Beruflich Dokumente
Kultur Dokumente
magazine
on the 18
cover
©ISTOCKPHOTO.COM/TRONAND
©ISTOCKPHOTO.COM/ERHUI1979
features
contents
18 The Tariffs of Tomorrow 49 China’s Solar Subsidy Policy
By Ahmad Faruqui and Cecile Bourbonnais By Houqi Dong, Bo Zeng,
Yuqing Wang, Yingxin Liu,
26 Flexible Network Pricing Encourages and Ming Zeng
Greater Sharing
By Furong Li, Chenghong Gu, and Ran Li 61 Utility Load Research
By Curt Puckett, Craig Williamson,
33 Distribution Network Rate Making Claude Godin, Will Gifford,
in Latin America Jonathan Farland, Tom Laing,
By Rodrigo Moreno, Bernardo Bezerra, and Tao Hong
Hugh Rudnick, Carlos Suazo-Martinez,
Martha Carvalho, Alejandro Navarro,
Carlos Silva, and Goran Strbac
columns &
departments
4 From the Editor
8 Leader’s Corner
14 Guest Editorial
72 History
80 Calendar
84 In My View
72
Digital Object Identifier 10.1109/MPE.2020.2971821
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.
H
HOW SHOULD ELECTRIC- hicles. The role of markets,
ity rates be designed? Many operating and maintenance
factors must be considered costs, and capital costs further
to properly balance reliabil- complicates the situation.
ity of service with the cost to
customers while also ensur- In This Issue
ing the economic viability of This issue of IEEE Power &
utility providers. Social and Energy Magazine presents a
political concerns add com- worldview of utility rate-mak-
plexity to regulatory decisions ing theory and implementa-
that require consideration of tion. The traditional two-part
cost causation, public policies, rates consist of fixed and vari-
and grid modernization. Rate able monthly energy charges
making was difficult enough, but require major reform to ad-
but now the transformational ©ISTOCKPHOTO.COM/SILVIONKA
dress the acceleration of grid
changes in the electric power transformation and make use
industry further complicate matters, given metering infrastructure (AMI) and of modern data collection and analysis
the growth of variable energy resources, other new technologies now provide techniques. The articles in this issue dis-
demand response, electrification, behind- highly precise data, but new analytic cuss lessons learned from rate-making
the-meter resources, microgrids, and the techniques are necessary for data inter- experience and provide an outlook for
introduction of newer technologies. Geo- pretation and analysis. This must be done future rate structures.
graphic diversity and types of demand in the backdrop of grid transformation, Guest Editors Ning Lu and Tao
served present other issues that affect the widespread addition of variable en- Hong compiled five articles that exam-
electricity rates. Many factors influence ergy resources, and greater penetrations ine the principles of rate making, the
the service provider’s revenue require- of new technologies such as electric ve- need for rate reform, proposals for new
ments, including the required rate of re-
turn; depreciation schedules; operations
and maintenance costs; other accounting IEEE PES Transmission & Distribution Conference and
factors; and financing, taxes, and sub- Exposition Postponed Until October
sidies. The rates provide incentives for
Following the declaration by the World Health Organization of COVID-19 as
desired behaviors, which influence reli-
a pandemic, local, regional, national, and international bodies have moved to
ability and economy of service.
restrict travel and mass gatherings. In light of these developments and for the
How can rate issues be addressed?
Data collection provides key information protection, health, and safety of our transmission and distribution (T&D) com-
for studying customer types and behav- munity, the IEEE Power & Energy Society made the decision to change the T&D
iors, system performance, and the need Conference and Exposition to 12–15 October 2020 at McCormick Place South.
for future capital investment. Advanced Please monitor the IEEE PES T&D website (https://www.ieeet-d.org/) for
updates on the reopening of registration and housing, exhibitor details, and
Digital Object Identifier 10.1109/MPE.2020.2972664 conference highlights.
Date of current version: 17 April 2020
Receive up to 60% off Fully integrated with Short Circuit, Over-Current Coordination,
and Equipment Evaluation modules
when you switch Evaluate alternatives quickly and easily to establish an optimal design
to SKM. Improve safety margins with user-definable arcing fault tolerance
Save time by automatically generating arc flash labels and work permits
You’ll also receive Determine required PPE levels instantly
free service to convert Increase flexibility by creating custom labels in any size with
competing software user-defined logos, text, comments, and field placement.
Avoid potential fines, post productivity, and increased insurance
projects over to SKM! and litigation costs.
Compare results from multiple project scenarios in a single table
based on the ANSI/IEEE C37 Standards. Separate solutions for low, DC SYSTEMS ANALYSIS Battery Sizing, Load Flow, & Short Circuit
different cases very easy.” medium and high voltage systems and for symmetrical, momentary Analysis. Evaluate all loading conditions for DC duty cycle loads and
and interrupting calculations. (Unbalanced Studies also available) AC emergency loads. Complies with IEEE std. 485, 1115, 399, 946,
and IEC std. 61660.
EQUIPMENT EVALUATION Equipment Evaluation Report. Automatically
compares short-circuit ratings, withstand ratings. Applies de-rating GROUND MAT Substation Ground Grid Design and Analysis. Optimizes
adjustments and user defined pass/marginal/failed criteria. Includes grid design using general purpose finite element algorithm for
error checking for input data and topology. potential analysis and graphical facilities to validate grounding
systems efficiency.
TMS Transient Motor Starting Simulation. Time-based motor starting
simulation with graphical output. Includes reduced voltage and Data Exchange Seamlessly share project data between
capacitor starting, graphical motor and load models. SKM projects and other third-party software. Exchange
format includes Tab Delimited, CSV, and SKM XML. Also includes
HI_WAVE Power Quality Investigation and Filter Design. Frequency
an interface to exchange data with Autodesk Revit and Intergraph
Scan, Harmonic Current, Voltage Distortion, Harmonic Load Flow and
SmartPlant Electrical.
Interactive Filter Design. Evaluate and correct harmonic distortions.
a focus on learning
PES University & Resource Center
V
VOLUNTEERING FOR THE IEEE Subcommittee, which facilitates and earned 1,691 scholarships, and
Power & Energy Society (PES) in any helps organize programs for students. more than 500 PES scholarship
capacity provides wonderful learning The subcommittee helps many gradu- recipients are working full time.
experiences, occasions for meeting ate and undergraduate power engineer- Similar initiatives have been ex-
knowledgeable colleagues, and the ing students attend the PES General panded to Canada, India, Italy,
opportunity to truly enrich our lives. Meeting, the North American Power and Puerto Rico.
On a bittersweet personal note, 2020 Symposium, and the T&D Conference ✔✔ The IEEE PES Erich Gunther
is my final year as the PES vice presi- and Exposition. For many students, Future Power Innovator Schol-
dent of Education after serving the these events are their first exposure arship: This fund was established
typical five-year term, which is ap- to PES and the power industry, offer- in memory of Erich Gunther to
proved annually by the PES president. ing them the chance to promote their recognize a student in power en-
PES Education provides a wide vari- technical interests, meet and collabo- gineering who pushes the bounds
ety of training venues for our members. rate with other students, and even find of current thought to improve the
Academics and industry professionals their first job. Students greatly benefit power and energy industry.
collaborate and generate educational from meeting lifelong PES members, ✔✔ IEEE PES Outstanding Student
materials in the form of webinars and who eagerly share information with Scholarships: These new awards
tutorials, fund and disseminate schol- upcoming generations of professionals recognize PES student members
arships for PES students, invite stu- and each other. worldwide who are pursuing
dents to attend PES conferences and master’s degrees and have cho-
cover their expenses, and participate PES Scholarships sen an academic path leading to
in many other activities that keep our PES now offers several scholarships to an electric power and energy en-
membership informed of the latest its student members, who are encour- gineering career. The criteria for
technical developments. The following aged to visit the PES website for details selecting recipients include their
sections provide information on several and eligibility criteria. The programs academic achievements, contri-
committees within PES Education. include the following: butions to meeting community
✔ ✔ IEEE PES Scholarship Plus: and humanitarian needs, and
Power and Energy This program defrays educa- leadership in advancing student
Education Committee tional expenses for qualified un- engagement within PES. In 2020,
The Power and Energy Education dergraduates who are interested PES will begin administering
Committee (PECC) promotes power in power and energy engineering this prestigious scholarship and
engineering and technology educa- careers. These undergraduates plans to award up to 20 student
tion and professional development in will, ultimately, replace the re- recipients with a plaque and prize
schools and industry by providing edu- tiring workforces in the United of US$10,000, which will be dis-
cational programs and advocating for States and Canada (throughout tributed directly to the student.
