Professional Documents
Culture Documents
AD917001EN-2015
magazine
58
items on stage—images licensed by ingram publishing
©istockphoto.com/Jamesgdesign
electric lights— ©istockphoto.com/lapandr,
stage—©istockphoto.com/Vectorpocket,
on the
cover
features
contents
24 Electrify Everything? 69 Heat Electrification
By Philip Sterchele, Andreas Palzer, By Steve Heinen, Pierluigi Mancarella,
and Hans-Martin Henning Ciara O’Dwyer, and Mark O’Malley
columns &
108 departments
4 From the Editor 108 History
10 Letters to the Editor 122 Society News
12 Leader’s Corner 127 Calendar
20 Guest Editorial 132 In My View
robert lobenstein
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.
PowerFactory 2018
DIgSILENT has set standards and trends in power system modelling,
analysis and simulation for more than 25 years. The proven advantages
of the PowerFactory software are its overall functional integration, its
applicability to the modelling of generation, transmission, distribution and
industrial grids and the analysis of the interactions of these grids.
W
W h at a n exciti ng as batteries and electrically
time to be in the electric produced fuels, and new de-
power industry! as mem- mand-response technolo-
bers of the ieee, we stand gies all playing key roles in
ready to deal with the rapid meeting requirements for
growth of state-of-the-art ancillary services, peak shav-
technologies, the major ing, and valley filling.
transformation of the grid
to one with ever-more re - In This Issue
newable resou rces, a nd this issue of IEEE Power &
challenges arising from Energy Magazine summa-
interactions between the rizes the scopes, assump-
bulk system grid and the tions, methods, and results
distribution system. given of scenario analyses achiev-
all this activity, how do we ing low-carbon futures. the
©istockphoto.com/John1179
Electricity
Contact us:
dsainfo@powertechlabs.com
81022-0005
sales@etap.com +1 949.900.1000
speakers at Student
Branches and mentor-
A common slav M. Begovic and
Miriam P. Sanders.
company that quickly succumbed to
completion, which implemented other
ing individuals. theme is the in “Society news,” groundbreaking advances in the state-
we also recognize the of-the-art of those times.
Vote! need for recent passing of Mike
t h i s issue fe a t u r e s
end-use adibi, who made ma- Thanks
st a t e m e nt s by can- jor cont r ibut ions to a special note of appreciation to Mel
didates for the ieee electrification system restoration. Olken, who continues to provide gui-
Division Vii delegate- dance and tutelage, and to ieee pub-
ele ct /d i re ctor- ele ct as a way of History lications staff who make this publi-
2019 and ieee Divi-
sion Vii delegate/di-
reducing carbon in the “history” column,
Joseph J. cunningham
cation possible. thanks to the many
contributors to this issue, especially
rector 2020–2021 (the emissions. shares information on our guest editor, Mackay Milller, and
elected individual will the early history of the authors. a particular note of apprecia-
serve both as both delegate and direc- electrification of Lower Manhattan, tion to associate editor John Paserba
tor). Please see the “Society news” new York. he discusses the rapid rise and Robert c. henderson, who pro-
column for an introduction to the two of innovative technology developed by vides editorial assistance.
p&e
highly qualified candidates, Miro- L e o Da f t a nd t he excelsior Power
R
READERS ARE ENCOURAGED tO Author’s Reply there could be two alternating sources
share their views on issues affecting In the ICS text, Electric Lighting and (top then bottom) of light. this may be
the electric power engineering profes- Railways (Scranton, Pennsylvania, the case but for such a brief interval that
sion. Send your letters via e-mail to 1901, pp. 18:6–10), the unknown author it would not be discernable to the hu-
Michael Henderson, editor-in-chief, mih states that for the open man eye. It should be
.psat@gmail.com. Letters may be ed- (no outer globe) “al- emphasized that the ac
ited for publication. ternating current arc, Your January/ arc lamps were not in
both carbons become widespread use. Pro-
When Electrodes Erode pointed or have very
February 2018 jecting light up was not
Your January/February 2018 issue has small craters, so that issue has a fine of much use in street
a fine article about arc lights (R. D. the light is thrown up- lighting, the realm of
Barnett, “Arc lighting systems,” IEEE wards much more than article about most arc lights, so not
Power & Energy Magazine, vol. 16, no. 1, with the direct-current much information was
pp. 56–64, 2018). I learned that in a dc lamp.” Presumably, the
arc lights. provided for students
arc, the positive electrode erodes in a lamps that produce cra- who would not fre-
concave shape, thus casting light down- ters differ from those that produce points quently encounter ac arc lights.
ward along the axis of the carbon rods. in some unspecified way. About the projector question,
the article did not say how the elec- With a closed arc lamp, the light pp. 18:89–90 from the same book deal
trodes erode when excited by ac. Do distribution is similar to the open ac with arc lamps used for “photo-en-
both electrodes develop hollow de- lamp—light is also thrown upward but, graving work, blueprinting, search-
pressions? Are there two light peaks in this case (closed), both carbons have lights” and for “projection work.” In the
along each electrode? I seem to remem- flat ends. the author says that this is latter type, dc arc lamps were used be-
ber that movie projectors (probably ac) a result of the arc’s tendency to “shift cause, in ac lamps, the arc hummed
mounted the rods or thogonal to the around over the ends,” but he offers no loudly and moved around more than
line of sight to the screen, implying a mechanism to explain the phenomenon. for a dc arc. Focusing mirrors were used
nondirectional arc. In closed lamps, the erosion of the with projector arcs, and the feed mecha-
Myron Kayton carbons is greatly reduced because of nism was more precise to keep the arc at
the partial exclusion of air, and perhaps the focal point of the mirrors. I hope this
this affects the shape in some way. With answers your questions. Let me know if
respect to the question regarding the two you need anything more.
Digital Object Identifier 10.1109/MPE.2018.2819027
light peaks, I believe the reader is imply- Bob Barnett
Date of publication: 18 June 2018 ing that, because of the shifting polarity, p&e
info@pscad.com
pscad.com
Available in accessible formats upon request. Powered by Manitoba Hydro International Ltd.
leader’s corner
Frank Lambert
T
The Ieee has been a very Im- launched our scholarship+ Program to the number of Pes sb Chapters had in-
portant part of my professional career attract students back to our profession. creased to 240. Pes sb Chapters num-
since my days as a student member Pes has been and continues to be bered 316 at the end of 2017, as shown
at Georgia Tech. after graduation in vital to our engineers working in in- in Figure 1.
1973, I began work at Georgia Power dustry. many engineers from my gen- The growth exper ienced in sb
Company as a distribution engineer eration are retired or soon will be. We Chapters in regions 1–7 (the United
and started attending our local atlanta need to work together with our local states and Canada) has been much less
Ieee Power & energy society (Pes) Pes professional and student branch than that enjoyed in other regions. To
Chapter meetings. These meetings were (sb) Chapters to help bring in the next bring the same level of excitement to
very beneficial—they helped me grow generation of power engineering pro- regions 1–7, a special one-time stu-
my professional network and ex- fessionals and get them involved in Pes dent Congress was held in boston in
pand areas of expertise for my career. upon graduation. august 2017.
The electric utility industry in the Pes has experienced phenomenal The Third Pes student Congress is
United states suffered a significant growth in our sb Chapters since we planned for 25–27 august 2018 in são
downturn in the early 1990s and all held our first Pes student Congress in bernado do Campo, são Paulo, brazil,
but stopped hiring new engineers. This ankara, Turkey, in august 2014. at the to continue the growth. This congress
trend continued until 2011 when Pes beginning of 2014, Pes only had 134 will have three pillars to prepare our fu-
sb Chapters. The second Pes student ture leaders: academic, industry, and
Digital Object Identifier 10.1109/MPE.2018.2823478
Congress was held in Kuala Lumpur in entrepreneurship. Our total Pes member-
Date of publication: 18 June 2018 august 2016. by the end of that year, ship was 38,925 at the end of December
350
250
200
150
100
50
0
R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 Total
2014 5 2 6 3 4 4 3 23 36 48 134
2015 10 2 8 3 4 8 3 25 49 63 175
2016 11 2 11 4 4 8 3 39 62 96 240
2017 12 2 11 5 5 11 5 57 90 118 316
IEEE Regions
AFL offers electric utilities a wide range of custom-designed solutions that will help increase efficiency of
networking operation centers, power-generating facilities and substations. With more than 80 years of
industry expertise, our substation solutions are developed for applications up to 765 kV and available for
Swage, bolted or welded connections in bronze or aluminum materials. At AFL, we aren’t just one piece of the
solution, we are the solution.
AFLglobal.com/IEEE
800-235-3423
This can include providing speakers
table 1. PES student members by Region.
for sb Chapter meetings, inviting sb
PES Student PES Graduate Chapter members to attend local Pes
IEEE Region Geographical Area Members Student Members
Chapter meetings free of charge, vol-
1–7 United States and 704 979 unteering to serve as mentors for stu-
Canada
dents, and sponsoring technical tours.
8 Africa, Europe, and 1,308 677 also encourage them to get involved
Middle East in some of the leading initiatives, like
9 Latin America 1,682 236 Ieee smart village, Ieee smart Cit-
ies, and Ieee smart Grid, to experience
10 Asia and Pacific 3,671 1,151
the value of membership while they are
Total 7,365 3,043 still students.
all Pes graduate students are Pes
young Professionals (yPs) and can
2017, with almost 27% of that students. for Pes membership. If they have not hold officer positions in local Pes
see Table 1. seen any value in Pes membership Chapters. We have success cases with
One of the biggest challenges facing as a student, chances are they will not Pes graduate students who were sb
Pes today is the transition from stu- make the transition to Pes member af- Chapter chairs and then transitioned
dent member to member after gradu- ter graduation. to officers in their local Pes Chapter.
ation, and we need your help! This is Consider inviting some past Pes sb
a particularly difficult time financially PES Chapters Chapter chairs and Pes graduate stu-
for many new hires, and often they We encourage all of our Pes Chapters dents (yPs) to get involved as an of-
don’t see room in their limited budget to adopt an sb Chapter in their area. ficer in your Chapter. This action can
5 Years
Single and 3-Phase
Transformer Ratiometers
bring energy and excitement to your and help them start one. It only takes a viduals who have not been a Pes mem-
Pes Chapter! Pes faculty advisor and six Pes stu- ber before 2018. more information is
If your local university doesn’t have dent members. Pes has a “first-year” available at https://www.ieee-pes.org/
a Pes sb Chapter, reach out to them free student membership valid for indi- images/files/pdf/student_flyer.pdf.
PES Young Professionals Chapter locator; see https://www The Bottom Line
We also encourage our yPs to volun- .ieee-pes.org/pes-communities/chapters/ Pes student members need to expe-
teer as speakers for our sb Chapters chapter-locator. rience the value of Pes membership
in their local area. experience has while they are still students. reach
shown that our student members very Other Ideas out to them, and show them the ben-
much like to hear from recent gradu- If you or your Chapter has experience efits of belonging to our Pes family.
ates who are now in the workforce. with other approaches to ease the tran- Today’s Pes student members are
yPs can also serve as mentors and be sition from student member to member, tomorrow’s leaders!
very helpful in providing advice for please send me an e-mail at flambert@
career decisions. T he nea rest Pes ieee.org, and we will schedule a call
p&e
sb chapter can be found in the Pes to discuss.
electrification
its role in deeply decarbonized energy systems
E
“ElEctrify EvErything” has in the first article, researchers from as when electrification is paired with
become a common slogan, an easy-to- the fraunhofer institute for solar En- energy efficiency, zero-carbon elec-
understand pathway promising a clean- ergy systems describe pathways to a tricity, or both. in this way, the authors
er and more efficient energy system. nearly carbon dioxide (cO2)-free german illuminate issues that can help decision
indeed, consensus is emerging that dra- energy system by 2050. the authors makers identify priorities for cost-
matic growth in electrification across lay out a rigorous simu- effective cO2 abatement.
transportation, buildings, and even lation approach to eval- in the third article,
some industrial uses will be a key strat- uate the technical and While the researchers from na-
egy for achieving deep decarbonization.
yet while the concept of electrification
economic feasibility of
achieving three pos-
concept of tional grid in the Unit-
ed Kingdom describe
is easy to understand, many of the key sible cO 2 em issions electrification h o w ele ct r i f icat ion
power system issues remain under- reductions targets by contributes to achiev-
appreciated and unresolved. the full 2050: 80, 90, and 95%. is easy to ing a deeply decarbon-
electrification of transportation, for ex-
ample, would require massive changes
in all three futures, the
transformation of nearly
understand, ized U.K. economy by
midcentury. Drawing
to grid design and operation, and full all economic sectors is many of the on their most recent
electrification of heat networks in cold required. But large and “future Energy scena-
climates would place dramatic new sometimes counterin- key power rios” report, which has
pressures on wintertime supply and de-
livery. in a highly electric economy with
tuitive differences also
emerge in meeting these
system issues been published annu-
ally since 2011, they lay
significant shares of variable wind and three decarbonization remain under- out what would need
solar electricity, systemic mismatches targets. By describing to happen in three key
between supply and demand would ne- and contrasting these appreciated and energy sectors (heat,
cessitate vast amounts of storage as well
as entirely new paradigms for capital in-
differences, the authors
help us understand how
unresolved. transport, and power)
to achieve the stated
vestment and cost recovery. a seemingly small differ- national target of 80%
the goal of this issue of IEEE Power ence in emissions targets can have a large cO2 reduction by 2050. across heat
& Energy Magazine is to provide a snap- system impact. and transport, they find rapid growth in
shot of state-of-the-art research fo- in the second article, researchers end-use electrification and identify cus-
cused on the role of electricity in deeply from the national renewable Energy tomer adoption as a key barrier to the
decarbonized energy systems. four of laboratory (nrEl) describe an almost pace needed for achieving the targets.
the articles take a geographic approach fully electrified U.s. energy system. sum- the fourth article compares and con-
and describe how national or regional marizing a recently published study, the trasts 2030 decarbonization pathways
electric sectors might evolve in highly authors detail scenarios for high elec- in two distinct regions of the United
decarbonized futures. two of the arti- trification in the United states, with states with similar carbon reduction
cles explore topics that cut across many and without accelerated investment goals: california and the seven-state
regions: electrification of heat and mar- in energy efficiency and zero-carbon northeast region. researchers from En-
ket design. the final article summarizes electricity. the authors explore how ergy and Environmental Economics, inc.
a wide range of recent electrification each scenario would impact electric- (E3), national grid U.s., and siemens
simulations and proposes principles for ity load shapes, generation capacity Power technologies international de-
advancing the quality of analysis. expansion, and cO 2 emissions. the scribe what would need to happen in
article also quantifies how much cO2 the two regions to achieve a 40% reduc-
Digital Object Identifier 10.1109/MPE.2018.2824099
reduction can be expected from end- tion by 2030. not surprisingly, given the
Date of publication: 18 June 2018 use electrification on its own, as well different climates, renewable resource
WWW.MEPPI.COM/HVDCDIAMOND
cigrE SeSSiOn
August 26 - 31, 2018
paris - Palais des congrès
47 A TechnicAl
eXhiBiTiOn
*registration *Dernier délai oN 3 floors
Electrify
Everything?
©istockphoto.com/tolokonov
T
The firsT efforTs To limiT greenhouse gas friendly, reliable, and affordable energy supply. By doing so,
(ghg) emissions on an international scale date back to 1997 the first steps for transforming the german energy system
with the ratification of the Kyoto Protocol. all contracting were taken. The primary goal of this transformation, called
states agreed to reduce their emissions by 18% in relation “energiewende,” is to significantly reduce ghg emissions. fig-
to 1990. on 12 september 2015, in the Conference of Par- ure 1 shows the historical emission values from 1990 to
ties 21, a follow-up agreement was elaborated to assure an 2015 in germany divided into energy-related and other emis-
ongoing pursuit of climate protection strategies. in total, 195 sions, such as from agriculture and specific industry pro-
parties to the Convention agreed on common goals to further cesses. The share of the energy-related emissions accounts
reduce their ghg emissions. This climate agreement came clearly for the majority of the total ghg emissions (approxi-
into force on 4 november 2016 and, as of november 2017, mately 85–90%). The blue line highlights the ghg emis-
was ratified by 170 states. sion reduction targets set by the german federal government:
By the end of september 2010, the german federal govern- −20% by 2010, −40% by 2020, −55% by 2030, −70% by
ment had already developed a strategy for an environmentally 2040, and at least −80% in 2050 or −95% wherever possible
(relative to 1990).
Digital Object Identifier 10.1109/MPE.2018.2824100
To investigate what technologies are needed for a techni-
Date of publication: 18 June 2018 cally possible and economically feasible transformation of the
german energy system in line of all energy carriers and consumer sectors—while
with the declared ghg reduction meeting the declared climate targets and ensuring a
targets, various studies by differ- secure energy supply at all times?
ent teams throughout germany 2) how does the resulting cost-optimized system con-
are being carried out. The energy figuration change, especially concerning its degree of
system models used as the basis electrification, depending on the investigated Co2
for those studies distinguish them- reduction targets?
selves by different features. The remix
model, for example, focuses heavily on Overview: The German Energy System
the electricity sector and its link to the Trends in primary and final energy consumption in germany
heating and mobility sector, considering, for have changed over time. Primary energy is defined as energy
example, the import and export of electricity that has not been subjected to any conversion or transforma-
with neighboring countries in a multiple-node tion process, for example, to heat or electricity. no conversion
approach. The aim of the model is to assess future or distribution losses are considered in calculating primary
energy-supply scenarios with high spatial and tempo- energy. final energy, in contrast, is the useful, secondary
ral resolution. like the remix model, the scope model is energy available to the final user, and its calculation includes
a linear optimization program used to assess different energy- conversion or distribution losses. Primary energy consump-
supply scenarios. its typical study area is europe (the Balkan tion per person in germany has decreased over the years from
regions are excluded), while the investigated period is defined 185 kWh in 1990 to 158 kWh in 2015. By comparison, aver-
by target years until 2050. The remod-D model, on the other age consumption in the united states in 2015 was about 80%
hand, is a nonlinear optimization program to assess national higher, at 285 kWh/person. however, while primary energy
energy systems through a single-node approach. its focus is consumption in germany decreased, total final energy con-
identifying a cost-optimized national energy system in line sumption has not changed significantly over time. even though
with set Co2 reduction targets as well as assessing interactions energy is increasingly used more efficiently, growth, both
between different energy carriers and consumption sectors economic and in consumption, prevent an overall decrease in
(see Palzer 2016 in the “for further reading” section for more energy consumption.
details). While these three models achieve a temporal resolu- figure 2 shows the development of final energy con-
tion of 1 h, the Times Paneu model relies on typical days for sumption divided into four consumption sectors: 1) low-tem-
its calculations. The strength of this optimization model is the perature heat, i.e., energy to supply heat for space heating
assessment of interactions among the power, mobility, heating and domestic hot water supply; 2) process heat, for example
and industry sectors, while simultaneously considering a high to dry paper or melt steel; 3) mobility, i.e., road traffic, ship-
geographical resolution. ping, aviation and rail traffic; and 4) electricity, where all
methodological differences as well as a different param- electricity applications are included.
eterization of the models can lead to divergent results. That figure 2 shows that the german energy system relied
is why, in a currently ongoing project called regmex, differ- heavily on fossil fuels such as natural gas, oil, and coal, while
ent energy-system models are compared and used to analyze the total amount of the final energy consumption provided by
the transformation toward a nearly Co2-neutral energy sys- other fuel types, such as firewood, sewage sludge, mine gas,
tem with the aim of identifying robust results, i.e., indispens- and waste heat, amounted to approximately 15% in 2015. The
able technologies to achieve such a transformation. While a distribution of final energy consumption among the four sec-
comparison of the individual models will not be discussed tors in 2015 is roughly evenly distributed (each sector contrib-
here, some of the main findings will be presented to address uting between 18 and 30%). however, the distribution of the
the following two research questions: energy-related Co2 emissions by sector shows a completely
1) how can a cost-optimized transformation of the ger- different picture [figure 2(b)]. The electricity sector alone,
man energy system be achieved—with consideration which accounts for 21% of the total final energy consumption,
1,200
in 2015.
1,000 still, even if the power-supply
–20 %
sector fundamentally transforms,
800 this alone will not be sufficient to
–40 %
achieve the ambitious climate pro-
600 tection targets shown in figure 1.
