverage the expertise of
market participants and other stakeholders to develop rtO/
IsO market rules that produce just and reasonable rates and
change or refine those rules if the record evidence suggest
that changes are warranted.
One consideration when evaluating a potential solution is
the degree to which that solution should be generic or provide
for regional flexibility. Ferc has long held that there can be
a variety of just and reasonable rtO/IsO market designs.
Furthermore, various regions have different resource mixes,
state regulatory structures, and institutional history that may
affect the need for a market design change and regional flex-
ibility in developing potential solutions. while regional flex-
ibility is often warranted, several factors can suggest when
more uniform market designs are appropriate, such as when
regional flexibility could create or exacerbate seams between
rtOs/IsOs or between an rtO/IsO and a non-rtO/IsO
region. In addition, uniform market designs have an advan-
tage relative to regional flexibility when one considers that
participants have to understand a variety of market designs
when regional flexibility is permitted.
Ferc’s price formation effort is a recent example of this
three-step approach (Docket aD14-14-000). In the price for-
mation effort, the commission examined price formation in
rtO/IsO energy and ancillary services markets in four areas:
1) the use of uplift payments
2) offer price mitigation and offer price caps
3) scarcity and shortage pricing
4) operator actions that affect price.
as part of this effort, in 2014, Ferc staff wrote four techni-
cal papers and held three technical conferences to discuss
these price formation issues.
to better define potential price formation problems in
rtO/IsO energy and ancillary services markets, Ferc issued
a request for comments after the technical conferences that
asked several detailed questions. Ferc also issued an order
directing rtOs/IsOs to submit reports about the pricing of
january/february 2019
fast-start resources, commitments to manage multiple contin-
gencies, look-ahead modeling, uplift allocation, and transpar-
ency. rtO/IsO stakeholders were also invited to comment on
these rtO/IsO reports. Ferc relied on the record developed
in the price formation proceeding to issue four proposed rules.
two of these rules have since become final (settlement inter-
vals and shortage pricing and offer caps). Both of the final rules
required uniform changes across the rtOs/IsOs while allow-
ing for some regional flexibility on implementation so that
changes were consistent with other marker design elements.
Ferc decided not to finalize a proposed rule on fast-start pric-
ing, choosing instead to permit regional variation. Ferc has
yet to act on the fourth proposed rule (uplift and transparency).
Our View of the Market Design
Challenges Ahead for RTOs/ISOs
the first challenge rtOs/IsOs in the United states will face in
the future is declining revenues in energy and ancillary services
markets because of low natural gas prices, a higher penetration
of resources that have low or zero marginal cost, and demand-
side investments to reduce peak load in response to high prices.
given these trends, locational marginal prices (lMPs) in some
rtOs/IsOs are at historic lows when viewed on an average
annual basis. Figure 1, prepared by the U.s. Department of
energy, plots electricity prices in both rtOs/IsOs and non-
rtO/IsO regions and shows the extent to which these prices
have fallen in recent years. It also compares those prices to the
natural gas price at henry hub (an important national bench-
mark for the U.s. natural gas market).
low wholesale energy prices are good for the U.s. econ-
omy, and rtO/IsO markets have been successful in translating
these energy market trends into savings for electricity custom-
ers in their respective footprints. however, these trends are
likely to change the way that some resources recover their
costs because, in the past, many resources recovered a por-
tion of their costs through inframarginal rents in rtO/IsO
energy markets. these inframarginal rents have declined over
time as lMPs have declined. Other revenue streams—such as
ancillary services and capacity market revenues—may need
to play a larger role in ensuring that the resources needed to
provide reliable electric service to load have the proper incen-
tives to provide this service. a challenge in addressing this
issue is differentiating between market fundamentals and
underlying concerns with rtO/IsO market designs in a low
lMP environment.
It is often instructive to examine rtO/IsO market design
issues in a mechanism design framework, which requires a clear
objective and a set of payments and penalties designed to maxi-
mize that objective in light of market participant incentives. For
example, in IsO-ne, MIsO, nYIsO, and PJM, capacity mar-
ket revenues are designed to recover the going-forward costs
of the resources needed to provide resource adequacy net of
energy and ancillary services revenues. accordingly, if lMPs
decline, capacity prices in these rtOs/IsOs are generally
expected to increase. however, a fixed annual payment in the
ieee power & energy magazine 69
 capacity market may mute the incentives resources have to per-
form during the delivery period (typically one calendar year). It
may be appropriate to revise resource compensation, add new
performance requirements, or develop new rtO/IsO products
to enhance the performance signal.
a second challenge rtOs/IsOs in the United states will
face is the need to replace installed generation capacity that
retires with new generation assets with the flexible operat-
ing characteristics needed to serve net load (i.e., load minus
distributed resource output) at least cost. this turnover could
be quite considerable in rtOs/IsOs that have a significant
amount of capacity at or near retirement age.
net load is expected to be more volatile in the future as
rtOs/IsOs integrate more distributed energy resources and
loads invest in equipment that enables them to be more price
responsive to wholesale electricity prices. the resulting net
load profiles will likely change the optimal mix of resources
that rtOs/IsOs will need to meet their respective reserve
requirements and balance the system in real time. For exam-
ple, as shown in Figure 2, wind resources were the marginal
resource, and thus set the energy price, in the sPP about 10%
of the time, on average.
while lower load growth and more price-responsive
loads are likely to reduce expected peak load, which will
reduce the capacity needed for resource adequacy planning
purposes, new generation assets will be needed, and some
of that capacity will have to be more flexible. resource flex-
ibility can be measured in several dimensions, including the
ability to start up quickly (e.g., a fast-start unit that can start
up within 10 min), the ability to get offline quickly (i.e., short
minimum run time), a quick ramp rate (having a fairly wide
140
120
Electricity Prices (US$/MWh)
100
80
60
40
20
0 0
2002 2004 2006 2008 2010 2012 2014 2016
figure 1. The average wholesale electric costs between 2002 and 2016. (Source: U.S. Department of Energy, “Staff
Report to the Secretary on Electricity Markets and Reliability,” Figure 6.2.)
70 ieee power & energy magazine
dispatchable range), or having a low or zero minimum operat-
ing level. a single resource need not possess each dimension
of flexibility. In the future, rtOs/IsOs will need to design
markets to ensure that sufficient investments occurs and that
they yield the rtO/IsO’s desired levels of flexibility.
some rtOs/IsOs have market designs that explicitly
procure and value resource flexibility. For example, the flex-
ible resource adequacy requirement in caIsO, which is
necessitated by the california Public Utilities commission’s
resource adequacy program, specifies that some capacity
resources have flexible capability. additionally, both caIsO
and MIsO explicitly value flexibility within their energy
markets; MIsO has a ramp capability product, and caIsO
has a flexible ramping product. these products are designed
to procure ramp capability based on the system’s expected
needs in future intervals. without a ramping product, rtOs/
IsOs may need to take out-of-market actions to posture
resources available to provide ramping capability in future
periods. the creation of a ramping product is one way to
move out-of-market actions into the market and reflect the
system need in prices.
a future challenge will be determining which market
(energy, ancillary services, or capacity, or some combina-
tion thereof) should provide the incentives for and compen-
sate the resource attributes that will be needed in the future.
an important consideration is whether short-term compen-
sation for flexibility will be sufficient to drive long-term
investments that enhance a resource’s flexibility. If multiple
markets are needed, market design changes may be needed
to improve coordination in such markets to avoid over-
or undercompensation.
New
England
Midwest New York
14 West
Mid-Atlantic
CAISO+ Central
12
Natural Gas Price (US$/MMBtu)
Southeast
10
ERCOT
8
6 New England (ISO-NE Mass Hub)
New York (NYISO NYC Zone J)
4 Mid-Atlantic (PJM Western Hub)
Midwest (MISO Indiana Hub)
2 Central (SPP South Hub)
ERCOT (ERCOT North Hub)
CAISO+ (CAISO SP15)
Northwest (Mid-Columbia)
Henry Hub Natural Gas Price
january/february 2019
 100

80
Intervals on the Margin (%)

60

40

20

0
Dec. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Jan. Feb. 2016 2017 2018
2016 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017 2018 2018 Winter

Other Gas, Simple Cycle Gas, Combined Cycle Coal Wind

figure 2. The SPP generation technology on the margin in the real-time electricity market. (Source: SPP Market Monitor-
ing Unit, “Winter 2018 State of the Market Report,” Figure 3-7, used with permission.)

a third challenge facing U.s. rtO/IsO markets today and

in the future is coordinating the operation of the interstate 1,600
natural gas pipeline system with the needs of natural gas- 1,400
Terawatt Hours (TWh)

fired generators, known as gas-electric coordination. this 1,200

issue will become increasingly important as the generation 1,000
fleet in rtOs/IsOs shifts toward natural gas-fired generation
800
resources. Figure 3 plots the U.s. energy Information admin-
600
istration’s reference case forecast of net generation from coal,
400
natural gas, nuclear, and renewable generation resources. nat- Coal Natural Gas
200 Nuclear Renewable
ural gas generation is expected to increase significantly.
Poor gas-electric coordination can place stress on certain 0
16
19
22
25
28
31
34
37
40
43
46
rtO/IsO markets in the northeastern United states, particu- 49
20
20
20
20
20
20
20
20
20
20
20
20

larly during extreme cold winter weather. natural gas gen-

erators in rtOs/IsOs often rely on interruptible natural gas
transportation delivery contracts as opposed to firm delivery
ones because sufficient capacity is typically available to serve
the needs of both a firm and the interruptible customers on
interstate pipelines. however, during the winter months, when
natural gas supplies are needed to serve both thermal heating
loads and natural gas generators, the interstate natural pipeline
system can become constrained and unable to meet demand
beyond firm commitments, which can threaten an rtO/IsO
market’s ability to provide reliable electric service. as shown
in Figure 4, the north american electric reliability corpora-
tion identified clusters of natural gas generators that would be
susceptible to a natural gas supply disruption.
Ferc has been investigating gas-electric coordination issues
since 2012 when Ferc held a technical conference on the mat-
ter (Docket aD12-12-000). after examining the record devel-
oped in the gas-electric coordination proceeding, Ferc issued
Order 809 in april 2015 that better coordinated the scheduling
figure 3. The U.S. Energy Information Administration’s 2018
annual energy outlook reference case forecast of net genera-
tion from coal, natural gas, nuclear, and renewable resourc-
es. (Source: U.S. Department of Energy, “Energy Information
Administration, 2018 Annual Energy Outlook, Reference
Case Electricity Supply, Disposition, Prices, and Emissions.”)
processes on interstate natural gas pipelines and public utilities
(Docket rM14-2-000). to some extent, gas-electric coordina-
tion presents many of the same issues as resource flexibility.
when concerns about the availability of natural gas are most
acute, rtOs/IsOs may take out-of-market action to be pre-
pared to respond if natural gas-fired resources are not avail-
able. Ideally, these out-of-market actions would be moved into
the market, for instance through performance penalties or new
reserve products. like flexibility, an important consideration is
what form of compensation will be sufficient to drive long-term

january/february 2019 ieee power & energy magazine 71

 Northwest New England
25 GW, 2016 11 GW, 2016
4 GW, Planned 5 GW, Planned

Mid-Atlantic
≤50 MW 96 GW, 2016
>50 MW 50 GW, Planned
>100 MW
>200 MW
>400 MW

South CA-AZ East TX-OK-LA Southeast Florida

60 GW, 2016 60 GW, 2016 55 GW, 2016 38 GW, 2016
17 GW, Planned 17 GW, Planned 7 GW, Planned 5 GW, Planned

figure 4. The North American Electric Reliability Corporation-identified natural gas generation clusters of at least 2 GW
that are vulnerable to a natural gas supply disruption. (Source: North American Electric Reliability Corporation, “Special
Reliability Assessment: Potential Bulk Power System Impacts due to Severe Disruptions on the Natural Gas System,”
Figure 5.1. This information from the North American Electric Reliability Corporation’s website is the property of the
North American Electric Reliability Corporation.)

investments that match the value consumers place on avoiding
the operational or cost concerns associated with the lack of natu-
ral gas-fired generation. resolving this issue will also require
Ferc, rtOs/IsOs, and other stakeholders to develop solutions
that both preserve reliability while at the same time minimizing
the costs to consumers in rtOs/IsOs.
Acknowledgments
the opinions and views offered here are those of the authors,
and not necessarily those of the United states, the Federal
energy regulatory commission, the individual commis-
sioners, or members of the commission staff.
For Further Reading
Federal energy regulatory commission, “electric storage par-
ticipation in markets operated by regional transmission organiza-
tions and independent system operators,” Federal energy regu-
latory commission, washington, D.c., Order 841, Feb. 15, 2018.
U.s. Department of energy, “staff report to the secretary
on electricity markets and reliability,” U.s. Department of
energy, washington, D.c., aug. 2017. [Online]. avail-
able: https://www.energy.gov/sites/prod/files/2017/08/f36/
staff%20report%20on%20electricity%20Markets%20
and%20reliability_0.pdf
energy Information administration, “2018 annual energy
outlook,” U.s. Department of energy, washington, D.c., Feb.
2018. [Online]. available: http://www.eia.gov/outlooks/aeo
north american electric reliability corporation, “spe-
cial reliability assessment: Potential bulk power system im-
pacts due to severe disruptions on the natural gas system,”
north american electric reliability corporation, atlanta,
ga, nov. 2017. [Online]. available: httpps://www.nerc.com/
pa/raPa/ra/reliability%20assessments%20Dl/nerc_
sPOD_11142017_Final.pdf
Biographies
Emma Nicholson is with the concentric energy advisors,
Inc., washington, D.c.
Arnie Quinn is with vistra energy, washington, D.c.
p&e