research opportunities. Several sub- IEEE Regions 1–6). PES pro-
committees fall under the PEEC, but vides seed money, and schol- IEEE PES High School
I would like to emphasize the impor- arships are then supported by Initiative
tance of the Student Meeting Activities the energy and power industry. The first PES High School Initiative
Digital Object Identifier 10.1109/MPE.2020.2972666
Since the first year of the pro- was held at the PES T&D Exposition
Date of current version: 17 April 2020 gram in 2011, 997 students have in 2018 in Denver, Colorado. This
U
UTILITY RATE MAKING IS BECOM- (considering appropriate govern- The electricity bills sent to most
ing increasingly critical. The high pen- ment subsidies and policy chang- American households today consist of
etration of distributed energy resources es) instead of making aggressive two parts: a fixed monthly charge and
changes not only the landscape of utili- transitions to new rate systems, an energy charge based on the total en-
ty operating and planning practices but such as establishing blockchain ergy usage of the month measured in
also the business models that govern platforms and real-time retail kilowatt-hours. The first article, “The
how utilities recover their service costs. markets? Tariffs of Tomorrow” by Faruqui and
The objective of power generation and ✔✔ What can be done in the policy Bourbonnais, explains the shortcom-
delivery has changed from considering paradigm to accelerate changes ings of such two-part rates. The authors
only the delivery of reliable and cost-ef- in the electric power industry? also offer an outlook on tomorrow’s
fective energy services to also account- ✔✔ How can a level playing field be tariffs structured in three-part rates
ing for sustainability and minimizing achieved for existing infrastruc- consisting of a fixed monthly charge,
adverse environmental impacts. Today’s tures, renewables, energy stor- a time-varying charge, and a demand
energy systems are evolving from tra- age technologies, and customer- charge. Moreover, they present many
ditional highly centralized suppliers to owned distributed generators to case studies from U.S. utilities, such
multiple distributed small suppliers or provide grid services? as Arizona Public Service, Oklahoma
even self-suppliers. This evolution could ✔✔ How can reliability be best fac- Gas and Electric, and Commonwealth
be further accelerated when major tech- tored into rate making? Edison, to illustrate various ways of
nology breakthroughs in energy stor- ✔✔ If self-generation becomes a ma- implementing the proposed rates.
age and photovoltaics (PVs) make these jor way of producing electricity, The United Kingdom’s energy reg-
systems less expensive, more reliable, how should transmission and ulator is actively reforming network
and more efficient. All of these electric distribution be priced? charges that facilitate a transition to a
system changes greatly affect many tra- ✔✔ What are the best rate-making low-carbon future. The existing pricing
ditional rate-making principles, meth- practices for regulated and non- methodologies reflect only the usage of
odologies, and processes. regulated utilities and for market- network assets by users but downplay
Many questions must be answered and nonmarket-based systems? the unused capacity. “Flexible Network
to address the challenges posed by this Pricing Encourages Greater Sharing”
paradigm shift. In This Issue by Li et al., the second article in this
✔ ✔ What rate-making principles The articles in this issue of IEEE issue, highlights the importance of
need to be abolished, adjusted, Power & Energy Magazine discuss ap- long-term uncertainty and long-term
replaced, or added? proaches, new trends, and future plans network investment in network pricing.
✔✔ How can a framework be created for understanding how utilities in dif- The authors propose new directions in
to better connect the wholesale ferent countries plan to answer these pricing principles that reward custom-
market with the retail one and questions. Five well-written articles ers who reduce network uncertainties
pass price signals to customers discuss various aspects of utility rate and network operators who best man-
in a timely manner? making under different operation and age the uncertainties.
✔✔ Should regulators maintain the regulation rules worldwide, including “Distribution Network Rate Mak-
status quo in retail rate making the perspectives of regulated and de- ing in Latin America” by Moreno et al.
Digital Object Identifier 10.1109/MPE.2020.2972665
regulated utilities in the United States, discusses the regulatory reforms imple-
Date of current version: 17 April 2020 China, Europe, and South America. mented by various r egulators in the
70% of fuse operations on overhead distribution transformers are for nuisance outages. While
there’s no trick for keeping the lights on at the end of lateral lines, there is technology. With the
VacuFuse™ Self-Resetting Interrupter, you’ll have advanced fault-testing technology where you need
it most at the outer reaches of the grid, improving customer satisfaction and saving you needless
maintenance expenses.
Plan for advanced protection with S&C, the worldwide leader in reliability.
See the difference at sandc.com/vacufuse19 © S&C Electric Company 2019. All rights reserved.
distribution sector in Latin American The rate-making process is non- retail rates, and implications for the de-
countries, including Colombia, Peru, trivial. No matter what rates are being ployment of distributed energy resources.
Chile, Brazil, and Argentina. Alterna- established and applied to the end users, The broad implications of rate reforms for
tive tariff structures currently being dis- they should be based on load research, a utilities, customers, and society are dis-
cussed in Chile and Brazil are introduced comprehensive study of how customers cussed. The authors recommend that reg-
as examples of the reformation efforts. use electricity. The last article, “Utility ulators and policy makers establish rate
In “China’s Solar Subsidy Policy,” Load Research” by Puckett et al., fo- structures that weigh these factors against
Dong et al. discuss the impact of sub- cuses on the origin of rate making. The each other while balancing other consid-
sidy policies on the nation’s PV indus- authors introduce the load research life- erations and stakeholder perspectives.
try over the past 15 years. The country’s cycle as well as the interaction between The articles in this issue would not
rapid growth of a new industry was load research and various utility busi- be made possible without the contribu-
achieved mainly by governmental ef- ness functions such as rate making, load tions of the volunteers who wrote and
forts. We hope that this article provides forecasting, and planning. They also of- reviewed them. We also thank the en-
you with a different perspective on fer their outlook on the next decade of tire IEEE Power & Energy Magazine
policy making for fostering new tech- load research areas and the key factors editorial board for their continued sup-
nologies and an alternative approach for load research to continue to thrive. port. We hope you enjoy reading the
for rate making, especially in countries In the “In My View” column, Satch- articles and find this issue informative
where utilities are closely regulated by well et al. discuss key drivers of retail and thought-provoking.
p&e
the government. rate reforms, design considerations of
Dynamic Security Assessment Software Advanced modeling and analysis features include:
• Comprehensive capabilities for voltage, transient, small signal, and
Offering a simple, user-friendly interface frequency stability assessment
with extensive analysis options, DSAToolsTM
is a suite of software tools for power system • Integration with EMS/WAMS for on-line DSA to assess security of
analysis. The software is designed for real-time system conditions
applications in both off-line studies and • Computation of secure operation regions (stability limits)
on-line dynamic security assessment.
• Recommendation of remedial actions
www.dsatools.com
Experience and support:
• Prominent provider of the on-line DSA technology with over 50
installations worldwide including 8 out of 9 ISOs in North America
• Consulting services for custom software and model development,
system studies, and training
Contact us:
dsainfo@powertechlabs.com
T
TODAY, MOST AMERICAN RESIDENTIAL CUSTOMERS have been carried out with time-varying rates. The empirical
pay for electric service via tariffs that are structured as evidence shows that customers can understand and respond
two-part rates consisting of a fixed monthly charge and a to incentives provided by new tariffs that convey the cost
volumetric energy charge expressed in U.S. dollars per kilo- structure of electricity to customers. Additionally, smart
watt-hour. The fixed-charge component generally comprises meters now track the energy use of half of all U.S. residen-
a small portion of the bill, which is dominated by the volu- tial customers, removing a major barrier to the deployment
metric charge. of modern tariffs.
California’s energy crisis in 2001–2002 triggered an The introduction of smart digital technologies, changing
important discussion about the need to better connect retail consumer tastes, and new state policies promoting renew-
and wholesale markets. Since then, scores of pilot programs able energy sources have cast doubt on the sustainability
of the utility business model based on traditional two-part
Digital Object Identifier 10.1109/MPE.2020.2972136
tariffs. This shift comes at a time when many utilities are
Date of current version: 17 April 2020 making significant investments to modernize the grid and
integrate distributed energy re- The second criterion, equity, refers to fairness among
sources (DERs). These capital in- customers and between the utility and customers. Although
vestments, along with most other rate design nearly always involves some degree of cross-sub-
utility costs, do not vary with the sidy, a utility should aim to remove unintentional subsidies
volume of electricity consumed. between customer types. According to Bonbright, a natural
They thus cannot accurately be way to achieve equity among customers with different load
reflected through a volumetric profiles and consumption values is through cost-reflective
charge. This has resulted in large rates. Under such rates, customers who incur high costs for
residential customers effectively the system will pay proportionally higher amounts than low-
subsidizing smaller customers, cost customers.
who, as a result of low energy us- Revenue stability refers to the utility’s ability to recover
age, pay less relative to the costs its costs through a sufficient and predictable level of reve-
they impose on the system. An nues. Bill stability, the fourth criterion, then stipulates that
additional limitation of flat volu- while the utility must recover its costs, ideally through cost-
metric charges is that they fail to reflective rates, it must also protect customers from unman-
capture the effects of temporal ageable fluctuations in their bills. Although new rates will
and seasonal variability, which nearly always result in bill increases for some customers,
result in peak periods that are utilities can take steps to minimize seriously adverse and
more expensive to utilities than unexpected impacts, for instance by gradually implementing
off-peak periods. changes to rates.