–55 %
The present situation shows that,
400 –70 % while the power sector is progres-
–80 % sively changing, the goals for other
200
sectors will most likely be missed
–95 % (for example, doubling the annual
0
renovation rate of buildings from 1
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
to 2%). To reduce ghg emissions
by at least 80% and, wherever
figure 1. GHG emissions in Germany from 1990 to 2015, and target values to 2050 possible, 95% below 1990 levels,
(blue dots). The green bars represent energy-related CO2 emissions and the red bars each one of the illustrated sectors
other GHG emissions. Percentage reductions refer to the 1990 value. (Source: (figure 2) will have to undergo
Henning and Palzer, 2015.) a substantial transformation itself,
i.e., follow a strict decarbonizing
is responsible for approximately 44% of the total energy-related strategy. To properly assess the transition of the energy sys-
Co2 emissions as power generation in germany relies heavily tem toward a climate friendly, reliable, and affordable supply
on fossil fuel-fired plants. in 2015, 71% of the gross electricity of energy, it is necessary to consider all sectors as well as
generation was provided by coal (42%), nuclear power (14%), their interactions.
natural gas (10%), and petroleum products and other fuels (5%),
while only 29% came from renewable energy resources. in Methodology
2015, the specific emission Co2 factor for electricity amounted The remod-D model is a bottom-up energy system model
to 535 g/kWh as a result of the extensive use of lignite and developed at the fraunhofer institute for solar energy systems
hard coal, with fuel Co2 emission factors of 337 g/kWhcoal and assesses the transformation of national energy systems to
and 407 g/kWhcoal, respectively, and overall power plant a more environmentally friendly and low-carbon economy.
efficiency below 40%. on the other hand, the share of final The fundamental idea behind the model is the identification
energy consumption for the supply of low-temperature heat of a cost-optimized system structure in line with the set Co2
in 2015 amounted to 30%, yet yielding only 18% of the total reduction targets by taking into account the highly complex
energy-related Co2 emissions. in this sector, gas boilers, not dependencies occurring between different energy carriers and
coal, are the dominant technology. The emission factor for sectors (sector coupling model). The objective is to dimension
natural gas amounts to 201 g/kWhgas, leading to lower emis- all system-relevant technologies in a cost-effective way, start-
sions when compared to the electricity sector. ing from 2015 until the target year 2050.
in this context, the changes in sources of electricity produc- The model calculations are based on a comprehensive
tion have been a central topic for the last few years. for exam- database in which weather data as well as technological and
ple, in July 2011, the german federal government declared a economic parameters are specified for each year from 1990
progressive reduction of nuclear power plants until a complete to 2050. The first part of the database, from 1990 to 2014,
shutdown of all reactors by 2022. similarly, in october 2017, describes technical parameters for all relevant components
germany’s environment minister declared that a withdrawal of the energy system. This includes installed capacities of
from coal-fired power is not a question of whether it should be power plants, storage facilities, heating technologies in build-
done but, rather, when and how it will be accomplished. While ings, building stock (as well as its physical characteristics),
the phase out from nuclear power is final, the retirement of number of vehicles, and transportation technologies. The sec-
coal-fired resources has not yet been legally embodied and is ond part, from 2015 to 2050, deals with assumptions regard-
the subject of controversial debate. at the same time, a further ing cost or performance projections of individual technolo-
increase in installed capacity of renewable energy resources is gies, technical restrictions (capacity limits per year), and other
expected. The government aims for a share of renewable elec- time-dependent characteristics. This information is essen-
tricity generation of 40–45% in 2025 and 55–60% in 2035. tial to determine the mix of generators, energy converters, and
The current development shows that the renewable energy consumers that meet environmental goals at minimal cost.
Low-Temperature Heat Fossil Fuels Low-Temperature Heat Others Industrial Process Heat Fossil Fuels
Industrial Process Heat Others Electricity Mobility Fossil Fuels Mobility Others
figure 2. The (a) development and share of the final energy consumption by sectors and (b) distribution of energy-related
CO2 emissions by sector in 2015. (Source data: BMWI Energy Data, 2017.)
here, the model distinguishes between two kinds of different energy can be supplied in the model using wind turbines (onshore
structures: open and closed systems. and offshore), photovoltaic systems, and hydropower plants.
When considering the technology of an open system, the Biomass can be used either directly or after conversion into a
model has the option to increase or decrease the specific secondary energy carrier. for example, wood can be burned
penetration of the technology as long as it doesn’t violate in boilers to provide process heat for industrial applications
the set expansion potential. for example, onshore wind tur- and for the generation of low-temperature heat in the building
bines can be installed as long as the total installed capacity sector. Biogas systems (gasification systems with subsequent
is below the cumulative technical potential (approximately synthesis into hydrogen, methane, or liquid fuels) and biodiesel
180 gWel in germany that, starting from 2015, is roughly systems are implemented as possible systems for the conversion
5 gWel per year). likewise, the model could also decide of biomass. electrical energy-storage systems in the form of
not to install any onshore wind turbines, such that the total stationary and mobile batteries (in vehicles) or pumped-storage
amount of onshore wind power decreases over the years power plants are implemented as storage systems. hydrogen
through retirements. While the installed capacity of each and thermal hot water storage systems in different orders of
technology can increase or decrease in open systems (i.e., magnitudes are considered as well.
renewable energy resources, power plants, and thermal and energy demand is divided into four groups, according to
chemical storage facilities), in closed systems, the technol- the different fields of use: mobility, intrinsic electricity applica-
ogies are not limited by technical constraints but by others. tions, heat for buildings (residential, nonresidential, and indus-
for example, the cumulative deployment of technologies trial buildings), and process heat in the industry. The mobil-
for space heating and domestic hot water supply is limited ity sector is mapped in detail concerning passenger cars and
by the total number of buildings in germany. here, the trucks, with seven vehicle classes each. The energy demand of
assumption is that every building is equipped with one heat aviation, shipping, and fuel-based railway traffic is considered
generator. Whenever this technology reaches the end of its in the balance, without temporal resolution. The basic elec-
expected lifetime, the model can decide whether to install tricity load is mapped using load profiles based on the data of
a heat generator of the same type or choose a different one. european transmission grid operators reduced by the weather-
This means that the model can optimize the share of each related electric load for heating systems. The load for heating
technology available within the closed system, but it can’t systems is calculated model endogenously and is not included
change the exogenously fixed number of technologies in in the basic load. Table 1 summarizes the main components of
each closed system. the energy system considered in remod-D.
Conventional lignite and hard coal power plants, nuclear
power plants, oil-fired power plants, gas turbines (gTs), com- Comparative Findings and Discussion
bined heat and power (ChP) plants as well as gas-fired and in the following section, the results of three cost-optimized
steam power plants are implemented as generators. renewable scenarios are analyzed in detail. The scenarios differ from
each other in terms of the set Co2 reduction goals by 2050, in the first scenario (80% Co2 reduction), this potential is
which are 80, 90, and 95% reductions in relation to 1990. it is exploited 61%; in the second (90% Co2 reduction), it is exploited
important to note that the energy system here is considered as 84%; and in the third scenario (95% Co2 reduction), it is exploited
an isolated island system to improve comparability with other completely. in concrete terms, starting from today’s installed capac-
involved energy system models. further assumptions will be ity of 91 gWel (nameplate wind and solar energy), this translates
documented in the final report of the regmex project. into a cumulative capacity in 2050 of 323, 440, and 540 gWel
respectively. The simulation results suggest the following:
Power Generation 1) The rise of the cumulative installed capacity of re-
figure 3 shows the cost-optimized development of the installed newable wind and photovoltaic resources is inversely
capacity for renewable energy sources, for each of the three proportional to the total allowable Co2 emissions bud-
scenarios from 2020 until the year 2050. The results show that get, i.e., the lower the allowable Co2 emissions in the
a significant increase of renewable energy resources is needed to energy system, the higher the necessary amount of en-
achieve the scenario reductions of energy-related Co2 emissions. ergy production from renewable resources.
in particular, photovoltaic systems and onshore wind turbines play 2) The relationship between the Co2 reduction target and
a key role when it comes to power generation. Their cumulative the required installed capacity of renewable resources
technical potential for germany is assumed to be 320 gWel and is not linear. moving from a 90% reduction to a 95%
180 gWel, respectively, while it amounts to 40 gWel for offshore reduction requires the same incremental increase in in-
wind turbines. This yields a total technical potential for the princi- stalled capacity (100 gWel) as moving from an 80% re-
pal variable renewable energies of approximately 540 gWel. duction to a 90% reduction.
PV
500 500 500
Onshore Wind
400 Offshore Wind 400 400
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)
figure 3. The annual development of the installed capacity of variable renewable energy resources. (a) 80% reduction,
(b) 90% reduction, and (c) 95% reduction.
figure 4. Temporal development of the installed capacity of thermal power plants. Combined cycle GTs (CCGT), hydro-
gen (H2), gas/methane (CH4). (a) 80% reduction, (b) 90% reduction, and (c) 95% reduction.
figure 4 shows the cost-optimized development of the required installed capacity of thermal power plants shown
installed capacity of thermal power plants for the three con- in figure 4.
sidered scenarios. for all three scenarios, a shift from higher figure 5 shows the cost-optimized development of differ-
ghg-emitting power plants (lignite and hard coal) to ent technologies for the conversion of electricity into syn-
lower emitting plants [Ch4-combined cycle gTs (CCgTs), thetic fuels, i.e., hydrogen, synthetic methane gas, and liquid
Ch4-gT, and h2-gT] becomes clear. major differences are fuels. These technologies are used when renewable energy sur-
evident, however, in the different levels of installed capacity. pluses occur, which is why the installed capacity of synthetic
The total cumulative installed capacity in the 95% reduction fuel technologies increases proportionally with the resulting
scenario in 2050 roughly doubles its initial value, while for expansion of renewable energies shown in figure 3.
the less-ambitious reduction targets (80 and 90%), the initial While the absolute value of the residual load for electric-
total installed capacity value doesn’t change as significantly ity surplus in 2050 amounts to 186 gWel, in the case of an
(+16%, +45%). 80% reduction target, it increases to 256 gWel when consid-
To better understand the illustrated development of the ering a reduction of 90% and to 328 gWel in the case of 95%
thermal power plant fleet, a closer look at the residual load is (Table 2). This nonlinear growth in residual load explains
necessary. The residual load describes the electric load of the why the cumulative capacity for power-conversion technolo-
energy system minus the nondispatchable power generation, gies increases nonlinearly with more restrictive Co2 reduc-
i.e., mainly variable renewable resources. This means that tion targets (figure 5).
whenever the residual load is positive, deficiency occurs and The illustrated power conversion technologies can be op-
power would need to be supplied from other sources. on the erated as flexible loads to reduce eventual electricity surplus
other hand, a negative value represents a surplus of renew- peaks, and they produce climate-friendly synthetic fuels.
able electricity. in either case, a range of different flexibility
options are required to balance the residual load.
in particular, the increasing use of variable renewable re- table 2. Maximum renewable power deficiencies
sources shown in figure 3 leads to higher peak values of the and surpluses in target years 2030 and 2050 for
three considered CO2 reduction targets.
residual load during the year and therefore to a higher need for The increase of the peak values of the residual
flexibility options. Technologies like power storage are effec- load corresponds with the increasing installed
tively used to balance small fluctuations of the residual load. capacity of variable energy resources.
When those fluctuations become more substantial, i.e., the power Power Power
requirement is higher, technologies such as gTs and CCgT Deficiency) Surplus
plants (for renewable power deficiency) or power conversion (in GWel) (in GWel)
technologies such as power-to-gas or power-to-heat (for renew- Reduction of
able power surplus) are mainly operated. Table 2 summarizes Energy-Related Emissions 2030 2050 2030 2050
the maximum and minimum values of the residual load in 2030 Scenario 1: −80% CO2 60 68 75 186
and 2050. (referred to 1990)
Table 2 shows that the maximum renewable power defi-
Scenario 2: −90% CO2 65 69 82 256
ciency (positive residual load) in 2050 for a 95% Co2 reduc- (referred to 1990)
tion target is almost 50% higher than the value of scenarios 1
Scenario 3: −95% CO2 62 101 134 328
and 2 (80% and 90%). This increase of renewable power (referred to 1990)
deficiency corresponds with the illustrated development of
80 Methanation 80 80
Electrolysis
60 60 60
40 40 40
20 20 20
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)
figure 5. The annual development of the installed capacity of technologies for the production of synthetic fuels: (a) 80%
reduction, (b) 90% reduction, and (c) 95% reduction.
These fuels become essential for the decarbonization of those boilers amounts to 21% for an 80% Co2 reduction target,
energy consumers in the energy system, where a complete 9% for a 90% Co2 reduction target, and roughly 4% for 95%
technological reorientation may be hard to achieve under Co2 reduction target.
today’s economic and political structures (for example, avia- electrical heat pumps exhibit the opposite trend. Their
tion or some industrial processes). number increases when considering more ambitious climate
protection goals. The results show that, in the third scenario
Consumption Sectors (−95% energy-related Co2 emissions), electrical heat pumps
figure 6 illustrates the scenario development of heating (air or brine) account for almost half of the heating technolo-
technologies for space heating and domestic hot water sup- gies in the energy system. These technologies bring two main
ply. While the total number of heating technologies is advantages for the overall energy system. first, their coeffi-
directly linked to the exogenously set number of buildings cient of performance allows an advantageous ratio between
in germany (boundary condition), the share of each tech- the required energy for operation and the provided heat
nology is a result of the cost optimization. note that every (for detailed model analysis on heat pumps, see sterchele
heating technology can be supplemented by hot water stor- et al. 2017 in the “for further reading” section). This benefit
age with solar heating and an electric heating rod. in efficiency contributes to the decarbonization of the low-
While gas and oil boilers clearly dominate today’s heat- temperature heat supply. second, electrical heat pumps can
ing technologies (approximately 70% of all installations), be operated flexibly to balance the residual load in times of
their share decreases in all three scenarios. The resulting renewable production surpluses (negative residual load), for
level of reduction is directly linked to the considered Co2- example, to charge thermal storage technologies. as shown
reduction target. Concretely, in 2050 the share of oil and gas for other power conversion technologies, here the need for
30 30 30
Technologies in Millions
25 25 25
20 20 20
Heating
15 15 15
10 10 10
5 5 5
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)
CHP + Fuel Cell Gas Heat Pump Oil and Gas Boiler
Electric Heat Pump (Air, Brine) Biomass Boiler Heating Grid and Geothermal Heat
figure 6. The annual development of heating technologies for space heating and domestic hot water supply. Heating
grids are considered as a combination of CCGT, electrical heat pumps, gas boilers, thermal storage, and solar thermal
technology. (a) 80% reduction, (b) 90% reduction, and (c) 95% reduction.
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)
Battery Hybrid H2-FC ICE Gas ICE Fuel
figure 7. The annual development of vehicles (cars, trucks, and buses) for three considered scenarios: (a) 80, (b) 90, and
(c) 95% CO2 reduction targets. Boundary condition assumption: Trucks can’t be fully electrified.
flexibility options increases due to the increasing installed methane-powered vehicles, iCes (gas), play a role in
capacity of variable energy resources. The number of con- each scenario. at present, this driving technology is less
nections to a heating grid increases as well. heating grids expensive than electric vehicles and emits lower Co2 emis-
in the model are represented as a combination of different sions compared to vehicles based on liquid fuels. however,
technologies, which are CCgT, electrical heat pumps, peak the results show that gas-powered vehicles will persist until
load gas boilers, and solar-assisted heat and hot water tanks 2050 in the first scenario, while being a transitional tech-
(including a heating rod). additionally, CCgT and micro- nology in the other scenarios with more ambitious climate
ChP plants can be operated flexibly by generating power in protection targets.
times of renewable power deficiency, while also charging hot figure 8 shows the technological development for
water tanks. the provision of process heat in the industrial sector. The
figure 7 shows the development of the road transportation parameterizing and calculation in the model distinguishes
sector. similar to the heating technologies, the total num- different temperature levels, allowing, for example, the use
ber of vehicles is set exogenously, while the share of each of electrical heat pumps only for temperature levels below
driving concept results from the cost optimization of the 180 °C.
model. The study assumed an upper limit for battery elec- The development of technologies for the provision of pro-
tric trucks of 5%. assumptions for driving technology and cess heat shows a slower-acting switching behavior when
range restrictions meant that only local electric transport for compared to other consumption sectors (space heating +
trucks was considered in the study. higher penetrations of domestic hot water supply and road transportation). even
electric transport could be considered under advanced tech- though gas and oil-based technologies decrease when consid-
nology scenarios. note that the number of vehicles in fig- ering more ambitious Co2 reduction targets, they are clearly
ure 7 includes both cars and trucks, and buses are included present in each scenario in the target year 2050. The share of
in the category of trucks. electric technologies, such as electric heat pumps and elec-
The development of the road transportation sector (fig- trode boilers, increases with increasing Co2 reduction tar-
ure 7) shows similarities to the development of the heating gets, while remaining fairly constant in the case of an 80%
technologies displayed in figure 6. in all three Co2 reduc- Co2 reduction target. at the same time, coal-fired boilers
tion scenarios, the share of today’s dominating transporta- phase out in the last years of the energy transition path (for a
tion technology [internal combustion engines (iCes) based 95% Co2 reduction).
on liquid fuels] decreases over the course of the years. a industrial ChP stations become less important over time,
faster phase out of this drive technology would be required beginning in the early simulated years. This trend partly arises
for more ambitious climate protection targets. simultane- from the model-assumption that all heating technologies in the
ously, like heating technologies, a shift to electric transport industry sector are not operated responsively to grid condi-
technologies, such as battery electric vehicles, fuel cell elec- tions, i.e., operation in the industry sector will, for business
tric vehicles, or hybrid vehicles (a combination of iCes with reasons, remain independent from the residual load. This leads
battery), emerges. electric vehicles have a higher efficiency to multiple hours during the year when those plants produce
than iCes, and battery-electric vehicles can be used to charge electricity, although a surplus of electricity already occurs
and supply the electric distribution network based on the (negative residual load). This leads to unfavorable operating
residual load. Therefore, they represent an alternative flexibil- conditions for ChP stations, explaining why they are driven
ity option for the energy system. out by other technologies.
60 60 60
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0 0 0
2020 2030 2040 2050 2020 2030 2040 2050 2020 2030 2040 2050
(a) (b) (c)
figure 8. The annual development of technologies for the provision of process heat. (a) 80% reduction, (b) 90% reduc-
tion, and (c) 95% reduction.
Low-Temperature Heat
80 Road Traffic 80 80
Industrial Process Heat
60 60 60
40 40 40
20 20 20
0 0 0
2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050 2020 2025 2030 2035 2040 2045 2050
(a) (b) (c)
figure 9. The degree of electrification of the consumption sectors: low-temperature heat (space heating and domestic
hot water supply), road transportation (cars, trucks, and buses), and industrial process heat for a reduction of energy-
related CO2 emissions of (a) 80, (b) 90, and (c) 95% below 1990 levels. The degree of electrification for the respective
sectors includes low-temperature heat: electric heat pumps; deep geothermic and heat rods road transportation: battery
electric vehicles, fuel cell electric vehicles; and partially hybrid vehicles industrial process heat: electric heat pumps
and electrode boilers.
