INDIA

Partnership to Advance Clean Energy - Deployment (PACE - D)

Technical Assistance Program

Smart Grids: An Approach to

Dynamic Pricing in India

April 2014
This report is made possible by the support of the American People through the United States Agency for
International Development (USAID). The contents of this report are the sole responsibility of Nexant, Inc. and do not
necessarily reflect the views of USAID or the United States Government. This report was prepared under Contract
Number AID-386-C-12-00001.

INDIA
Partnership to Advance Clean Energy - Deployment (PACE - D)
Technical Assistance Program

Smart Grids: An Approach to

Dynamic Pricing in India

April 2014
This report is made possible by the support of the American People through the United States Agency for International
Development (USAID). The contents of this report are the sole responsibility of Nexant, Inc. and do not necessarily reflect the views
of USAID or the United States Government. This report was prepared under Contract Number AID-386-C-12-00001.

Table of Contents
Executive Summary ......................................................................... 1
An Approach to Dynamic Pricing in India ...................................... 3
Acronyms ....................................................................................... 22
References ...................................................................................... 23

Acknowledgements
This report was prepared under the Partnership to
Advance Clean Energy - Deployment (PACE-D)
Technical Assistance Program, which is funded by the
United States Agency for International Development
(USAID).
The PACE-D Technical Assistance Program would like to
express its sincere appreciation and gratitude to the
numerous experts in India who were interviewed by
the program and who provided useful information on
their experience and insights on the subject. The team
would particularly like to thank B N Sharma, Joint
Secretary, Ministry of Power; and the Indian Smart Grid
Task Force Secretariat, in particular N S Sodha and
Sachin Shukla.
The report substantially benefited from the overall
guidance provided by Kumud Wadhwa of Power Grid
Corporation of India. We would like to thank Monali
Zeya Hazra of USAID/India who provided extensive
input during the design and preparation of the report.
We are also grateful to Jeremy Gustafson, Apurva
Chaturvedi and Anurag Mishra of USAID/India for their
overall guidance to the team and the program.
Principal Author
Dr. Ajit Kulkarni
Contributing Authors
Anish De
Vikas Gaba
V Bharath
Technical Lead
Bhaskar Natarajan
Editorial Team
Kavita Kaur
Jessica Pelham
Peter du Pont
Rahul Kumar

Executive Summary
This report develops an approach to dynamic pricing in support of the Ministry of Power (MOP)
initiative for smart grid pilot projects in India. Dynamic pricing experiments are intended to reveal
how consumers respond to electricity prices and identify the most promising mechanisms
suitable for wide-scale deployment. Traditional electricity rate structures do not reflect the cost
difference of supplying electricity in peak versus off-peak hours. Therefore, the customer has no
market incentive to adjust their pattern of electricity consumption.
Dynamic pricing options such as time of use (TOU), critical peak pricing (CPP), critical peak rebate
(CPR), real time pricing (RTP), and variable peak pricing (VPP), that reflect time-varying cost of
electricity supply, have been in use worldwide to encourage peak load management and
demand reduction.
A global review of dynamic pricing pilots reveal that its success lies in customer engagement
together with the type of pricing options, enabling technology, and market segment chosen, as
well as the reliability of service. While success rates have been higher in smaller size test
markets, larger sized test markets have also been successful with the help of strong coordination
and stakeholder commitment.
Based on the research and pilots reviewed, it is recommended that a smaller size pilot be initially
selected for a residential test market with homogeneous income group (and similar consumption
patterns) in urban areas with TOU, CPP, and CPR pricing options.
Successful introduction of smart meters and time-based dynamic electricity pricing requires a
well-planned social marketing campaign to help raise awareness and give customers the
information and support they need to become more energy efficient. More sophisticated
dynamic pricing experiments will require the same levels, if not more, of infrastructure and
customer education. The infrastructure requirements can be met by the anticipated smart grid
pilots in India. The benefits of such pricing programs may be larger than TOU pricing. In the long
run, it may be more efficient to move directly to such a program rather than first moving to TOU
pricing and then to a more sophisticated dynamic pricing program.

Smart Grids: An Approach to

Dynamic Pricing in India

Currently, the power system in many developing countries is ripe for Advanced Metering
Infrastructure (AMI), supporting efforts to reach 100 percent metering of loads. With progress in
metering of loads, AMI can be supplemented by implementing relatively straight-forward and
transparent TOU pricing.
For any such experiment, a costbenefit analysis and full-scale rollout should be performed and
updated as better information becomes available. If the cost-benefit analysis does not show an
overall benefit for full-scale implementation, it may not make sense to spend the effort on the
smaller scale experiment.

PACE-D Technical Assistance Program

An Approach to Dynamic Pricing in India

Background
The Governments of the U.S. and India signed a Memorandum of Understanding on
November 24, 2009, creating the Partnership to Advance Clean Energy (PACE) to enhance
cooperation on clean energy, energy security and climate change. PACE has two interlinked
components: a Research Component, known as PACE-R, and a Deployment Component,
known as PACE-D.
USAID's PACE-D Technical Assistance Program is a part of the overall PACE-D initiative and
builds upon USAID/India's previous energy programs. It aims to accelerate India's transition to
a high-performing, low-emissions, and energy-secure economy. The five-year program will
support institutional strengthening and development of an enabling environment by providing
technical assistance to formulate and implement policies, regulations, and strategies for clean
energy deployment.
A key component of the program focuses on deployment of smart grid electric system. Under
this program, the activities are closely aligned with the work being undertaken by the MOP
through the India Smart Grid Task Force (ISGTF). The program aims to support and strengthen
smart grid initiatives in the country through focused technical assistance, technology exchange
and access, knowledge sharing, training and capacity building, and access to financing. The
technical assistance provided under PACE-D TA Program supports the goals of establishing
smart grids in India to increase power availability, reduce aggregate technical and commercial
losses (AT&C), and improve utilization of renewable resources for sustainable growth.
The Government of India has selected 14 electricity distribution utilities to implement smart
grid pilot projects. These projects will contribute toward India's pressing priorities of providing
broader access to electric power supply, reducing energy losses due to infrastructure and
power theft, integrating renewable energy, and improving energy efficiency and reliability.
Pilots will be evaluated for technological and commercial benefits and for the potential of
being scaled up.