Consequently, a m isa lign- Finally, customer satisfaction is needed for the successful
ment between utilities’ cost and implementation of any changes to the pricing structure. If
rate structures currently exists, not properly explained or rolled out, even simple rates can
particularly with respect to the cause confusion and subsequently trigger a backlash from
residential class. In response to customers. Regulators and utility companies who anticipate
this challenge, many utilities such an adverse reaction from customers will resist imple-
©ISTOCKPHOTO.COM/TRONAND
The efficient pricing frontier reflects Bonbright’s crite- charged more in peak hours when the cost of energy is high.
rion of equity in that, while it is good to offer choices to cus- Figure 2 illustrates this tradeoff.
tomers, each rate must still recover the cost of serving each This tradeoff applies to all customers, including those
customer correctly to avoid creating intercustomer subsidies. who are disadvantaged or have a low income. In fact, low-
As a result, to capture the impact of customer load volatil- income customers tend to have flatter load profiles and thus
ity on wholesale prices, low-risk rates should yield lower are likely to benefit from rates that charge higher prices in
potential savings than higher-risk rates that better reflect the peak hours. On the other hand, critics of dynamic new rates
time-varying cost of energy. For instance, the average price argue that disadvantaged and low-income customers gener-
implicit in a standard tariff should be higher than the aver- ally have more difficulty responding to new rates, either due
age price implicit in a TOU rate, under which a customer is to physical or flexibility constraints.
Such apprehension about cus-
tomer reactions largely explains
why the adoption of smart rates has
not kept up with the adoption of
smart meters. As important as eco-
nomic efficiency may be, it remains
only one of Bonbright’s key crite-
ria, and concerns about customer
Reward (Bill Savings)
RTP
TE satisfaction can immediately
DSS
VPP halt any changes to the pricing
PTR CPP
TOU structure. However, as described in
Demand Charge more detail later in this article, pro-
Standard Tariff tections can be applied for vulner-
Higher FC able customers to mitigate risk and
ease the transition.
GB With DR
Another concern surrounding
three-part tariffs is that, by reduc-
GB Bill ing the volumetric portion of cus-
tomers’ bills, they may discourage
Risk (Bill Volatility)
energy efficiency. However, cost-
reflective tariffs are the best way
figure 2. Efficient pricing frontier. FC: fixed charge. to simultaneously promote equity
C
CONSU M ERS I N T H E U N I T ED K I NG -
dom typically spend, on average, £33 billion/
year for electricity, of which a significant portion
(21% in 2017) is paid to distribution companies for deliv-
©ISTOCKPHOTO.COM/ERHUI1979
ering electricity to customer premises. To protect consum-
ers’ interest, Ofgem (Office of Gas and Electricity Markets),
the energy regulator in the United Kingdom, is actively
reforming network charges so that they remain usable for
By Furong Li, the purpose of facilitating the industry’s transition to low-
carbon emissions.
Chenghong Gu, Network charges are applied to generators and suppli-
and Ran Li ers for their use of the system to transport energy over a
distance. Utilities employ network charges as commercial
instruments that influence network users’ behavior, which
Flexible
promotes the most efficient use of the network and achieves
cost recovery. The fundamental principle for traditional net-
work pricing methodologies also reflects the use of network
assets attributable to network customers, such as distributed
Network Pricing
Encourages
Greater Sharing
Digital Object Identifier 10.1109/MPE.2020.2972137
Date of current version: 17 April 2020
table 1. Great Britain’s network charges reform timeline. The orange color background denotes the
transition from a postage stamp to cost-reflective pricing.
1990–1992 First Major Reform Second Major Reform
Status Quo (1993) (2006) Third Major Reform (2013)
Transmission Postage stamp ICRP ICRP Improved ICRP
• The same price • Reflects the distance • Differentiates the
and network power must travel generation technologies,
from generation i.e., intermittent renewable
to the distribution generation and fossil
center generation and their
• The same treatment impact on long-term
between generation network development
and demand
Distribution Postage stamp for each DRM (EHV, HV, LRIC/FCP (EHV) LRIC (EHV)
voltage level–DRM and LV) • Reflects the distance CDCM (HV and LV)
(EHV, HV, and LV) power must travel
• The same price and to the customers’
voltage level premises
• Downstream • Reflects the degree of
customers assumed to utilization along the
use all the upstream distribution paths
assets • Treats generation and
demand the same–
CDCM (HV and LV)
• DRM pricing
• Time of use (TOU)
• Treats generation and
demand differently;
generation always
given a credit
DRM: distribution reinforcement model; EHV: extra high voltage; LV: low voltage; MV: medium voltage; LRIC: long-run
incremental cost pricing; FCP: forward cost pricing; CDCM: common distribution charging methodology.
0.4 0.15
0.3
0.1
Density
Density
0.2
0.05
0.1
0 0
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
(Hour) (Hour)
(a) (b)
0.3 0.25
0.25 0.2
0.2
0.15
Density
Density
0.15
0.1
0.1
0.05 0.05
0 0
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24
(Hour) (Hour)
(c) (d)
figure 1. The temporal distribution of residential consumers’ peak demands. (a) R1: The residential customers without stim-
ulus (demand response) or special tariffs. They are billed an existing flat rate and currently the most representative residen-
tial consumers. (b) R2: Consumers billed using TOU tariffs reflecting different unit prices through the day. (c) R3: Consumers
with PVs. (d) R4: Consumers with EVs.
Density
0.2
Reduce Future Uncertainties
and Risk Investment
0.1
There is a huge diversity in energy usage across a wide cus-
tomer base. Uncertainties vary substantially among customer
types, both in terms of the magnitude and time of peak use. 0
Likewise, uncertainties across different voltage levels vary 0 2 4 6 8 10 12 14 16 18 20 22 24
significantly, with lower voltage levels having much higher (Hour)
uncertainties. However, traditional network pricing meth- (a)
odologies do not distinguish between load uncertainties and 0.5
future projections for different loads, which could lead to
unnecessary network investments, as is currently experienced 0.4
in parts of Great Britain’s network. As a result, the new round
0.3
of reforms in network charges that began in 2019 in the United
Density
0.3
0.3
Density
0.2
Density
0.2
0.1
0.1
0
0
0 2 4 6 8 10 12 14 16 18 20 22 24
0 2 4 6 8 10 12 14 16 18 20 22 24 (Hour)
(Hour) (c)
figure 2. The temporal distribution of SME consumers’ figure 3. The peak demand temporal distribution at (a) HV,
peak demand. (b) MV, and (c) LV levels.
table 2. The forecasting accuracy for different table 3. The forecasting accuracy across different
customer types. voltage levels.
Customer Type R1 R2 R3 R4 SME Voltage Level LV MV HV
MAE 0.0608 0.0602 0.1178 0.2624 0.0673 MAPE (%) 55.78 12.83 4.21
One approach to managing uncertainty could be by offer- access to mitigate uncertainties. In doing so, the risk of
ing flexible network access, recognizing that future distri- making mistakes in costly investment can be substantially
bution network operators would have two core businesses: reduced, and network operators can be rewarded for their
traditional long-term network investment and short-term improved ability to mitigate uncertainties.
flexible operation that mobilize inherent resources at the
customer and network levels to manage uncertainties. For For Further Reading
the latter, it is important to understand that resources have “Digest of United Kingdom Energy Statistics (DUKES)
finite capacity over limited time periods; therefore, network 2016: Main chapters and annexes,” Department for Business,
operators could be rewarded if they are willing to take risks Energy and Industrial Strategy, U.K. Government, London,
and develop greater capability in understanding and miti- Sept. 9, 2016. [Online]. Available: https://www.gov.uk/
gating uncertainties through short-term operational intelli- government/statistics/digest-of-united-kingdom-energy
gence, thus minimizing investments in new assets. -statistics-dukes-2016-main-chapters-and-annexes
As more customers opt for less-expensive intermittent net- F. Li et al., “Distribution pricing: Are we ready for the
work access, there could be less of a need for customers to smart grid?,” IEEE Power Energy Mag., vol. 13, no. 4, pp.
request firm capacity, and the cost of accessing firm capacity to 76–86, July 2015. doi: 10.1109/MPE.2015.2416112.
serve fewer customers would become excessively expensive. In “Structure of electricity distribution charges: Consulta-
such a situation, local energy provisions may become competi- tion on the longer term charging framework,” Ofgem, 2005.
tive when compared to their distant counterparts. This pricing [Online]. Available: https://www.ofgem.gov.uk/publications
framework would drive an efficient future-proof network over -and-updates/structure-electricity-distribution-charges
time, which would have less network investment and rely more -consultation-longer-term-charging-framework
on network operations that reduce congestion and uncertainties. F. Li and D. L. Tolley, “Long-run incremental cost pricing
Such networks would provide the capacity to deal with short- based on unused capacity,” IEEE Trans. Power Syst., vol. 22,
term uncertainties rather than a network dominated by longer- no. 4, pp. 1683–1689, 2007. doi: 10.1109/TPWRS.2007.908469.
term plans for network asset additions; however, regulators F. Li, D. Tolley, N. P. Padhy, and J. Wang, “Framework
would need to consider social justice issues arising from users for assessing the economic efficiencies of long-run network
requiring set electrical service, such as the elderly, the disabled, pricing models,” IEEE Trans. Power Syst., vol. 24, no. 4, pp.
and caregivers. 1641–1648, 2009. doi: 10.1109/TPWRS.2009.2030283.