Electrified
Daniel Steinberg,
Jeffrey Logan,
David Bielen,
Kelly Eurek, and
A
as an energy source, electricity
benefits from a number of desirable characteris-
tics: it can be transported at nearly the speed of light
with transmission infrastructure, it has zero end-
use emissions, it is highly flexible and controllable,
it is now storable at rapidly declining costs, and it can
offer improved service quality relative to conven-
tional fuels. as such, electrification—the conver-
sion of previously fossil-fueled end-use processes
to electricity—has been identified as a key pathway
to a clean, reliable, and secure energy future. elec-
tric vehicles are the most widely cited application
of electrification, but technology improvements in
electrically driven devices for buildings and indus-
trial end uses, including heat pumps for space and
water heating needs, induction stoves for cooking,
infrared or ultraviolet curing processes, and electric
arc furnaces for process heating, could lead to more
widespread electrification across these sectors.
in this article, we report results from a recently
published initial analysis conducted by the national
renewable energy laboratory (nrel) that simu-
lated widespread electrification from present day
through 2050 in the united states. the study focused
on 1) levels of technically achievable end-use electri-
fication, 2) power-sector capacity expansion needs
required to meet the growing demand for electricity
under an electrified future, 3) the impacts of electrification 2050 compared to 2005 levels. ongoing work to analyze the
on electricity demand load shapes, and 4) the implications for economic potential of electrification, the impact on electric-
economy-wide carbon dioxide (co2) emissions. compared to ity demand load profiles, and the ability to operate the grid
a reference case, very high levels of electrification can lead reliably, given high levels of variable generation and smart
to over twice the total electricity demand by 2050, with most demand responsiveness, will add important new insights on
of that in the transportation sector. such high levels of elec- the potential future of electrification in the united states
trification, when combined with additional improvements and beyond.
in overall efficiency and grid decarbonization, can yield
economy-wide carbon emission reductions of 72–75% by Background and Motivation
the u.s. economy relies on a diverse range of fuel sources,
intermediate energy carriers, and end-use and supply tech-
nologies. in 2015, the u.s. energy system consumed 98 qua-
drillion British thermal units (quads) of primary energy, of
which 39% (38 quads) was used for electricity production
and the remainder consumed primarily through the direct
combustion of fossil fuels in the other sectors, such as trans-
portation, industry, and (commercial and residential) build-
ings, as shown in Figure 1. this electricity generation and
direct fossil consumption resulted in 5,259 million metric
tons (MMt) of co2 emissions from the 2015 energy sys-
tem, comprising a large majority of all u.s. greenhouse gas
(gHg) emissions. energy-related emissions represents about
80% of total gHg emissions on a co2-equivalent basis and
about 90% of net gHg emissions when both sources and
sinks are considered. in Figure 1, emissions from electricity
generation are allocated to the sectors and end uses using the
national and annual average emission factor.
the electrification of end-use services coupled with the
decarbonization of electricity generation has long been
identified as a means to achieve, among other things, a
low-carbon future. this low-carbon pathway would reduce
both energy and carbon intensities for the economy along
two primary dimensions. First, end-use electric technolo-
gies are often more energy efficient, requiring less energy
to provide the same service, than conventional fossil-based
options. electric vehicles, for example, are typically three to
four times more efficient than internal combustion engines
in converting on-board energy into power at the wheels,
although assessing differences in overall well-to-wheel
efficiencies requires accounting for factors such as the
©istockphoto.com/metamorworks
Nonelectricity
Other Other
Other
Water Heating
Other Water Heating
Commercial Light Trucks
Air Space Heating
Total 2015 Primary Energy Consumption (%)
Space Heating
75 Paper
Electricity
Mining
Freight Trucks
Refining Other
Other
50
Bulk Chemicals
Refrigeration
Light-Duty Vehicles
25 Lighting Lighting
Other
Space Cooling Space Cooling
Paper
Mining Water Heating Refrigeration
Refining
Bulk Chemicals Space Heating Space Heating
0
(a)
Nonelectricity
Other Water Heating Other
Water Heating
Air Other
Total 2015 Energy-Related CO2 Emissions (%)
Space Heating
Commercial Light Trucks Space Heating
75
Freight Trucks
Paper Electricity
Mining
Refining
Other
Other
50
Bulk Chemicals
25 Lighting Lighting
Other
figure 1. The 2015 U.S. (a) primary energy consumption and (b) 2015 CO 2 emissions by sector and end use. [Data from
“Annual Energy Outlook 2017,” U.S. Energy Information Administration, and figure design courtesy of Jadun et al. (2017).]
Fuel Oil
1.5
Natural Gas
District Heat
1.0
0.5
0.0
Space Lighting Refrigeration Ventilation Space Cooking Water Computing Office Other
Heating Cooling Heating Equipment
End Use
(b)
4
Purchased Steam Coal (Excluding Coke and Breeze)
By-Products Natural Gas and Gas Liquids
3
Fuel Oil and Diesel Electricity
(Quads)
0
CHP and/or
Cogeneration Process
Conventional Boiler
Use
Process Heating
Machine Drive
Facility HVAC
Electrochemical
Processes
Facility Lighting
On-Site Transportation
Conventional Electricity
Generation
End Use
(c)
figure 2. Final energy consumption in (a) residential buildings, (b) commercial buildings, and (c) industry sectors by
end use and fuel. The data for industry exclude energy consumption as a nonfuel (feedstock). (Data from the EIA 2009
Residential Energy Consumption Survey, EIA 2012 Commercial Buildings Energy Consumption Survey, and EIA 2010
Manufacturing Energy Consumption Survey.)
4500 4500
4000 4000
3500 3500
HDV: Hydrogen
3000 3000 LDV: Hydrogen
HDV: Electricity
2500 2500
LDV: Electricity
2000 HDV: Oil/Gas/Bio
2000
LDV: Oil/Gas/Bio
1500 1500
1000 1000
500 500
0 0
2010 2020 2030 2040 2050 2010 2020 2030 2040 2050
figure 3. Vehicle miles traveled from 2010 to 2050 under the reference and high-electrification scenarios. LDV: light-duty
vehicles; HDV: heavy-duty vehicles.
table 1. The industrial end uses and electric technologies selected for analysis.
Percent Electrified
Electrotechnology Industry End-Use Service by 2050
Electrolytic reduction Nonferrous metals, excluding aluminum Process heating 100
Induction heating Metal fabrication Process heating 100
Electric boilers All manufacturing industries Conventional boiler use 100
Resistance heating and melting Glass Process heating 100
Direct arc melting Iron and steel Process heating 21
Industrial process heat pump Food, pulp and paper, and chemicals Process heating 100
cost-effectively serve processes with high-temperature and we simply assume an efficiency improvement of 0.8%/year.
large energy demands. in addition, our analysis does not this estimate was developed by weighting the current oppor-
consider industrial equipment turnover, relative economics tunity for state-of-the-art energy reductions identified by
of electric and nonelectric technologies, and nonenergy ben- u.s. Department of energy energy bandwidth studies by
efits and impacts such as productivity and product quality. their relative fraction of 2010 manufacturing primary energy
in addition to the high-electrification view, we also con- use, assuming these reductions would be achieved by 2050.
sider separate scenarios where significant increases in energy all remaining manufacturing industries not covered by the
efficiency are achieved, in combination with high electrifica- energy bandwidth studies were assumed to reduce energy
tion, across all sectors. We refer to this as the high-combined use by 10% by 2050. these top-down sets of assumptions
scenario, in which transportation efficiency (in terms of miles do not specify the efficiency measures to achieve these rates
per gallon of gasoline equivalent) is assumed to increase but, instead, rely on a wide range of measures such as vehicle
beyond u.s. federal vehicle efficiency standards even after lightweighting, building insulation improvements, and pro-
the standards sunset. For buildings, we assume efficiency cess improvements.
improvements in electric end-use devices from 2017 to 2050 Figure 4 shows annual electricity consumption estimated
at various rates ranging from 0.5%/year to nearly 2%/year. for the high-electrification and high-combined scenarios.
efficiency measures are assumed for services provided by under high electrification, annual electricity demand grows
new electric technologies as well as in other end uses includ- by 2.6%/year resulting in nearly 11,000 tWh of electricity
ing lighting, refrigeration, and other plug loads. For industry, demand by 2050. this growth rate is roughly in line with
figure 4. The reference and incremental electricity consumption from electrification by sector, 2010–2050.
historical growth rates during much of the 1900s and differs seasons, days, and hours play an important role for elec-
significantly from more recent growth (flat) and projected tric system resource adequacy and, ultimately, reliability.
growth rate (0.7%/year) in our reference scenario (shown resource adequacy refers to the ability to meet aggregate
in gray). as shown in Figure 4, most of the incremental elec- electrical demand and is usually considered met when there
tricity consumption growth is in the transportation sector is sufficient firm capacity resources (primarily from the
through the increased use of electric and fuel cell vehicles. supply side but also including demand-side resources) to
For the latter vehicle types, we assume, for simplicity, that meet projected peak demands plus a reserve margin (~15%).
hydrogen production occurs only through electrolysis, which resource adequacy is only one component of reliability but
because of the large share of fuel-cell vehicles and the low plays an important role in utility planning. changes in popu-
efficiency of electrolysis-based hydrogen production results lation and energy use, including new electrical loads from
in very large electricity requirements. in fact, in 2050 under electrification, can impact the timing and magnitude of peak
the high-electrification scenario, about half of the incre- demands and affect supply-side planning.
mental electricity consumption in transportation is used for to demonstrate the effects of electrification on load shapes,
electrolysis-based hydrogen production, whereas the ratio of we calculate the ratio of national peak-to-average demand in our
vehicle miles traveled from direct electric vehicles to hydro- scenarios (Figure 5). in the reference scenario, we assume limited
gen-based fuel-cell vehicles is 3.2 to 1. electricity demand
growth in buildings and industry is also significant in the
high-electrification scenario. as a result, in this scenario, 2050
u.s. electricity consumption is more than twice 2050 demand 1.8
Ratio of Peak to Average Demand
changes to rate structures, end-use technologies, weather, and between variable generation profiles and demand. the amount
other demand-side changes. as a result, the peak-to-average of renewable curtailment can also be affected by how electri-
demand ratio remains constant through time in this scenario, fication impacts load profiles. Whether regulations, market
with peak load 70% higher than average load (59% load factor). designs, and rates enable electrification to support efficient grid
in Figure 5, the “peak” is defined as the top 40 load hours. evolution such as through greater demand-side participation is
this simple definition does not capture all implications, as an important research area.
discussed later, but illustrates the large impact that electri-
fication can play on load shapes. in contrast, the peak-to- Electricity Supply-Side Evolution
average ratio decreases substantially through time in both with Widespread Electrification
the high-electrification and high-combined scenarios. By the impacts of u.s. electrification depend on how the
2050, peak demand is less than 9–13% higher than aver- demand side might evolve but also on the supply-side future
age demand in these scenarios (89–91% load factor). this of the electricity system. We develop various electricity sup-
result is driven in large part by our bounding assumption that ply-side scenarios—responding to the demand-side changes
electric vehicle charging and hydrogen production loads are envisioned—by employing the nrel regional energy
highly flexible. specifically, we assume zero charging dur- Deployment system (reeDs) model. the model simulates
ing the top 40 load hours of the year and that load associated the operation and expansion of the u.s. power system,
with hydrogen production is completely flexible within a day. including power plants, transmission, and storage from pres-
our implicit assumption is that rate structures incentivize ent day through 2050 by choosing the cost-optimal mix of
smart charging and the existence of sufficient infrastructure technologies. (utility stationary storage, as opposed to stor-
to avoid vehicle charging during the peak hours and ade- age in vehicles, is treated as a resource in the modeling.) the
quate hydrogen storage and other infrastructure exist to shift least-cost solutions found by the model are constrained to
hydrogen production to nonpeak hours. given that load asso- meet all regional electric power demand requirements (with
ciated with hydrogen production represents approximately and without electrification), planning and operating reserve
50% of total annual transportation electricity consumption requirements, technology resource constraints, and any pol-
under this scenario (in 2050), the diurnal flexibility ascribed icy requirements. in all scenarios, we use technology cost
to hydrogen production results in highly flexible load in the and performance assumptions from the mid-cost projection
vehicle sector. our analysis does not evaluate the costs to of the nrel 2016 annual technology Baseline and refer-
achieve this level of flexibility, which is highlighted as a ence case fuel price assumptions from the eia 2016 annual
future research need. Without intelligent policies and con- energy outlook.
sumer behavior, the peak-to-average demand ratio is likely Figure 6 shows the annual generation and installed
to remain higher than shown in the outer years of the analy- capacity from four scenarios modeled using reeDs. in the
sis, with important implications for build-out and operation reference scenario, which includes only current policies,
of the grid. demand for new capacity resulting from business-as-usual
How electrification drives electricity consumption pat- growth in load and end-of-life retirement of existing genera-
terns can have other impacts to electric system planning and tors is met predominantly by new wind, solar, and natural
operations beyond resource adequacy. For example, the flat- gas generation. total generation in 2050 under such a sce-
tening of demand profiles can have important implications for nario is 5,300 tWh, of which about 33% comes from wind
the amount and type of supply-side resources; flatter demands and solar (compared to about 7% in 2016), 28% from natural
might drive the preference for technologies that provide energy gas, and 22% from nuclear, hydropower, and other renewable
resources over capacity resources (e.g., resources designed for technologies. While coal generation declines in the long run,
efficient performance and low-cost energy but with higher it still provides about 17% of 2050 generation in this refer-
capital costs; these have traditionally been referred to as base- ence scenario. on a capacity basis, we find 16 gW of average
load resources). in another example, changing demand profiles net annual additions of wind and solar capacity from 2017 to
can impact renewable integration challenges and opportuni- 2050 and about 9 gW of net annual natural gas capacity addi-
ties. the capacity and energy values of wind and solar gen- tions, figures that are roughly in line with u.s. trends since
eration are closely tied to the correlation (or lack thereof) 2010. in other words, without widespread electrification
10,000
Annual Generation (TWh)
8,000
6,000
4,000
2,000
0
2010 2030 2050 2010 2030 2050 2010 2030 2050 2010 2030 2050
(a)
Reference + High Electrification + Grid Decarbonization + High Efficiency
4,000
3,500
3,000
Capacity (GW)
2,500
2,000
1,500
1,000
500
0
2010 2030 2050 2010 2030 2050 2010 2030 2050 2010 2030 2050
(b)
figure 6. The (a) generation and (b) capacity mix over time for various supply- and demand-side scenarios. The first (far
left) panel includes both reference demand- and supply-side evolution. The second panel is end-use electrification only.
The third panel includes electrification and grid decarbonization (an 83% reduction from 2005 levels in 2050), and the
far-right panel combines electrification, grid decarbonization, and end-use energy efficiency.
development in the u.s. electricity sector. these estimates the last column of Figure 6 shows results from a scenario
suggest growth rates in new capacity investments more aligned that combines high electrification, grid decarbonization, and
with those currently found in large developing economies, high energy efficiency. the generation mix in this scenario
such as china and india, as opposed to the development is similar to that from the previous one in percentage terms
from recent history in the united states and europe. (63% wind and solar, 13% ccs, less than 10% fossil with-
combining high electrification with grid decarbonization out ccs, and 14% other); however, the absolute magnitudes
(the third column in Figure 6), the changes in generation are much lower for all technologies. For example, average
and capacity needs are even more pronounced. We define net wind and solar additions are estimated to be 57 gW per
grid decarbonization as 83% reductions in combustion- year from 2017 to 2050 and lower than 5 gW per year of ccs
related co2 emissions from 2005 levels by 2050 (for elec- capacity is needed. this result suggests that efficiency can
tricity production only) and model scenarios with a national help reduce infrastructure needs (and potentially bottlenecks)
electric sector emissions cap achieving this reduction level. that might arise from high electrification and stringent emis-
specifically, we model a co2 combustion cap starting in sions requirements.
2022 based on emissions levels in the reference scenario and
decreasing linearly to the 83% reduction (of 2005 levels) by CO2 Emissions Implications
2050. We include the cap in scenarios with and without high of Widespread Electrification
levels of electrification. in this scenario, 2050 wind and solar the future of energy utilization and supply has obvious
generation comprise nearly 7,400 tWh (65%), which would implications for gHg emissions. our scenarios can be used
likely introduce significant new challenges for grid integra- to estimate the individual and combined impacts of electrifi-
tion. (We note that 600 tWh of this generation is from con- cation, grid decarbonization, and energy efficiency for direct
centrating solar power with thermal energy storage.) total u.s. electricity and energy system emissions. Figure 7 shows
stationary storage capacity from pumped hydro, batteries, how electrification can lead to increases in direct co2 emis-
and compressed air energy storage sources exceeds 130 gW sions from the power system relative to the reference scenario
by 2050 compared with about 30 gW in the reference sce- where emissions are estimated to remain largely flat or even
nario. [storage nameplate capacity is shown in Figure 6(b). decline slightly in the long term. although electrification
the small amount of negative storage net generation is not increases absolute emissions from the power sector, emis-
shown in the figure.] in the modeling, stationary storage sions intensities are lower in 2050 (23 MMtco2/MWh versus
resources are treated as a utility resource and dispatched 28 MMtco2/MWh) as the incremental demand growth is
optimally to meet system needs for energy, capacity, and largely met by low-emitting renewable energy and natural
ancillary services. gas sources. these average emission intensities are about half
renewable energy deployment and associated supply- that of the current u.s. generation mix. More importantly,
chain considerations would also be substantial under this electrification displaces emissions at the end use in the other
scenario with average net annual growth of about 78 gW. economic sectors, transportation, industry, and buildings,
With the stringent hypothetical carbon cap modeled in this resulting in energy systemwide emissions savings. We esti-
scenario, 2050 generation from fossil technologies without mate that electrification in isolation can avoid almost 1 billion
carbon capture would be reduced dramatically (7% for natu- metric tons of co2 in 2050 relative to the reference scenario,
ral gas and <1% for coal). However, we find carbon capture which is equivalent to a reduction of 20% in 2050 energy sys-
and (underground) sequestration (ccs) technologies, pri- tem emissions.
marily using natural gas, become economically competitive Figure 7 shows that electrification alone has essentially
with nearly 17% of total 2050 generation from ccs, which the same impact on total energy system emissions as grid
has an average capacity deployment rate of about 8 gW per decarbonization alone (without electrification), both of which
year through the full period but increases to over 25 gW per result in measurable amounts of avoided emissions relative
year during the 2040s. other resources, including nuclear to the reference case. in either case, however, total emissions
and hydropower, comprise about 11% of total generation remain significantly higher than the levels some have esti-
in 2050 under this combined high-electrification and grid mated are needed to reduce the probability of global tem-
decarbonization scenario. perature increases above 2 °c. With either electrification or
MMT CO2
emissions significantly. For co2, an niza
tion
d Single Measure
we estimate nearly 3.2 billion met- E D
ec le Impact
ar trc
ric tons fewer emissions in 2050 3,000 bo ific
+ n a
(66% lower than the reference sce- En iz tio
er at n
gy io
nario) when both these measures Ef
n
are achieved to the degree envi- 2,000 fic
ie
nc
sioned. Furthermore, implement- y
Combined Impact
ing high end-use energy efficiency 83% Reduction from 2005 Levels
1,000
would help avoid an additional
160 MMt. We find that even with
these combined measures—near- 0
complete electrification of trans- 2010 2020 2030 2040 2050
port and buildings, substantial
6,000
electrification of industry, and Power System Emissions
drastic decarbonization of electric-
ity generation—are insufficient to 5,000
achieve the 2050 emissions levels
indicated by climate scientists to
reduce the most-severe impacts 4,000
of climate change. u.s. energy
MMT CO2
Future Work
and Remaining 0
Research Gaps 2010 2020 2030 2040 2050
the results presented here reflect
an initial attempt to explore the figure 7. Projected energy system and electricity system emissions.
potential implications of wide-
spread electrification of energy
services in the buildings, transportation, and industrial sec- recognizing the shortcomings of our preliminary analy-
tors in the future evolution of the u.s. electricity system. this anal- sis, nrel, in collaboration with other researchers, has initi-
ysis provides a quantitative but approximate indication of the ated a new study to employ higher fidelity modeling to better
impacts that electrification might have on u.s. emissions and understand the impacts of widespread electrification in the
power sector needs. given its exploratory nature, many gaps united states. the new study, referred to as the electrifica-
remain. the completed analysis only considers a limited set of tion Futures study (eFs) (www.nrel.gov/efs), is designed to
electrification, efficiency, and power sector scenarios, whereas employ three new modeling components missing from the
large uncertainties exist for many assumptions made for each existing study. First, the eFs will use a consistent cross-sector
scenario. in addition, the electrification potential estimates for framework (the energyPatHWays model) to track stock
each sector are based on relatively aggregated analysis and do turnover, costs, and performance for 366 end-use technolo-
not consider explicit technologies, adoption patterns, or other gies serving all major end uses. second, the eFs will develop
details that could ultimately affect anticipated electrification, a new bottom-up engineering model to project hourly elec-
future energy use, and consumption profiles, all of which would tricity consumption from future buildings, transportation,
have material impact on electricity supply and emissions. and industrial equipment stock. referred to as demand-side
grid (dsgrid), this new model will rely on multiple bottom- new technologies. they can be used to highlight where
up sector models to enable an unprecedented assessment of research and development would be most effective. For
future electricity load patterns. Finally, for the eFs, we plan to electric technologies, which are typically more capi-
employ unit commitment and economic dispatch modeling to tal intensive but with relatively lower operating costs,
examine, at hourly and subhourly resolution, how power sys- technology road maps can indicate needed efficiency
tems might operate with the new electrified load patterns and improvements and cost reductions. these road maps
greater shares of variable renewable electricity. this modeling can be applied to electric technologies themselves as
can also be used to assess the potential value of demand-side well as enabling technologies, e.g., charging infrastruc-
resources, particularly from new electricity loads, to help sup- ture or information and communication technologies to
port grid reliability. as with the study described in this article, support grid flexibility. road maps for industrial elec-
the eFs will also rely on the nrel reeDs model to develop tric technologies, which recently have not typically re-
future electric sector expansion scenarios. ceived as much attention as transportation and building
in addition to the more detailed modeling to be employed technologies, could fill a large research gap.
for eFs, the new study will also include a wider range of ✔ Data on current energy use. For the united states,
experts and a deeper exploration of available data, includ- the energy information administration and other
ing data for projected cost and performance of electric tech- agencies provide valuable data on current and histori-
nologies. the models and data will be used to develop more cal energy use. However, data gaps remain for many
realistic scenarios than the “technical potential” scenarios subsectors, particularly when a complex array of fac-
that are the focus of the results presented previously. We also tors can impact energy consumption and technology
plan to consider more comprehensively system and house- alternatives. For example, the diversity and complex-
hold impacts in the eFs as well as a broader suite of envi- ity of the u.s. industrial sector makes any electrifi-
ronmental and other impacts. cation assessment difficult. Further, while energy
Despite ongoing and planned studies of electrification in consumption surveys provide invaluable data on elec-
the united states, including the eFs, many research gaps tricity consumption, they are updated only periodi-
will remain. next, we present a list of key research needs for cally and can be further disaggregated by technology,
a more complete understanding of electrification potential geography, and end use.
and implications. ✔ Cost and value of grid flexibility. our scenario analysis
✔ End-use technology adoption modeling. consumer highlights how flexible vehicle charging and hydrogen
adoption models exist for certain technologies, e.g., production can reduce peak capacity needs; however, an
distributed photovoltaics and passenger vehicles, but evaluation of the cost and benefits of this added flex-
are less developed for other technologies and end-use ibility is also needed. More generally, research is needed
subsectors, including end uses in buildings, heavy-duty on the enablement costs associated with increased flex-
vehicles, and in industry. a more specific analysis ibility, including inconvenience, rate structure design,
of industry could benefit from expanding the account- and electric charging and hydrogen infrastructure costs
ing of physical equipment stocks and the development and needs. similarly, greater high-fidelity transmission
of associated stock turnover models. advanced adoption and distribution system modeling is needed to assess the
modeling considers multiple layers, including resource value of this flexibility.
and technology characteristics, demographics, behav- ✔ Macroeconomic modeling and analysis. Models that
ioral economics, capital investment of industrial firms, focus on broader economic impacts, such as comput-
and consumer preferences. the development of such a able general equilibrium or integrated assessment
model to span many end uses or the combination of mul- models, can be used to explore widespread electrifi-
tiple end-use specific models would illuminate the eco- cation, including potential rebound and leakage im-
nomic and market potential for electrification as well as pacts. typically, by themselves, these models lack the
highlight potential challenges and policy interventions. technology and end-use fidelity needed to fully assess
✔ Electric technology road maps. technology road maps electrification. Model improvements or efficient link-
are helpful to elucidate advancement opportunities for ages between models would be needed.