72 ieee power & energy magazine january/february 2019

 Demand-Side,
Renewable,
Economic and Distributed
Efficiency and Resources
Debunking Incremental
the Natural Improvements
Monopoly Myth

By Benjamin F. Hobbs
and Shmuel S. Oren
©istockphoto.com/siberianart

Three Waves of
U.S. Reforms
N
New geNeratioN, storage, demaNd maNagemeNt, aNd smart
Following grid technologies have the potential to make the grid of 2050 very different from
today’s power system. Can market designers keep up, and what do they need
the Path of to pay attention to now? Before offering a few answers to these questions, we first
give some historical perspective on the creation of markets for electric power and
Wholesale how those markets have been adapted to new needs.
the power industry is now undergoing dramatic changes, following many
Electricity transitions that we have already experienced. in the 1960s, nearly all power was
provided by vertically integrated utilities. they confidently planned ever-larger
Market coal and nuclear generators (including floating nuclear plants), whose prudently
incurred costs would be guaranteed by ratepayers. at least one industry group
Restructuring concluded that its greatest environmental challenge was the ugliness of overhead
distribution lines; climate change meant global cooling and was irrelevant to plan-
ners and regulators.
all of that has now changed beyond recognition in most (but not all) of the
world. Future changes in the industry will differ in type but not in magnitude
Digital Object Identifier 10.1109/MPE.2018.2873952
from those past changes. if market designs adhere to the principles of open access,
Date of publication: 7 January 2019 technology neutrality, recognition of network realities, and financial rewards for

january/february 2019 1540-7977/19©2019IEEE ieee power & energy magazine 73

 providing what the system really needs when and where it
needs it, then the power system will be well placed to adapt
to whatever technology and social changes are coming.
The First Wave of Reforms:
Unbundling and Debunking the
Natural Monopoly Myth
the first introduction of competition into the electric power
sector of the U.s. economy is often attributed to the 1978
Public Utilities regulatory Policy act (PUrPa). that act
required electric utilities, which were most commonly verti-
cally integrated monopolies within their service territories, to
buy power produced by small (lower than 50 mw) and renew-
able power sources.
PUrPa came about at a time when washington, d.C., was
abuzz about deregulation. economist Fred Kahn of Cornell,
who was President Jimmy Carter’s chief economic advisor,
is largely credited with first promoting the liberalization of
markets away from heavily regulated or publicly owned firms.
the basic idea was that freer markets could spur companies to
offer a wider range of products and lower costs. Policy makers
deregulated and restructured railroads, telecommunications,
air travel, shipping, securities markets, and gas and water utili-
ties. in the 1990s, policy makers, consumers, traders, and aca-
demics sought to bring the same revolution to electric power.
However, PUrPa and the subsequent reforms of the
1990s were not the first time that competition was intro-
duced to power provision. Britain’s Central electricity Board
created an auction in the 1930s in which power was con-
tracted from the providers who could offer it most efficiently,
contributing to a halving of heat rates in that decade. in the
United states, despite samuel insull’s success in promoting
the regulatory compact between monopoly power providers
and states that agreed to regulate them, distribution-level
competition was not completely eliminated. in some cities,
rival distribution companies ran lines in parallel to poach
each other’s customers; there is some evidence that lower
rates resulted where there was such competition. although
post–world war ii nationalization displaced Britain’s com-
petitive market and distribution-level competition eventually
disappeared in the United states, these examples were an
inspiration to U.s. reformers in the 1990s.
Prior to the 1990s, free-market ideology in the United King-
dom and Chile led to the first wholesale restructurings of the
power sector, unburdened by any felt need for careful analy-
sis of benefits and costs. in British Prime minister margaret
thatcher’s view, there was simply “no other alternative,” given,
e.g., how the nationalized industry failed to keep the lights on
during the coal strikes earlier in the 1970s. at the same time, in
the United states, academics were establishing the intellectual
underpinnings for restructuring. Fred schweppe of the massa-
chusetts institute of technology (mit) and his students math-
ematically described how prices could vary over time and from
bus to bus based on the instantaneous marginal cost of deliver-
ing power and how such spot prices could incentivize building
74 ieee power & energy magazine
the right amount, type, and place of generation capacity and
even transmission (schweppe et al., 1988). a great irony is that
schweppe first proposed that this be used to motivate consum-
ers to modify their behavior; only later was this idea extended
to supply. the irony is that, in today’s markets, the demand
side of the market remains largely insulated from prices that
reflect actual system conditions in time and space. mean-
while, schweppe’s mit colleagues Paul Joskow and richard
schmalensee (1983) described how institutions could be
defined to promote competition in distribution, transmission,
and especially generation.
But it was not the earlier examples of competition or the
intellectual frameworks proposed by schweppe et al. that
were chiefly responsible for spurring the U.s. restructuring
efforts of the 1990s. instead, the large and obvious decline
in the cost of rail shipments, natural gas, telephone calls,
and air travel was a major inspiration. another driver was
that large customers chafed under rising electric rates occur-
ring in some jurisdictions due to generation construction cost
overruns; they demanded access to the cheaper power being
sold in other areas or produced by efficient and quickly built
plants burning now-cheap natural gas. rates in neighboring
states sometimes differed by a factor of two. in some areas,
the difference between the low marginal costs of producing
power by gas and the high embedded cost rates of vertically
integrated supply created a strong incentive for consumers to
access cheaper power.
the 1990s witnessed the debunking of a long-standing
myth that the electricity industry is a natural monopoly, with
competition being introduced in many jurisdictions in two
forms. one was PUrPa-style competition among indepen-
dent generators to sell to utilities at a regulated avoided cost-
based price. to the extent that these avoided costs accurately
reflected the cost of utility-based supply, this type of compe-
tition reduced the cost of supply. However, regulators often
overestimated avoided costs, which sometimes resulted in
overpayment and expansion of uneconomical supply, which,
in turn, inflated electric rates. this increased pressure for a
second type of competition: access by consumers to cheaper
generation, facilitated by marketers.
at first, this access was through wheeling arrangements.
But to spread widely the benefits of access, many states turned
to a restructuring framework in which distribution, transmis-
sion, and generation were unbundled. generation assets
were then to be spun off to independent supply companies that
would then compete to sell power to consumers, using the grid
as a common carrier. this second type of competition also did
not necessarily decrease the overall societal costs of power.
it has been argued that it was primarily an effort by some
customers to shift the fixed costs of stranded assets (such as
uneconomical PUrPa contracts or utility generation whose
construction costs were underestimated) either to other cus-
tomers or investors (Borenstein and Bushnell, 2015). today,
similar efforts by consumers to shift costs to other parties
by avoiding average cost–based rates are continuing to drive
january/february 2019
 Today, similar efforts by consumers to shift costs to other parties by
avoiding average cost–based rates are continuing to drive radical
changes in the industry.
radical changes in the industry—but in a very different way,
as we explain later.
Genesis of the Second Wave
in the 1990s, the enthusiasm of many states for restructuring
was fed by overestimated savings from competition; opti-
mistic projections of the competitiveness of new, cleaner
technology; and the prospect that the some of the benefits
of competition could be tapped for popular programs sub-
sidizing energy efficiency and renewable technology. the
belief that restructuring was a cornucopia to benefit every-
one was epitomized by the unanimous vote of the Califor-
nia assembly in favor of assembly Bill 1890 in 1996, the
electricity restructuring act. other states quickly followed
suit. within the next few years, nearly half of the states had
passed laws or instituted regulations to restructure the power
industry. many of those states intended to allow retail access
for all customers, optimistically anticipating wide participa-
tion and dramatic cost savings. these state actions were fol-
lowed in many cases, most notably in California, by wide-
spread voluntary or forced divestment of generation assets
by the former monopoly utilities.
Unfortunately, if such a dramatic change to an industry
is favored by everyone, at least some people must be overly
optimistic about the benefits they will receive. any such
change, even if it improves efficiency and lowers prices, will
inevitably produce some losers as well as winners, which, in
fact, became apparent later. Furthermore, as many industry
writers warned at the time, the technical intricacies of power
systems meant that creating a market for power would not
be nearly as easy as for much simpler systems. the inspira-
tion provided by the successful creation or restructuring of
markets for simple commodities, such as the 1990 Clean
air act’s continent-scale pollution-trading scheme, was, in
part, misleading.
at the same time that states were acting, the Federal
energy regulatory Commission (FerC) also moved the
restructuring agenda forward. in april 1996, it adopted order
888, which was intended to “remove impediments to competi-
tion in the wholesale bulk power marketplace and to bring more
efficient, lower cost power to the Nation’s electricity consum-
ers.” all public utilities that owned assets transmitting electric-
ity in interstate commerce had to allow market parties to use
their facilities on a nondiscriminatory basis, based on published
transmission tariffs. making the high-voltage grid available
to all comers would, it was hoped, facilitate the states’ retail
markets and give access to the independent generators that
january/february 2019
the states facilitated through their unbundling and divestment
actions. in december 1999, FerC issued order 2000, which
explicitly encourages the establishment of regional transmis-
sion organizations (rtos) to facilitate markets and further the
goals of order 888. some of those rtos have since become
full-fledged independent system operators (isos) that oper-
ate day-ahead and real-time electricity spot markets, although
many parts of the country honor order 888 simply by com-
plying with rules requiring nondiscriminatory access to their
transmission facilities.
The First Wave Disappoints:
Genesis of the Second Wave
The California Crisis
the 2000–2001 California crisis, however, stopped state
restructuring efforts in their tracks. the causes of the crisis
have been chronicled and debated extensively in this maga-
zine. we believe that the problems arose, first, from an exer-
cise of market power right out of an economics 101 class—
the withdrawal of supply to push up prices. this exercise was
facilitated by supply problems, such as drought, a natural gas
compressor explosion, and kelp sucked into a nuclear station’s
cooling system and was also enabled by state regulators who
actively discouraged forward contracting.
second, the market design also disregarded important phys-
ical characteristics of electricity. these naïve designs included
the definition of uniform pricing zones in the day-ahead market
that ignored local transmission congestion. as a result, opera-
tors had to scramble to resolve congestion in real time, endan-
gering reliability and exacerbating local market power.
another issue was the creation of isolated day-ahead and
real-time markets for energy and operating reserves that over-
looked their strong interconnections. enron and other mar-
ket parties resorted to disguising actions or outright lying to
arbitrage between those markets. such arbitrage is, in fact, an
essential function in all other markets and arguably improved
rather than harmed power market functioning at the time.
there was also the failure to promote forward contract-
ing, which enabled any exercise of market power to immedi-
ately benefit the whole portfolio of a company’s generators.
the consequences of poor market designs and market power
were unambiguous: rolling blackouts, a threefold increase in
wholesale power costs, and loss of public confidence in the
ability of restructuring to deliver lower costs. since the cri-
sis, no additional states have joined the original 23 to allow
for retail choice in their electricity markets, and, of those 23,
ieee power & energy magazine 75
 eight have suspended such choice. of the states that origi-
nally encouraged retail markets, only texas has a vibrant
and active retail markets for all consumers: industrial, com-
mercial, and residential.
despite the crisis, FerC gamely pressed on. after an unsuc-
cessful attempt to develop a standard market design to impose
on all restructured wholesale markets, a voluntary wholesale
market “platform” was promoted, with elements that FerC
believed would prevent California-style problems from occur-
ring in the future. these included location-specific (nodal)
prices that reflected consideration of the network’s actual con-
straints, before-the-market identification and mitigation of
bids that appear to attempt to exercise market power, forward
financial contracts and resource-adequacy markets, financial
transmission rights (Ftrs), cooptimized spot energy and
operating reserves markets, financial arbitrage between day-
ahead and real-time markets, and a hybrid market that oper-
ates like a so-called power pool company (PooLCo) but also
facilitates bilateral transactions.
a PooLCo in a particular region is defined as a single
market operator that would run an auction to buy power from
generators and then resell it to consumers or load-serving
entities. PooLCos can also enable bilateral deals to be made
between generators and consumers or load-serving entities
by moving the power from the former to the latter and charg-
ing for the short-term use of the network based on the differ-
ences in the energy spot prices it calculates at the locations
of sellers and buyers. this allows trading to take place, while
accounting for all of the physical constraints in the network
that must be satisfied.
Lessons Learned from the First Wave
to summarize, the first wave of restructuring was driven by a
global move toward privatization of infrastructure industries,
vertical and horizontal unbundling, and implementation of
competitive wholesale markets as means for achieving eco-
nomic efficiency in investment, production, and consump-
tion. we now describe several of the lessons learned from the
first wave of restructuring and the California crisis, lessons
that are, more or less, embodied in FerC’s platform and in
the market redesigns actively promoted in PJm interconnec-
tion, California, and the other U.s. isos.
one lesson concerned how to unbundle and then distribute
the resulting economic benefits and costs. Not everyone auto-
matically becomes better off. the initial main challenges
were the divestiture of assets and recovery of the stranded
costs resulting from the regulatory compact. some thought
that electricity under deregulation would be much cheaper,
which would, therefore, significantly depreciate the value of
the divested assets. in California, the Competitive transition
Charge mechanism was designed to make the shareholders
of the investor-owned utilities whole. But the arrangement
was asymmetric and did not protect customers against wind-
fall profits for the asset buyers due to unexpected events and
the exercise of market power. No one thought of protecting
76 ieee power & energy magazine
consumers’ stranded benefits (earned through depreciation
payments embedded in the regulated rates) through vesting
call options (imposing bid caps) on divested plants. such vest-
ing contracts were imposed, e.g., in the earlier U.K. divesti-
ture. the managers of future restructuring, which may even-
tually come to the northwestern and southeastern parts of the
United states that have resisted it, should consider smoothing
the transition through such contracts.
another lesson was the need for consumers to forward-
contract for supply to prevent the exercise of market power.
the California regulators discouraged long-term contracting
between generators and load-serving entities, believing that
clearing all electricity in the day-ahead spot market would
eliminate the need for regulatory intervention in price setting.
after the crisis, California regulators reversed course and
instead mandated month- and year-ahead forward contract-
ing and showings of adequate physical resources. FerC, in
turn, has encouraged, but not mandated, so-called resource-
adequacy mechanisms in part to give more assurance that
enough supply and demand resources would be available to
prevent shortages. after the crisis, these mechanisms helped
assure nervous regulators that another California blowup
would not happen and provided a source for missing money
for resource investors who would not earn adequate returns
from price-capped energy markets alone.
the adopted approaches, however, differ in different
areas. PJm and iso New england have capacity markets in
which the isos procure forward capacity (subject to specific
must-offer obligations) through annual auctions that allow
participation by existing and planned capacity as well as
demand-side resources willing to curtail load. texas, how-
ever, has opted for an energy-only market, with spot prices
allowed to reach Us$9,000/mwh. it has also complemented
the raising of price caps with an administrative operating
reserve demand curve that imposes a price adder on energy
to reflect reserve depletion, so that energy prices better reflect
scarcity rents in times of shortage.
a third lesson was learned when both state regulators and
FerC underestimated the market power and other abuses
that could take advantage of inconsistencies between the
commercial model (e.g., uniform price zones and separate
energy and ancillary service markets) and the technical real-
ity. FerC authorization of market-based wholesale rates
was based on standard department of Justice criteria—
e.g., the Herfindahl–Hirschman concentration index—that
proved to be inadequate for the electric power industry. the
commercial models initially implemented in California and
texas emphasized trading simplicity, as is still the case in
europe. the California crisis and widespread gaming (par-
ticularly the so-called “dec game”) in PJm, California, and
texas demonstrated the need for consistency between the
commercial and technical models, which eventually led to
further reforms of these markets, such as California’s mar-
ket restructuring and technical upgrade and the texas nodal
market. (in the dec game, a generator in a location with
january/february 2019
excess power would schedule a power sale into a zonal day-
ahead market at a regional high price that disregards conges-
tion and then buy the power back in real time at a lower price
that reflects local congestion.)
in particular, nodal markets, whose locational marginal
prices would be determined through uniform price-clearing
auctions, were instituted in both the day-ahead and real-time
markets. initially, real-time markets were hourly, but to address
the need for greater flexibility in response to renewable output
variations, FerC order 764 reduced the settlement intervals to
15-min markets, with reoptimization of dispatch every 5 min.
Contrary to fears that having separate prices at each bus would
increase the ability to manipulate prices, the markets diminished
opportunities to exercise market power by making constraints
transparent and managing congestion day ahead. However,
these measures are not enough to eliminate market power in
load pockets with limited supply and transmission bottlenecks.
From the California crisis, we learned that there was a
need for active market monitoring together with automated
detection and mitigation of inflated supply bids that attempt
to exercise market power in local markets. trading energy at
nodal prices has provided accurate price signals reflecting con-
gestion that facilitate efficiency in production, consumption,
and investment. However, it quickly became apparent that the
variability and uncertainty of such locational prices exposes
market participants to high risks. thus, another lesson was
that there was a need for financial hedges for these risks. Bill
Hogan of Harvard complemented schweppe’s vision of nodal
energy markets with a proposal that isos create hedging and
property rights instruments in the form of Ftrs. these are
financial instruments that entitle their holders to the net gain
resulting from nodal price difference between two specific
nodes. they allow energy market participants and transmis-
sion owners to hedge the locational price risk associated with
energy trading and provide means for monetizing some of the
property rights resulting from the ownership of transmission
assets that restrict power flows. markets for Ftrs, with peri-
odic auctions, are managed by all isos as a complement to the
where most energy is traded hourly for the next 24 h with
financially binding quantities and prices determined by a
security-constrained, bid-based economic dispatch, imple-
mented through unit commitment optimization employing
mixed-integer programming.
deviations from the day-ahead hourly commitments are
reoptimized every 5 min based on rebidding every 15 min. Vir-
tual bidding allows arbitrage by both energy traders and spec-
ulators and results in a closer convergence of day-ahead and
real-time prices. it also eliminates the incentive to disguise arbi-
trage through implicit virtual bidding, in which physical energy
offers or bids to buy are deferred until real time to manipulate
day-ahead prices. Both the supply and demand sides engaged
in implicit virtual bidding during the California crisis. Figure 1
illustrates a typical day-ahead and real-time dispatch versus the
hourly day-ahead load forecast and actual load.
The Second Wave:
Focus on Economic Efficiency
and Incremental Improvements
what was learned in the decade after the California crisis is
that market design is not a destination but a journey. as sup-
ply and demand technologies improve, computational capa-
bilities are enhanced, and more is learned about the func-
tioning and needs of markets, further changes are proposed
and implemented. in recent years, these changes have been
primarily incremental in nature (which we call the second
wave), but as we will discuss in the next section, more dra-
matic reforms may lie ahead in response to the smart grid,
renewable energy, and distributed resources revolutions.
27,500
26,500
25,500
24,500
Power (MW)