“Delivering the electricity distribution structure of charg-
Summary es project,” Ofgem, 2008. [Online]. Available: https://www
The degree of sophistication in network pricing has improved .ofgem.gov.uk/publications-and-updates/delivering-electricity
dramatically in the last 30 years, moving away from pure -distribution-structure-charges-project
cost allocation to cost-reflective pricing, indicating both the “Project TransmiT,” Ofgem, 2011. [Online]. Available:
characteristics of the network and its users to improve overall https://www.ofgem.gov.uk/electricity/transmission-networks/
network utilization. Decarbonization and digitalization have charging/project-transmit
fundamentally changed the energy landscape and created “Charging futures forum,” Ofgem, 2017. [Online]. Avail-
major uncertainties in customer demand and network use as able: https://www.ofgem.gov.uk/publications-and-updates/
well as the future outlook. These changes challenge the present charging-futures-forum
principles and charging methodologies of electric rates, which
do not adequately consider uncertainties in network customer Biographies
use. Many types of regulatory regimes can address this issue. Furong Li is with the University of Bath, United Kingdom.
In this article, we quantified the diversity in network and Chenghong Gu is with the University of Bath, United
customer uncertainty, identified opportunities for mitigat- Kingdom.
ing uncertainty through sharing network use and customer Ran Li is with the University of Bath, United Kingdom.
demand response and generator assets, and proposed future
p&e
directions for network pricing that offer flexible network
F
F O L L OW I N G THE TREND OB-
served in developed economies, vari-
ous Latin American governments are
committed to reducing greenhouse gas
emissions, particularly in the power
sector. In countries such as Chile, Peru,
Colombia, Brazil, and Mexico, various
regulatory policies have been issued
to meet renewable-generation integra-
tion targets and satisfy the increasing
demand from consumers for supply
quality. Meanwhile, the integration of
distributed generation (DG) in rural
and urban areas as well as the increas-
ing need to integrate electric vehicles
(EVs) in urban areas are driving impor-
tant reforms in the distribution sector.
Distribution networks are expected
to increase reliability and integrate
various distributed energy resources
(DERs) by using an array of newly
available technologies (including smart
meters, online monitoring and control,
©ISTOCKPHOTO.COM/LEOLINTANG
and other smart grid and IT advances), enabling utilities to in the context of DERs, and it poses challenges related to the
implement an active (rather than the historical passive or deployed meter technology and the fact that tariffs, to which
“fit-and-forget”) approach to network operation, advanced network users react, present a limited level of spatiotem-
operational measures, and nonwire solutions that will effec- poral granularity and cost reflectiveness. Other important
tively displace costly investments. This environment poses concerns include equity and fairness in tariff and reliability
challenges, but there are opportunities too. Understanding levels, especially in rural and other underdeveloped areas.
those opportunities and designing regulations accordingly is Thus, regulators consider various policy concerns and objec-
a priority in Latin American countries, where maintaining tives when deciding on network regulations, which are sum-
affordable energy is essential. marized in Figure 1. In the rest of this article, we focus on
To ease the transition toward a more reliable and modern distribution network price controls, especially remuneration
electricity network, regulators need to reshape frameworks and tariffs. We summarize rate making in Latin America
that have historically focused on low costs (rather than the countries and discuss the Chilean and Brazilian experiences.
best value for money) and expand electricity coverage and
access. Regulators in countries including Colombia, Peru, Rate Making in Latin America
Chile, Brazil, and Argentina have implemented or are under-
taking regulatory reforms in the distribution sector. The most Overview
important topics of debate (relevant to rate making) include Determining the regulated revenues for a distribution net-
enhancing reliability for consumers, connecting DERs, and work company is a complex task with multiple conflicting
providing secure and flexible services between distribution objectives as well as determining rates, charges, or tariffs
and transmission networks within an incentive- or perfor- that the company is allowed to impose for billing and col-
mance-based price-control regime. lecting fees from consumers. The design of such processes
Well-designed incentive- and performance-based regula- should consider that, in broad terms, remuneration will affect
tions use financial enticements to encourage distribution net- network owners’ decisions (investments, operation, mainte-
work owners to provide their (multiple) services securely and nance, administration, and so forth), while tariffs will impact
cost-effectively. Apart from encouraging network owners to network users (the locations they choose, their consump-
be efficient, network users (that is, producers, consumers, tion and production profiles, and so on). While traditional
and prosumers) should be motivated to improve their opera- objectives for appropriate remuneration center on revenue
tional and deployment profiles. This is particularly relevant adequacy [to ensure that income is adequate to cover capital
expenses (CAPEX) and operating expenses (OPEX), includ-
ing a reasonable return on capital] and productive efficiency
(for the regulated service to be provided at the most efficient
cost), the objectives for tariff design focus on allocative effi-
ciency (guaranteeing that effective charge signals are sent
Quality and
Reliability Decarbonization so that users can make cost-effective decisions about their
consumption and production) and equity (ensuring that net-
work tariffs feature desirable distributional characteristics,
Economic
especially in terms of protecting vulnerable consumers). To
Efficiency and achieve this, there are numerous regulations that emphasize/
Equity and
Revenue
Fairness weigh conflicting objectives differently.
Adequacy Latin America employs cost-of-service and incentive- or
performance-based regulations (with a preference for the
latter) to determine allowed revenues. In countries using
the cost-of-service approach, such as Ecuador and Costa
Rica, planners determine future network investments, and
regulators ensure that a company’s costs are fully recov-
figure 1. The tetralemma for distribution network regula- ered through appropriate charges. In countries applying
tion in Latin America. incentive- or performance-based approaches, such as Chile,
105 90
80
100 70
SAIDI (h/year)
Access (%)
60
95 50
90 40
30
85 20
10
80 0
Chile
Cuba
Uruguay
Brazil
Argentina
Venezuela
Paraguay
Costa Rica
Mexico
Ecuador
Colombia
Dominican Republic
El Salvador
Panama
Peru
Guatemala
Bolivia
Nicaragua
Honduras
Cuba
Brazil
Argentina
Chile
Panama
El Salvador
Colombia
Peru
Costa Rica
Dominican Republic
(a) (b)
figure 2. The (a) electrification and (b) reliability levels in Latin America.
table 2. The end consumer’s bills and their breakdown in three Latin American countries (in US$). All are low voltage.
Santiago, Chile Quito,* Ecuador Rio de Janeiro, Brazil
Sector Unit Residential Industrial Residential Industrial Residential Industrial
Generation US$/month — — — — — —
US$/kWh 0.1 0.08 0.03 0.03 0.06 0.06
US$/kW/month — 6.99 — — — —
Transmission US$/month — — — — — —
US$/kWh 0.02 0.02 0.01 0.01 0.03 0.03
US$/kW/month — — — — — —
Distribution US$/month 0.85 0.99 1.41 — — —
(including
metering US$/kWh 0.02 — 0.05 0.03 0.03 0.03
and billing)
US$/kW/month — 9.2 — 4.18 — —
Others US$/month — — — — — —
US$/kWh 0.001 0.001 0.01 0.01 0.03 0.03
US$/kW/month — — — — — —
Total US$/month 0.85 0.99 1.41 1.41 — —
US$/kWh 0.14 0.1 0.1 0.08 0.16 0.16
US$/kW/month — 16.2 — 4.18 — —
Time kWh rate kWh and No time kWh rate No time No time
resolution with up to kW rates variations with up to variations variations
(maximum) three levels with three levels
in one day no time in one day
variations
*In Ecuador, there is a government subsidy of roughly US$0.05/kWh that reduces the overall bill to the values shown here.
price caps have been limited to relatively small values, low- ✔✔ innovation and network modernization, evolving to-
ering costs significantly in comparison with other devel- ward the smart grid
oped countries. However, as illustrated in Figure 3, lower ✔✔ decarbonization and renewables.
costs have an impact on service quality because of tradeoffs These objectives were initially de-emphasized by regulators to
between network performance and expenses. In general, prioritize electrification, coverage, and access at a reasonable
low-cost networks tend to be less reliable. Understanding cost (notice that access is still a problem in countries such as
this cost-performance balance is critical in developing coun- Bolivia, Nicaragua, and Honduras, as indicated in Figure 2).
tries because higher reliability levels might require funding Remuneration should evolve, in our opinion, to enable dis-
increased levels of network investment and, hence, higher tribution networks to innovate and become more active, such
electricity bills. as by operating with smart grid technologies. This would
After nearly 40 years of distribution network rate mak- improve the optimization of CAPEX and OPEX in network
ing in Latin America, the quality of service is becoming companies, driving new operational measures and nonwire
more critical. In addition, policy makers are pushing regu- solutions to deliver reliability and other services more effec-
lators to include more sophisticated objectives in remunera- tively. With the increasing penetration of DERs, price sig-
tion and tariff regimes (apart from the historical ones focus- nals will become more relevant for them to be adequately
ing mainly on cost reductions). The goals include deployed and operated, potentially providing services to
✔✔ quality of service, even beyond reliability and voltage distribution and transmission networks. Increased levels of
quality, including resiliency against hazards and cus- coordination will be needed at the interface between trans-
tomer satisfaction mission and distribution systems. It is envisioned that net-
✔✔ equity and fairness in tariffs and quality work operators will be able to trade various security and
0.09 AT Austria
LU PT
Electricity Bill (Distribution Network Cost) (US$/kWh)
CL Chile
0.08 DE CZ Czech Republic
SE DE Germany
0.07 DK Denmark
NL EE Estonia
UK IE ES Spain
0.06 FR EE FR France
DK AT CZ
SI GR Greece
0.05 ES HR Croatia
HU HU Hungary
0.04 HR IE Ireland
CL IT Italy
LT LV RO LT Lithuania
0.03 IT GR LU Luxembourg
MT LV Latvia
0.02 MT Malta
NL The Netherlands
0.01 PT Portugal
RO Romania
SE Sweden
0
0 100 200 300 400 500 600 SI Slovenia
SAIDI Distribution Network (min) UK United Kingdom
figure 3. Distribution network costs in capital cities versus SAIDI in all distribution networks. Due to the lower voltage
levels in Chile’s distribution networks (up to 23 kV), we have considered part of the subtransmission systems (which cover
up to 110 kV) as distribution networks to properly compare values among countries.
table 3. The SAIDI in hours per region and per year in Chile.