F
For the United Kingdom, one oF the main 2 °C above preindustrial levels and to pursue efforts to limit
drivers for a green ambition is the Climate Change act of warming to 1.5 °C.
2008. this forms the basis for the country’s approach achieving these goals is only one potential outcome of
to responding to climate change and legally commits the the energy future. Within national grid (United Kingdom),
U.K. government to reducing greenhouse gas emissions we produce four credible future energy scenarios that look
by at least 80% below 1990 levels by 2050. this level out to 2050. these are contained in an annual publica-
of commitment to reducing emissions was further con- tion, Future Energy Scenarios (FES). only one of the sce-
firmed in the Paris agreement, the aim of which is to hold narios within this publication meets the 2050 target, two
the increase in global average temperature to well below degrees, which is named to reflect the target adopted by the
Paris agreement.
Digital Object Identifier 10.1109/MPE.2018.2822864
Within the carbon reduction arena, there are three areas
Date of publication: 18 June 2018 that will have the greatest impact: power generation, heating,
and transport. Currently, significant movement has been made in decarbonizing power genera-
tion, and some movement has been made in transport, but there has been little advancement in
residential heating. the two degrees scenario relies significantly (but not completely) on electri-
fication of the transport and heat sectors. in this article, we discuss what will need to happen for
the United Kingdom to meet its legally binding commitments and the corresponding implications
for the power sector.
A World That Is Relatively Sustainability Is Top Priority decade. But since 2011, continued
Wealthy and Market Driven reductions in the cost of photovol-
taic panels, alongside the intro-
duction of a feed-in tariff subsidy
regime, have led to huge growth
(see Figure 2).
Less Money
Available
(smaller projects). at the same time, feed-in tariffs rates were of electricity with an increasing number of market players.
decreased. this led to a “solar rush” in the United Kingdom decentralized generation (located not on the high-voltage
as developers tried to install and accredit projects before the transmission network but, rather, on the lower-voltage dis-
relevant closing dates for support, followed by a slowing tribution network) is changing the pattern of electricity
of growth after these points. this rush was largely unan- flows (see Figure 3). also, interconnectors are allowing
ticipated by government and industry parties alike. By 2017, access to sources of supply from other countries. this is an
solar capacity stood at over 12 gW, nearly achieving the gov- additional source of flexibility that is required as domestic
ernment’s forecast of 13 gW by 2020 several years early. generation becomes more intermittent with the increase in
Similarly to solar capacity, onshore wind generation in the renewable capacity.
United Kingdom also saw large growth in the early parts of
this decade, with subsidy mechanisms supporting continued What Could the Future Look Like?
installations and research and development. however, this our two degrees scenario sees continued rapid decarbon-
growth has slowed as support mechanisms are being with- ization of power generation achieved by significant growth
drawn between 2016 and 2019 because this technology is now
considered to be established by the U.K. government.
in contrast, offshore wind generators continue to receive 14
support via the contracts for difference (Cfd) mechanism, >25 MW
12
which guarantees a power price to offshore wind generation. 5 to ≤ 25 MW
in September 2017, an auction round saw future build off- 10 50 kW to ≤ 5 MW
shore wind farms bid into the Cfd mechanism at a price 8 10 to ≤ 50 kW
TWh
From the 1970s until early in this century, the United King- What Needs to Happen?
dom was self-sufficient in natural gas, which it was able to to reach the 2050 carbon reduction target in a cost-effec-
extract from the UKCS in the north Sea. however, since the tive manner, the two degrees scenario includes a signifi-
early 2000s, gas extraction from the UKCS has been declining. cant drive toward the electrification of heating. this drive
today, about 80% of homes have gas heating, and all is unlikely to occur spontaneously, and it will require incen-
together these households use 332.6 tWh of gas a year for tives to persuade households to move from gas and at the
heating and cooking, representing 36% of U.K. annual gas same time make their homes much more energy efficient.
consumption, the largest gas demand category. Commer- if the drive was a simple switch from gas boilers to elec-
cial heating accounts for 48 tWh per year, or about 4% of tric resistance heat, then the amount of electricity required
total annual energy consumption. industry uses 187 tWh or to heat the housing stock would be such that costly network
approximately 17% of total energy consumption, predomi- infrastructural changes would be required. this would be the
nately for high-temperature industrial processing. case not only for the generation requirements but also for the
of the 29 million residential homes in the United King- transmission and distribution systems reinforcement, which
dom, around 150,000 (~0.5%) have low-carbon heating would have to deliver the energy. FES does not attach a cost
installations such as heat pumps, biomass boilers, and micro to its scenarios, as the intention is to look at an unconstrained
combined heat and power systems. this relatively low adop- world, except where this is obviously unreasonable. Subse-
tion is despite government attempts to support their installa- quent financial analysis is carried out by dependent model-
tions in the form of the following: ing and the outcomes are reported in the network options
✔ the domestic renewable heat incentive, a govern- assessment. therefore, for the most cost-optimal system solu-
ment financial incentive to promote the use of renew- tion, the two degrees scenario sees a need for the demand to
able heat. People who join the program and stick to its drop. this drop will be in the face of a rising population that
rules receive quarterly payments for seven years for will result in 4 million additional homes being built.
the estimated amount of clean, green renewable heat the two degrees scenario forecasts a large-scale
their system produces. move away from gas boilers: from today’s 21 million to
✔ the green deal, a government financial loan program approximately 7 million by 2050, largely (but not exclusively)
to make energy-saving improvements to one’s home. through conversion to air-source heat pumps (aShPs). the
this would typically include measures such as wall journey toward electrified, or low-carbon, heating takes
insulation, draught-proofing, and double glazing. the place in three phases, the product of an optimization model
original program ran from 2013 to mid-2015 and is that is applied solely to the two degrees scenario. Phase 1
scheduled to be relaunched in 2018. sees some rollout of aShPs by those who have the properties
Larger schemes such as district heating account for fewer that can accommodate and benefit from such heating. Phase 2
than 500,000 buildings. Such programs provide around 2% sees hybrid heat pumps becoming more popular, especially
of heating demand in the United Kingdom. in properties that are not particularly energy efficient and will
however, this is not to say that the heat sector is devoid still require the extra heat boost that a hybrid accommodates.
of progress toward decarbonization goals. Currently in Based on stakeholder feedback, hybrid systems are only a step-
the United Kingdom, research is underway to consider ping stone away from full gas heating and, therefore, more
a number of low-carbon heating solutions, including elec- attractive for consumers. in the third phase, aShPs dominate
tric heating (ground and air source heat pumps), the use of once again, reflecting an assumption of significant techno-
hydrogen instead of natural gas, and district heating pro- logical improvements that, in part, drive up their efficiency.
grams. the Leeds h21 Citygate project is a study with the From now to 2050 is effectively the timescale for two
aim of determining the feasibility of converting an entire boiler replacements. a persistent challenge with conversion
city’s natural gas network to 100% hydrogen. at the same is that boiler replacements are often distress purchases, so
time Cadent, the United Kingdom’s largest gas distribution there needs to be both a clear cost advantage and a simple
company, is researching the introduction of hydrogen into switch-over process in place to make such changes the pre-
the gas network (as a blend with natural gas). the con- ferred option, possibly in advance of product breakdowns.
cepts being explored would use gas, in conjunction with the two degrees scenario also includes other forms of
CCS, to generate hydrogen. heating such as fuel cells, district heating, and micro heat and
2011
2012
2013
2014
2015
2016
2010 2012 2014 2016
eV. there are more powerful char-
(a) (b)
gers available, of above 100 kW,
but they are the exception. recent
announcements have indicated figure 6. (a) EV numbers and (b) in the context of all car numbers.
plans to build a network of at least
400 rapid eV chargers across europe. these are likely to be how much traveling that is undertaken varies in each sce-
150-kW units and may be capable of supporting 350-kW char- nario. in the United Kingdom, individuals’ annual average
gers in the future. trip distances have decreased over recent years. this trend
Some policy impetus has been given by recent U.K. gov- will continue in two degrees, driven by more working at
ernment announcements. the Secretary of State for environ- home and less traveling to meetings. this will come about
ment, Food and rural affairs has said that the government through information and technology communications im-
has an ambition to end the sale of new conventional petrol provements as the United Kingdom moves toward a more
and diesel cars and vans by 2040. in the same press release, service-driven economy.
the transport secretary goes on to say that the expectation is two degrees sees shared autonomous vehicle use growing
for nearly every car and van on U.K. roads to be zero emis- from today’s 0% to 50% by 2050. With the adoption of auton-
sion by 2050. omous vehicles, an individual’s overall mileage per car will be
these announcements have been supported by £80 million reduced as sharing will become commonplace in this greener
to improve charging infrastructure for eVs and £246 million society; in other FES scenarios, mileage will increase. these
to assist businesses in designing, developing, and manufactur- vehicles will predominately charge at central depots rather
ing batteries for the electrification of vehicles. additionally, than at domestic sites. this also means fewer cars on the road,
the 2017 autumn budget included £400 million for electric potentially 25 million by 2050, which is about 5 million fewer
car charging points across the nation, an extra £100 million than today, in spite of a growing population. importantly,
for plug-in car grants to help people to buy eVs, and £40 mil- the energy requirements of PeVs will improve from today’s
lion for eV charging research and development. added to this 0.266 kWh/mi to 0.123 kWh/mi by 2050, brought about by
is the ever-declining cost of batteries, which constitutes the more efficient engines and a move toward smaller vehicles, as
most expensive element of an eV. the combination of gov- opposed to larger SUV types.
ernment policy and falling prices should enable accelerated no one means of charging will be universal in any of the
uptake of eVs. scenarios; there will be a mosaic of solutions, with adop-
tion depending on the car owner’s circumstances. there will
What Needs to Happen? be a place for residential trickle chargers and, at the other
For low-carbon transport to play its required role while lim- extreme, a national network of superfast chargers (250 kW
iting the disruptive impact it would have on the network, a or more). nondomestic charging will be important. in 2011,
number of steps are required. these steps were assumed for a nationwide survey by the U.K. department of transport
two degrees prior to the government’s announcements, and, established that only 57% of households had access to off-
fortunately, they are aligned. street parking. assuming that 43% do not, this will equate to
Currently, PheVs represent 83% of the eVs on the road, 8.6 million vehicles by 2050.
and this will steadily decline to 43% in 2030, and by 2040 it will as discussed earlier, time-of-use tariffs will be in opera-
be 9%. in 2040, new PheVs sales will be phased out. PheVs tion, and 85% of car owners will take advantage of the dif-
currently use either diesel or petrol engines and, therefore, ferential charging rates. as a result, they will move away
contribute to carbon emissions. as a result, in our two degrees from charging at peak times. two degrees assumes that
scenario, all eV sales will be PeVs by 2045. smart chargers will become standard post-2020, partly as a
4
Engaged Consumers Charge Their
Lower-Demanding EVs at Off-Peak Time
2
as Do Shared Autonomous Vehicles
(85% of Consumers Are Engaged)
0
2015 2020 2025 2030 2035 2040 2045 2050
70 More-Efficient ASHPs Replace
Current GB Annual Hybrids, but Less Heat Required
Demand Is 306 TWh Due to Thermal Efficiency of
60 Houses
50
Annual Demand (TWh)
30
10
All Cars Sold Are PEVs Post 2040
0
2015 2020 2025 2030 2035 2040 2045 2050
figure 7. Additional annual and peak demands from EVs and heat pumps.
given the sizable role of electricity in the two degrees scenario, northern gas networks. (July, 2016). h21 leeds city gate.
these and many more questions will need to be addressed [online]. available: https://www.northerngasnetworks
in the future as society moves toward a less carbon-inten- .co.uk/wp-content/uploads/2017/04/h21-report-interactive-
sive future. PdF-July-2016.compressed.pdf
national grid. (Jan., 2018). network options assessment.
For Further Reading [online]. available: https://www.nationalgrid.com/uk/
national grid. (July, 2017). Future energy scenarios 2017. publications/network-options-assessment-noa
[online]. available: http://fes.nationalgrid.com/fes-document/
national grid. (July, 2017). Future energy scenarios mod- Biographies
eling methods. [online]. available: http://fes.nationalgrid Russell Fowler is with national grid, United Kingdom.
.com/fes-document/ Orlando Elmhirst is with national grid, United Kingdom.
national grid. (July, 2017). Future energy scenarios chart Juliette Richards is with national grid, United Kingdom.
workbook. [online]. available: http://fes.nationalgrid.com/
p&e
fes-document/
C
Climate Change threatens our quality of life and the
habitability of planet earth for many species. the intergovernmental Panel
on Climate Change estimates that, to reduce the risk that global temperature
increases more than 2 °C above preindustrial levels, greenhouse gas (ghg)
emissions in developed countries must fall by approximately 80% below 1990
levels by 2050. a number of states and regions in the united states have com-
mitted to reducing long-term ghg emissions by this level, including Califor-
nia, new york, and new england. several studies have evaluated the u.s. and
state-specific long-term paths to achieving an 80% reduction in ghg emis-
sions by 2050. less attention, however, has been given to the midterm transi-
tion challenges associated with meeting these long-term climate goals.
this article compares the midterm (2030) electricity-sector challenges
and implications of two recent studies exploring decarbonization strategies
in California and the northeast united states. (in this article, the northeast
refers to new york, massachusetts, Connecticut, rhode island, Vermont, new
©istockphoto.com/JamesGdesiGn
350 benefited as much from the trend toward natural gas generation.
300 Historical GHG the 80 × 50 target implies an annual 2050 fossil fuel
Emissions 2030 GHG
250
Target
emissions budget of 77 mmtCo2 in the northeast and 73
200 mmtCo2 in California, assuming commensurate reduc-
150 tions in non-fossil fuel ghg emissions (such as methane and
100 California fluorinated gases) are achieved (figure 1). in the California
Northeast (NE + NY) 2050 GHG
50 analysis, deeper than 80% reductions in fossil fuel emissions
Target
0 are modeled, achieving 50 mmtCo2 by 2050 (an 87% reduc-
1990 2000 2010 2020 2030 2040 2050
Year
tion from 1990 levels). deeper reductions in the energy sector
are modeled due to anticipated limitations on the mitigation
potential for methane and other agricultural ghg emissions.
figure 1. In-state fossil-fuel CO2 emissions in California
and the U.S. Northeast, historical trends, and future goals. meeting these goals will require a step change in the
pace of emissions reductions in both regions. in California,
the pace of ghg reductions needed to meet the 2030 ghg
help to both lower retail electricity costs and to mitigate grid goal is even faster than the already ambitious rate of change
integration challenges relative to a low electrification future. required in the northeast, requiring a more ambitious suite
We also identify three key challenges warranting further of emissions reduction measures.
research: 1) accelerating customer adoption of building and
transportation electrification, which is lagging the pace needed Emissions by Sector and the Need for
to achieve stated targets; 2) reforming electricity market de- Transportation and Building Electrification
sign, especially to ensure long-term resource adequacy; and in both the northeast and California, the electric power sec-
3) managing distribution-level grid impacts, including larger tor is broadly on track for meeting 80 × 50 goals, but the
instantaneous loads, multidirectional flows, and increased transport and heat sectors are not. transportation emissions
voltage fluctuations. are the largest single component in both regions, exceeding
this article first describes the sectoral emissions land- 42% of emissions in both new york and new england and
scape in both regions. then we describe the study methods representing 37% of California’s total emissions. moreover,
for both areas and results for California and the northeast transportation emissions are higher today than in 1990 in
region, respectively. finally, we conclude with a discussion both regions.
of findings and areas for further research. in the northeast, buildings are the second largest source
of direct ghg emissions, representing over 30% of total
A Comparison of U.S. Northeast ghg emissions. ee programs and fuel conversions from oil
and California Emissions to natural gas have reduced heating ghg emissions from
ghg emissions from fossil fuel combustion in the seven-state buildings by 12% from 1990 levels. still, delivered fuels
u.s. northeast region and California are of a similar magni- (mainly oil) retain a 30% market share. in contrast, direct
tude. according to the u.s. energy information administra- ghg emissions from buildings in California represent only
tion (eia), 1990 Co2 emissions from fossil fuel combustion about 9% of total emissions, due to the milder climate and
(excluding imported electricity) were 382 million metric higher reliance (over 90%) on natural gas heat. figure 2 com-
tons Co2 (mmtCo2) in the northeast and 364 mmtCo2 pares in-state Co2 emissions from fossil fuel combustion, by
in California. for consistency across states and regions, emis- sector, for both regions.
sions data are based on the u.s. eia’s state carbon dioxide given this current landscape, transportation electrification
emissions data, which excludes emissions associated with is widely viewed as a central decarbonization strategy in both
imported electricity as well as non-fossil fuel combustion Co2 regions. in the northeast, support for building electrification
and other ghgs. is growing due to the region’s high reliance on heating oil
fossil fuel emissions in both regions peaked around 2004. and higher share of emissions from buildings. in California,
in the northeast, by 2015, annual Co2 emissions had fallen building electrification has not been policy priority to date, as
by approximately 16% from 1990 levels, driven mainly by buildings represent a much smaller share of total emissions
a shift away from coal- and petroleum-fueled electricity and because the prevalence of low-cost natural gas for heating
to natural gas-fired generation. reduced production from presents greater economic hurdles to switching.
heavy, energy-intensive industries and ee investment played
a role as well. Methods
California’s fossil fuel emissions have fallen less dramati- this section describes the analytical methods and scenarios
cally and are now approximately equal to 1990 levels. the state’s applied in both regions.
California Methodology raXmP and gas Pipeline Competition model (gPCm) simu-
to simulate the impact of ghg reduction strategies on Cali- lations to simulate investment and operational impacts across
fornia’s electric sector and energy infrastructure needs, the region. auroraXmP is an electricity dispatch and capac-
energy and environmental economics, inc. (e3) used the ity expansion model of the electric grid. gPCm is a model
California PathWays model. the model is an economy- that incorporates natural gas supply, demand, and infrastruc-
wide, scenario-based energy technology model representing ture inputs, including pipelines, to solve for expected prices
California’s energy system. this model incorporates infor- and flows throughout north america.
mation about the lifetime and rate of turnover of ghg-emit- electrification load forecasts were developed exogenously
ting equipment and technologies (e.g., buildings, vehicles, from the simulation based on projections of end-use perfor-
water heaters, and light bulbs), allowing users to specify the mance and stock turnover of key technologies. the study
rates of sales of new equipment, the re buildout, and other team created hourly, temperature-dependent, heating load
electric sector assumptions, as well as biofuel use, other ee shapes based on the most up-to-date available data, includ-
improvements, and changes in noncombustion ghg emis- ing a 2016 massachusetts and rhode island study that gen-
sions. the model calculates total economy-wide ghg emis- erated field-validated, air-source heat pump operating data.
sions, energy supply and demand, incremental capital costs hourly ldV electrification load shapes were likewise based
associated with the sale of new technology and equipment, on field-validated data from the u.s. department of energy
and fuel costs associated with each scenario. “eV Project” from 2011 to 2013.
using this tool, we estimate the hourly electricity demand for eVs and electric heating equipment adoption trends, and
each year between 2017 and 2050, reflecting changes over time the resulting electric load profiles, were created for simu-
to end-use energy demands due to economic and population lation in auroraXmP, which performed capacity expansion
growth, ee, and electrification. dispatchable electric generation and hourly, zonal economic dispatch spanning the new
is dispatched against the hourly electric loads net of renewable york independent system operator (nyiso) and iso new
generation, allowing the model to calculate hours of renewable england (iso-ne) footprints from 2017 to 2030, while
overgeneration (i.e., when renewable supply exceeds demand in maintaining minimum (15%) reserve margins. the study
any given hour). renewable integration solutions, such as flex- also modeled monthly natural gas fuel price dynamics based
ible eV charging, energy storage, and flexible loads in build-
ings, are added to the scenario to reduce renewable curtailment.