23,500
model energy markets (Hogan, 1992).
everyone also learned from the California debacle that 22,500
markets should not be arbitrarily isolated and that arbitrage 21,500
is important for improving efficiency, providing informa- 20,500
tion, and making it harder to exercise market power. one 19,500
response is to design markets so that energy and ancillary 18,500
service interdependencies are recognized; this is now done 17,500
by cooptimizing those markets, committing resources to
0:00

6:00

12:00

18:00

24:00

provide whichever commodity is most valuable to the sys-

tem and paying in a way that also makes that schedule the Time of Day
most profitable course for individual suppliers. another Day-Ahead Generation Schedule
response is the two-settlement energy market design, with Day-Ahead Load Forecast
virtual bidding, which is now standard in U.s. organized Hour-Ahead Load Forecast
markets. thus, most energy trading is contracted in forward Actual Load
bilateral and over-the-counter markets, complemented by
long-run financial contracts. the isos settle day-ahead and figure 1. A typical two-settlement dispatch, from 8 Janu-
spot energy transactions through a two-settlement approach ary 2018. (Data from California ISO.)

january/february 2019 ieee power & energy magazine 77

 one fundamental challenge is the divergence between
the economist’s ideal of a convex, smooth supply cost func-
tion and the reality of lumpy commitments, minimum output
constraints, and nonconvex heat rate curves. the U.s. mar-
ket designs allow generation resources to provide economic
offers in the day-ahead and real-time markets in the form of
various combinations of an energy supply function, start-up
cost, and no-load cost. such offers are cleared using unit com-
mitment optimization mixed-integer programming software.
alternatively, generation resources can submit physical bilat-
eral transactions that are scheduled outside the optimization
as price-taking transactions.
one of the consequences of the nonconvex cost structure
of generation is that generation units may not cover their
lumpy costs through energy payments they receive over a
24-h period. But because of another widely adopted market
feature, generators who participate in the market by submit-
ting economic offers receive make-whole payments, covered
by an uplift known as bid cost recovery (BCR) payments that
ensure no losses over each 24-h period. the BCr payments
provide an incentive for generators to submit economic offers
rather than self-schedule.
the rules for bidding and paying for lumpy costs have
remained an important item on the market reform agenda,
with ongoing dissatisfaction over the basic BCr model. to
minimize uplifts and recover as much of the cost as possible
through energy charges, an approach known variously as
extended locational marginal pricing and convex hull pric-
ing has been partially implemented at the mid-Continent
iso and is being considered by other isos.
another area of incremental improvement has been
the geographic expansion of markets to take advantage of
complementary fuel mixes in neighboring regions and short-
term diversity in load and, increasingly, renewable output.
PJm’s expansion westward to Chicago immediately resulted
in increased west-to-east power flows and in fuel cost sav-
ings of over Us$100 million per year. the integration of the
day-ahead markets in europe has achieved similar results.
more recently, the desire for more efficiency in real-time
balancing in western North america has led to the rapid
expansion of the California iso’s energy imbalance market
(eim) involving eight U.s. states and even the Canadian
province of British Columbia. a large motivator is the nega-
tive real-time prices that often occur midday in the Califor-
nia market, which means that neighboring states would be
paid to take California’s surplus. the eim eliminates the
hurdle rate of transmission access and other charges that
hinder trade between neighboring balancing areas. Unfortu-
nately, governance issues have prevented the California iso
from expanding this model to the day-ahead market. sensing
an opportunity, the southwest Power Pool/mountain west
partnership is competing to enlist balancing areas that have
not yet joined the eim. it also seeks to expand day-ahead
markets, perhaps including closer coordination of the east-
ern and western interconnections through expanded dc links.
78 ieee power & energy magazine
another trend made possible by computational software
and hardware improvements is the explicit consideration of
postcontingency corrective actions in market software. Pre-
viously, such risks were handled in the markets by preventive
prepositioning of supply, through zonal operating reserve
requirements, N-1 constraints on dispatch, or minimum
online constraints. the hope was that this excess supply
would be sufficient to allow the system to securely manage
unexpected equipment outages or fluctuations in net loads.
For some time, academics have pointed out that if postcon-
tingency redispatch could be modeled and optimized, then
the cost of prepositioning the generation could be reduced.
this is because the software would automatically fine-tune
the location and quantity of reserves that are acquired.
the California iso is implementing corrective action
optimization for certain contingencies by explicitly representing
how dispatches and commitments could be adjusted optimally
after an event. similar approaches could be used to optimize
reserves to cover, e.g., ramp events, avoiding the need for
highly conservative and blunt zonal or system requirements
for flexible ramping products. Ultimately, this could lead to full-
blown stochastic optimization for unit commitment, in which a
large number of possible contingency, load, and supply scenarios
could be considered when optimizing which units to put online
and which fast-start units to hold in reserve.
the incremental improvements associated with the second
wave have resulted in or will soon yield significant operat-
ing cost savings. Not all parties, however, are happy with this
trend toward making market software more complex. some
complain that price drivers are not transparent and perhaps
have contributed to decreases in trading activity. an evolu-
tion toward stochastic market clearing could accelerate that
trend. some observers call for a reversal of this trend, e.g.,
toward simplifying supply offers by eliminating lumpy cost
and minimum output bidding. But this discussion has received
far less attention than other looming issues that the industry in
general and market operations in particular are facing.
The Third Wave:
Making Way for Demand-Side,
Renewable, and Distributed Resources
the third wave of market and industry reform, which is largely
still a work in progress, is associated with the rapid penetra-
tion of renewables and distributed energy resources (ders)
into the electricity generation mix and the deployment of smart
grid technologies, such as smart metering, phasor measure-
ment units, and storage (sioshansi, 2016). also important are
changes in consumption patterns due to demand response, the
electrification of the transportation sector, and increased cus-
tomer participation in electricity production (so-called prosum-
age). as we discussed earlier, the first two waves were largely
motivated by the move toward vertical and horizontal unbun-
dling and privatization of the electricity infrastructure and
by increased efficiency of investment and operation through
competition. in contrast, the third wave is largely driven by
january/february 2019
an environmental agenda for the decarbonization of electric-
ity generation to reduce global warming, a social movement
toward more exercise of customer choice and democratization
of the energy supply, and rapid technological innovation in sup-
ply, storage, metering, and control in the energy area.
Political forces advocating for these trends have cham-
pioned policies that subsidize technological innovation on
ideological grounds rather than economic efficiency. this
tilting of the playing field is counter to the fundamental prin-
ciples of market design. the result is perverse incentives and
rent-seeking behavior on the supply and demand sides alike
and proposals for increasingly complicated market mecha-
nisms that attempt to mitigate the technical challenges and
economic distortions resulting from those incentives.
in the rest of this article, we consider some of the impli-
cations for the third wave of market design of the accelerat-
ing penetration of renewable and distributed sources. For
example, policies such as net metering have become pervasive
in the United states and abroad. these policies compensate
rooftop photovoltaic (PV) solar production at the retail rate,
even when household production of solar energy exceeds con-
sumption. to illustrate, in California, net metering payments
under Pacific gas and electric’s retail tariff e6 can reach
Us$0.45/kwh at peak hours, while the wholesale energy cost
may be as low as Us$0.04/Kwh. such rates fail to recognize
that the high volumetric retail rates were designed to distribute
the infrastructure cost among retail customers in an equitable
manner so that consumers that use more energy pay a markup
reflecting a higher share of the infrastructure cost. Under the
net metering paradigm, however, such high-end consum-
ers can offset their payments by installing solar panels and
shifting their share of the infrastructure cost to low-end con-
sumers. the distortion created by net metering has resulted
in rapid expansion of rooftop PV systems in states such as
California, Nevada, and Hawaii, even though their costs are
roughly twice that of grid-scale solar.
this expansion, in turn, has led to a legislative rebellion
that discontinued the original net metering policy in Nevada
and Hawaii and produced a close three to two vote in the Cali-
fornia Public Utility Commission that revised net metering
practices and rates. Hawaii has opted for a major reduction
in payments for the excess production of rooftop PV systems
above customers’ self-use and an increase in fixed connection
charges for new rooftop solar installations. Nevada went even
further along that path by retroactively changing the terms
originally offered to customers with rooftop solar installa-
tions. these moves were fought in the political arena by the
rooftop solar power industry.
while net metering may distort economic incentives in
favor of rooftop solar installations, it should be noted that cen-
tralized PV solar power is becoming increasingly competitive,
and the U.s. department of energy has recently announced
that centralized PV solar installations have reached long-term
costs of Us$0.06/kwh. Consistent with that, in israel, the
Public Utility authority has recently signed a 20-year supply
january/february 2019
contract for PV solar power at Us$0.057/kwh, while in Chile’s
atacama desert, contracts for PV solar at Us$0.033/kwh were
recently signed. such supply for renewable energy would eas-
ily compete in current electricity markets without the need for
perverse incentives inducing inefficient rent chasing, espe-
cially if a carbon tax or a cap-and-trade policy is implemented.
wind power is another major renewable energy source that
could be competitive without targeted subsidies if appropri-
ate carbon pricing were introduced. Unfortunately, in many
systems, wind power has been implicitly subsidized through
feed-in tariffs and must-take policies imposed on system oper-
ators. such policies are often supported by the false premises
that renewable energy is free, is valuable for its own sake, and
should not be spilled except for emergencies or overgeneration
situations. these suppositions do not account for many of the
complexities associated with providing reliable power consid-
ered in security-constrained, economical dispatch. it can be
easily demonstrated that the strategic curtailment of the renew-
ables supply (which is technically feasible) enables the more
efficient and even less polluting dispatch of conventional slow-
ramping resources. Hence, in some systems, such as the Cali-
fornia iso, negative bids (down to −Us$150/mwh), and thus
prices, have been incorporated into the market design to negate
some of the perverse production subsidies. Further incentives
for the curtailment or withholding of renewable energy can
be used to correct the distortions resulting from suboptimal
deployment of conventional standby resources.
in europe, and particularly in germany, the rapid pen-
etration of wind generation and solar power through perverse
economic incentives has resulted in the virtual collapse of the
wholesale energy market. as shown in Figure 2, such a col-
lapse in wholesale energy prices resulted in a missing money