Seven-Year
Region Population 2012 2013 2014 2015 2016 2017 2018 Average
Aysen 103,158 23.1 27.9 26.1 29.9 19.7 31.2 14 24.6
Magallanes y Antartica 166,533 6.2 8.8 8.2 9.2 5.1 6 6.9 7.2
Chilena
Arica y Parinacota 226,068 21 14.3 33.9 12.3 10.7 15.3 23.2 18.6
Atacama 286,168 25.9 19.5 22.8 53.6 11.1 22.6 16.4 24.6
Tarapaca 330,558 29.8 24.4 59.9 23.8 20.2 18.1 14.6 27.3
Los Rios 384,837 27.9 25.6 25.1 26.6 22.3 24.7 19.5 24.5
Antofagasta 607,534 18.6 14.9 25.2 22.9 15.9 16.3 11.7 17.9
Coquimbo 757,586 10 11.5 9.8 44 11.5 10.5 10.1 15.3
Los Lagos 828,708 30.1 24.2 25.9 23.9 18.4 22.3 17.2 23.2
O’Higgins 914,555 16.6 18.2 18 20.4 17.9 23.2 16.8 18.7
La Araucania 957,224 34.1 34.6 30.7 32.3 31.5 51 28.3 34.6
Maule 1,044,950 20.1 14.1 16.9 26 20.8 33.1 14.7 20.8
Valparaiso 1,815,902 12.4 9.3 10.1 17.2 9.4 10 7.1 10.8
Biobio y Nuble 2,037,414 28.6 19.3 20.6 19.3 16.9 20.5 13.2 19.8
Metropolitana de Santiago 7,112,808 8.9 7.7 8.4 8.8 8.2 13.6 8.5 9.2
Country level 17,574,003 16.7 13.9 15.7 18.1 13.4 18.6 12.1 15.5
Exceptional events are included. Values higher than 50 h are in red. Regions are sorted according to population.
100
0.2 is planned. Chile has an increas-
SAIDI (h/year)
800
700
Number of EVs and Plug-In EVs
600
500
400
300
200
100
0
January
April
July
October
January
April
July
October
January
April
July
October
January
April
July
October
January
April
July
October
January
April
July
October
January
April
July
October
January
April
July
October
2012 2013 2014 2015 2016 2017 2018 2019
(a)
6,000
Number of PV Rooftop Installations
5,000
4,000
3,000
2,000
1,000
0
January
March
May
July
September
November
January
March
May
July
September
November
January
March
May
July
September
November
January
March
May
July
September
November
January
March
May
July
September
November
that their investments will be adequately remunerated, com- This relates to the discussion about the interface be-
panies are likely to prefer conventional solutions that may tween transmission and distribution and the need to
be less expensive during the short term but compromise the have a whole-system view of the planning and regula-
distribution system’s economic performance (and the entire tory problem.
electricity system) through the longer term. ✔✔ The set of prices used to calculate the model firms’ costs
The debate about modernization and innovation includes follows the concept of the replacement cost, valuing
the uncertainty related to planning network expansions. assets as new according to current market prices. This
This is becoming increasingly important due to the many can cause fluctuations in a company’s valuation and
factors that must be considered, including the rapidly evolv- produce higher risks for investors.
ing costs of technologies, market prices, policies, and DER ✔✔ The real costs of capital to investors remain hidden
deployment. The need to find technical solutions to network during the regulatory process, since companies’
uncertainty becomes a challenge, and regulators must prop- expenses are valued by the replacement cost. Most
erly scrutinize and incentivize proposals. With uncertainty, importantly, this blindness includes the return that in-
innovative solutions (such as battery energy systems and vestors will ultimately receive and is a failure of mo-
open soft points, which apply power-electronic devices to nopoly regulation.
normally open portions of the distribution system) become ✔ ✔ A referential cost of capital equal to 10% (before
more attractive as a means to more easily adapt to, cost- taxes and assuming that there is no leverage) that is
effectively deal with, and hedge against a number of sce- fixed by law (for model firms) may not appropriately
narios that might happen in the future. In other words, rec- reflect 1) market conditions at the time of the control
ognizing uncertainty in the network-expansion-planning and 2) real firms’ risks, including those imposed by
problem increases the value of smart grid technologies, the remuneration approach. This problem should be
which network planners and regulators must acknowledge. fixed by calculating appropriate costs of capital every
The second important aspect of planning under uncer- time tariffs are determined.
tainty relates to regulatory scrutiny. It is important to rec- ✔✔ Within the control period, there is no differentiation
ognize that an optimal-investment decision that was made between the costs that distribution companies can and
under uncertainty cannot be evaluated and justified ex post cannot control. Therefore, all variations in real costs,
facto, when complete information is available. From a sto- even those that cannot be managed by the companies,
chastic-programming viewpoint, it is clear that an ex ante are passed to network owners. This increases the risk
decision made under uncertainty may seem suboptimal at a to investors, without a clear benefit.
later date. This argument has special relevance for incentive- ✔✔ Despite its complexity, the remuneration mechanism is,
based remuneration methods that face after-the-fact regu- in practical terms, too approximated (it relies on a small
latory scrutiny and correction of the regulatory asset base sample of companies, averages results from different
(RAB), where companies face the risk of a backlash if invest- sources in a two-thirds and one-third fashion, uses sim-
ments are deemed unjustified. This is also important in Chile, plistic greenfield model firms, and so forth). Even for
where real companies’ infrastructure, in every tariff period, the same company, results can vary too much, depend-
is valued through a greenfield model firm that possesses per- ing on who determined them (the authority or the firm;
fect information, an approach that disregards legacies built see the values and biases in Figure 6). This implies that
under uncertainty and the optimal infrastructure needed to a network owner can realize a lower or higher cost of
face future unknowns. Incorporating tools to deal with uncer- capital depending on the inherent randomness of the
tainty in network regulation is becoming a must. process, not on corporate efforts to reduce expenses.
✔✔ There is no coherence between the distribution of net-
Other Concerns About Remuneration work owners’ capital costs and efficiencies. Following
There are other important concerns that arise in the debate the principle of comparing similar companies through
surrounding the model-firm approach and are common to yardstick regulation, firms with larger efficiencies
other remuneration methods. Among them the following: should receive larger returns. This does not occur under
✔✔ Model firms minimize costs within the distribution Chile’s remuneration approach, which does not measure
sector, in isolation from the rest of the power system. firms’ relative inefficiencies to mimic real competition.
Company Regulator
complexity and require a greater 40
effort by regulators to understand 35
the costs and decisions of real
30
firms. Chile’s answer probably lies
25
between the two extremes. The
debate involves questions such as 20
the following: Up to which point 15
do real costs have to be recognized 10
in the remuneration process? What 5
would the new role, set of tasks, 0
and burden be for regulators? 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6
Some principles underlying 2004 2008 2012 2016
Chile’s remuneration method may Representative Company and Year
need to be maintained to smooth
any transition. Part of companies’ figure 6. The distribution charge per representative firm under the past four tariff
real costs may need to be rec- controls, according to company and regulator studies, before the weighting process.
ognized (for example, by using Representative firms are sorted by connection density (number 1 presents the highest
a “brownfield” model firm) to and corresponds to Santiago, Chile).
encourage investments in quality
and reliability and other desir- table 4. Tariff modalities in Brazil.
able distribution-service outputs
Time-Dependent Network
as well as to compensate investors Tariff User Structure Pricing Remuneration
for part of the risk they face. Rec-
ognizing companies’ real needs Conventional Low voltage kWh only No Volumetric
and costs would increase regula- Blue High voltage kWh and kW Peak/off peak Peak demand
tors’ burden of scrutiny and per-
Green High voltage kWh and kW Peak/off peak Peak demand
haps their responsibility in future
investment decisions. A major White Optional for kWh only Peak/intermediate/ Volumetric
low voltage off peak
concern involves the extent of the
table 5. Yearly revenues, costs, and profits of three investment options (in generic currency).
Option 1 Option 2 Option 3
Invest Profit in Invest Profit in Invest Profit in
Yearly RAB ment Present ment Present ment Present
Year Depreciation (Option 1) Revenue Cost Value Revenue Cost Value Revenue Cost Value
100 100 –100
1 5 95 13 12.15 13 12.15 13 12.15
2 5 90 12.6 11.01 12.6 11.01 12.6 11.01
3 5 85 12.2 9.96 12.2 9.96 12.2 9.96
4 5 80 11.8 9 11.8 9 11.8 100 –67.29
5 5 75 11.4 8.13 13 (as in 9.27
year 1)
6 5 70 11 7.33 12.6 8.4
7 5 65 10.6 6.6 12.2 7.6
8 5 60 10.2 5.94 11.8 6.87
9 5 55 9.8 5.33 11.4 6.2
10 5 50 9.4 4.78 11 5.59
11 5 45 9 4.28 10.6 5.04
12 5 40 8.6 3.82 10.2 4.53
13 5 35 8.2 3.4 9.8 4.07
14 5 30 7.8 3.02 9.4 3.65
15 5 25 7.4 2.68 9 3.26
16 5 20 7 2.37 8.6 2.91
17 5 15 6.6 2.09 8.2 2.6
18 5 10 6.2 1.83 7.8 2.31
19 5 5 5.8 1.6 7.4 2.05
20 5 5.4 1.4 7 1.81
21 6.6 1.59
22 6.2 1.4
23 5.8 1.22
24 5.4 1.06
Total 6.72 42.12 47.24
19
20
21
22
23
24
25
26
27
system expansion.