California’s electric system is modeled as a single zone, with
import and export constraints imposed at the borders. 400
(Million Metric Tons CO2)
300
major end-use category [e.g., light-duty vehicle (ldV) elec-
trification and building electrification] were then fed into a 200
second electric-sector-only model as a check on the electricity
dispatch results and to evaluate optimal electric sector capacity 100
expansion and renewable integration solutions. this electric-
0
sector model, resolVe, minimizes total electric sector capi- 1990 2015 1990 2015
tal expenditures and fuel costs in California over the planning CA NE
horizon, subject to an electric sector ghg constraint. the elec-
tric sector ghg constraint is derived from the economy-wide Residential Commercial
pathways scenario, creating a set of assumptions for the elec- Industrial Transportation
tric sector that are internally consistent with meeting the state’s Electric Power—In State
economy-wide ghg reduction targets in 2030 and 2050.
figure 2. In-state fossil-fuel CO2 emissions in California
Northeast Methodology and the U.S. Northeast by sector, 1990 and 2015. (Data
the northeast study, carried out by national grid and sie- source: U.S. EIA State Carbon Dioxide Emissions Data,
mens Power technology international, used iterative auro- released October 2017.)
250
offsetting the impacts of transport and building electrifica-
200
tion. after 2030, electricity demand increases dramatically as
150
100 higher levels of electrification and new electricity demands for
50 hydrogen production for fuel-cell vehicles begin to outweigh
0 the impacts of electric ee. as a result, by 2050, total electric-
2015 2020 2025 2030 2035 2040 2045 2050 ity demand in the high-electrification scenario is 60% higher
Hydrogen Production than current levels (figure 3). in the long term, the largest
Other Transportation Electrification new sources of electricity demand are from the transportation
Light-Duty EVs sector, including electrified light-duty eVs, electrified trucks,
Incremental Building Electrification freight, and other transportation vehicles, which represent 20%
Buildings of total electricity demand in 2050.
Industrial another new source of electricity demand in this sce-
Agriculture and Other nario is grid-connected electrolysis for hydrogen production.
hydrogen produced using low-carbon electricity reduces the
figure 3. California electric loads by sector and selected need for renewable curtailment and/or energy storage and
end use in the high-electrification scenario (TWh, 2015–2050). represents a low-carbon energy carrier to help decarbonize
Generation (TWh)
tion mix is 95% zero carbon (figure 4). 400
the renewable capacity build-out required to achieve these
300
levels of zero-carbon generation is impressive, nearly tripling in
fewer than 15 years. in the high-electrification scenario, utility- 200
scale solar PVs increase from fewer than 10 gW today to more
than 30 gW in 2030, accompanied by an expansion in rooftop 100
solar PVs from 5 gW today to 18 gW by 2030. in-state wind
resources stay relatively constant between 8 and 9 gW through 0
2015
2018
2021
2024
2027
2030
2033
2036
2039
2042
2045
2048
2030, while out-of-state wind development is only assumed
to occur after 2030 due to challenges with building interstate
CHP Nuclear Natural Gas Imports Hydro
transmission lines to deliver the wind to loads.
Biomass Geothermal Solar Rooftop PV Wind
100
implementation of energy efficiency, controlled eV charg-
ing, and flexible loads in buildings. Without the benefit of 90
higher ee (reference efficiency) or flexible loads in build- 80
ings or controlled charging of eVs (no flexible loads), sys-
tem peak demands in the high-electrification scenario could 70
be 10–20% higher than in the reference scenario by 2050 60
(figure 5). California’s new renewable and energy-stor-
50
age capacity, in combination with existing levels of gas and 2015 2020 2025 2030 2035 2040 2045 2050
hydroelectric capacity, is expected to be sufficient to meet
Reference
the state’s 2030 planning reserve margin, although more High-Electrification Scenario
detailed studies are needed. High-Electrification Scenario
this finding emphasizes the importance of flexible build- with Reference Efficiency
ing loads and controlled eV charging for managing future High-Electrification Scenario
system peak demands. these scenarios assume widespread with No Flexible Loads
adoption of lower voltage (240 V) eV charging during day-
time periods. none of the scenarios evaluated the impact of figure 5. California system peak demand by scenario (in
dc fast charging (480 V) of eVs on system peak demand, gigawatts, net of behind-the-meter generation).
scenario.
310,000 64
300,000 62
290,000 60
40
280,000 58 35 Offshore Wind
30 Onshore Wind
270,000 56 25 Solar
GW
20
260,000 54 15
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
10
5
High RE Scenario - Annual 0
High RE
40 × 30
High RE
40 × 30
High RE
40 × 30
40 × 30 Scenario - Annual
High RE Scenario - Peak
40 × 30 Scenario - Peak
2020 2025 2030
Curtailment
450 in the 40 × 30 scenario, curtailment is estimated to
1990–2015
High RE Scenario reach 4% by 2030 but reaches 7% in the high re scenario,
400
40 × 30 Scenario despite similar levels of Vre penetration. the general find-
40% Below 1990 ing is that higher levels of Vre adoption in both regions
MMTCO2e
350
may be complemented by new electric loads, particularly if
300 there is some flexibility to eV charging. in both scenarios,
increased transmission would be required, although the
250 limited spatial resolution of the simulation does not provide
specific estimates.
200
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2018
2020
2022
2024
2026
2028
2030
Cost
northeast electricity rates today are currently higher than the
national average, although widespread ee investment means that
Figure 9. Northeast economy-wide emissions in the high total annual electricity costs per customer are near the national
RE and 40 × 30 scenarios. average. Both the 40 × 30 and high re scenarios were found
to raise electricity supply costs
by approximately $0.03/kWh, or
table 1. The 40 × 30 scenario assumptions for California approximately 20% above today’s
and the Northeast in 2030, compared to present day.
rates, before inflation. While the
Today 2030 40 × 30 scenario triggered higher
Category California Northeast California Northeast capacity additions, as these incre-
mental costs are spread over more
Solar, including behind the meter 13% <1% 38% 16% demand, the cost impacts were
(generation)
on par with the high re scenario.
Wind (generation) 9% 2% 11% 18% from a customer point of view,
Total RE generation, excluding 29% 5% 60% 35% the 40 × 30 cost increases would
large hydro (generation) be partially or wholly mitigated
Total RE generation, including 39% 21% 70% 51% by reduced expenditures on petro-
large hydro (generation) leum products.
Light-duty ZEV adoption 2% <2% 64% 100%
(annual sales) Regional Synthesis
ZEV penetration (light-duty fleet) <1% <2% 20% 50%
of Study Results
Key scenario assumptions are
ZEV penetration (medium- 0% 0% 3% 0% summarized in table 1, includ-
duty fleet)
ing a comparison between today
ZEV penetration (heavy- 0% 0% 4% 0% and 2030 for California and the
duty fleet)
northeast. By 2030, the Cali-
Other off-road and hydrogen 0% 0% 5% 0% fornia scenario includes more
production electric loads (load) renewable and large hydroelec-
Total transportation electrification 0% 0% 9% 8% tric generation than in the north-
loads (load) east scenario (70% versus 51%).
Electric heat penetration (heating 4% <2% 7% 25% the northeast 40 × 30 scenario
demand in all buildings) includes more aggressive light-
Total electric heat loads (load) 3% 2% 4% 7% duty eV adoption assumptions but
does not assume the electrification
H
Heat represents rougHly 50% of tHe euro- planning. the high seasonality of heat demand in many
pean union’s (eu) final energy demand (figure 1). But in countries, coupled with the increased supply variability due
the past, heat was largely absent in the energy debate, given to the parallel deployment of variable renewable generation,
the traditional focus on energy supply data, which only is increasing the need for flexible operation and planning
shows heat fuels, mainly fossil fuels and a small fraction of to ensure supply and demand balance. electric heating, if
electricity. the potential synergies between heat and elec- deployed in an uncoordinated manner, results in propor-
tricity in the transition to a clean energy system have been tionally stronger winter peak growth than average demand
recognized in the eu’s 2016 heat strategy. the increasingly growth and could further decrease asset utilization. How-
decarbonized electricity system powered by renewable elec- ever, heat can be stored more efficiently and economically
tricity can provide clean heat supply, while the flexibility of than electricity, which offers new opportunities for energy
heat demand can support electricity peak management and system integration solutions. an intelligent or controlled
the integration of variable renewable energies. integration of electric heat can draw on the flexibility of the
Heat electrification represents a growth area for the elec- heating sector (thermal storage and inertia) to facilitate the
tricity industry, but balancing challenges are a major con- integration of renewables and manage peak loads. the use of
cern for short-term operations as well as long-term capacity information and communication technology (ICt) in electric
Heat/Electricity (GW)
months (figure 3). space-heating demand is highest in the
250
winter months, and on the coldest days the opportunities to
provide power-system flexibility may be limited due to the 200
high energy demands and the reduced potential to shift the 150
timing of the demand. During the shoulder seasons of spring 100
and autumn, heating loads are reduced, and opportunities for
50
heat load shifting and storage in thermal stores are increased.
0
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Determinants of Heat Demand
residential space-heating demand varies considerably on a
regional basis. While heating demand is heavily dependent figure 3. The hourly electricity demand and low-grade
on climatic conditions (heating degree days), other signifi- heat demand in the United Kingdom for 2012. [Source:
cant factors include typical dwelling size, household size DECC (2012). The Future of Heating: A Strategic Frame-
(number of occupants), average indoor temperatures, build- work for Low Carbon Heat in the UK. U.K. Department of
ing insulation levels, and the type and efficiency of adopted Energy and Climate Change.]
heating technologies. for future heating demand estimation,
these trends need to be examined.
Changes in population will have a large impact on the total
demand for heat in the residential sector. While different regions 1
are displaying opposing population trends, for the eu overall, 0.9
the population is slowly increasing. While more stringent build-
0.8
ing regulations are driving down demand for heat, increasing
Coincidence Factor
0.7
average dwelling sizes, declining average household sizes (i.e.,
fewer occupants per household), and behavioral changes with 0.6
increased requirements for thermal comfort (average indoor 0.5
temperature) are moving in the opposite direction. While build- 0.4
ing regulations still have an enormous impact on space heating 0.3
demand for new buildings that only use a fraction of the heating 0.2
demand of older dwellings, overall residential heating demand 0.1
will depend heavily on the uptake of retrofit activities. retrofits 0
are being widely promoted in the eu given that the majority 0 100 200 300 400
Number of Houses
of homes that will be present by 2050 have already been built.
(a)
aggregate water heating demand is influenced by pop-
ulation size. However, there are also strong regional varia-
tions in water heating demand per capita. typical end uses, 1
heating technology, and consumer behavior drive these dif- 0.9 Gas Boiler
Old Building
ferences. for example, frequent sauna bathing in finland 0.8
Modern Building
Coincidence Factor
Thermal Storage
thermal storage in the building ena-
10 2,500
9 bles the optimization of the elec-
8 2,000 tricity consumption and charging
7 based on electricity market condi-
Load (kW)
Load (kW)
6 1,500
5
tions while still providing thermal
4 1,000 comfort to the user. temporal decou-
3 pling of electrical power demand
2 500
1 from the heating demand can be
0 0 achieved when electrical heating
with a sufficient storage capacity
0
0
0
0
0
1, 0
1, 0
1, 0
0
0
0
0
0
0
1, 0
1, 0
1, 0
0
18
36
54
72
90
08
26
44
18
36
54
72
90
08
26
44
is deployed.
(a) (b)
electrical space heaters with
internal storage have been widely
figure 5. (a) Individual and (b) aggregated electricity profiles for 1,000 customers used for decades in many regions
with electric heating and domestic hot water, over 24 h, with 1-min resolution. with temperate climates, including
Ireland, the united Kingdom, germany, france, and new compared to stand-alone Hp systems. However, hybrid heat-
Zealand. traditional nighttime storage heaters were used to ers could offer electricity system benefits if connected in a
reduce systemwide peak loads and fill in the night “valley” smart manner that enables the electricity system to access
and allowed customers to use less-expensive off-peak elec- the flexibility of the gas system by switching from the Hp to
tricity tariffs to meet their heating demand. storage heaters the gas boiler whenever the electricity system is under stress,
contain a core of high thermal capacity bricks, heated which can include an extended period over several days.
by a resistive heating element and surrounded by a highly Hybrid gas boiler-resistance heater systems could also pro-
insulated enclosure. they are designed to be used on a 24-h vide the option to use excess renewable electricity by switch-
cycle (i.e., charged at night and heat released during the day), ing from gas to electricity.
although older storage heaters could not always keep suf-
ficient charge levels to maintain thermal comfort on colder System Integration
winter days. Modern storage heaters, through improved insu- Widespread electrical heating can have a significant impact
lation and controls (the release of the heat during a 24-h cycle on the overall system energy demand as well as on the shape
can be more accurately controlled with the use of a variable of that demand. High shares of direct electrical space heating
speed circulation fan), offer a higher degree of comfort to end increase the winter system peak demand. on the other hand,
users. Moreover, with adequate controls and communication, when combined with thermal storage, the electrical demand
smart electrical thermal storage (sets) can participate in can be shifted to off-peak hours, with minimal impact on
active demand-side management (DsM) and enhance power the peak, and with the additional advantage of increasing
system flexibility. the eu-funded Horizon 2020 realValue baseload plant utilization in the off-peak hours. However,
project is trialing sets systems at a household level, provid- as shares of variable renewable energy increase and the net
ing a range of system services while maintaining the thermal load becomes more variable and harder to forecast, smarter
comfort of customers. controls and two-way communication are required to fully
electricity is also widely used in the residential sector take advantage of the thermal energy storage capacity.
for water heating, which is typically combined with thermal
storage (a water cylinder). as with space heating, this stor- Coupling Small-Scale Electric Heating
age capacity decouples the power demand from the heating and Thermal Storage
demand, and when equipped with communication and con- Cooptimizing heating (combined with storage) and power
trol architecture, electrical water heaters can also participate system scheduling could reduce overall system costs and
in DsM. increase shares of variable renewable generation that can
the building envelope itself also provides thermal inertia be integrated. as discussed earlier, sets for both space
depending on the insulation level. In a well-insulated house, and water heating (i.e., resistive electrical heating com-
electric load can be shifted (around 5–12 h), depending on bined with thermal storage) is currently being trialed in
the building insulation level, while consumer comfort is still the eu Horizon 2020 realValue project. a combination of
met. preheating increases flexibility but typically increases physical demonstrations in three winter peaking systems
energy usage, depending on the insulation level. thermal (Ireland, germany, and latvia) along with detailed build-
storage and building preheating enables large demand shift- ing and power system modeling will demonstrate how local
ing potential at very low cost, especially when compared to small-scale energy storage with advanced ICt, optimized
electricity storage. across the eu energy system, could bring benefits to all
market participants.
Hybridization enormous benefits can be realized from the integration
Hybrid heaters combine different heating appliances in one sys- of the power and heat energy systems. However, to ade-
tem and can switch between those appliances during opera- quately assess the scale and value of the flexibility that can be
tion, providing a very flexible demand. possible configura- accessed from the heat sector, detailed modeling is required
tions include Hp-gas boiler (Hp-B) and gas boiler-resistance to consider the impact of any load shifting and demand
heaters. Hp-B systems have been available commercially response activities on the end users. Cooptimized building-
for several years and were developed to lower investment to-grid models have been developed and refined as part of
Cost
Sav
Cost
Sav
Cost
Sav
Cost
Sav
Cost
Sav
Cost
Sav
Cost
Sav
0 16
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 networks and particularly at the
(a) lV level where most such technol-
24 ogies are connected.
20
22
In general, electricity demand
15 exhibits a relatively high degree
20 of diversity that makes the after-
10
5 18 diversity maximum demand quite
smaller than the sums of the indi-
0 16
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 vidual peaks, with a coincidence
(b) factor that is relatively low and on
the order of 15–20% for the united
HP Indoor Temperature (°C)
Kingdom. However, heat demand
has a comparatively lower diver-
figure 9. The standard and flexible operation of different heaters and building sity with a coincidence factor of up
indoor air temperature for the windiest week of an average year. (a) HP and (b) HP to 35% (figure 4). this has impli-
Flex. (Source: Heinen et al., 2017.) cations in terms of network impact
and design, as it effectively means
opposed to other storage technologies. If real-time or smart that larger infrastructure is required compared to a situation
control is enabled, then an Hp or another electric heater where the peaks were driven by nonheating consumption.
could be operated in a flexible manner according to power Because of this loss of diversity due to heating electrifica-
system conditions without impacting the thermal comfort of tion, the impact due to an increasing penetration of electrified
building occupants. heating could manifest itself on the lV network even at low
a systemwide cost study for Ireland (see Heinen et al., 2016 systemwide penetration levels. In practice, lV transformers
in the “for further reading” section) has shown that utiliza- specifically could be affected first by this loss of diversity,
tion of thermal inertia enables cost-effective heat electrification due to thermal issues (i.e., sheer increase in after-diversity
using Hps. Hp operation during the windiest week exempli- maximum demand) (figure 10), followed by lV conductors
fies the peak load reduction benefits of flexible operation ver- also driven by thermal issues. Voltage drop impacts are likely
sus standard operation. In standard operation, the Hp is only to appear only at higher penetration levels (see also navarro-
allowed to operate as indicated by the user thermostat settings, espinosa and Mancarella, 2014 in the “for further read-
which in Ireland is typically a few hours in the morning and ing” section). In figure 10, reinforcement will be required to
after working hours until late evening. this creates high morn- integrate electric Hps when transformer utilization exceeds
ing and afternoon peaks as the heater needs to bring the building 100% (or earlier if a safety margin is considered).
indoor temperature to the desired indoor temperature, assumed given the role of peak demand growth on potential lV network
to be 20 °C. In the flexible operation mode, the Hp can oper- impacts, thermal storage can reduce network impact, in partic-
ate at any time as long as the indoor temperature is 20 °C dur- ular in the form of hot water tanks for both domestic hot water
ing the thermostat settings and never higher than 23 °C. this and space heating (in the case of hydronic systems). However,
operational flexibility enables the heater to preheat the building the system needs to be carefully designed to override the peak
ahead of an expected peak event or when excess wind is on the consumption on extreme cold days, where again diversity is
system. for the week in figure 9, the building indoor tempera- likely to decrease substantially (figure 4). If the thermal stor-
ture is preheated very notably at least three times (days 2, 5, and age system is adequately designed for extreme days, then it can
7) and reduces peak load compared to the standard/nonflexible yield a substantial level of flexibility under normal conditions
Hp operation. In practice, rolling out these systems will require that can be utilized for demand response or more in general
detailed building sensing equipment and may not be acceptable DsM, in this case to deal with both thermal and voltage issues.
for all customers. this can, for example, happen in the presence of electric heat-
ing paired to photovoltaic (pV) systems, where pV produc-
Distribution-Level Impact of Heat Electrification tion and self-consumption can be maximized by making use
a variety of drivers that impact heat electrification have been of the flexibility enabled by thermal storage. More generally,
discussed so far, in particular heat consumption, outdoor and this distributed storage could be used to support active network
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
90
100
distribution network. In addition, the
potential of using the resource for HP Penetration (%) HP Penetration (%)
congestion management and poten- (a) (b)
tial upgrade deferral is also explored. 800-kVA Transformer
Building models are included in a 500-kVA Transformer
three-phase unbalanced ac optimal One Standard Deviation
power flow that allows heat demand
to be treated as a variable. Differ-
ent demand-response strategies are
figure 10. The average transformer utilization level (with one-standard deviation
examined, considering both the im- range) for (a) ASHP and (b) GSHP. Note that he penetration level refers to percentage
pact on the network and the thermal of houses with HPs. (Source: Navarro-Espinosa and Mancarella, 2014.)
comfort of customers.