phenomenon that has motivated the capacity remuneration
of conventional power plants. ironically, the abundance of
renewable power has also resulted in a sharp drop in carbon
prices and led to the increased use of lignite, which negates
much of the environmental benefit provided by renewables.
From a market design perspective, the rapid penetration of
renewables and ders has created several major challenges.
70
Cost of Energy (€/MWh)
60
50
40
30
20
10
0
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Year
figure 2. The collapse of the German wholesale market.
(Data from energy-charts.de.)
ieee power & energy magazine 79
 the inherent systematic variability and uncertainty of renew-
able resources results in extreme ramping and a need for flex-
ibility. the famous California duck curve, shown in Figure 3,
illustrates the impact of high solar penetration on the need for
fast-ramping resources in the early morning and late afternoon
and the overgeneration potential that would require curtailing
renewable resources. to address these problems, the California
Public Utility Commission has modified its resource adequacy
(ra) mechanism to require about a third of the capacity that
load-serving entities must show for achieving their obligation
needs to meet flexibility criteria. Flexible resources must be
economically dispatchable and able to be ramped continuously
over a 3-h period. Unlike regular ra resources, which must
offer obligations by being dispatched or available for dispatch,
flexible ra must be offered at a price that enables the iso to
dispatch them economically.
For several reasons, however, the ra mechanism is a
blunt tool to encourage flexibility. First, is it 3-h, 1-h, or 5-min
ramps that must be managed? different resources are best
for each. it also matters whether the ramps are predictable
or uncertain. second, how does one compare, for instance,
the flexibility contributed by combustion turbines that are
limited to one start a day, storage with 15 min of energy,
demand response that can be called upon only a handful of
times per month, and plants with different ramp rates? this
suggests that, rather than develop a generic definition of flex-
ible capacity, it would be better to reform energy markets to
reward capacity that is available precisely when needed and
can ramp up quickly when prices spike and then disappear
when they collapse.
But reforming energy and ancillary services markets to
provide more appropriate rewards for flexibility is not neces-
sarily simple either. Flexibility of energy resources is essen-
tial for the integration of renewables and ders at various
time scales, and hence such flexibility must be priced in the
various market horizons so that it provides the appropri-
ate price signals for investment and operation of resources.
this is accomplished by creating revenue streams and
market products that enable the remuneration of resources
that provide flexibility. such flexibility can be mobilized
through shorter settlement intervals, demand-side manage-
ment, the flexible use of transmission assets (e.g., trans-
mission switching and dynamic ratings of thermal limits),
and flexible ramping products. the value/cost of flexibility
must be reflected in the payments to renewable and distrib-
uted resources that impose such costs on the system, and
resources that help resolve issues associated with flexibility
should be remunerated.
For example, California iso has introduced a new flexible
ramping product (sometimes called flexiramp) that enables
remuneration of resources that are being held in reserve
out of merit order instead of being dispatched to produce
energy which they offer below the market clearing price.
such reserves differ from regulation-type reserves, which

30

25 Distributed Solar
Utility-Scale Solar
Energy (GWh)

20
Imports
15 Other Renewables
Thermal
10
Nuclear
5 Hydroelectric
0
0 2 4 6 8 10 12 14 16 18 20 22 24
Hour of the Day
(a)
Cost of Energy ($/MWh)

60
40 Real-Time
Average
20
Hourly
0 Price
–20
0 2 4 6 8 10 12 14 16 18 20 22 24
Hour of the Day
(b)

figure 3. An illustration of (a) the California ISO net generation duck curve and (b) negative wholesale prices, both for 11
March 2017. (Note: Distributed solar generation is estimated based on December 2016 installed net-metered capacity as
reported in form EIA-826, Monthly Electric Utility Sales and Revenue Report with State Distributions, U.S. Energy Informa-
tion Administration.)

80 ieee power & energy magazine january/february 2019

are a capacity-based ancillary service product used for load
following within each 5-min interval. in contrast, flexiramp
serves to assure the availability of energy and regulation in
future intervals. Hence, these flexiramp reserves are remu-
nerated for lost opportunity costs, which is the profit that is
forgone when capacity is held back out of merit rather than
dispatched (i.e., opportunity cost = clearing price − bid).
Besides the growth of renewables, the other dramatic
change in the industry is the proliferation of many small
ders on both the demand and supply sides. this reflects a
general trend toward diverse participation in the produc-
tion of energy and the above-described need to address the
many new challenges posed by the introduction of renew-
ables. such proliferation is largely enabled by innovation in
the smart grid technologies area. there are many ways to
integrate such resources and organize market platforms that
will facilitate the procurement of ders, enable peer-to-peer
transactions at the distribution level, and create new products.
these changes will redefine the role of future utilities to sup-
port their ability to assure retail supply reliability in a world
where their volumetric revenue base is shrinking because of
increased self-supply of energy behind the meter.
one market design choice is whether to make the market
iso-centric, where ders are integrated into the system through
distribution system operators and third-party aggregators who
participate in the iso wholesale market. alternatively, the mar-
ket organization could be distribution-level–centric, with the
iso performing limited coordination functions among distri-
bution system operators that accomplish most of the der inte-
gration through local distribution system market platforms. the
California iso and PJm have opted for the first through-market
initiatives that carve a role for der aggregators (in California)
and curtailment service providers (PJm) that behave as virtual
power plants that can participate as regular plants in the iso
wholesale market. such participation broadens the scope of the
conventional isos and adapts them to the brave new world of
renewables and distributed resources. New York state, on the
other hand, has been promoting its reform the energy Vision
initiative, which emphasizes customer choice and active partici-
pation by ders.
Conclusions
From thomas edison’s Pearl street station (and his earlier
Holborn Viaduct plant in London) to samuel insull’s regu-
lated utility and the vertically integrated monopolies of the
late 1900s, the power industry saw tremendous innovation
in technology and institutions during its first century. the
last three decades have seen further tremendous changes,
with environmental concerns coming to the fore and renew-
able generation now dominating new capacity additions.
New computation and communications technologies have
facilitated a revolution in markets, allowing much bigger
geographic coverage and efficiently coordinating an array
of energy, ancillary service, and transmission products. this
growth in the scope of markets lowers costs by increasing
january/february 2019
competition and taking advantage of diversity of load and
generation patterns across large regions. ironically, while
the scale of markets has greatly increased, at the same time
the scale of who can compete has become much smaller, so
that household-scale technologies are now revolutionizing
the market.
market designs have not kept up with these changes.
this means that we have not realized the full economic and
environmental benefits of new technologies. Present chal-
lenges for market designers include further expansion of the
geographic scope of markets; reconciling the wildly incon-
sistent price signals in retail and wholesale power prices;
incentivizing an optimal mix, location, and scale of flexible
generation, storage, and demand-side resources to firm up
renewable generation; and avoiding confusing means for
ends by keeping a firm eye on the goals we want to accom-
plish with renewable and other policies. it is lower system
costs, greater reliability, and lower pollution we want, not
particular technologies; the best means to achieve those
goals are likely to surprise us. therefore, markets must be
designed to allow the best solutions to emerge and succeed
without policy makers imposing their prejudices about what
they should look like.
For Further Reading
s. Borenstein and J. Bushnell, “the Us electricity industry
after 20 years of restructuring,” Annu. Rev. Economics, vol.
7, no. 1, pp. 437–463, 2015.
w. w. Hogan, “Contract markets for electric power trans-
mission,” J. Regulatory Econ., vol. 4, no. 3, pp. 211–242, 1992.
P. L. Joskow and r. schmalensee, Markets for Power: An
Analysis of Electric Utility Deregulation. Cambridge, ma:
mit Press, 1983.
e. t. mansur and m. w. white. (2012, Jan. 13). market
organization and efficiency in electricity markets. dart-
mouth College, Hanover, NH. [online]. available: www
.dartmouth.edu/~mansur/papers/mansur_white_pjmaep.pdf
F. C. schweppe, m. C. Caramanis, r. d. tabors, and r.
e. Bohn, Spot Pricing of Electricity. Norwell, ma: Kluwer,
1988.
F. P. sioshansi, ed., Future of Utilities, Utilities of the Fu-
ture: How Technological Innovations in Distributed Energy
Resources Will Reshape the Electric Power Sector. New
York: academic, 2016.
s. stoft, “economic feedback: working with the insa-
tiable profit motive,” IEEE Power Energy Mag., vol. 99, no.
3, pp. 14–16, 2003.
Biographies
Benjamin F. Hobbs is with Johns Hopkins University, Bal-
timore, maryland.
Shmuel S. Oren is with the University of California,
Berkeley.
p&e
ieee power & energy magazine 81
 PV-Battery
Systems
for Critical
Loads During
Emergencies
By Chanaka Keerthisinghe, Mareldi Ahumada-Paras,
Lilo D. Pozzo, Daniel S. Kirschen, Hugo Pontes,
Wesley K. Tatum, and Marvi A. Matos

photos courtesy of: university of Washington

Digital Object Identifier 10.1109/MPE.2018.2877896

Date of publication: 7 January 2019

82 ieee power & energy magazine 1540-7977/19©2019IEEE january/february 2019

 H
Hurricane Maria struck Puerto rico on 20 sePteMber
2017 and left large parts of the island without electricity for months. as
Figure 1 shows, restoration in remote mountainous regions took consider-
ably longer.
Long-term power outages can be lethal to individuals who rely on
electrically powered medical devices or require medicines that must
be refrigerated. although the initial death toll of Hurricane Maria was
64, one study concluded that actual number of deaths caused by the
hurricane is closer to 2,975 due to the lack of utilities, such as electric-
ity. researchers calculated this number by comparing the total medi-
cal-related deaths during 2017 with the average numbers in the past four
years. another study esti-
mated the death toll at 4,645.
it is likely that remote areas were
more severely affected because dam-
A Case Study from
aged roads not only prevented access to med-
ical facilities but also hampered the regular delivery of medical supplies
Puerto Rico After
and fuel for portable generators. communities were deprived of electric-
ity and water, hospitals and clinics were inoperable, and communication
Hurricane Maria
and transportation were unavailable for months; 200 days after the hur-
ricane, 160,000 people still lacked electricity. the devastating destruction
from the hurricane is highlighted in Figure 2.
small systems combining photovoltaic (PV) generation and battery energy
storage could be deployed during such emergencies to help affected indi-
viduals cope until grid-supplied power is restored. However, very little
reliable data are available on how such systems would actually
be used to meet critical medical needs and, thus, on what
their design requirements should be. to gather data on
how emergency systems would actually work and pro-
vide some relief to a severely affected community,
our group from the university of Washington in
seattle made three field trips to Jayuya, a remote
small town in the mountainous center of Puerto
rico, to collect preliminary information about
medical needs, install PV battery systems,
and gather the data collected by these sys-
tems after several months of use. Figure 3
shows some photos highlighting the work.