20
20
20
20
20
20
20
20
20
20
Year
Tariff-Structure Alternatives
figure 7. The binomial tariff’s impact on the penetration of During 2018 and 2019, ANEEL held two public hearings to
DG for low-voltage consumers. (Data source: EPE, 2018.) discuss new tariff structures for low-voltage consumers and
changes to the net-metering mechanism. For the low-voltage new meters would represent a 13% increase for a typical elec-
consumers, ANEEL defined five alternatives to the purely vol- tricity bill. In alternatives 3 and 4, a fixed payment to recover
umetric tariff, carrying out a regulatory analysis for representa- the distribution network costs would be applied. The differ-
tive distribution companies based on the following principles: ence is that, in alternative 4, the amount to be paid is based on
1) Revenue adequacy: guaranteeing sufficient service a consumer’s average consumption during the past 12 months,
remuneration while in alternative 3 the fixed payment does not vary with
2) Allocative efficiency: encouraging productive con- consumption. According to the regulator, alternative 4 pres-
sumer behavior during the short and long term ents the better tradeoff between costs and benefits (especially
3) Transparency and intelligibility: simplicity of adop- when considering the transaction cost), since there would be
tion and understanding no need to change the existing meters.
4) Justice and nondiscrimination: no competitive advan- At the second hearing, the regulator proposed five alter-
tages to any consumer. natives for net metering. The mechanism would continue
The alternatives under discussion are listed in Table 6. applying volumetric tariffs, but, contrary to today’s prac-
Alternative 1 proposes to increase the minimum amount of tice, rates would vary for withdrawals and injections, dif-
power prosumers must consume to use the distribution net- ferentiating prosumers’ production and consumption. While
work (even if they consume zero or negative kWh), from 100 consumption would continue paying for the entire electricity
to 219 kWh, which is necessary to recover distribution net- supply chain, production would pay for only part of it. Table 7
work costs. This would reduce the distribution companies’ presents the alternatives. The first, for example, corresponds
cost-recovery risk. Still, consumers without DG would sub- to the current practice, where the energy produced by DG is
sidize prosumers, since both would be exposed to volumetric remunerated at a price that contains all components. Instead,
tariffs. In alternative 2, the minimal consumption limit would alternative 5 remunerates DG production at the wholesale
be replaced by a fixed tariff per consumer. The analysis car- price (without including network and other sector charges),
ried out by the regulator showed that cross subsidies remained which is significantly lower than that of the first option.
in this approach, since all consumers would be charged by the The chosen alternative will function temporarily, until a
same tariff regardless of their
peak load levels.
Alternative 5, with a pay- table 7. Components included in the volumetric tariff to remunerate DG
ment in $/kW, proved to be the production.
most efficient in terms of cost Sector Sector
allocation, since all consum- Distribution Transmission Charges Distribution Charges
Network Network Applied to Network Applied to Energy
ers would be charged based Alternative Tariff Tariff Peak Load Losses Consumption Tariff
on their (approximate) use of
the network (although the time Status quo X X X X X X
element is still missing, that is, One X X X X X
when the peak load occurs).
Two X X X X
This approach’s main problem
concerns the cost of replac- Three X X X
ing all the low-voltage-level Four X X
consumers’ meters. Initial es-
Five X
timates show that the cost of
©ISTOCKPHOTO.COM/GYN9038
T
By Houqi Dong, THE CHINESE GOVERNMENT HAS ISSUED NUMEROUS REGULA-
tions that significantly affect the number of photovoltaic (PV) installa-
Bo Zeng, tions in the country and the subsidies for their use. This article summa-
Yuqing Wang, rizes the internal and external environment of China’s PV industry and
describes its future trends and prospects and also discusses a proposed
Yingxin Liu, rate-making process and renewable energy supervision.
In 2017, China added 53 GW of PV power generation, an increase of
and Ming Zeng 18.5 GW compared to 2016, reaching a new historical high year-on-year
growth of 53.62%. However, prior to 2018, the national PV policy had many
issues, such as high presubsidy prices, a lack of incentives for technological
innovation, and the absence of installation performance monitoring.
Digital Object Identifier 10.1109/MPE.2020.2971824
Against a backdrop of excess production of PV panels and a waning
Date of current version: 17 April 2020 desire to provide financial support to the manufacturers, the government
PV Power-Generation
power were added, down 19.7% from a similar period in
PV Application
Downstream
Products
Products
System
duction decreased by 48.9% during the third quarter, and
the output of batteries and associated components fell by
more than 20%. Downstream enterprises were directly
affected by the policy change and sluggish demand, and
their profit margins shrank sharply through the second
half of 2018.
To ensure the sustainable development of PV panel
production, the State Energy Administration established
reduced subsidies in June 2018, extending them until
2022 while also postponing a planned feed-in tariff
PV Modules
Solar Cell
Modules
Polycrystalline
Solar Cells
PV Cells
Polycrystalline
Silicon Wafers
Silicon Wafers
Polycrystalline
Silicon Ingots
Silicon Rods
China’s PV Industry
Crystalline Silicon
Development Status
figure 1. The PV industry chain.
Furnace
Feedstock
Silicon
140
120
100
Production
80
60
40
20
0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Polycrystalline
4.5 8.4 7.1 8.4 13.6 16.5 19.4 24.2 25.6 28
Silicon (10,000 Tons)
Silicon Wafers (GW) 11 20 26 29.5 38 48 64.8 91.7 109.2 120
Solar Cell Pieces (GW) 10.8 21 23 25.1 33 41 51 72 87.2 95
Modules (GW) 10.8 21 23 27.4 35.6 45.8 53.7 75 85.7 93
figure 2. China’s production of polycrystalline silicon, silicon wafers, solar cell pieces, and modules.
120 450
425%
Setback 400
Phase Resulting 102
Initial Recovery and Rapid
100 From Antidumping and
Demonstration Development Phase 94.3 350
Antibribery Investigation
Phase
300
80 269.64%
73
250
206.72% (%)
60 200
53 53.06
0.16
Resource Region 1
0.14 US$0.1453 Resource Region 2
US$0.1380 US$0.1424 Resource Region 3
Benchmark Electricity Prices (US$/kWh)
0.02
0
2013–2015 2016 2017 2018 2018.6
Resource Region 1
Resource Region 2
Resource Region 3
Other Resource Region
• Clear Targets
• Follow the Reform Principles of Electricity Price and Subsidy Rates
Ensure the Same Targets • Adopt a Gradual Price-Reform Implementation Path
Extension
©ISTOCKPHOTO.COM/Z_WEI
Utility Load
Research
The Future of Load Research Is Now
Digital Object Identifier 10.1109/MPE.2020.2972668
Date of current version: 17 April 2020
The Load Research Lifecycle stratification include identifying the variables to use, such as
Figure 1 shows the load research lifecycle. We start in the upper seasonal and annual use, peak demand, and load factor, and
left corner with a “need for information.” Load researchers the key sample design planning variables of interest, such as
should work with internal clients to develop a preliminary plan system and class peaks.
that specifies the goals, objectives, and analytical approach for Next, decisions are made with respect to the confidence
fulfilling clients’ needs. level, desired relative precision, and anticipated variability
Next, load researchers are responsible for securing or of the target variables. These decisions dictate the sample
creating an appropriate population frame. For classical load size requirements. This often becomes an iterative process,
research, this comes in the form of the population billing with tradeoffs necessitated by the available budget. Finally,
data known for all customers on the system. The customers having settled on an appropriate sample design, the load
are typically segmented by rate class or other domains of researchers guide the selection of the sample.
interest. For some domains, such as customers with rooftop For some projects, such as an analysis of demand
solar or electric vehicles (EVs), the construct of the popu- response, technology assessments, and pricing experiments,
lation can be more complex and may require combining the sampling plans will need to be developed inside an
multiple internal and/or external sources of data to create an experimental design paradigm. The load researchers provide
appropriate framework for analysis. guidance and governance on the appropriate experimental
After identifying the measures of customer usage to be design to use, potentially randomized control trial, within-
used as the focus of the research study and the populations sample comparison, test versus comparison, preparticipation
for which the measures need to be reported, load research- versus postparticipation, and so forth.
ers develop an appropriate sample design, selecting the Historically, the next step has been to work with the
minimum number of customers needed for estimating a metering department to procure and install pulse-initiat-
population-level measure constrained by a desired level of ing meters with load recorders. This includes determining
precision of the estimate. Load researchers are responsible an appropriate data collection strategy, such as handheld
for making decisions with respect to the stratification strat- devices, modems, or a fixed network. Due to the high cost
egy for the sample. Sample stratification seeks to divide a of equipment, equipment installation, and data collection,
population into groups that are similar within the individual load researchers had to rely on interval data collected from
segments exhibiting characteristics related to the measure a limited sample of customers in the population of interest.
being estimated in the study. The greater the similarity of However, in today’s fast-approaching advanced metering
the customers within the strata, the smaller the number of infrastructure (AMI) world, these constraints are fading
customers needed to be sampled. The particular features of away, as we are seeing utilities deploying “super samples” or
even “census” representations.