E
End-usE ElEctrification is onE of thE thrEE pillars
of deep decarbonization, which are fundamental strategies necessary
to eliminate energy-system carbon emissions. as a low-carbon strategy,
electrification is only effective when paired with reductions in the emis-
sions intensity of electricity generation. thus, any discussion of electri-
fication’s role in reducing carbon emissions, either through the direct
electrification of end uses but also fuel switching to electricity derived
fuels, must include supply–demand interactions within the electricity
system. this article highlights what these interactions may look like in
the transition to a very low-carbon energy system in the united states
and some potential implications for the future of the grid and wholesale
electricity markets. the concepts are primarily illustrated using updated
scenarios from the 2014 u.s. deep decarbonization pathways project
(ddpp) technical report.
©istockphoto.com/gorica
500
but can still solve shorter times-
400 cale energy imbalance and, criti-
cally, mitigate their own impact
300 on the need for new system infra-
structure build.
200
New Loads Can Have Very
100 High and Unpredictable
0
Instantaneous Demand
1 24 1 24 1 24 1 24
Electric vehicle rapid-charging has
become a selling point for new elec-
figure 1. The month-hour average load growth from 2018 to 2050 in the DDPP tric vehicles, and as battery charge
high-renewables scenario by end use. The vehicle-charging profile shown before management continues to improve,
flexible charging is applied. public dc charging of light-duty
43% 110% 39% 37% 37% 38% 40% 43% 47% 53% 59% 65% of using intermittently available
39% 100% 35% 32% 32% 33% 36% 39% 44% 50% 56% 62%
energy economically.
35% 90% 30% 27% 27% 28% 31% 35% 40% 46% 52% 59%
30% 80% 26% 22% 21% 23% 26% 31% 36% 42% 48% 55%
Wind and Solar Alone
24% 70% 20% 17% 16% 17% 21% 25% 31% 37% 43% 50%
18% 60% 14% 11% 10% 11% 15% 19% 25% 31% 37% 43%
Do Not Provide
11% 50% 9% 7% 6% 6% 9% 13% 18% 23% 29% 35% Resource Adequacy
6% 40% 4% 3% 2% 2% 3% 6% 10% 15% 20% 25% systems with high variable gen-
2% 30% 1% 1% 0% 0% 1% 1% 3% 5% 9% 13% eration have large needs for dis-
0% 20% 0% 0% 0% 0% 0% 0% 0% 0% 1% 2% patchable capacity, called residual
0% 10% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% capacity, to provide reliability, fre-
0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% quently at times of the year when
Solar Fraction 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% energy deficits make it more dif-
Wind Fraction 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% ficult for short duration storage to
fill this need. figure 4 shows the
residual capacity need in a 100%
figure 3. The percentage of renewables curtailed without storage or flexible load
renewable scenario for the u.s.
for a simulated system in New York State. Renewable delivery respects a must-run
requirement of nuclear, cogeneration, and hydro (16% by energy). A renewable port- Eastern interconnection using a
folio standard of greater than 100% implies that renewables are overbuilt to increase single year of data from 2011. in
the amount of delivered renewables, with excess energy curtailed. This type of chart an ideal scenario, the net load (load
can be generated for other geographies with the exact pattern depending on load minus renewables) would be zero,
and renewable profiles. meaning that no residual capacity
800
123
179 190
700 227
261 280 298 317 323 322 300
600
500
Capacity (GW)
400
691
635 624
300 587
553 534 516 497 491 492 514
200
100
-
100% Solar,
0% Wind
90% Solar,
10% Wind
80% Solar,
20% Wind
70% Solar,
30% Wind
60% Solar,
40% Wind
50% Solar,
50% Wind
40% Solar,
60% Wind
30% Solar,
70% Wind
20% Solar,
80% Wind
10% Solar,
90% Wind
0% Solar,
100% Wind
figure 4. The capacity met with 100% renewables by energy in the U.S. Eastern Interconnection with different wind/
solar compositions.
as sustained peaking capability, may need to become an inte- direct access customers are not ppa off-takers, they become
gral part of all markets. this is illustrated by the 21st century a fixed-cost free rider and receive energy subsidized by other
screening curve in the “capacity Market Must have Multiple market participants. this could be desirable from a public
products that represent capacity of different time dura- policy perspective, for example, to encourage zero-carbon
tions” section. transportation using hydrogen generated from marginally
free electricity, but it raises legitimate concerns about cross-
How Will Fixed Costs Be Allocated subsidies between customer classes.
Among Electricity Consumers?
today, variable costs already make up a relatively small How Can Flexible Load
share of total electricity system costs (encompassing genera- Be Induced to Participate?
tion, transmission, and distribution), and their share of the flexible load has a large potential to solve energy imbal-
total electricity revenue requirement is set to plummet in a ance at low cost compared to many other solutions. antici-
high-renewables scenario, as shown in figure 6 using data pated new electric loads have inherent flexibility, either as
from the u.s. ddpp scenarios. in the high-renewables sce- a function of their thermal storage (electric space and water
nario, only 2.5% of costs are variable by 2050, thus 97.5% heating) or chemical storage (electric vehicle batteries or
of the revenue requirement must be collected regardless of hydrogen electrolysis), which can be leveraged to manage
how, when, or whether electricity gets used. supply–demand imbalances.
it can be argued that many new electric loads, including the role that flexible load could play in a future system is
electric fuel production, will be able to balance the system illustrated in figure 7, showing average load for three simu-
by consuming energy that is free at the margin. however, lated months in 2050 in a high-renewables scenario. the first
this ignores a serious cost-allocation problem, namely that panel shows load net of all nonthermal generation technolo-
other market participants must pay for the renewable gen- gies, including dispatchable hydro and nuclear. as seen, the
erator providing that marginally free electricity at the price average net load is frequently negative, indicating a surplus of
of a long-term power purchase agreement (ppa). thus, if energy on the system. Given the seasonal features shown in
figure 2, the month of april has the greatest surplus energy.
the second panel shows the average hourly operation of dif-
ferent balancing solutions in each month. Balancing solutions
20%
above the x-axis effectively increase load and absorb surplus
US$110B
energy, while those below the x-axis reduce load. Electric fuel
of Revenue Requirement
US$85B
production plays a large role in reducing curtailment in the
Variable Cost Share
15%
spring, and flexible end-use demand, including vehicle charg-
ing flexibility and flexible thermal loads, plays a large role in
10% reducing daily net load peaks. the final panel shows the resid-
ual net load (summing panel one with panel two) that in this
5% Scenario scenario is met with thermal dispatchable generation. these
AEO 2017 Reference three panels demonstrate the importance of flexible loads
High Renewables US$34B
among the balancing solutions. in these scenarios, energy
0%
storage plays a relatively small role because it does not have
2020 2025 2030 2035 2040 2045 2050 the storage duration to replicate the role of electric fuel pro-
Year
duction in seasonal balancing, and it is more expensive than
flexible load for solving diurnal imbalance.
figure 6. Variable costs as a share of the total U.S. elec-
tricity revenue requirement. This is created using generator conceptually, it is useful to divide flexible load into two
fuel costs rather than market prices. In electricity markets types: short duration, which is primarily traditional firm load
where a marginal price is paid to all generators, a greater made flexible through the use of control technology, and long
portion of fixed costs are recovered on a variable basis duration, which includes new nonfirm loads with low marginal
than shown. energy-storage costs, for instance, grid-scale electrolysis. the
400
200
Average GW
–200
Balancing Resource
–400
Renewable Curtailment
Electric Fuel Production
–600 Flexible End-Use Demand
Energy Storage
–800
1 241 241 24 1 24 1 24 1 24 1 24 1 24 1 24
figure 7. An example of the magnitude of flexible demand that operates in a high-renewables system taken from the U.S.
DDPP high-renewables scenario.
first type of flexible load can solve diurnal energy imbalance the quantity of electric fuel production could be substantial
and reduce load at peak to avoid capital investments (i.e., and thus will have a systemic impact on markets.
substations and transmission lines). the second type of flex-
ible load provides a productive use for excess energy during How Can Developers Be Induced
seasons with extensive overgeneration and allows for further to Make Long-Term Investments
renewable build at lower marginal curtailment by reducing Given Uncertain Revenues?
energy deficits during other times of year. ultimately for both today, most renewable generators are compensated through
types of flexible load, the goal is to flatten net load. bilateral ppas, which guarantee fixed volumetric rates over
short- and long-duration flexible load each face differ- many years and insulate project developers from variable mar-
ent barriers to adoption. interruptible load for peak shav- ket prices. few merchant nondispatchable generators exist
ing is a mature market for large commercial and industrial today, and it seems unlikely that merchant exposure is a robust
applications; recently, interruptible programs have expanded model for future development given the long payback times
into small commercial and residential applications, with for renewable generators and the significant risk of curtail-
many small interruptible loads aggregated into large ones. ment and market price depression over the project lifetime.
More recent still are load-shifting programs, which are an thus, bilateral ppas are likely to remain the dominant struc-
important step past traditional demand response. for most ture for paying renewable generators.
customers, though, even given the necessary smart controls technologies that provide balancing, such as energy stor-
in demand equipment, the price signals sent do not have the age or electric fuel production, face similar risks in energy
temporal or spatial granularity to unlock the full cost-of-ser- markets due to unknown future patterns of renewable devel-
vice reductions possible from utilizing flexible load. opment and/or the deployment of competing balancing
the barriers to long-duration flexible load also include resources. While risk is a normal part of the merchant elec-
its exclusion from full participation in energy markets. tricity equation, the rapidity of changes on the supply side
today the impact that price-responsive flexible load has on may produce so much uncertainty in long-term revenues that
the shape of load must be accounted for in load forecasts. no project is bankable or the deployment of balancing sub-
this works fine when flexible load is marginal, but if very optimally lags the build of variable generation.
large nonfirm flexible loads such as electric fuel production
develop, symmetry between supply and demand in electric- How Will Future Electricity-System
ity markets will be critical for optimal outcomes, for two Planning Be Conducted?
reasons. first, wind and solar vary enough day to day that if the goal is reaching 2050 electricity emission rate targets
a prearranged schedule will always be suboptimal. second, consistent with deep decarbonization, it is necessary for the
electricity system to go through several decades of unprec- Energy Market Compensates Balancing
edented change. Widespread electrification will lead to load Services, with Full Symmetry Between
growth not seen in decades. planning in a system dominated Supply- and Demand-Side Balancing
by variable generation is more complex because it intro- in a world with near-zero marginal cost for energy and uncon-
duces a temporal component into every planning problem. trollable supply, energy markets primarily become a place to
the timing and complementarity of the many different but compensate and schedule balancing services. as seen in fig-
necessary electricity investment decisions are critical, and ure 7, these balancing services can take many forms, including
many 20th century planning paradigms and tools are likely flexible end uses, new nonfirm loads, energy storage, thermal
insufficient to address the challenges ahead. resources, and renewable curtailment. such resource types
compounding the planning challenges, deregulation has would provide the bulk of balancing in a highly electrified,
decentralized planning by putting many planning activities in high-variable generation system, but today these resources are
the hands of market participants. in current competitive mar- not integrated into markets in a way that would allow them to
kets, no stakeholder has visibility into all parts of the system, operate in the way pictured.
which will seriously hamper execution of an overall long-term in such a system, large loads would bid a demand curve
vision. if uncertainty about policy, market rules, or timing of in which different amounts of energy consumption would be
complementary resource deployment calls future revenue offered at different prices, in a way that is completely analo-
streams into question, then carbon targets may not be met, the gous to how generators bid an energy supply curve today.
cost of financing may increase, and sub-optimal development for small customers, this could be done by an aggregator on
strategies pursued. their behalf that then sends control signals to building loads
Electrification exacerbates planning challenges by involv- after considering a customer’s expressed preferences.
ing sectors that historically have no strong links to electric- one technical challenge when scheduling flexible load
ity, most notably transportation. as mentioned earlier, or energy storage is state-of-charge constraints, which are
transportation loads represent significant energy but, per- not yet integrated into markets. today, most resources
haps even more importantly, may have very high and unpre- that track state of charge, such as hydro or energy stor-
dictable instantaneous power demand. the new spatial age, self-schedule generation, which works fine when those
planning challenge, of developing transportation charg- resource are marginal, when net-load is predictable and
ing infrastructure in which the highest demand is located well understood by the scheduler, or when few competing
where the electricity grid is best able to accommodate it, balancing resources exist (there is no real need to exchange
is considerable. information). as supply becomes less flexible and the
opportunity cost of not using newly electrified flexible
Concepts for Future Market Design load increases, markets will play a critical role as a clear-
Given the challenges posed by rapid electrification and inghouse for balancing resources. it will be important for
renewable build, markets will need to keep pace on sev- regional transmission operators (rtos) to gain experience
eral different fronts simultaneously. it can be argued that scheduling flexible loads at a small scale before balancing
a vertically integrated, rather than deregulated, structure needs grow massive. the alternative is more reliance on
is best suited for coordinating the scope of electricity conventional resources, which in the context of deep decar-
system changes brought about by deep decarboniza- bonization could add significantly to cost or make carbon
tion. But this argument neglects functions that markets goals unattainable.
are better suited to accomplish, namely communicating
information between participants and driving efficiency Capacity Market Must Have Multiple Products
improvements. the challenge is thus how to get the best That Represent Capacity of Different Time
of both: 1) the flow of information and efficient opera- Durations (“21st Century Screening Curve”)
tions from markets and 2) effective long-term planning capacity markets serve two key functions in electricity today.
from a central decision maker. Below are some concepts, the first is as a longer-term planning mechanism to ensure
challenges, and principles that such an optimal approach that sufficient resources are built and maintained to meet
should consider. expected loads, months or years in advance. the second is
400
AEO 2017
Reference
200
0
600
Residual GWh
400
High
Renewables
200
figure 8. The residual capacity need across simulated 2050 scenarios. Residual capacity is equal to total load minus all
zero-carbon resources, including nuclear and hydroelectricity. In the modeled scenarios, this residual was met with ther-
mal resources.
486 GW
422 GW
and regulators have a large role to
400 376 GW play in deep decarbonization by
383 GW providing stability in future revenue
278 GW streams for projects deemed optimal
200
by long-term planning exercises.
141 GW Market mechanisms, such as reverse
12 GW
0
auctions, should be employed within
1
10
100
1,000
10,000
1
10
100
1,000
10,000
likely produce less-optimal planning decisions than would more valuable than the ability to perfectly follow a blue-
otherwise happen under a slower transition. the main risk is print created years before.
that too many poor decisions could derail the entire process,
driving up costs while missing the carbon goals. this high- For Further Reading
lights the need for expanded long-term planning. J. h. Williams, B. haley, f. Kahrl, J. Moore, a. d. Jones,
M. s. torn, and h. McJeon. (2014). pathways to deep decar-
Concluding Thoughts: An Elevated bonization in the united states: the u.s. report of the deep
Role for Planning Is Needed to Ensure decarbonization pathways project of the sustainable devel-
Prudency and Guide Sequencing opment solutions network and the institute for sustainable
Toward Deep Decarbonization Goals development and international relations. revision with tech-
Establishing and frequently updating a clear road map to deep nical supplement, nov. 16, 2015. [online]. available: http://
decarbonization in electricity is critical to ensure goals are met usddpp.org/downloads/2014-technical-report.pdf
on time and at reasonable cost. in creating and administering J. h. Williams, B. haley, and r. Jones. (2015, nov. 17).
such road maps, utilities and their regulators play an indispens- policy implications of deep decarbonization in the united
able role and must rise to the occasion for the enterprise to states: a report of the deep decarbonization pathways proj-
succeed. We conclude by offering some guiding principles for ect of the sustainable development solutions network and the
success over the coming decades. institute for sustainable development and international rela-
✔ a key to future planning will be the efficient and trans- tions. [online]. available: http://usddpp.org/downloads/2015-
parent sharing of information among stakeholders, report-on-policy-implications.pdf
some of which have not coordinated historically. J. Williams, r. Jones, G. Kwok, and B. haley, “deep de-
✔ regulators should strive to understand which changes carbonization in the northeastern united states and expand-
will be best accomplished through markets and which ed coordination with hydro-Quebéc,” 2018.
through regulations. G. Kwok and B. haley. (2018). portland general electric
✔ it is better to delay reaching 2050 goals than to make decarbonization study. [online]. available: https://www
poor decisions because the goals may end up even .evolved.energy/single-post/2018/02/23/portland-General-
further delayed and have a damping effect on achiev- Electric-decarbonization-study
ing long-term decarbonization. deliberation must be r. Jones. (2017). conference presentation: realities
properly balanced against urgency. of balancing electricity systems with 100% renewables.
✔ all near-term decisions should always be viewed through [online]. available: https://www.evolved.energy/single-
the lens of long-term goals. this will help minimize post/2017/07/11/realities-of-balancing-electricity-systems-
carbon lock-in or stranded assets. with-100-renewables
✔ regulators have an important role to play in democra-
tizing risk and in transforming markets by deploying Biographies
novel technologies. Ryan Jones is with Evolved Energy research, denver,
this article raises many questions regarding markets colorado.
and regulatory structures at large associated with a highly Ben Haley is with Evolved Energy research, denver,
electrified electricity system with significant amounts of colorado.
renewables. due to the pace of change in technology and Gabe Kwok is with Evolved Energy research, portland,
the speed at which those technologies need to be deployed, oregon.
agility in planning and market design is paramount. this Jeremy Hargreaves is with Evolved Energy research,
is perhaps the single most important concept for regulators san francisco, california.
and system planners in the coming decades: maintaining Jim Williams is with the university of san francisco,
the ability to correct course quickly and frequently will be california. p&e
H
Historically, tHe term electri-
fication has meant providing electricity to
customers who lack access to electric infra
structure. From the earliest urban electric
services in the 1890s to the electrification
era in rural america from the 1930s to
the 1950s, access to electricity drastically
improved quality of life, including thermal
comfort, hygiene, food availability, illumi
nation, mobility, commerce, communica
tion, and entertainment. as a result of these
early electrification efforts, almost all U.s.
homes are fully or partially electrified, with
notable exceptions in some Native ameri
can and alaska Native communities.
an expanded definition of electrifica
tion allows us to contemplate a new era of
electrification that is now in its early stages,
driven by economic opportunities, new tech
nologies, consumer demands, environmental
issues, and changing societal and industrial
structures. in this context, electrification can
be broadly understood to include three means
of increasing electricity demand.
1) Expanded adoption of existing ser-
vices: increasing the extent to which
electricity is used to provide existing
energy services (e.g., growth in the use
of electric cooling technologies).
2) Fuel switching: substituting electric
ity for another fuel to provide the same
energy service (e.g., using electric heat
pumps for space heating instead of fuel
oil or using electric vehicles for trans
portation instead of gasoline vehicles).
90 ieee power & energy magazine U.S. Government work not protected by U.S. Copyright. july/august 2018
A Review of Electrification Study Scenarios
and Future Analysis Needs for the United States
3) New end uses: adopting electric technologies that they include a focus on consumer value (choice, convenience,
provide an entirely new service demand for electricity productivity, and environmental impacts) and business value
(e.g., smartphones, data centers, and vertical farming). (expanded opportunities for electric providers). electrifica
a range of recent studies examining strategic, beneficial, tion’s economic potential has primarily been studied through
or efficient electrification show substantial growth potential the lens of reducing energyrelated greenhouse gas (GHG)
for electricity use in transportation, buildings, and industry. emissions. substantial electrification, however, is not solely
driven by decarbonization and merits addi
tional analysis.
in the context of current and historical
electrification trends, this article summarizes
recent studies that model future pathways
that increase electricity’s share of energy
consumption in U.s. buildings, industry, and
transportation. We discuss a range of output
metrics mentioned in the studies, including
total electricity demand, electrification by
sector, environmental impacts, costs, eco
nomic growth, and energy productivity, to
identify broad trends and possible impacts
of electrification. in addition, we compare
methodologies among studies, identify areas
for additional research, and discuss policy
implications of common outcomes of elec
trification across studies.
16%
Industrial tial heating, electricity’s share of energy demand grew in all
Electricity as %
14%
12% regions of the United states between 2005 and 2013, partly due
10%
Commercial to increased electric heat and partly due to population migra
8% tion to milder climate regions. in the industrial sector, electricity
6%
4% has fueled improved manufacturing services, such as induction
Residential
2% heating and the electric arc furnace. indoor or vertical farming
0% is a growing area of commercialindustrial electricity demand,
1950 1960 1970 1980 1990 2000 2010 2016
and the deployment of information and communication tech
nologies is affecting electricity use across all sectors.
figure 1. Electricity as a percentage of delivered energy in
the United States. (Source: The EIA’s Monthly Energy Review.)