Field Results
Figure 4 outlines the time line and pur-
poses of the three field trips to Puerto rico.

First Field Trip:

Preliminary Needs Assessment
the goal of the first trip was to identify, based
on interviews with patients and their families,
the critical medical needs that require electric
power at the household level and carry out a pre-
liminary assessment of the power and energy require-
ments associated with these needs. the critical medical
conditions and associated electrical devices that were iden-
tified during these interviews were
✔ feeding machines and electrical bedding for patients with
percutaneous endoscopic gastrostomy (PeG)

january/february 2019 ieee power & energy magazine 83

 ✔ nebulizers or oxygen concentrators for patients with ✔ refrigeration for medicines, such as insulin for dia-
asthma and/or chronic obstructive pulmonary disease betic patients, and food for patients with special di-
(coPD) etary requirements.
✔ continuous positive airway pressure (cPaP) machines Dialysis and treatments for other acute medical needs were
for patients with sleep apnea not included in this assessment because they are usually not
carried out at home and patients with critical needs are fre-
quently evacuated before or after emergencies. interviews also
revealed how individuals coped with the lack of electricity.
100
PeG patients used gravity feeding instead of powered peri-
Customers with Power (%)

Entire Puerto Rico

80 San Juan (Capital) staltic pumps, which resulted in substantially faster feeding.
Caguas (Mountainous) Many diabetic patients kept their insulin cool either by using
60 water and ice or moving their refrigerators to nearby busi-
nesses that had emergency generators; however, some stopped
40 18–23 March
19–24 November 2018 (Trip 2) taking insulin out of fear that it had degraded. the local clinic
20 2017 (Trip 1) was without power until a backup generator arrived, and they
lost medicine requiring refrigeration as well as all vaccines.
0 other loads common to all households were perceived
0 20 40 60 80 100 120 140 160 180 200
Time (Days) as important but less critical: refrigeration for regular food,
fans, lighting, tV, and washing machines (particularly
figure 1. The restoration of power to customers in Puerto important in households with bedridden patients). cooking
Rico after Hurricane Maria, beginning on 20 September appliances do not appear on this list because most house-
2017. Note the significant difference between the San Juan holds in this part of Puerto rico use gas for cooking.
urban area and the mountainous and directly hit region basic information about the critical loads was also gath-
of Caguas. ered by reading manufacturers’ labels and asking families

figure 2. The devastation caused by Hurricane Maria in Puerto Rico. (Photos courtesy of the University of Washington,
used with permission.)

84 ieee power & energy magazine january/february 2019

how often each appliance was used and for how long. table 1
summarizes time-of-use and duration data. because there
was no electricity during the first field trip, it was not pos-
sible to measure actual demands or load profiles, so the mea-
sured power values in table 1 were obtained during the third
trip. the maximum power drawn by the devices was used
for sizing the inverter, whereas the entire load profile was
required to size the PV and battery systems.
Diesel generators are commonly used in emergencies but
were perceived as having the following drawbacks:
✔ our interviews revealed that the condition of some asth-
ma and coPD patients had worsened because of the
exhaust gases (such as carbon monoxide) or the aero-
solized soot particles produced by the diesel engines.
✔ because of the noise, generators cannot be used at night
to power the cPaP machines of sleep apnea patients.

figure 3. Some of the PV-battery systems installed in Puerto Rico by researchers at the University of Washington in
Seattle. (Photos courtesy of the University of Washington, used with permission.)

Trip 1: Trip 2: Trip 3:

19–24 November 18–23 March 2018 2–5 July 2018
Hurricane 2017
Maria Hit
Puerto Rico on
20 September Identify Critical Donate Extract Data
2017 Loads and Value PV-Battery (i.e., Electrical
of a PV-Battery Systems Load Profiles)
System

figure 4. The time line and purposes of the field trips.

january/february 2019 ieee power & energy magazine 85

 table 1. The maximum and average measured power, time of use, and duration of critical electrical loads.
Device Name Duration, Time of Use Average Power Maximum Power
PEG feeding machine 30 min, four times <10 W 120 W (manufacturer)
Electrically actuated bed 30 s, ten times 18 W 53–65 W
Inflatable mattress 8–24 h Lower than 10 W <10 W
Nebulizers 5–25 min, two–four times 23–52 W 23–66 W
Oxygen concentrators Up to 24 h 350 W 428 W
CPAP 9 h, 10 p.m.–7 a.m. 34 W 47 W
Refrigerator 8–24 h 136–352 W 140–392 W
Refrigerator (small) 8–24 h 20 W 97 W
TV 2–4 h, 3–9 p.m. 28–292 W 31–392 W

table 2. A summary of the various systems installed.

Number of
PV Size (W) Battery MPPT Inverter (kW) Installations
Type A 260 160 Ah (lead acid) Yes 1 Six
Type B 100–200 80–100 Ah (lead acid) No 1 Six
Type C 400 1.1 kWh (lithium ion) Yes 1.1 Five
Type D 100 100 Ah (lead acid) No dc system Four

✔ Maintaining a generator (e.g., changing the oil) is

challenging for some users because it requires time,
knowledge, and money.

Second Field Trip: PV-Battery

To PV Panels
MPPT Charge
Controller
Inverter Monitor
figure 5. The top view of a PV-battery system designed
at the University of Washington. The batteries, data logger,
and majority of the cables are inside the box.
✔ reliable access to a supply of diesel fuel during an
emergency can be difficult or impossible, particularly
in a remote area.
✔ using a generator can cost up to us$10 per day. over
an extended period of time, this can be excessive for
low-income households.
✔ emergency diesel generators are often not designed for con-
tinuous use. individuals would have to use them for periods
of 2–6 h. For example, some households ran diesel genera-
tors up to 8 h per day (4 h in the morning and 4 h in the
evening) to keep medicines cool. there were also many in-
stances in which generators had broken because of overuse.
Systems Deployment
the goal of the second field trip was to install 17 stand-alone
PV-battery systems of four different types and power ratings
in households with individuals relying on electrically powered
medical devices or medically related refrigeration needs. a
summary of the installed systems is given in table 2; four sys-
tems had been installed in the first trip. because systems were
not available for all households with critical medical needs,
priority was given to those who lacked the resources to buy
or obtain another source of electricity and those living in the
more remote mountain areas.
Figure 5 shows a top view of a type a system, which
consists of an inverter, a maximum power point tracking
(MPPt) solar-charge controller, a display monitor, lead-acid
batteries, and data loggers to record the electrical load pro-
files, PV output, battery state of charge (soc), and tempera-
ture. system types a, b, and D were designed at the univer-
sity of Washington, and system type c was a commercially
available system.
because of budget and time restrictions, the systems were
limited in capacity. nevertheless, we anticipated that they
would be able to meet at least some of the critical loads of indi-
vidual households, either fully on their own or in tandem with
diesel generators, if households were already using a generator.

86 ieee power & energy magazine january/february 2019

 Long-term power outages can be lethal to individuals
who rely on electrically powered medical devices
or require medicines that must be refrigerated.

because these systems were sized from inaccurate load pro-
files based on the limited information collected during the
first field trip, some of them were undersized. although we
encouraged the users to try not to discharge the batteries below
recommended safety limits to maximize battery life, some bat-
teries degraded prematurely because of excessive depth of
discharge (DoD).
Third Field Trip: Data Collection and Analysis
the purpose of the third trip was to obtain the data col-
lected by the data loggers in the installed PV-battery sys-
tems. satisfaction surveys were also administered to gauge
user experience and perceptions of solar energy as a form
of emergency energy supply. these quantitative and quali-
tative data are analyzed in the next section.
Analysis of Field Data
We choose to analyze the six households that received type
a PV-battery systems because they were the only systems
equipped with data loggers that could record all of the nec-
essary data. systems in five other households recorded only
the load profiles.
Energy Consumption and Generation
Figure 6 shows the total daily energy consumption and PV
generation over 47 days for six households where type a PV-
battery systems were installed. the energy-use pattern for
household 6 is similar to the PV generation pattern, which
shows that users managed to consume the PV generation on
the same day. this energy-consumption pattern contributed to
keeping the battery in good condition. another important

2 2
1.5 1.5
1 1
0.5 0.5
0 0
5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45
Days Days Stopped Using the Load
(a) (b) Logger but Continued to
Use the System
Electrical Energy (kWh)

Electrical Energy (kWh)

2 2
1.5 1.5
Got Electricity Back,
1 1 So Stopped Using the
0.5 0.5 System and Load Logger
0 0
5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45
Days Days
(c) (d)
2 2
1.5 Got Electricity Back 1.5
1 Started Using Stopped Using 1
the Load the Load Logger
0.5 Logger Late 0.5
0 0
5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45
Days Days
(e) (f)
PV Output Load

figure 6. The total daily load and PV output for six households over 47 days. Households (a) 1, (b) 2, (c) 3, (d) 4, (e) 5, and
(f) 6. The load was measured after the inverter.

january/february 2019 ieee power & energy magazine 87

 The local clinic was without power until a backup generator
arrived, and they lost medicine requiring refrigeration as
well as all vaccines.

finding is that households 4 and 5 stopped using the systems
after they regained grid power within one month after instal-
lation, whereas the other users continued to use the systems
throughout the whole period.
Load Profiles
Figure 7 illustrates the load profiles for the days with the
highest energy consumption for eight households. the
load measured corresponds to different devices specific to
each household connected at each moment. For example,
household 1 used the system to power a refrigerator (~150 W)
from 7 a.m. to 3 p.m., an entertainment system (~375 W)
for periods of an hour at a time, and a nebulizer (23–52 W)
a few times a day for 45 min. Households 1–6 used most of
their devices during the day when there was sun to maxi-
mize battery life. nevertheless, they used the battery when

600 200
Entertainment System Refrigerator
150
400
TV
Refrigerator 100
Nebulizer
200 50
0 0
4 8 12 16 20 24 4 8 12 16 20 24
Time Time
(a) (b)
600 600
400 Refrigerator 400 Refrigerator
200 200
Electrical Power (W)

Electrical Power (W)

0 0
4 8 12 16 20 24 4 8 12 16 20 24
Time Time
(c) (d)
400 200
300 150
Nebulizer
200 Refrigerator 100
100 50
0 0
4 8 12 16 20 24 4 8 12 16 20 24
Time Time
(e) (f)
100 8
Small Refrigerator PEG Patient
6
50 4
2
Feeding Machine Electrical Bed
0 0
4 8 12 16 20 24 4 8 12 16 20 24
Time Time
(g) (h)

Day 1 Day 2 Day 3

figure 7. The electrical load profiles of eight different households over one to three days. Households (a) 1, (b) 2, (c) 3, (d)
4, (e) 5, (f) 6, (g) 7, and (h) 8. The load profiles consist of all of the critical devices; however, we were able to identify some
of these devices through interviews, as noted in the figure.

88 ieee power & energy magazine january/february 2019

 Because battery use in emergency situations is likely to be poorly
controlled, it is important to collect data about charge/discharge cycles
so that information can inform the design of longer-life systems.

consumption was greater than PV generation, thereby
increasing battery degradation. the PV systems given to
households 1–6 are rated at only 260 W.
the days with the highest consumption from Figure 7
are used to construct the critical load percentages shown
in table 3. For households 1 and 2, the critical loads repre-
sent 17.2 and 15.4%, respectively, of the average daily load
before the hurricane. this information was obtained from
electricity bills dated before the hurricane. unfortunately,
it was not possible to obtain the loads before the hurri-
cane for some households. these percentages can be used
to create critical load profiles for other households in the
future when designing microgrid systems for resilience.
be inevitable, because critical needs might not occur during the
daylight hours when PV generation is sufficient. We were able
to confirm during the third field trip that the lead-acid battery of
household 6 was still in excellent condition, but the residents of
household 2 said that their battery was not performing as well as
it did originally.
because a battery’s temperature also affects its useful life, the
data loggers recorded battery temperature. as Figure 8(b) shows,
even though Puerto rico has a tropical climate, the battery
temperatures generally remained within the expected limit of
20−25 °c, with the exception of the battery of household 5, where
the temperature increased up to 30 °c. a possible explanation
is that the household 5 is at a lower elevation than the others.