As we transition to full AMI, sample data present several
unique challenges, such as storage and how to conduct vali-
dation, editing, and estimation (VEE). Added to the chal-
Need for Population
Data
lenge of storage is the communication bandwidth needed
Information Preliminary
Plan to transmit interval data from the field. Utilities often find
the need to compromise load interval granularity (5 min,
15 min, hourly, and daily) to ensure that they have the capac-
Population ity to reliably transmit data while maintaining bandwidth
Model Load Research
Data
Lifecycle for other operational priorities. Data quality issues arise
from any number of sources, including communication
problems, database malfunctions, and human errors. These
Expand issues can affect the data capture and its eventual analysis,
Results Sample
Data if not handled properly.
Some simple sanity checks such as ensuring consistency
with the correct time stamp, interval orientation (begin-
figure 1. The load research lifecycle. ning or end of period), units of measure, and frequency
26 July 2016
4
System Peak Hour
3.5
T1: Treatment Group
3 C1: Comparison Group
kW Per Customer
2.5
2
VPP Impact
1.5 Battery Charge
0.5
-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour Ending
(a) (b)
figure 2. The residential grid-side battery storage impacts: (a) the storage battery and (b) the results of the impacts when
the batteries were charged in the early morning and discharged during the on-peak period. VPP: virtual power plant.
All 100
HSMW HSHW Residential 90
80
0.47 5.37 0.65 9.63 0.36 5.14
% of Max kW
kW kW kW 70
0 6 12 18 24 0 6 12 18 24 0 6 12 18 24 60
M M M 50
40
A A A
30
20
J J J 100 80 60 40 20 0
A /C J J J Percent of Time Value Is Below
A A A
HSMW HSHW
S S S
O O O
Residential high winter load
N N N customers have a poorer load
factor than their medium winter
counterparts . . . this is an
Heating J J J indication that they likely
have a higher cost to serve.
figure 5. Clustering employing seasonal billing use determinants. A/C: air conditioning; HSMW: high summer medium
winter; HSHW: high summer high winter.
Most PES University courses are eligible for continuing education credits.
Visit ieee-pes.org/pes-university to learn more
U
UTILITIES STUDY HOW THEIR
customers use electricity across time Load research allows utilities to examine their customers’ use of electricity,
through load research. The outcomes of either in total or by individual end uses, and requires numerous disciplines in-
load research play an important role in cluding engineering, statistics, computer programming, and marketing. In this
cost-of-service studies and rate design issue’s “History” column, we examine utility load research, with a look at the
as well as many other activities, such as past from the first days of load research in the 1930s up to the rapidly advancing
transformer sizing, demand-side man-
future with automated metering infrastructure.
agement, load settlement, load forecast-
We welcome Curt Puckett and Craig Williamson, both with DNV GL, to
ing, and distribution and resource plan-
these “History” pages of IEEE Power & Electronics Magazine.
ning. Despite its existence for nearly
a century, load research is not a well- John Paserba
known subject in the power engineering Associate Editor, “History”
community. Here, we discuss the histo-
ry of load research and its current prac-
tice and applications. In another article
in this issue of IEEE Power & Energy table 1. The members of the AEIC Special Committee on Load Studies
Magazine, “Utility Load Research: The (1938–1943).
Future of Load Research Is Now,” we • S.W. Andrews, American Gas and Electric Service Corporation
provide a look into the future of load • W.E. Barbour Jr., Boston Edison Company
research as utilities increase their in-
• C.W. Bary, chairman, Philadelphia Electric Company
terval load data-collection capabilities
through the use of automated metering • A.D. Caskey, Public Service of Northern Illinois
infrastructure (AMI). • H.A. Enos, American Gas and Electric Service Corporation
• E.J. Fowler, Commonwealth Edison Company
Association of Edison
• J.R. Gardner, Central Hudson Gas and Electric Corporation
Illuminating Companies
Load Research Committee • R.E. Ginna, Rochester Gas and Electric Corporation
Load research got its start in the United • H.L. Harrington, Niagara Hudson Power Corporation
States in the late 1930s as a way to better
• L.V. Nelson, Union Electric Company
understand electric customers and their
contribution to an ever-expanding base • E.H. Schmidtman, Wisconsin Electric Power Company
and unprecedented growth. Interesting- • H.A. Snow, Detroit Edison Company
ly, the Association of Edison Illuminat- • E.T. Steel, Potomac Electric Power Company
ing Companies (AEIC) has had a Spe-
• A.H. Sweetnam, Boston Edison Company
cial Committee on Load Studies dating
back to 1938. Table 1 lists those who • F.M. Terry, Consolidated Edison of New York
were members of this committee from • C.M. Turner, Ebasco Services
• W.R. Waggoner, Commonwealth and Southern Corporation
Digital Object Identifier 10.1109/MPE.2020.2971845
• R.R. Hermann, correspondent member, Northern States Power Company
Date of current version: 17 April 2020
figure 1. The evolution of metering technologies since the 1950s. AMR: automatic meter reading.
Stop The
1943 and held its first organizational
meeting in 1944. Soon afterward, mem-
PAIN!
ber utilities began conducting load stud-
ies to develop a better understanding of
major energy-using appliances, heating
and cooling equipment, and commercial
and industrial processes.
Following World War II, the elec-
tric utility industry in the United States Get Back to Your
underwent an extended period of rapid Oneline Faster With
growth, and new facilities needed to be
built. Load forecasts projected that the Revit Integration
high rate of growth would continue into
the foreseeable future; however, U.S. ex- Get to your analysis faster. Use our efficient data
pansion slowed in the 1970s resulting in integration between Autodesk® Revit® and EasyPower
the utility industry overforecasting the power system analysis. Stop the pain of wasted time
need for capacity. In 1978, the Public in your day.
Utilities Regulatory Policy Act (PUR-
PA) established requirements for utili- This stand-alone module automatically creates a
ties to collect load research data, such one-line diagram in EasyPower with one mouse-click
as daily kilowatt (kW) demand load in Revit. Get back to what you want to work on.
curves, for large utility rate classes that
were at least 10% of the overall sales
· Map Revit families and properties to EasyPower
equipment
of the utility. PURPA helped establish
confidence (90%) and precision (±10%) · Bring EasyPower analysis results back into Revit
requirements on estimates of selected Autodesk EasyPower
cost-allocation factors, such as coin- Revit Model System One-line
cident (system) peak demand. In turn,
T-SCV
500 kVA
.48-0.48 kV
4%
kV
T-SCV-S
8
.4
3-#250, 1#250, 1-#3
PANEL
Automated
600 600
500 500
0.15
400 400
Protective
80 80
TX-2
-2
1000 / 1288 kVA 1600/1760
SWG-4
24
40 40
EasyPower LLC 30 BL-3 30
Comments: GE AKR-30H
20 800/400 20
1.90
d
Risk Hazar
BL-1
GE MVT
MV T-Plus
MVT-Plus
and Shock
Device
10 10
17.58
TIME IN SECONDS
Plug = 1600
6 6
Appro
5 5
LT
T Band = 1
3 PHASE Fault Total Fault Fault Equipment Duties M-1
19
4 4
Required
STPU
STPU = 2 2.5
.5 (4400A) BL-5
ST
S T Delay
Dela
elayy = Int
Incident Energy coverall
3 3
Boundary C-H HFD
Flash Hazard inches - Arc Flash4’ - 0”or arc rated
ST
S T Delayy I²t
I² = Out
Sym X/R Mult Asym
Coordination
Energy
M-1 0.480 19480.3 3.69 1.19 23188.3 LVPCB 19480.3 cover is removed 1 1
PNL-1 0.480 8882.5 3.73 1.19 10597.4 LVPCB 8882.5 t Name: MCC-23A CONFIGURATION Equipment Name:
SWG-4 (Fed By: .2
LT Band = 1
IInst
nst = 4 (3200A)
C-6
.2
Equipmen SYSTEM VALID FOR NORMAL 18A) 1 - 400 kcmil CU
NORMAL
PNL-2 0.208 5961.4 2.11 1.07 6377.3 LVPCB 5961.4 VALID FOR SYSTEM CONFIGURATIO
N ONLY TX-2
.1 1000 / 1288 kVA .1
.06 .06
C uttler H
Cutler am
m
Hammer Series C
.05 .05
HF
HFD
.04 BL-3 .04
Frame
Frame = 225A (150A
(150AT)
Trip = 150 29249A
.03 .03
Short Circuit
.02 .02
BL-5
20574A
.01 .01
.5 .6 .8 1 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8 9 100 2 3 4 5 6 7 8 9 1000 2 3 4 5 6 7 8 9 10000
Transformer
Bags In the
early days,
instrumentation
and data
collection were
extremely
substation
liners challenging.