Comparison of Recent
Electrification Scenarios
that uses a different fuel source (e.g., electric heat pumps the studies summarized here highlight potential pathways
replacing heating oil for space heating). electricity’s share that accelerate the current electrification trends and explore
of delivered energy shown in Figure 1 does not account for what it might look like if the United states were to enter a
increasing efficiency of electric technologies. For example, new era of electrification. Here we build on and update a 2015
substituting lightemitting diode and compact fluorescent review by Peter loftus and coauthors that surveyed lowcar
lightbulbs for older incandescent bulbs does not decrease bon pathway studies, which all project increased demand for
demand for the lighting service, but it does reduce energy electricity, particularly by passenger vehicles and heat.
consumption from lighting on a perunit basis. the studies reviewed are recent ones that generally use
currently, energy consumption by fuel varies greatly by sec national energysector models to explore the electrification of
tor, as shown in Figure 2. electricity constitutes approximately U.s. building, transportation, and industrial sectors. each study
half of energy consumed on site in buildings in the residential includes multiple scenarios that model energy, economic, and
and commercial sectors. By contrast, electricity makes up just environmental impacts out to the year 2050 based on a set of
13% of industrial and less than 1% of transportation energy con assumptions regarding fuel and technology costs, economic
sumption. subsectors, geographic areas, and time trends show growth, policy goals, and other factors. in this context, a sce
additional variation. For example, electricity’s share of building nario is the set of assumptions and constraints modeled, while
energy use varies significantly by climate region. a projection of a particular metric is the output of the modeled
Within these sectors, both new technologies and the increased scenario for a future year. scenarios and their resulting projec
deployment of existing technologies contribute to trends po tions are meant to reflect economically and technically plau
tentially driving electricity consumption growth. technol sible pathways, but they are not forecasts or predictions.
ogy deployment is driven by economics, technology maturity the exception is the 2017 annual energy outlook (aeo)
and availability, consumer preferences, changing population pat reference case from the energy information agency (eia);
terns, and policies. in the transportation sector, for example, this projects 2050 energy use under the assumption that cur
rent policies remain this same and is used here as a bench
mark to evaluate the impact of the other study scenarios.
30 many of these studies incorporate electrification as only one
Direct Renewable
of several critical elements, such as land use, that can shape
Delivered Energy (Quads)
25 Biomass
Petroleum future energy pathways. For each study, we review the scenario
20 Natural Gas that specifically focuses on achieving high electrification if
Coal one is provided; otherwise, we examine results from the core
15 Electricity
or main scenario. all of the scenarios share certain attributes:
10 they assume rapid electrification of end uses, rely on a balanced
portfolio of low and zerocarbon electric generation technolo
5
gies, achieve highdecarbonization goals, and assess the impli
0 cations for the energy system to meet those goals.
Residential Commercial Industrial Transportation
While many of the scenarios are conceptually similar, the
Percent
Electricity: 44%
models used to generate the scenarios or examine the impacts
52% 13% <1%
of exogenously defined scenarios differ in terms of resolu
figure 2. The total delivered energy by fuel type and sec- tion, scope, and operation. For example, the Global change
tor for the United States in 2016, not including electricity- assessment model (GcamUsa) and the regional energy
related losses. (Source: The EIA’s Monthly Energy Review.) Deployment system (reeDs) model are both optimization
Historic (EIA)
from around 25% in the reference AEO 2017 Reference
MCS (2016)
AEO (2017)
july/august 2018
Transportation Sector 2050 Buildings Sector 2050 Industrial Sector 2050
35
Delivered Energy in 2050 (EJ)
30
25
20
15
10
5
0
O
er
er
er
g
ss
ss
am
am
am
er
er
er
C
C
AE
AE
AE
Iy
Iy
Iy
M
ei
ei
M
nb
nb
nb
illi
illi
illi
W
W
ei
ei
ei
W
W
St
St
St
Electricity Other Energy
figure 4. Electrification in 2050 by sector as projected by various studies. The total height of the bar is delivered energy
in 2050; the darker-shaded portion is amount of delivered energy derived from electricity. Note: *Weiss et al. assump-
tions of full electrification for residential and commercial heating and light-duty vehicles and trucks applied here to the
2017 AEO. Weiss et al. used the 2015 AEO and do not address industrial uses.
Other Metrics a significant impact. the Weiss study discusses the potential
in addition to the commonly used metrics described previously, for utility sales growth.
there are many additional metrics that could inform decision
makers and define the potential opportunities and challenges Societal and Consumer Costs
that could arise from a new era of electrification. table 2 builds the electrification of enduse services will certainly impact
on the loftus approach to show a variety of such metrics that consumer energy expenditures, but the magnitude and direc
are either quantitatively or qualitatively discussed in at least tion of the impact will depend on regional electricity rates,
one study. a broader consideration of these metrics in future retail rate structures, the relative efficiencies and prices of
analyses will improve the community’s understanding of the enduse technologies, and the relative price of displaced fuels
economic and environmental impacts of electrification across (oil, natural gas, and propane). some studies quantify expen
multiple geographic and temporal scales. ditures but vary in their treatment of costs; some include
price impacts, others include bill impacts, and still others
Pace of Change and Economic Growth include societal costs. For example, steinberg et al. explore
the availability and rate of technology replacement are key the wholesale price of electricity and Williams et. al include
metrics for determining the degree of electrification in a both household and overall costs.
scenario. a few studies in this review explore the pace of current results emphasize the importance of conventional
change required to meet a specific policy target, assessing technologies’ capital and fuelcost assumptions and energy
how quickly the electricity system and supporting techno efficiency technology and policy assumptions, in addition to
logies would need to evolve
across regions to avoid put table 2. Treatment of alternative metrics across studies.
ting investments at risk. For
AEO
example, the Williams study 2017 Iyer Steinberg MCS Williams Weiss
notes that the rateoftech
Pace of change and
nology replacement limits how economic growth
quickly and to what extent sig Societal and consumer
nificant electrification could energy costs
be achieved in certain end Energy productivity
uses with longlived assets
Resilience and reliability
(e.g., industrial boilers). How
ever, even without forcing an Energy security
early retirement of longlived Other environmental
assets in modeled scenarios, impacts
the increased penetration of Key: No treatment; Qualitative treatment ; Quantitative treatment.
electric technologies can have
Beyond the economic parameters that impact a consum moving beyond energy price to incorporate a richer assess
er’s decision to electrify (e.g., upfront technology costs and ment of consumer value will be an important part of assessing
operating costs), other parameters within the model architec the benefits of electrification. First, the value of electrification
ture impact the extent to which a given model pursues elec should be modeled to include the location and timedependent
trification, including consumer choice parameters, which can benefits and costs of increased electricity consumption. sec
be oversimplified or based on historical preferences for elec ond, the value of electrification should be expanded to include
trification that may not reflect future preferences; barriers to noneconomic energy service characteristics that drive choice
switching to electricity from direct fossil fuel use; support (e.g., noise pollution, convenience, safety, and performance).
ing infrastructure requirements, such as the ratio of electric in addition to systemlevel costs, the economic interests
vehicles to charging stations; restrictions on the pace of addi of different stakeholders need to be more fully examined.
tional generating capacity; and limited ability to incorporate For example, stress to existing utility business models, caused
research, development, and demonstrationstage technolo in part by lack of electricity demand growth, could be alle
gies such as electrified heavyduty vehicles. Understanding viated by increased demand associated with electrification.
the sensitivities of results to these parameters and how dif changing job opportunities and availability could be a con
ferent models treat them is important for contextualizing and sequence. Jobs have not been addressed frequently in the
evaluating model outputs. current literature, perhaps because most studies look across
We therefore recommend that studies prioritize and report the energy system and would not capture movements among
sensitivity analysis on key parameters and a new model in subsectors. Future electrification studies could explore pace
tercomparison study be carried out that explores the repre of change, economic growth, and associated stakeholder
sentation and treatment of key parameters for electrification impacts through investment needed and avoided along with
across models. other job and economic activity related metrics.
©istockphoto.com/theerapong28
System
T
TwenTy-five years ago, J.D.
willson was the control center manager
with one of the United states’s largest
utilities—the Pennsylvania, new Jersey,
Restoration
Maryland interconnection (PJM)—
and also served as chair of the ieee
Power system restoration Training
Task force. M.e. Long was a training
manager at Pepco and chair of the ieee
working group on operator Training
Readiness
(wgoT). PJM was a pioneer in study-
ing ways in which utilities could inter-
connect their power systems to increase
reliability while reducing overall costs
and also led efforts to improve system
operator training.
Interconnected Power
The Evolution in North America Systems Drove Need
for Restoration Plans
interconnected power systems origi-
Digital Object Identifier 10.1109/MPE.2018.2824101
nally operated so that they could man-
Date of publication: 18 June 2018 age the loss of the single largest unit,
100% for
Does your plan include actions of other companies? 50%
RC and TOP
100% for
Do you have a training program to support your restoration plan? Fewer than 50%
RC and TOP
Do you provide annual system restoration training to your system operators? About 50% 100%
100% if identified in
Do you involve plant operators in your system restoration training? About 40%
the RC or TOP plan
100% if identified in
Do you involve neighboring personnel in your system restoration training? About 25%
the RC or TOP plan
Do you use a system-specific simulator for your system restoration training/drills? About 25% 100%
What is the average length of your system restoration drills? About 4–8 h About 8 h
Many in the industry did not wait for mandatory stan- erence Document”) listing several technical issues to address
dards to develop detailed system restoration plans or system in system restoration plans, only 67% of the respondents with
restoration training for system operators: they began devel- restoration plans indicated they used the information in that
oping/testing restoration plans and devising/implementing document as a basis for their restoration plan.
restoration training long before the nerC reliability stan- early restoration plans focused primarily on the actions
dards became enforceable. These entities recognized that, of the entity’s own system operators. while approximately
with the complexity of interconnected power systems, real- 25% of the plans included some actions of system operators
time operations are not linear, making it is essential that system in other companies, only about 40% included the actions of
operators understand the principles associated with isolated personnel such as plant operators in the same company. and,
power systems and then apply those principles to a variety of although only 20% involved corporate communications, most
possible restoration scenarios. (76%) did include mobilization plans.
early restoration training programs used a variety of train-
Restoration Plans Circa 1992 ing methods, with almost half using some self-study as well
in 1991, the ieee published the paper “Power system res- as lectures led by system operators. approximately 25%
toration issues,” emphasizing the need to perform detailed used a company-specific simulator, and fewer than 10% used
analytical studies to ensure a sound technical basis for res- a generic simulator. a little over 30% incorporated drills into
toration plans. in 1992, the ieee wgoT and the Psrwg the training (figure 3 shows uses of drill evaluations). while
surveyed the industry to assess the state of system restoration
training provided to system operators. almost 100 entities
representing different-sized entities from various regions of
Plan Tools
both the United states and Canada returned this survey (see
the column headed “1992” in figure 2). Most respondents
(84%) indicated that they did have a restoration plan; however, Evaluate
“lessons learned through the experience” of system operators,
system engineers, and system managers provided the foun- Processes Staff
dation for these plans rather than detailed technical studies.
Many respondents (about 60%) indicated that they revised
their plans based on lessons learned from system restoration Training
exercises or drills. while nerC had published its first version
of the guidance document (“electric system restoration ref- figure 3. Drill evaluations identify needed changes.
PJM RC: MISO, what’s your direct number out there? TVA and ourselves are having October 2015
trouble finding a good one.
serc joined the chat October 2015
serc joined the chat October 2015
PJM RC: As of 14:15 EPT PJM’s EKPC zone has lost all of its generation and appears October 2015
to be fully blacked out. EKPC is performing a thorough assessment, updates to follow.
MISO RC: joined the chat October 2015
peve joined the chat October 2015
PJM RC joined the chat October 2015
serc joined the chat October 2015
TVA_RC joined the chat October 2015
TVA_RC: MISO RC are you on the line October 2015
TVA_RC joined the chat October 2015
TVA_RC: TVA reported approx. 2500 MWs lost in the SE region of TVA. Crews in route October 2015
TVA_RC: LGEE has loss all generation and load in their area. (Total system blackout.) October 2015
MISO RC joined the chat October 2015
MISO RC joined the chat October 2015
NERC joined the chat October 2015
MISO RC joined the chat October 2015
MISO RC: MEPA has currently Moselle #3, #4 & Sylvarena #3 online serving 101 MWs October 2015
total. 161 kV in service. Moselle to Hintonville Moselle to Columbia Columbia
to Plant Morrow Morrow to Purvis Bulk Purvis Bulk to Lumberton Hintonville
to Plant Morrow Hintonville to Bennedale Hitonville to Waynesborrow.
Recommendations
year, fewer than half the system operators in an organiza- system operators with near real-time experience in restoring
tion participate in the exercises. Therefore, it is important to systems from a state in which their own system is disconnected
incorporate lessons learned from each drill in the organiza- from the grid; they are thus better prepared in the event they
tion’s ongoing training programs. face the need to restore their systems.
Commercial airline pilots practice a wide variety of
scenarios on high-fidelity simulators before sitting in a Conclusions
cockpit ready to address real-life emergencies. By com- Many in the industry are doing much more than nerC and
parison, most system operators have limited opportuni- ferC require: nerC reports and ieee technical papers
ties to experience a wide variety of emergency scenarios serve as a sound foundation for both generic and company-
before possibly facing a real restoration. More frequent specific system restoration training and drills. Because power
and more varied training exercises with high-fidelity simu- systems are complex and power system operations are not
lation would support a greater degree of system operator linear, it is essential that system operators understand the
readiness. The fact that most entities found a need to modify principles associated with a restoration. Completely accu-
their system restoration plans based on drill results is a clear rate prediction of exactly how, when, or where a system
indication that the industry hasn’t yet explored all possible will disconnect from the rest of the grid is unlikely. we must
ways of efficiently separating and restoring portions of the depend on system operators to apply their knowledge of sys-
power grid. tem restoration principles to a wide variety of scenarios—
many of which are unanticipated.
Improve the Efficiency of Internal Multiple advances have been made in the development of
and External Communication system restoration training and drills over the past 25 years.
with the proliferation of cyber and physical attacks, it is Today’s system operators are better prepared than ever before,
essential that system operators recognize and report these but there is still an opportunity for improvement. for more
events in a timely manner. There is an opportunity for information about wgoT, contact Patrick everly at peverly@
more information sharing with respect to what cyberattacks oesna.com.
look like to a system operator. entities are required to have
policies in place to ensure as-soon-as possible reporting of For Further Reading
cyberattacks to various authorities. The intent is to have the M. M. adibi, Power System Restoration: Methodologies and
authorities analyze and share that information so other enti- Implementation Strategies. new york: wiley, 2000.
ties will become aware of the event and to increase aware- J. D. willson, “system restoration guidelines: How to set
ness that an attack has taken place in one entity and could up, conduct, and evaluate a drill: a report by the operator
expand or be duplicated elsewhere. such quick exchange of training working group,” IEEE Trans. Power Syst., vol. 11,
information is possible only if a system operator can deter- no. 3, pp. 1619–1629, aug. 1996.
mine that an unusual operation or event is reportable, so “grid security exercise gridex iii report,” nerC, at-
sharing information about what these events might look like lanta, ga, Mar. 2016.
is critical. ferC, nerC, and regional entities, “report on the
ferC-nerC-regional entity joint review of restoration and
Benefits of System Restoration recovery plans,” Jan. 2016.
Training and Drills
Many benefits accrue from system restoration training and Biographies
drills. Perhaps most importantly, entities have used the lessons J.D. Willson, retired, was with the Pennsylvania, new Jer-
learned to refine their company-specific restoration plans, and sey, Maryland interconnection.
many have shared these lessons learned with others in the M.E. Long, retired, was with the north american elec-
industry. a second outcome has been the development of job tric reliability Corporation.
aids for supporting restoration drills that would also be helpful
p&e
during a real restoration. a third benefit has been providing
forgotten pioneer
leo daft and the excelsior power company
G
Gold Street in lower Man
hattan is a narrow passage laid out Early electrification has a rich history in a wide range of diverse geographic loca-
when horse traffic was the predomi tions, many of which have been highlighted in IEEE Power & Energy Magazine in the
nant mode of travel. Just south of Ful pages of this “History” column. One of the most active areas is lower Manhattan,
ton Street, a large converted residential
New York, in the latter part of the 19th century, when it was a crossroads of inven-
structure at #33–43 bears a plaque with
tors, technology, entrepreneurs, and finance and with a large population of people
the words Excelsior Power Co. and
and businesses to be served.
the date A.D. 1888 (Figure 1). while
it may be a curiosity to the casual ob Within this history, we don’t often hear the stories of the pioneers and com-
server, it is the city’s last remaining pub panies that experienced a rapid rise and an equally swift fall, as technology was
lic sign from the pioneer days of electric eclipsed due to rapid progress in the new fields of the day, as was the case with
power. two short blocks west of thom electrification in the late 1800s. This column presents us with such a story on the
as edison’s famous Pearl Street station intersection of electrical pioneer Leo Daft and the Excelsior Power Company.
of 1882, the excelsior Company was Joseph J. Cunningham returns for his eighth visit to these pages. He has re-
a pioneer in providing power for me searched and authored numerous booklets, articles, and books on topics such as
chanical tasks, as opposed to simple industrial electrification, electric utility power systems, and electric rail transporta-
lighting. it began with the provision of tion. His latest book, New York Power, was published in 2013 by IEEE History Cen-
steam power and then adopted electric
ter Press. The author and the book were significant references stated in the N.Y.C.
power as it was more efficient. the de
Landmarks Preservation Commission decision to approve the Excelsior building for
velopment of the excelsior Company
landmark status. We welcome back Joe as our guest history author for this issue of
paralleled the career of inventor leo
daft until mutual business interests IEEE Power & Energy Magazine.
brought them together. John Paserba
Associate Editor, History
Leo Daft and Company
leo daft was born in Birmingham,
england. the son of a civil engineer,
daft showed an early aptitude for me gineer, drafting, photography, and, fi to Greenville, new Jersey, in 1881. His
chanical devices. at 15, he joined his nally, electric lighting and propulsion. initial successes included elevators in
father’s firm as a draftsman and then in electric lighting and propulsion, Boston and industrial motors and light
entered university studies and subse daft opened a small lighting company ing; he then tackled street railways and
quently worked with his father’s firm on Centre Street in lower Manhattan, experimented with electric power on
as an electrician. intrigued by develop which expanded to become the daft the ninth avenue elevated railway
ments in the United States, daft left electric light Company headquartered in Manhattan.
england for new York City in 1866 and on lower Broadway near wall Street. By the late 1880s, daft’s firm was
worked in a variety of jobs and fields as arc lighting became popular for at its peak with a wide variety of prod
that included a steam locomotive en large spaces, a number of companies ucts, many that were a very sophis
flourished amid considerable competi ticated design for their time. those
Digital Object Identifier 10.1109/MPE.2018.2819038 tion. daft also pursued motors, which included motors, generators, and con
Date of publication: 18 June 2018 required more space, and he relocated trols (Figures 2 and 3). the range of
Ensuring
The Excelsior
Power Company
the excelsior Power Company began
as the excelsior Steam Power Com
pany and, in 1873, acquired the assets the Reliable
Performance
of previous steamplant companies in
lower Manhattan. through a system
of Your
that would be considered bizarre and
arcane by modern standards, excelsior
transmitted steam power to industrial
customers by spinning shafts located
in tunnels beneath the city’s streets.
the company flourished, and, by 1887,
Assets
it operated three large steam engines
with a combined 800 hp in a plant at
• High Voltage Testing
13 Spruce Street. those engines supplied
power over a fourblock area to busi
nesses and newspaper printing plants • High Current Testing
on Frankfort, Spruce, william, and
Beekman Streets. Given the cumber
• Mobile Cable Testing &
some nature of such a system, the con
cept of power transmission by electric Commissioning
wire was alluring. See Figure 5 for a
geographic perspective of the area. • Asset Management &
the excelsior electric Company
Life Extension
was formed to enter the electric light
business through the patents of wil
liam Hochhausen, an inventor and lo • Rotating Machines
cal businessman. Hochhausen had de
veloped incandescent light systems to a
• Substation Automation
degree of perfection that exceeded edi
son’s contemporary system. the com & Equipment Analysis
pany offices at 66–68 duane Street of
fered a display of his lighting systems. • Advanced Transformer
By 1884 the company was billing itself
Testing
as the sole owner of the Hochhausen
patents, “the most perfect system of
electric lighting in the world.” Such in • Overhead Line
flated rhetoric was common in adver Inspection
tising of the day but, in this case, was
not unrealistic.
the system provided for the automat
ic switching of faulted lamps and could
be combined with arc lights for a mix of
small and largescale illumination, an www.kinectrics.com
approach not practical with the edison
system. as was the case with most of the /KINECTRICS /KINECTRICS @KINECTRICS
early pioneers, lighting was the primary
focus of excelsior, but the provision of
power was not ignored as lighting pro
duced insufficient return to amortize the
capital investment required.
daft installed a combined power
and light system in Boston in 1884, an
other in rhode island, electric railways
in Baltimore and upstate new York,
and the first electric elevator in Man
hattan. a demonstration of his motors
at the 1884 international electrical
exposition in Philadelphia was used to
print an entire edition of the prestigious
trade journal The Electrical World. ex
celsior’s prime steampower customers
were printing and publishing houses,
thus excelsior had a market for con
version to electric power. daft products
figure 1. An intriguing and mysterious plaque in 2018. (Photo used by permis- and excelsior’s system made a per
sion of Robert Lobenstein.) fect combination.