Battery Degradation Survey

repeatedly discharging lead-acid batteries below a recommend- as shown in table 4, all 15 households interviewed said that
ed soc progressively degrades their energy capacity. because they would prefer a solar system over a diesel or gas genera-
battery use in emergency situations is likely to be poorly con- tor for their energy needs. among those, 14 said that they are
trolled, it is important to collect data
about charge/discharge cycles so
table 3. Critical and normal energy consumption for six households.
that information can inform the
design of longer-life systems. all Load Before Cost of
recipients of PV-battery systems Critical Load Hurricane Critical Load Generator
Household (kWh/day) (kWh/day) Percentage (%) Fuel (US$/day)
were instructed to operate the battery
within safe limits. as Figure 8(a) 1 1.5 8.67 17.2 10
shows, households 5 and 6 used the 2 1.2 7.77 15.4 7
systems such that the DoD of the 3 1.2 — — 10
battery rarely exceeded 50%. on the 4 1.1 — — —
other hand, most of the discharge
5 0.7 7.5 9.2 —
cycles of the batteries in households
6 1 — — —
1–3 exceeded a 50% DoD. this may

200 32
Battery Temperature (°C)

Above 50% DoD 30

150 Below 50% DoD 28
Battery Cycles

26
100 24
22
50 Households: 1 2 3
20
4 5 6
0 18
1 2 3 4 5 6 10 20 30 40 50 60 70 80 90 100
Households Days
(a) (b)

figure 8. The battery operation for six households over 105 days: (a) the number of battery cycles above and below 50%
DoD and (b) the average daily battery temperature.

january/february 2019 ieee power & energy magazine 89

 TV was critical to most families because it was a distraction
from their current difficulties, was a source of entertainment,
and provided information crucial to their well-being.

more comfortable using a PV-battery system. the household batteries are assumed to last 200 cycles when the DoD is
that answered “same” is household 4, in which electricity was above 50% and 500 cycles when the DoD is below 50%.
restored within 18 days (Figure 6). eleven households had used table 5 summarizes the operation of different combina-
a generator before, and nine of the 11 that were actively using tions of battery and PV ratings for each case. in particular, it
or had previously used a diesel/gas generator mentioned that shows how much energy would be produced and consumed
the PV-battery system improved their health. one household and how many charge/discharge cycles the battery would
was unfortunately provided with a defective commercial solar undergo in one, three, five, and ten years. the last four col-
generator system that quickly broke down. they were given a umns indicate the cost of each PV battery combination (capi-
replacement for use in future emergencies, but they gave our tal cost), including battery replacement costs (maintenance
survey a low rating based on their experience. cost), for a lifetime of one, three, five, and ten years. the cost
of a system depends more on the number of battery replace-
Simulation Results and Discussion ments needed than on duration of the system. in some cases,
a linear optimization method was used to minimize PV-bat- the cost is the same for initial installations of larger batteries
tery system costs that are able to supply critical loads for a year versus smaller ones that need replacing because of more fre-
considering PV and demand variations. battery degradation cost quent, deeper DoD operations.
over the lifetime of these systems was considered when choosing the systems designed for cases a, b, and c provide
the optimal design. We consider the following scenarios: some guidelines regarding the sizing of PV-battery sys-
✔ case a involves household 1’s critical load profile as illus- tems for emergencies. Figure 9(a) shows how the optimal
trated in Figure 7(a) and includes an efficiently managed battery capacity decreases as the PV rating increases for
refrigerator, a nebulizer, and an entertainment system. case a’s system. because we are interested only in the
✔ case b represents a household with a PeG patient and minimum battery capacity required to provide continuous
a small refrigerator. the critical load profile is a com- power to critical loads, extra PV generation is curtailed
bination of Figure 7(g) and (h). to avoid having to increase the battery size. Figure 9(b) shows
✔ case c is a household with a sleep apnea patient. the that, because the unit cost of batteries (us$192/kWh for
critical loads include the cPaP machine, a refrigera- lead-acid batteries and us$800/kWh for lithium-ion bat-
tor, and a small tV. the critical load profile for this teries) is much higher than the per-unit cost of PV gen-
scenario was created using the data in table 1. eration (us$1.5/W), the cost of the system decreases with
✔ case D has the largest critical load, which consists of decreasing battery size, even though the required PV rat-
a continuously running oxygen concentrator, a refrig- ing increases.
erator, and a large tV. using lead-acid batteries leads to the lowest cost systems for
an interesting finding was that tV was critical to most fami- cases a, b, and c for up to a three-year system life because,
lies because it was a distraction from their current difficulties, although lithium-ion batteries can be discharged completely and
was a source of entertainment, and provided information cru- therefore have a smaller rating, their cost is still much higher. in
cial to their well-being. particular, a system that lets lead-acid batteries discharge com-
PV output data for the optimizations are from the national pletely to meet critical loads is the least expensive solution
renewable energy Laboratory’s PVwatts calculator. Lead-acid for a one-year system life because such a system is effectively
sized to minimize the number of
table 4. A summary of survey results. battery cycles. When the expected
system lifetime increases, limit-
Solar Generator Same
ing the lead-acid battery to 50%
Do you prefer solar or generator? 100% (15) 0% 0% DoD is more economical because
Which are you most comfortable using? 93% (14) 0% 6.7% (one) it reduces the required number of
Yes No bat tery replacements.
on the other hand, case D re-
Did you have a generator before? 11 Four —
quires such a large system that it
If yes, did solar improve your health? 82% (nine) 9% (one) 9% (one)
would probably be economically

90 ieee power & energy magazine january/february 2019

 table 5. A comparison of various PV battery system designs.
Total Total
PV Load Battery Cycles (n) Cost (US$)
Battery PV (kWh), (kWh),
Battery Type Size (Ah) Size One One One Three Five Ten One Three Five Ten
and DoD at 12 V (Wp) Year Year Year Years Years Years Year Years Years Years
Case A: Refrigerator, nebulizer, and TV
Lithium ion 100 800 1,291 484 78 235 392 783 2,150, 2,150, 2,150, 2,150,
(100% DoD) r=0 r=0 r=0 r=0
Lead acid 170 600 968 63 188 313 626 1,291, 1,291, 1,682, 2,464,
(100% DoD) r=0 r=0 r=1 r=3
Lead acid 260 700 1,130 36 107 178 356 1,648, 1,648, 1,648, 1,648,
(50% DoD) r=0 r=0 r=0 r=0
Case B: PEG patient and small refrigerator
Lithium ion 60 300 484 197 166 497 828 1,655 1,020, 1,020, 1,020, 1,020,
(100% DoD) r=0 r=0 r=0 r=0
Lead acid 120 200 323 92 275 458 916 576, 852, 1,128, 1,680,
(100% DoD) r=0 r=1 r=2 r=4
Lead acid 120 300 484 88 263 438 876 726, 726, 726, 1,002,
(50% DoD) r=0 r=0 r=0 r=1
Case C: CPAP, refrigerator, and small TV
Lithium ion 120 700 1,130 459 245 735 1,225 2,450 2,190, 2,190, 2,190, 2,190,
(100% DoD) r=0 r=0 r=0 r=0
Lead acid 270 500 807 117 351 585 1,170 1,371, 1,992, 2,613, 4,476,
(100% DoD) r=0 r=1 r=2 r=5
Lead acid 290 700 1,130 108 324 540 1,080 1,717, 1,717, 2,384, 3,051,
(50% DoD) r=0 r=0 r=1 r=2
Case D: Oxygen concentrator, refrigerator, and TV
Lithium ion 950 4,900 7,907 3,594 219 656 1,093 2,185 5,900, 5,900, 5,900, 5,900,
(100% DoD) r=0 r=0 r=0 r=0
Lead acid 2,070 3,800 6,132 108 324 540 1,080 r=0 r=1 r=2 r=5
(100% DoD)
Lead acid 1,980 5,500 8,875 110 331 552 1,103 r=0 r=0 r=1 r=2
(50% DoD)
The cost consists of only the PV and batteries, and r is the number of battery replacements. Wp is the nameplate value, which is a measure of
watts at peak production.

12 4k
Battery Capacity (kWh)

Case A
10
3k
8
Cost (US$)

6 2k
4
1k
2
0 0
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
PV System Size (kWp) PV System Size (kWp)
(a) (b)

Lithium Ion (100% DoD) Lead Acid (100% DoD) Lead Acid (50% DoD)

figure 9. (a) The minimum battery capacity required to power critical loads over a year for case A as a function of the PV
rating. (b) The total system cost as a function of the PV rating.