Workforce Development
Following the publication of the first edi-
tion of the manual, the Load Research
Committee launched an educational se-
ries built around it. The Fundamentals
of Load Research Workshop continues
as the cornerstone of the education se-
1.800.231.6074 ries, helping to educate and introduced
reefindustries.com
new professionals to the world of load sample design, regression, and advanced comprehensive analytics review while
research. This course is held annually analytics. The course was taught by this focusing on new and advanced appli-
in New Orleans, Louisiana. team until late in the 1990s. cations of load research. Curt Puckett
In the late 1970s, Bill The early training was and Tim Hennessy serve as the lead
Hines from the Georgia
Institute of Technology
The directed by a team of in
dustry practitioners and
instructors for the advanced course.
The intermediate and advanced cours-
worked as a consultant Fundamentals academics. Today, the es are taught every other year on a ro-
to several Southeast- AEIC load research work- tating basis.
ern utilities, supporting of Load shops and training are In addition, the AEIC load research
the implementation of
load research efforts as
Research led by long-term indus-
try professionals com-
committee organizes an annual con-
ference where load researchers make
well as building the ini- Workshop prising consultants and presentations and share their analyti-
tial intermediate-level practitioners. The intro- cal experiences. In addition to load re-
training session for the continues as the ductory (fundamentals) searchers, the conference also attracted
AEIC, along with Doug
Montgomery. H i n e s
cornerstone of course is taught each year
and was recently moved
many load forecasters from utilities
as well as load forecasting vendors
and Montgomery were the education to a fall offering. The in- and consultants. Tao Hong, the for-
joined by Roger Wright, termediate course was mer chair of the IEEE Working Group
forming the original in- series. redesigned by Craig Wil- on Energy Forecasting, is a frequent
structor group for the liamson and Kelly Marin speaker. Many attendees of the AEIC
AEIC’s “advanced” course. This AEIC and focuses on the mathematics of load research conference are also ac-
course was a week-long immersion in the load research. Similarly, the advanced tive participants in IEEE Power & En-
mathematics of load research, including course was redesigned and includes a ergy Society conferences.
800-722-8078 • pr-tech.com
6874
& IAS
PES meetings
for more information, www.ieee-pes.org
T
TH E I EEE POW ER & EN ERGY contact Alina Schneiders, a.schneiders@ December 2020
Society’s (PES’s) website (http://www academy.rwth-aachen.de IEEE International Conference on
.ieee-pes.org) features a meetings sec- Power Electronics, Drives, and En-
tion, which includes calls for papers IEEE PES Asia-Pacific Power & ergy System (PEDES 2020), 16–19
and additional information about each Energy Engineering Conference December, Jaipur, India, contact Sury-
of the PES-sponsored meetings. Please (APPEEC 2020), 20 –23 September, anarayana Doolla, doolla@gmail.com,
check the website for the most current Nanjing, China, contact Li Zhang, http://pedes2020.com
information. zhanglinuaa@hhu.edu.cn, https://ieee
-appeec.org/ January 2021
June 2020 IEEE PES 2021 Joint Technical
IEEE Transportation Electrification IEEE PES Transmission and Dis Committee Meeting (JTCM 2021),
Conference & Expo (ITEC 2020), 24–26 tribution Conference and Exposi- 7–15 January, New Orleans, Louisiana,
June, Chicago, Illinois, United States, con- tion Latin America (T&D LA 2020), United States, contact Solveig Ward,
tact Rebecca Krishnamurthy, rebecca.k@ 28 September–2 October, Montevideo, sward@quanta-technology.com, www
rna-associates.com, https://itec-conf.com Uruguay, contact Kathy Heilman, kathy .pestechnical.org
.heilman@ieee.org, https://www.ieee
August 2020 -tdla2020.org/ June 2021
IEEE PES General Meeting (GM 2020), I EEE PES GT&D International
2–6 August, Montréal, Canada, con- October 2020 Conference & Exposition, Istanbul
tact Roseanne Jones, roseanne.jones@ IEEE PES Transmission and Distri- (GTD), 1– 4 June, Istanbul, Turkey,
ieee.org, http://www.pes-gm.org/2020/ bution Conference and Exposition contact Omer Usta, usta@ieee.org
(T&D 2020), 12–15 October, Chicago,
IEEE PES/IAS PowerAfrica (Power- Illinois, United States, contact Carl IEEE PES Innovative Smart Grid
Africa 2020), 25–28 August, Nairobi, Ken Segneri, csegneri@mjelectric.com, Technologies Asia (ISGT Asia 2021),
ya, contact Mercy Chelanget, chelangat http://www.ieeet-d.org/ 8–11 June, Brisbane, Australia, contact
_ke@ieee.org, https://ieee-powerafrica.org/ Tapan Saha, saha@itee.uq.edu.au
IEEE PES Innovative Smart Grid
September 2020 Technologies Europe (ISGT Europe IEEE PowerTech Milan (PowerTech
IEEE International Conference on 2020), 25–28 October, Delft, The 2021), 28 June–2 July, Madrid, Spain,
Power Systems Technology (Power- Netherlands, contact Kathy Heilman, contact Tomas GomezSan-Roman,
Con 2020), 9–12 September, Banga- kathy.heilman@ieee.org, https://www tomas.gomez@comillas.edu
lore, India, contact Kasi Viswanadha .isgt-europe-2020.nl/
Raju Gadiraju, kasiViswanadhaRaju For more information on additional tech-
.Gadiraju@ge.com November 2020 nical committee meetings, webinars, and
IEEE PES Innovative Smart Grid events, please visit our IEEE PES calen-
IEEE International Forum on Smart Technologies Asia (ISGT Asia 2020), dar: https://www.ieee-pes.org/meetings
Grids for Smart Cities (SG4SC 2020), 23 –26 Novemb er, Per t h, West er n -and-conferences/conference-calendar.
16–18 September, Aachen, Germany, Australia, contact Farhad Shahnia,
Digital Object Identifier 10.1109/MPE.2020.2971823
f.shahnia@murdoch.edu.au, https://
p&e
Date of current version: 17 April 2020 ieee-isgt-asia.org/
R
RETAIL ELECTRICITY PRICING as well as by favorable incentive poli- Designing Retail Rates for
is evolving in the context of broader cies and electric utility rate designs, a Changing Grid
shifts in how electricity customers pay such as net energy metering (NEM). Although numerous changes to retail
for grid services and are compensated Increased DER adoption has put pres- rates have occurred across the United
for customer-sited generation. These sure on regulators and utilities to re- States in recent years, five particularly
retail rate reforms are occurring most- vise retail rate designs prompted by salient trends have emerged, at least
ly, though not exclusively, among resi- NEM program caps and optimize the in part, in response to these drivers.
dential customer classes and are driven size, location, and operation of DERs. The depth and longevity of experience
by a variety of technology- and policy- Third, utilities have increasingly across these five rate designs vary sig-
related trends. raised concerns about fixed-cost recov- nificantly, but each represents some
ery and revenue sufficiency, an issue incremental progression from fully
Key Drivers partly associated with increased DER bundled rates based on average util-
of Retail Rate Reforms deployment but related more generally ity costs toward rates with a greater
Four particular drivers stand out. First, to flat or declining load growth and unbundling of electricity services and
many utilities across the United States compounded by cost increases related temporal differentiation and location-
have deployed residential advanced to grid modernization. This motivated al granularity.
metering infrastructure (AMI), allow- some utilities and states to reexamine
ing for two-way communication and their historically heavy reliance on The Increased Pursuit of
a more granular recording of energy volumetric energy charges to cover Residential Time-Based Rates
consumption. According to the Edison both fixed and variable utility costs Despite more than four decades of ex-
Foundation Institute for Electric Inno- and also prompted efforts to revise net- perience, only roughly 3% of residen-
vation, more than 55% of U.S. house- metering tariffs. tial electricity customers were on time-
holds now have advanced meters, the Fourth, at the bulk power system based rates in 2017. As AMI deployment
majority of which were installed be- level, utility system planners and op- progresses, broader and stronger regula-
tween 2010 and 2015. One of the key erators are experiencing or anticipat- tory support for time-based rate options
rationales for justifying AMI deploy- ing significant changes to hourly net is beginning to emerge. The implemen-
ment has been its potential to enable a load as a result of the broader deploy- tation of time-based rates is primarily
more widespread application of time- ment of variable renewable energy re- motivated by economic efficiency (i.e.,
based rate designs. sources, both utility and smaller scale to better align electricity prices with
Second, residential and commercial distributed resources. In particular, marginal system costs) and the need to
electricity customers continue to in- the surplus generation of electricity lower utility costs by reducing peak de-
crease their investment in distributed by these resources during the middle mand and improving system load fac-
energy resources (DERs). For exam- of the day is causing significant de- tors by encouraging customers to shift
ple, U.S. residential solar photovoltaic clines in energy supply prices. This electricity consumption from peak to
(PV) capacity has grown by 50% per creates an opportunity for customers off-peak periods. Recent and notable
year on average since 2010, spurred to capture value in shifting load to examples of this include the Sacra-
by substantial installed-cost declines those periods and away from morn- mento Municipal Utility District in Cali-
ing and/or afternoon peaks if rate de- fornia, which is transitioning all of its
Digital Object Identifier 10.1109/MPE.2020.2971844
signs reflective of these price profiles
Date of current version: 17 April 2020 are implemented. (continued on p. 81)