Save Energy.
Rent Test Equipment.
Advanced Test Equipment Rentals makes renting easy so you can focus your energy elsewhere.
Rent: • Regenerative Grid Simulators • AC / DC Loads
• Programmable Power Supplies • Cable Fault Locators
• Relay Testers • and more...
(b)
(c)
(a)
(b)
ART Series
Current Sensors
ART Series sensors are the first
to combine easy-to-use flexible
Rogowski coil technology with
Class 0.5s accuracy. When it comes figure 4. The Daft factory in Marion, New Jersey.
to measuring energy production, (Source: Daft Electric Light Company, “Electrical
Power Plants, Daft System,” New York City, 1888.)
monitoring substation transformer
health or real-time allocation of energy
costs and usage in smart buildings,
the daft Company (through what appears to have
ART Series sensors set the standard.
been a utility company named the electric Power
Company of new York) began operating an electric
At the heart of power electronics. light plant on the excelsior Company’s Spruce Street
property on 1 January 1884, nine months before the
Philadelphia exposition. two months later, excelsior
www.lem.com introduced electric motors for printers and other in
dustries. demand was strong, and in the fall of 1887,
excelsior began the construction of an addition to the
figure 5. An 1899 map of the territory supplied by the Excelsior/Daft system. Competition
(The author would like to thank Robert Colburn of the IEEE History Center for while the daft/excelsior motors were
locating the map of the area, source: Wikimedia Commons.) a “quantum” improvement over the use
evolved. Self-Illuminated
Power Line Marker
Peak
SOLTEX
COMPOUNDS
AND FLUIDS:
Reliability
With your reputation on the line, get the compounds and
fluids trusted by major cable manufacturers, utilities and
electrical contractors. Soltex gives you high-performance
products, technical expertise, responsive and agile service,
plus the consistent quality you can only get from a world-
class manufacturing facility.
orderentry@soltexinc.com | soltexinc.com
DF and Electrifill Dielectric Fluids • Transformer Fluids • Solguard ACSR Anti-Corrosion Greases
Acetylene Black for Semi-Conductive Compounds • Naptel and Soltexfill Telecom Filling/Flooding Compounds
Daft/Excelsior Eclipsed
the daft/excelsior Company held its
position for a time. then edison’s sys
tem expanded rapidly as the Sprague
motor made it fully competitive with
excelsior. in 1886, edison opened an
annex station on liberty Street and,
five years later, a mammoth (for the
time) 50,000light (about 1,810kw)
station at 55 duane Street. By 1895,
excelsior was absorbed into a succes
sion of companies in a combination cre
ated by utility magnate anthony Brady
and his system expert and inventor
thomas e. Murray. that combination
evolved into the new York edison Com
pany in 1901, by which time 120/240
Vdc utility systems had become firmly
established. those were supplied by
neighborhood substations that conver
ted highvoltage ac from largescale
central stations to dc for local distribu
tion systems in an era when ac motors
needed much development and ac dis
tribution in dense urban areas was still
problematic. dC motors had undergone
some 15 years of development, and dc
distribution allowed the use of battery
assemblies for peak loads and protec
tion against outages.
the excelsior plant became a sub
station in april 1899, although it re
tained standby arclamp generators for
another 23 years. industrial customers
that had occupied space on the upper
floors of the Gold Street building re
Europe
This year's conference will bring together researchers and industry experts from Europe and
all over the world to provide an international forum to share and discuss issues and
developments in the field of smart grid technologies and applications.
You won't want to miss the keynotes, plenary sessions, panels, industry exhibits, paper and
poster presentation, and also tutorials by experts on smart grid applications and systems
integration.
B
Beginning in August, elec member of the Brook Byers institute ✔ Pes president and chair, Pes
tions will be held for the position for sustainable systems and universi governing Board and Pes ex
of ieee Division Vii delegateelect/ ty center of excellence in Photovoltaic ecutive committee (2014–2015),
directorelect 2019 and ieee Divi Research. His research and ieee Pes Past
sion Vii delegate/director 2020–2021 activities are in the moni President (2016–
(the elected individual will serve both toring, analysis, and con There are two 2017). As president,
as both delegate and director). Divi trol of power systems as he strongly supported
sion Vii consists of the ieee Power well as the development candidates, both • membership and
and energy society (Pes) only. there
are two candidates, both nominated
and applications of re
newable and sustain
nominated by chapter growth by
supporting chapter
by Division Vii, vying for this office: able energy systems. He Division VII, chair orientation
Miroslav M. Begovic and Miriam P. was editor of the sec trainings in all Re
sanders. the successful candidate will tion on transmission vying for this gions every year
serve as the ieee Division Vii direc
tor in 2020–2021. to learn more about
systems and smart grids
in Springer Encyclope-
office: Miroslav (student chapters
grew 40% yearly
the candidates before casting your bal dia on Sustainability M. Begovic during the period)
lot, read the biographies and candidate (2012), guest editor of • global outreach, es
statements that follow. the IET Generation, and Miriam pecially in Regions
The Candidates
Transmission and Dis-
tribution special issue
P. Sanders. 8 and 10, strength
ening relations with
on widearea monitor sister associations,
Miroslav M. Begovic ing and control. He has authored near such as cigRÉ and the chinese
Miroslav M. Be ly 200 journal and conference papers society of electrical engineers
govic is a Fellow and ieee standards and technical • coordinating the first major reor
of the ieee. He is reports and has contributed over 100 ganization of the Pes technical
t he d e p a r t m e nt keynote and invited presentations in all committees and strongly sup
head of the elec ieee Regions. porting close coordination of
tr ical and com strategic relations of Pes with
puter engineering IEEE Activities and ieee sister societies, the ieee
(ece) Department and carolyn e. and Accomplishments standards Association, and isO.
tommie s. lohman ‘59 Professor at ✔ nominated in 2018 for ieee ✔ Member, ieee Publication ser
texas A&M university. Prior to that, Division V i i d i re ctor ele ct vices and Products Board com
he was a professor and chair of the (2019) by the ieee Pes nomi mittee (2015).
electric energy Research group in the n a t ion s a nd Ap p oi nt m ent s ✔ Pes presidentelect, member
school of ece, georgia institute of committee. of the governing Board and its
technology, and an affiliated faculty ✔ Member, ieee nom i nat ions executive and Finance commit
and Appointments committee tees, and chair of the ieee Pes
Digital Object Identifier 10.1109/MPE.2018.2836838
(2017–2018) and ieee Fellows longRange Planning committee
Date of publication: 18 June 2018 committee (2017). (2012–2013), which developed
2018
IEEE PES
GENERAL MEETING
5-9 AUGUST
PORTLAND, OR USA
The 2018 IEEE Power & Energy Society General Meeting will be held Mark Ahlstrom
from August 5-9, 2018 at the Convention Center and Crown Plaza.
The PES General Meeting attracts over 3,400 professionals from every
segment of the electric power and energy industries. It features a
comprehensive technical program with paper presentations, poster
and panel sessions, a number of technical tours, a student program Larry Bekkedahl
and companion activities.
The 10th IEEE PES Asia-Pacific Power and Energy Engineering Conference 2018
(APPEEC) will be held at Hilton, Kota Kinabalu, Sabah, MALAYSIA on 7-10 October 2018.
The aim of the conference is to provide a premier platform for electrical
engineers and researchers to present their works and to share experiences
and ideas in power and energy engineering with experts and scholars
from around the world. Started in Wuhan in 2009, APPEEC is now an
annual power engineering conference organized in Asia-Pacific Region.
p&e
Are You
Don’t miss an issue of this magazine—
update your contact information now!
Moving?
Update your information by:
E-MAIL: address-change@ieee.org
PHONE: +1 800 678 4333 in the United States
or +1 732 981 0060 outside
the United States
If you require additional assistance
regarding your IEEE mailings,
visit the IEEE Support Center
at supportcenter.ieee.org.
© ISTOCKPHOTO.COM/BRIANAJACKSON
PES meetings
for more information, www.ieee-pes.org
T
T H E I E E E P ow E r & E n E rgy November 2018 May 2019
Society’s (PES’s) website (http://www IEEE Electronic Power Grid (eGRID IEEE International Con ference
.ieee-pes.org) features a meetings sec- 2018), 12–14 november, Charleston, on Electrical Machines and Drives
tion, which includes calls for papers South Carolina, United States, contact (IEMDC 2019), 11–15 May, San Di-
and additional information about each Johan Enslin, jenslin@clemson.edu, ego, California, United States, contact
of the PES-sponsored meetings. http://ieee-et-d.org/ Avoki omekanda, avoki.omekanda@
ieee.org
August 2018 IEEE International Forum on Smart
IEEE PES General Meeting (GM Grids for Smart Cities (SG4SC June 2019
2018), 5–9 August, Portland, oregon, 2018), 26–28 november, genk, Bel- 14th International Conference on
United States, contact richard god- gium, contact geert Deconinck, geert. Transmission & Distribution Con-
dard, richard.goddard@pgn.com, or deconinck@kuleuven.be, http://www struction, Operation & Live-Line
Shane Freepons, Shane.Freepons@pgn .ieeesg4sc.org/ Maintenance (ESMO 2019), 24–27
.com, http://pes-gm.org/2018/ June, Columbus, ohio, United States,
December 2018 contact Eriks Surmanis, e.surmanis@
September 2018 IEEE International Conference on pdc-cables.com, http://ieee-esmo.com/
IEEE PES Transmission and Distri- Power Electronics, Drives and Ener-
bution Conference and Exposition – gy System (PEDES 2018), 18–21 De- IEEE PowerTech Milan (PowerTech
Latin America (T&D LA 2018), 18– cember, Chennai, India, contact Arun 2019), 23–27 June, Milan, Italy, contact
21 September, Lima, Peru, contact Karuppaswamy, akp@ee.iitm.ac.in, Federica Foiadelli, federica.foiadelli@
Alex Arquinego Paz, alex.arquinego@ www.ee.iitm.ac.in/PEDES2018/ polimi.it, http://ieee-powertech.org/
ieee.org, http://ieee-tdla2018.org/
January 2019 August 2019
October 2018 IEEE PES 201 Joint Technical Com- IEEE PES General Meeting (GM
IEEE PES Asia-Pacific Power & En- mittee Meeting (JTCM 2019), 13–17 2019), 4–9 August, Atlanta, georgia,
ergy Engineering Conference (AP- January, orange County, California, United States, contact Matt Stryjewski,
PEEC 2018), 7–10 october, Saba h, United States, contact Solveig ward, matthew.stryjewski@ieee.org
Malaysia, contact Zuhaina Zakaria, sward@quanta-technology.com, www p&e
zuhaina@ieee.org, http://sites.ieee.org/ .pestechnical.org
appeec-2018/
March 2019
IEEE PES Innovative Smart Grid IEEE PES GTD Grand Internation-
Technologies Europe (ISGT Europe al Conference and Exposition Asia
2018), 21–25 october, Sarajevo, Bos- 2019 (GTD Asia 2019), 20–23 March,
nia and Herzegovina, contact Senad Bangkok, Thailand, contact nopbhorn
Huseinbegovic, shuseinbegovic@etf Leeprechanon, nopbhorn@engr.tu.ac
.unsa.ba, http://sites.ieee.org/isgt- .th, http://www.ieeegt-d.org/
europe-2018/
and industry, and there is some historical cost. More important, even when the di- type lanes and not change the game by
animosity toward it that needs to be ac- rect combustion of fossil fuels is more ef- switching to new energy supplies.
knowledged and laid to rest. ficient in a given end use from a primary There’s no question that energy effi-
In the U.S. utility context, electri- energy standpoint, it is generally a worse ciency remains essential for decarbon-
fication has a close historical associa- option from an emissions standpoint if ization. Some say that natural gas is the
tion with load building. Since the early the electricity is low carbon. Carbon, not bridge to a low-carbon future, but the
days, utilities have had strong incen- energy per se, dictates the logic of energy real bridge is energy efficiency. It will
tives to grow load to increase profits, systems in a climate-friendly future. play an outsized role in sectors with lim-
expand rate bases, and reduce average It is critical for industry, regulators, ited fuel-switching potential, for exam-
rates. During the first half of the 20th and policy makers to recognize that deep ple, freight trucking and industrial pro-
century, when access to inexpensive decarbonization cannot be achieved cess heat. In parts of the United States
electricity transformed first urban and through energy efficiency alone or even with no history of efficiency programs,
then rural life in the United States, elec- a combination of energy efficiency plus grossly inefficient building shells, over-
trification served a powerful social pur- renewable electricity. Electrification is sized HVAC systems, and antiquated in-
pose. But since the 1970s, in the face absolutely required, and in many ap- frastructure, efficiency will still be the
of energy security and environmental plications, it may complement or even first tool out of the clean-energy tool-
concerns, electric load growth has often displace a focus on energy efficiency. box. Efficiency provides a brake on irre-
been viewed negatively. Beyond a certain level, conventional en- sponsible, high-carbon load building in
In the energy efficiency paradigm ergy-efficiency investments can produce coal-based power systems with no tran-
pioneered by people like Art Rosenfeld diminishing returns for carbon reduc- sition plan. Even in a decarbonized sys-
and Amory Lovins decades ago, using tion compared to a similar investment in tem, energy efficiency can help reduce
less primary energy to provide the same electrification, as long as the electricity the scale, cost, and land use impacts of a
energy services made compelling sense is low carbon. Electrifying end uses is low-carbon infrastructure buildout.
on many levels: lower fuel demand for not counter to the fundamental purposes
thermal generation, less pollution from that motivate investments in energy effi- Long-Term Policy in
burning that fuel, and less need for gen- ciency. Indeed, modeling shows that, in the Electric Economy
erating capacity and, with it, lower cap- a U.S. low-carbon transition, the largest Article 4.19 of the Paris Agreement
ital requirements and fewer conflicts source of energy efficiency will be elec- calls on all countries to develop mid-
over licensing and land use, as was (and trification itself, due to the thermody- century strategies for decarbonizing
is) often the case for large hydro dams namic superiority of electric drive trains their economies. The commitments
and nuclear power plants. and heat pumps over their combustion- made at Paris are near term, out to 2025
In thermal-dominated systems where based alternatives. or 2030, and only promised modest
the variable cost of fuel was the lion’s Getting energy efficiency and elec- emission reductions. But the Paris em-
share of generation costs, saving BTUs trification to play nice with each other phasis on the long-term future provides
meant saving money, and in the 1970s in the regulatory and policy a renas a platform for strategizing and enacting
when much of the generation fleet was oil may be challenging. Clean-energy ad- transformational changes. Given the
powered, it meant less economic depen- vocates have fought hard to incorpo- multidecade lifetimes of the most criti-
dence on that volatile and conflict-fraught rate renewable generation and building cal infrastructure on both the supply and
commodity. Using electricity to heat water energy efficiency in utility plans, and demand sides of the energy system (e.g.
and space in thermal-dominated systems many will be skeptical about electri- power plants, buildings, industrial boil-
was about three times as energy intensive, fication if they see it as threatening ers, and cars and trucks) and the poten-
counting thermodynamic and line losses, decades of hard-won gains. That may tial for emissions lock-in and stranded
as bypassing the conversion to electricity change as the primary energy-efficien- assets, the long-term perspective must
and using natural gas directly in furnaces cy paradigm is reconsidered in the light be a factor in near-term decision making
and water heaters. of deep decarbonization. and investment. A key revelation that
While much of the logic of primary But there is also no clear mandate emerges from long-term planning—and
energy efficiency remains valid today, for promoting electrification in cur- is seldom visible in shorter-term analy-
it has limitations when seen through the rent policy. Indeed, if anything, there sis—is the need for electrification as the
lens of a low-carbon transformation. In are formidable barriers to fuel switch- third pillar accompanying energy effi-
a power system dominated by renew- ing, not least of which are the inter- ciency and low-carbon generation.
able energy, with near-zero variable cost, ests of oil companies and gas utilities. Deep decarbonization requires a
saving primary energy does not translate Current energy-efficiency programs suite of policies to transform infrastruc-
directly into lower marginal or average are designed to stay within their fuel- ture and markets, covering all three pillars.
ESMO is the premier event for electric utility professionals, contracting, construction and
consulting companies who are interested in hands-on solutions for the safe engineering,
construction, operation and maintenance of the world's power delivery systems.
EXHIBITORS: ESMO 2019 will feature two days of outdoor field demonstrations and
exhibits and a two-day technical program combined with an indoor exhibit area.
For more information, please visit ieee-esmo.com/exhibitors.
OMICRON 16 www.omicronenergy.com
R
REDUCInG CARBOn DIOxIDE tally constrained in quantity, given ex- al practice in integrating renewables at
(CO2) emissions to the level needed to isting technology and land-use practic- levels of around one-third of total gener-
stabilize the climate will require very es. Carbon capture and storage (CCS), ation, and there is now widespread con-
deep reductions in emissions from en- which could allow the continued use fidence in the ability of many systems
ergy suppliers and end users. In Europe, of fossil fuels in power generation and to reach well above this share. However,
the United States, and other industrial- industry, are constrained by the slow above about a two-thirds share of inter-
ized countries, this means reducing CO2 pace of technology development and, mittent power, especially with a limited
from fossil-fuel combustion at least 80% frankly, a lack of interest among policy carbon budget for gas generation, a new
below present levels by mid-century and makers and the fossil-fuel industry. mix of solutions for addressing energy
eliminating the emissions altogether a The upshot is that currently there is imbalances will be required, including
decade or two after that. The technology no feasible alternative to electricity in thermal, curtailment, storage, and in-
pathways for achieving such reductions, most end-use applications—automo- creased load and resource diversity from
while maintaining an energy system that biles, buildings, industry—for supply- expanded regional integration. Flexible
supports a modern economy, have re- ing the low-carbon energy required. electric loads, including the production
cently begun to be spelled out in consid- A mostly fossil-fuel-free energy sys- of fuels like hydrogen and synthetic nat-
erable detail in many countries, across tem means a major increase in elec- ural gas, may become key parts of the
sectors, and over time. tricity generation in the United States, balancing mix by mid-century.
roughly a doubling from current levels
Electricity and by mid-century, even accounting for Electrification
Deep Decarbonization the offset of some load by behind-the- and Energy Efficiency
Everywhere, the basic formula is the meter solar photovoltaics (PVs). This Integrating high levels of renewable gen-
same. There are three pillars required to realization has not yet been absorbed in eration poses some engineering-econom-
support a decarbonized energy system: some quarters. Many utilities see low ic challenges, but it has a market driver
highly efficient energy use, electric- or negative load growth at present and in falling wind turbine and PV module
ity produced with virtually no carbon project that into their long-term load prices, and it is widely understood to be
emissions (much lower than possible forecasts. Yet an economy-wide low- the path forward for power systems in a
with a natural-gas-based generation carbon transformation by mid-century climate-friendly world. The next frontier
fleet), and end uses that mostly run on will require light-duty vehicles and of decarbonization is widespread elec-
low-carbon electricity. These pillars buildings to be more than 90% electri- trification. This will be driven, in part,
stand up across geographies and stages fied and for industry to double its cur- by consumer demand and technological
of economic development because they rent electrification rate. This means a improvements, as we are beginning to
embody basic physics and chemistry. lot of new load to be met by a lot of new witness with electric vehicles (EVs) and
There is a role for other types of low-carbon generation. battery prices. But markets often need to
low-carbon fuels; for example, biofuels Without a revival of nuclear power, be kick-started by policy, and that will
may prove essential in aviation and or an unexpected emergence of CCS, require broad-based support. The need
shipping. But biofuels are fundamen- most of the low-carbon electricity for electrification hasn’t yet been univer-
needed will be provided by renewable sally embraced, especially in buildings
Digital Object Identifier 10.1109/MPE.2018.2820446
energy, especially wind and solar. Great
Date of publication: 18 June 2018 progress has already been made in actu- (continued on p. 128)
Envision a digital
future
Partner with Siemens PTI to
optimize your grid
siemens.com/power-technologies
ASPEN
OneLiner
Short circuit and relay coordination program
trusted by utilities worldwide
ASPEN OneLiner ™
Powerful, fast, intuitive and easy to use