january/february 2019 ieee power & energy magazine 91


2,500
PV-Battery System Day 65–66
2,000 Diesel Generator
Slope Is the
Cost (US$)
1,500 Fuel, Oil
Change, and
1,000 Depreciation
Slope Is the Total Costs Per Day
Per-Day Cost of
500 Battery Cycle and PV
Capital and Electronic
Costs Depreciation
0
10 20 30 40 50 60 70 80 90 100
Days
figure 10. A comparison of the cost of using a diesel genera-
tor versus a PV battery system during an emergency for case A.
beneficial to install a grid-connected system and operate it
for nonemergencies as well. in this case, lithium-ion batteries
would be less expensive because their cost is lower at larger rat-
ings. the cost of the system in case D is based on a us$0.75/W
for PV and us$195/kWh for lithium-ion batteries. although
lithium-ion batteries are currently more expensive than lead-
acid ones, they are more convenient because they are smaller
and lighter. they are easier for a user to move around the house
to connect to various critical loads and also easier to deploy
after a natural disaster.
Figure 10 compares the total cost of using a diesel gen-
erator and PV-battery system as a function of the duration
of a power outage for case a (i.e., household 1). the capital
cost of a generator is lower than for a PV-battery system.
However, the ongoing costs of fuel, oil changes, and depre-
ciation rise rapidly. on the other hand, the ongoing cost
of running a PV-battery system largely depends on battery
degradation because the PV panels and the other elec-
tronics are expected to last ten–15 years. these results sug-
gest that any place likely to experience more than 66 days
of power outages (cumulative over multiple events) would
benefit from a PV-battery system. Longer usage loads (e.g.,
cPaP, air mattress, or oxygen concentrator) may inevita-
bly need PV-battery systems or, at the very least, batteries
that can be charged with a generator during its use because
small diesel/gas generators are noisy when used at night and
are often not designed for continuous use.
Acknowledgments
the authors would like to thank the university of Washington’s
clean energy institute and office of Global affairs’ Global
innovation Fund for providing travel funds. the solar-powered
battery devices installed in March 2018 were purchased thanks
to generous support from the clean energy Group through its
resilient Power Project and from gifts from numerous donors.
the group would like to thank all of the participants in the
study and the community organizers in Jayuya, Puerto rico,
who helped organize and identify participants. the authors
92 ieee power & energy magazine
also would like to thank Dennis r. Wise from the university of
Washington for some of the photographs.
For Further Reading
clean energy institute. (2018). Puerto rico: clean energy so-
lutions for public health. clean energy institute, university
of Washington. [online]. available: https://www.cei.washington
.edu/education/global-engagement/puerto-rico/
u.s. Department of energy, office of cybersecurity,
energy security, and emergency response. (2018, apr.)
Hurricanes nate, Maria, irma, and Harvey situation reports.
[online]. available: https://www.energy.gov/ceser/downloads/
hurricanes-nate-maria-irma-and-harvey-situation-reports
k. anderson, k. burman, t. simpkins, e. Helson, and L.
Lisell. (2016, June). new York solar smart DG hub-resilient
solar project: economic and resiliency impact of PV and stor-
age on new York critical infrastructure. national renewable
energy Laboratory (nreL) and city university of new York
(cunY). [online]. available: https://www.nrel.gov/docs/
fy16osti/66617.pdf
L. santiago, c. e. shoichet and J. kravarik. (2018, aug.)
Puerto rico’s new Hurricane Maria death toll is 46 times
higher than the government’s previous count. cnn. [online].
available: https://www.cnn.com/2018/08/28/health/puerto-
rico-gw-report-excess-deaths/index.html
n. kishore, D. Marqués, a. Mahmud, M. V. kiang, i. ro-
driguez, a. Fuller, P. ebner, c. sorensen, F. racy, J. Lemery,
L. Maas, J. Leaning, r. a. irizarry, s. balsari, and c. o.
buckee, “Mortality in Puerto rico after Hurricane Maria,”
New Eng. J. Med., vol. 379, no. 2, pp. 162–170, 2018. doi:
10.1056/neJMsa1803972.
c. keerthisinghe, M. ahumada-Paras, L. D. Pozzo, D. s.
kirschen, H. Pontes, W. k. tatum, and M. a. Matos, “PV-
battery systems for critical loads during emergencies: case
study from Puerto rico after Hurricane Maria,” 2019. [on-
line]. available: https://www.researchgate.net/project/PV-
battery-systems-for-critical-Loads-during-emergencies-case-
study-from-Puerto-rico-after-Hurricane-Maria
Biographies
Chanaka Keerthisinghe is with the university of Washing-
ton, seattle.
Mareldi Ahumada-Paras is with the university of
Washington, seattle.
Lilo D. Pozzo is with the university of Washington,
seattle.
Daniel S. Kirschen is with the university of Washington,
seattle.
Hugo Pontes is with the university of Washington,
seattle.
Wesley K. Tatum is with the university of Washington,
seattle.
Marvi A. Matos is with blue origin, kent, Washington.
p&e
january/february 2019
Digital Object Identifier 10.1109/MPE.2018.2885101
 society news
M
MeMbers of the Ieee Power
& energy society (Pes) and other in-
dustry enthusiasts came together for the
Pes General Meeting 5–9 August 2018.
More than 3,200 attendees converged
on Portland, oregon, for new insights,
innovative ideas, and answers to some
of the most intriguing and important
questions facing the power industry today.
During the course of the week, members
were able to attend informational, techni-
cal, and tutorial sessions; committee and
working group meetings; poster sessions;
a job fair; networking events; tours of tech-
nical facilities; and companion activities.
As they say, “You can, in Portland,” and
we certainly did! whether it was enjoying
Portland’s eclectic visual arts scene, great
meals from hundreds of food carts, or fan-
tastic tax-free vintage shopping, Portland
provided a truly unique environment for
this year’s conference.
Digital Object Identifier 10.1109/MPE.2018.2877915
Date of publication: 7 January 2019
The plenary session panel discussed shifting trends in the electric industry.
94 ieee power & energy magazine
M.J. Stryjewski
2018 General Meeting
reimagining the electric grid
the week’s events kicked off at the
oregon Museum of science and Indus-
try (oMsI) with a delicious cat ere d
welcome reception, where friends old
and new were united. the oMsI had
many interesting exhibits that captured
the imaginations of attendees. As luck
would have it, the premier attraction
was the robot revolution, a display that
demonstrated how, in the near future,
robots will become our companions and
colleagues—changing the way we play,
live, and work together.
the following morning, the all-im-
portant Members’ Meeting featured an
update on society activities from Pes
President saifur rahman, including a
new interactive portion where attend-
ees were able to respond to questions in
real time and help drive the conversa-
tion toward their interests.
At the Plenary session, hosted by
richard Goddard of the host utility Port- PES President Saifur Rahman explains
land General electric, attendees were the new interactive portion of the
officially welcomed to Portland and opening session.
january/february 2019
 In Memoriam: Prabha Shankar Kundur
On 9 October 2018, Prabha Shankar Kundur passed away
unexpectedly at the age of 79. A true technical leader in the
electric power industry, Prabha chaired numerous commit-
tees and working groups of the IEEE Power & Energy Soci-
ety (PES), and he was elected as a Fellow
of the IEEE in 1985 “for contributions to
modeling techniques in power system
control and stability analysis and their
applications.” He served as the chair of
the IEEE Power System Dynamic Perfor-
mance Committee from 2002 to 2004.
From 2006 to 2010, he was a member of
the IEEE PES Governing Board as the vice
president for Education. He received sev-
eral IEEE awards, including the 1997 IEEE
Nikola Tesla Award, the 2005 IEEE PES
Charles Concordia Power System Engi-
neering Award, and the 2010 IEEE Medal
in Power Engineering. Prabha Shankar Kundur
Prabha was also active in CIGRE for
many years. He first served on Study Committee (SC) 38 as
the Canadian representative and, later, as chair of Adviso-
ry Group 2. During the restructuring of CIGRE in 2002, he
served as the first chair of SC C4 on System Technical Per-
formance from 2002 to 2006 and, subsequently, as a mem-
ber of the Administrative Council from 2006 to 2010. He re-
ceived the CIGRE Technical Committee Award in 1999, the
CIGRE Honorary Member Award in 2006, and the CIGRE
Medal in 2014.
A s president of Kundur Power System Solutions
I n c . , Prabha performed international consulting and
delivered advanced-level technical courses for utilities,
manufacturers, and universities around the world. He
Digital Object Identifier 10.1109/MPE.2018.2877874
Date of publication: 7 January 2019
previously served as president and CEO of Powertech
Labs Inc., and, before that, he worked for nearly 25 years
at Ontario Hydro, where he held senior positions involv-
ing advanced power system planning and design. Much
of his work focused on modeling and
measurement tools, analysis methods,
and control techniques that enhance
power system stability.
Prabha was elected as a fellow of
the Canadian Academy of Engineering
in 2003 and a foreign member of the
U.S. National Academy of Engineering
in 2011. He was awarded two honorary
degrees: doctor honoris causa by the
Politehnica University of Bucharest, Ro-
mania, in 2003 and doctor of engineer-
ing, honoris causa, by the University of
Waterloo, Canada, in 2004.
After earning his Ph.D. degree in
1967 in electrical engineering from the
University of Toronto, Prabha returned to India to teach at
Bangalore University before accepting a position as analyti-
cal engineer at Ontario Hydro in 1969. He also served as an
adjunct professor at the University of Toronto from 1979 to
2017 and the University of British Columbia from 1994 to
2006. He authored Power System Stability and Control (Mc-
Graw-Hill, 1994), a comprehensive book used worldwide by
industry practitioners and students alike.
To recognize his technical leadership, the IEEE PES Prab-
ha S. Kundur Power System Dynamics and Control Award
was established in his honor in 2012 to acknowledge out-
standing contributions to the understanding and control of
the dynamics of the interconnected bulk power system.
Prabha is survived by his wife of 55 years, Geetha, his
daughter, Deepa, and his two grandchildren.

shifting trends in the electric industry
were discussed as well as strategy im-
plications. six speakers from an interna-
tional cross section of organizations deal-
ing with diverse aspects of technology and
energy integration provided insight into
initiatives that are currently underway
and their visions of the future. the excit-
ing lineup of participants included
✔ Maria Pope, president and chief
executive officer, Portland Gen-
eral electric
✔ Larry bekkedahl, vice president,
transmission and Distribution,
Portland General electric
✔ Peter Jørgensen, vice president,
Associated Activities, energinet
✔ Mark Ahlstrom, vice president,
renewable energy Policy, Nextera
✔ heather rosentrater, vice president,
energy Delivery, Avista Corp.
✔ David Danner, chair, washing-
ton Utilities and transporta-
tion Commission.
timely and emerging topics were ex-
plored in various session formats during
the week, including the highly attended
super sessions, where topics closely fol-
lowed this year’s theme of reimagining
the electric grid:
✔ data science and data quality as
applied to power systems
✔ lower inertia and weaker grids
✔ new application of technology in
the power industry
✔ emergency response.

january/february 2019 ieee power & energy magazine 95

Long-time and new friends enjoyed the reception at the OMSI.

The poster sessions stimulated interesting discussions.

Many attendees congratulated award winners at the dinner.

before the week was done, more than
500 tutorials, technical sessions, pa-
t he annual Pes
Awards Gala once again
The creativity Although it seems
that the 2018 General
per sessions, poster sessions, commit- recognized our peers and pioneering Meeting just happened,
tee meetings, working group meetings, by applauding their we are hard at work on
and task force meetings were held, vision, leadership, and technological the 2019 General Meet-
propagating the latest technical infor-
mation and developing leading prac-
innovations. the cre-
ativity and pioneering
excellence ing, which will be held
4–8 August in Atlanta,
tices and standards for the power and technological excel- of the award Georgia. the steering
energy industry. lence of the award win- and Local organizing
the annual student/Industry/faculty ners was celebrated by winners was Committies plan to raise
Luncheon was a hit with students and
employers alike, who dined together
a smooth jazz band and
a delicious meal. we
celebrated by the bar again as we ex-
plore the theme expect
and then participated in a well-attended hope you’ll consider a smooth jazz Uncertainty/Prepare
job fair. this, along with the scholar- nominating a colleague to Adapt. we hope to
ship Plus reception and student poster in this upcoming year band and a see you in Atlanta in
session, highlighted a compelling stu-
dent Program.
and attend the 2019
Awards Gala.
delicious meal. August 2019!
p&e

january/february 2019 ieee power & energy magazine 97

 calendar
T
T H E I E E E P ow E r & E n E rg y
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.
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
February 2019
IEEE PES Innovative Smart Grid
Technologies (ISGT 2019), 18–21
February, washington, D.C., United
States, contact 2019ISgT@ieee.org,
http://ieee-isgt.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/
Digital Object Identifier 10.1109/MPE.2018.2873954
Date of publication: 7 January 2019
98 ieee power & energy magazine
PES meetings
for more information, www.ieee-pes.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, http://iemdc.org/
IEEE PES Innovative Smart Grid
Technologies Asia (ISGT Asia 2019),
21–24 May, Chengdu, China, contact
Qi Huang, hwong@uestc.edu.cn, http://
sites.ieee.org/isgt-asia-2019/
June 2019
14th International Conference on
Transmission and Distribution Con-
struction, Operation and 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),
4–8 August, Atlanta, georgia, United States,
contact Matt Stryjewski, matthew.stry
jewski@ieee.org, http://pes-gm.org/2019/
IEEE Electric Ship Technologies
Symposium (ESTS 2019), 13–16 Au-
gust, washington, D.C., United States,
contact Scott Sudhoff, sudhoff@purdue
.edu, https://ests19.mit.edu/
September 2019
IEEE PES Innovative Smart Grid
Technologies Latin America (ISGT LA
2019), 15–18 September, gramado, Bra-
zil, contact gabriel Arguello, garguello@
cenace.org.ec, http://ieee-isgt-latam.org/
I EEE PE S Innovative Smart Grid
Technologies Europe (ISGT Europe
2019), 29 September–2 october, Bu-
charest, romania, contact george Cris-
tian Lazaroiu, cristian.lazaroiu@upb.ro,
http://sites.ieee.org/isgt-europe-2019/
December 2019
IEEE PES Asia-Pacific Power and
Energy Engineering Conference
(APPEEC 2019), 1–4 December, Macao,
China, contact Man-Chung wong,
cmwong@umac.mo
p&e
january/february 2019
Digital Object Identifier 10.1109/MPE.2018.2885104
 in my view (continued from p. 104)
problem and the search for alternatives,
usually some form of long-term capac-
ity payment, to restore the economics of
investment. These alternative approaches
typically postulated some traditional
generating-capacity type to serve as the
benchmark for the cost of new entry to
identify and pay the missing money.
The early and growing penetration
of renewables has been driven by man-
dates and subsidies. At some point, the
likely importance of these top-down ap-
proaches will diminish for at least one
of two reasons. The cost of the subsidies
and mandates will be too great, and the
political economy will shift away from
this approach. Or the cost of renewables
will decline to the point where they are
fully economic without the mandates and
subsidies, which will then atrophy and
disappear. The interesting questions
are how can economically efficient re-
newables alter the fundamentals and how
will this dictate changes in the basics of
efficient electricity market design?
Going forward, the increasing role
of intermittent renewables and the tech-
nological potential of distributed energy
resources, especially including demand
participation, might seem to turn over
the assumptions of 30 years ago and
call for an entirely new approach. From
this perspective, large-scale thermal
generation, with relatively high short-
Short-Run Electricity Market
with Zero Marginal Cost Renewables
Price at
Energy Price (¢/kWh)
7–7:30 p.m.
Price at
9–9:30 a.m.
Demand
Price at
2–2:30 a.m.
2–2:30 a.m.
Q1
figure 2. The renewable-only spot market.
100 ieee power & energy magazine
run operating costs, will be replaced by
capital-intensive but low or zero-vari-
able-cost resources. Distributed demand
participation will be evermore impor-
tant in managing overall system balance
and security. The result will be a much
less important role for spot markets and
the associated spot locational marginal
prices. So the argument goes that new
revenue models will be required, and
the fundamentals of the basic electricity
market will have to change.
Despite the appeal of this logic, it de-
pends on implicit assumptions that are
incorrect and point in the wrong direc-
tion for market reform. One implicit as-
sumption is that the basic analysis cannot
accommodate a world with a large frac-
tion of energy supplied by zero-variable-
cost renewable energy. The argument is
that zero-variable-cost supply will pro-
duce low spot prices that could never
support efficient investment.
The flaw in the argument is revealed
by looking again at the implied spot
market. Consider the extreme case in
Figure 2. The short-run supply curve,
or dispatch stack, is now replaced with
a green supply that is completely zero-
variable-cost renewables. In most hours,
the implied efficient price is zero, and
demand participation would expand to
take advantage of the free energy. In
some hours, when the available capacity
Short-Run
Marginal
Cost
Scarcity Price
Demand
7–7:30 p.m.
Demand
9–9:30 a.m.
Q2 Qmax
(MW)
is fully utilized, the price rises to limit
demand. In these periods, the short-run
price is entirely determined by the scar-
city value. But in all other respects, the
analysis of basic principles is the same
as it is for the Figure 1. The dispatch
quantity changes over time, but the spot
price is always equal to the short-run
variable cost, including the marginal
cost of scarcity. Efficient spot prices are
part of the solution derived from basic
principles. And with greater reliance
on zero-variable-cost resources that can
change availability over short horizons,
the scarcity component of spot pricing
becomes even more important. In the
extreme case, scarcity pricing is the only
short-run incentive that matters.
Another implicit assumption of the
flawed argument is that demand partici-
pation can be managed without efficient
pricing mechanisms. Under the current
supply configuration with large thermal
generators, it is possible to imagine cen-
tralized control without the benefits of
spot prices. But even for the present case,
it is commonplace for system operators
to observe how much better the system
works when locational prices provide the
right incentives and reinforce the choic-
es of economic dispatch. Moving to a
greater reliance on distributed resources
(many and small) and demand participa-
tors (many and small) leads inexorably
to a greater need to have real-time spot
prices that send the right price signals.
The central control of distributed re-
sources would not be feasible, and prices
must provide the needed incentives. Fail-
ure to provide the right price signals will
lead to distributed decisions that would
undermine efficient operations.
Hence, the increased arrival of renew-
ables and a greater reliance on distributed
resources both point to fundamentals that
reinforce rather than invalidate the fun-
damental logic of electricity spot-market
operation and pricing. This overview
reveals that the problem lies not in the
fundamentals of market design but rather
in the flaws of implementation that have
led to the distracting focus on capacity
markets and other ad hoc approaches for
january/february 2019
 2019

10 Years of ISGT!
Innovation for a Flexible and Resilient Grid
REGISTER NOW
2019 Innovative Smart Grid Technologies (ISGT 2019)
18-21 February 2019 | Grand Hyatt Washington, Washington DC

Join experts from around the world as we gather at the ISGT North America Conference to
share experiences and lessons learned, and to raise awareness and understanding of the latest
concepts, applications, and technologies.

ISGT 2019 provides a forum to discuss the latest issues, trends, and emerging and innovative
technologies for grid modernization in the face of challenges of a rapidly changing environment
resulting from the dramatic increase in deployments of renewable and Distributed Energy
Resources (DERs) and the emergence of new business and operating concepts and services.
The topics discovered include prosumers, microgrids, aggregators, distribution markets, and
platforms for the generation, commercialization, and management of electricity.

The conference is organized along three tracks:

• Transmission and distribution systems planning and analysis
• Technology applications and supporting tools for transmission and
distribution system operations
• Market and policy considerations in facilitating innovation and enabling
a flexible and resilient grid

For more information visit: ieee-isgt.org

Digital Object Identifier 10.1109/MPE.2018.2885103

 correcting the defects of the spot-market
implementation. There are small and
large changes in spot-market implemen-
tation that have been long needed, and
these reforms will be even more impor-
tant for the future electricity system.
The first reform is better scarcity pric-
ing. The example of the Electric Reli-
ability Council of Texas market stands
out as guiding the way. Embedded in
the original spot-market analysis, an as-
sumption was that demand participation
would handle the needed scarcity pric-
ing. But there is a chicken-and-egg prob-
lem: without scarcity pricing, there is not
enough incentive for demand participa-
tion. A missing element is in the pricing
of operating reserves, i.e., the short-term
capacity set aside to deal with unexpected
changes over the next dispatch intervals.
This necessarily administrative construct
can be addressed with the operating re-
serve demand curve, based on the value
of lost load and the probability that load
will be curtailed. As in Texas, this cre-
ates the right incentives for generation to
be available and, through the associated
energy price, sends the virtuous signal
to demand participation and distributed
resources to respond to the needs of ef-
ficient dispatch. This separates scarcity
pricing from the exercise of market power,
adjusts the implied scarcity price over all
hours, and provides good incentives with-
out requiring the system operator to de-
clare an emergency, and an alternative to
the administrative determination of long-
term capacity payments.
A second reform would be to give
greater attention to multiperiod dispatch
and pricing. In particular, ramping con-
straints can and do limit the ability to
respond to short-term changes in net de-
mand that will be of increasing importance
with the expansion of intermittent resourc-
es. This sometimes goes under the head-
ing of flexibility requirements. In practice,
much of the problem relates to actual
dispatch models using something close to
a single-period formulation with the as-
sociated prices. This inherently ignores
the value of ramping. The computational
problems of a multiperiod dispatch are
relevant but should not be determinative.
With a multiperiod dispatch longer than
102 ieee power & energy magazine
the range of the ramping constraints, spot
prices in all periods can reflect the value
of ramping flexibility and provide market
rewards for the generators and demand
participants that can and do respond, with-
out the need to discriminate by creating
special categories of dispatch resources.
One of the simplifying assumptions
of the basic spot-market model is that
everything is so flexible that there are no
start-up costs, no minimum run times,
and so on. Some markets are premised
on the assumption that these real-world
complications are so minor that they can
be ignored. But ignoring these complica-
tions becomes more important in a system
where the remaining thermal resources are
transformed from base-load facilities that
are always running to dispatched resourc-
es that keep changing output to meet the
system requirements. In this model of the
future, the costs of startup and minimum
load complicate the pricing analysis and
lead to the requirement for uplift payments
to ensure participation in the dispatch. A
natural extension of the basic analysis of
efficient spot prices is found under the
heading of extended locational marginal
prices (ELMPs) that preserve, as much as
possible, the basic pricing arguments and
minimize the need for uplift payments.
Although approximations of this pricing
reform have been in use for many years,
the more general application is receiving
increased attention with the increased pen-
etration of intermittent resources. Further-
more, implementing the ELMPs in spot
markets interacts in a natural way with the
need for multiperiod dispatch and pricing,
so the reforms can proceed together.
The central role of the real-time spot-
market design is one of the key findings in
the analysis of the fundamentals of elec-
tricity markets. The associated locational
pricing model resolved the longstanding
dilemma of how to charge for the op-
portunity costs of transmission, provides
the foundations for financial transmission
rights (FTRs), and creates the benchmark
for the design of day-ahead markets that
must be compatible with real-time pric-
ing. The forward markets allow for a
reconfiguration of FTRs and the intro-
duction of virtual bidding for day-ahead
financial contracts that will be settled
against the real-time spot price. The value
of these prominent best-practice features
in existing electricity markets has been
well established in actual implementation.
Reconsidering of these best practices in
light of the increase in intermittent gener-
ation and distributed resources reinforces
the importance of the fundamentals.
The problem is not the underlying
structure of the theory of electricity spot
markets. The basic model works in the-
ory and performs well in practice. But
the practical implementations have used
certain shortcuts, or averted eyes from
the basics, and created imperfect market
implementations that seem to be incom-
patible with the needs of the electricity
markets of the immediate future.
The argument here is that the funda-
mentals continue to point in the same
direction. The defects in practical short-
term markets are not minor issues that can
be ignored. But a comprehensive rethink-
ing of the market design is not required.
The importance of the fundamentals and
the need to continue to improve on the
market designs already in place must be
recognized. The expansion of intermit-
tent and distributed resources makes this
more urgent, but it would be a good idea
in any event. The big mistake would be to
continue to create new products and sub-
sidies in the futile attempt to replace mar-
ket incentives with central procurement
directives. The proper challenge is to take
the fundamentals seriously, and follow
where they lead: get the prices right.
For Further Reading
ERCOT. (2014). About the operating re-
serve demand curve and wholesale electric
prices. [Online]. Available: http://www
.ercot.com/content/news/presentations/
2014/ORDCUpdate-FINAL.pdf
PJM Interconnection. (2017). Pro-
posed enhancements to energy price for-
mation. [Online]. Available: http://www
.pjm.com/-/media/library/reports-notices/
special-reports/20171115-proposed-enhance
ments-to-energy-price-formation.ashx
F. C. Schweppe, M. C. Caramanis,
R. D. Tabors, and R. E. Bohn, Spot Pric-
ing of Electricity. Norwell, MA: Kluwer
Academic, 1988.
p&e
january/february 2019
advertisers index
The Advertisers Index contained in this issue is compiled as a service to our readers and advertisers: the
publisher is not liable for errors or omissions although every effort is made to ensure its accuracy. Be sure
to let our advertisers know you found them through IEEE Power & Energy Magazine.

Company page# URL Phone

ASPEN, Inc. CVR4 www.aspeninc.com +1 650 347 3997
Bigwood Systems, Inc. 6 www.bigwood-systems.com +1 607 257 0915
CYME CVR 2 www.eaton.com/cyme +1 800 361 3627
Delta Star 96 deltastar.com
DIgSILENT GmbH 3 www.digsilent.de
EasyPower LLC 17 www.EasyPower.com/1584
Electrical Studies, Corp. 97 www.electrical-studies.com
ETAP 15 etap.com/arcflash +1 949 900 1000
Manitoba Hydro International Ltd. 7 www.pscad.com
Massachusetts Water Resources Authority 96 https://mwra.applicantpro.com/jobs +1 617 242 6000
NR Electric Co., Ltd 12 www.nrec.com +86 25 8717 8888
P & R Technologies 19 www.pr-tech.com +1 800 722 8078
Powertech Labs, Inc. 11 www.dsatools.com
PowerWorld Corporation 8 www.powerworld.com +1 217 384 6330
RTDS Technologies, Inc. 9 www.rtds.com
Siemens Power Technologies International CVR 3 siemens.com/electrical-digital-twin
SKM Systems Analysis, Inc. 5 www.skm.com +1 800 500 4SKM

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.2018.2884804

Reach Decision Makers in the Power & Energy Market With

IEEE PES Support Opportunities!

Advertise in the Following Publications:

Power and Energy Magazine
- The global source of information for power and energy professionals

PES Monthly eNews Update

- Build your brand with high-impact digital ads on our eNewsletter and website,
reaching more than 38,000 subscribers monthly

IEEE Electrification Magazine

- A publication for industry practitioners focused on advanced concepts, technologies,
and practices associated with all aspects of electrification

Support Upcoming Conferences:

IEEE PES General Meeting 2019
- August 4-8, 2019 in Atlanta, GA. Raise the profile of your organization among more IEEE PES
GENERAL MEETING

than 3500 electric energy professionals who are your current and potential customers. 4-8 AUGUST
ATLANTA, GA USA

For more information contact: Erik Henson | Media Director | (352) 333-3443
Naylor Association Solutions | ehenson@naylor.com
Digital Object Identifier 10.1109/MPE.2018.2885098
 in my view
O
ORGANIzED WHOLESALE ELEC-
tricity markets in the United States fol-
low the principles of bid-based, securi-
ty-constrained, economic dispatch with
locational marginal prices. The basic
elements build on analyses done when
large thermal generators dominated
the structure of the electricity market in
most countries. Notable exceptions were
countries like Brazil that utilized large-
scale pondage hydro systems. For such
systems, the critical problem centered on
managing a multiyear inventory of stored
water. But for most developed electric-
ity systems, the dominance of thermal
generation implied that the major inter-
actions in unit commitment decisions
would be measured in hours to days, and
the interactions in operating decisions
would occur over minutes to hours. As a
result, single-period economic dispatch
became the dominant model for analyz-
ing the underlying basic principles.
The structure of this analysis inte-
grated the terminology of economics
and engineering. As shown in Figure 1,
the increasing short-run marginal cost
of generation defined the dispatch stack
or supply curve. Thermal efficiencies
and fuel costs were the primary sources
of short-run generation cost differences.
The introduction of markets added the
demand perspective, where lower prices
induced higher loads. As demand shift-
ed over the day, prices would rise and
fall. The rents earned by generators, in
the periods when prices were higher
Digital Object Identifier 10.1109/MPE.2018.2871736
Date of publication: 7 January 2019
104 ieee power & energy magazine
market design practices
which ones are best?
Short-Run Electricity Market
Price at
7–7:30 p.m.
Energy Price (¢/kWh)
Price at
9–9:30 a.m.
Price at
2–2:30 a.m.
Demand
2–2:30 a.m.
Q1
figure 1. The spot market.
than dispatch costs, would provide the
contribution for covering the cost of
investment. With the accompanying
simplifying assumptions, this efficient
dispatch and pricing model would have
been all that was be needed to support
operating and investment decisions.
More than 30 years ago, Schweppe et al.
showed how this basic efficient market
could be expanded to include the effects
of transmission and the associated loca-
tional prices. In principle, extending this
model to include multipart bids, look-
ahead, and multiperiod optimization
presented no major challenges.
The road to success with this market
design was littered with prominent failed
attempts to avoid the basic analysis and
dismiss the special characteristics of
security-constrained economic dispatch.
Eventually, the expensive failures showed
William W. Hogan
Short-Run
Marginal
Cost
Demand
7–7:30 p.m.
Demand
9–9:30 a.m.
Q2 Qmax
(MW)
the importance of the basic principles and
good market design. Not every electricity
market in the world has made this transi-
tion. However, it is clear that if open ac-
cess and nondiscrimination principles set
the constraints, there is only one market
design that meets the test, and the critical
elements of this efficient market design
have been adopted by every organized
wholesale market in the United States.
The practical application of the ba-
sic principles has been highly successful
but not always perfect. A major problem
occurred with implementation details that
tended to depress the spot energy price,
particularly during periods of constrained
capacity. The defect of the twin absences
of demand participation and explicit
scarcity pricing created the missing-money
(continued on p. 100)
january/february 2019
Streamlined renewable
integration analysis
Leverage existing GIS data to optimize
processes and planning studies

The grid is becoming increasingly connected at the distribution level

due to the growing penetration of Distributed Energy Resources
(DERs). Power distribution planners are experiencing more
interconnection requests – causing longer queues and additional
manual efforts to perform studies. Future ready utilities need a
solution to help reduce efforts, automate processes, and maximize
ROI. With the Siemens Electrical Digital Twin solution, distribution
planners can streamline interconnection processes and leverage
data from existing GIS systems for more accurate planning studies.

siemens.com/electrical-digital-twin
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