Smart Grids: An Approach to

Dynamic Pricing in India

Dynamic Pricing in Smart Grid Pilots

Time-based or dynamic pricing refers to the provision of a service or commodity in which the price
depends on the time when the service is provided or the commodity is delivered. The rationale of
dynamic time-varying pricing is to reflect changes (expected or observed) in supply and demand over
time and their impact on costs. Time-based pricing includes: (i) fixed time-of use rates for electricity
and public transport,(ii) dynamic pricing reflecting current supply-demand situation; or (iii)
differentiated offers for delivery of a commodity depending on the date of delivery (futures contract).
Traditional electricity rate structures do not reflect the cost difference of supplying electricity in peak
versus off-peak hours. Therefore, the customer has no market incentive to adjust their pattern of
electricity consumption. This report develops an approach to dynamic pricing to support the MOP
initiative for smart grid pilot projects in India. Dynamic pricing experiments are intended to gather
information about how consumers respond to electricity prices and identify the most promising
mechanisms suitable for wide-scale deployment.
For example, dynamic pricing programs have the potential to make longer term impacts as plug-in
hybrid electric vehicles (PHEVs) become prevalent. These types of pricing programs and the charging
times for PHEVs provide another opportunity to address the significant disparity between peak and
off-peak demand. In an analysis of the potential impacts of PHEVs in 2020 and 2030 in 13 regions of
the U.S., researchers at the Oak Ridge National Laboratory found that charging PHEVs at 10 p.m.
instead of 5 p.m. would make a significant difference in total generation costs (reference #1).
In the U.S., the Energy Policy Act of 2005 (EPAct) adds several new federal standards to the Public
Utility Regulatory Policies Act of 1978, including time-based metering and communications. Section
1252 of EPAct, smart metering, covers these federal standards and defines the following time-based
pricing methods:
Time of use (TOU) pricing: Under TOU pricing, the electricity prices are rates set for specific
hourly time periods on an advance or forward basis. Prices paid for energy consumed during
these periods are pre-established and known to consumers in advance, thus allowing them to
vary their usage in response to these prices and manage their energy costs by shifting usage
to a lower cost period or reducing their consumption overall.
Critical peak pricing (CPP): TOU prices are in effect except for certain critical peak days when
prices may reflect the exceptionally high costs of generating and/or purchasing electricity at
the wholesale level.
Real time pricing (RTP): Electricity prices may change hourly, or even sub-hourly, with price
signals provided to the user shortly in advance, reflecting the utility's cost of generating and/or
purchasing electricity at the wholesale level.

PACE-D Technical Assistance Program

 Peak load reduction credits: For consumers with large loads who enter into
pre-established peak load reduction agreements that reduce a utility's planned
capacity obligations.
As a result of EPAct 2005, the U.S. Federal Energy Regulatory Commission report Assessment of
Demand Response and Advanced Metering was published in 2006 and updated in 2008. This
report assesses the status and potential electric demand response resources of the U.S. (by region)
from all consumer classes. Among other topics, it reviews the following:
The saturation and penetration rate of advanced meters and communication technologies,
devices and systems;
Existing demand response programs and time-based rate programs.
Dynamic-pricing programs that charge higher electricity prices during peak demand periods are
conceived as an effective tool to shift electricity consumption from peak to off-peak hours. In this
report, dynamic pricing refers to any of the time-based pricing methods for electricity, including but
not limited to those listed above. Many utilities, worldwide, have incorporated some form of
dynamic pricing (such as TOU pricing) in their rate structures and are experimenting with the
incorporation of further dynamic pricing (such as RTP and associated demand response programs) in
smart grid initiatives in order to create a mechanism for improving market efficiency through
demand response. Dynamic pricing leads to different electricity prices at different times of the day
and year to reflect the time-varying cost of supplying electricity.

Dynamic Pricing in the Indian Context

The power market structure in India can be categorized based on the tenure of the contract
between the distribution utility and the generator. These categories are as follows:
Long term (7-25 years)
Medium term (1-7 years)
Short term (less than 1 year)
Long term procurement constitutes the majority of the utilities' power purchase portfolio and serves
as its base load. However, due to increased demand, utilities are inevitably exposed to the power
market to procure electricity to meet their medium and short term needs. In several cases this
procured power is at very higher cost, or often subject to a high level of volatility.
As seen in Graph 1 and 2, procurement from short term has increased significantly over the past
few years at the national level.

Smart Grids: An Approach to

Dynamic Pricing in India

Graph 1: Volume of Electricity Transacted through Over the Counter (OTC) and
Power Exchange
80

Volume (In Billion Units)

70
60

24

16

50

16

40

30
3

20

33

10
12

14

15

2004-05

2005-06

2006-07

21

22

2007-08

2008-09

52

51

2011-12

2012-13

40

Electricity transacted through Trading Licensees

2009-10

2010-11

Electricity transacted through Power Exchange

Source: CERC Monthly Market Monitoring Reports

Similarly, prices in the short term (ST) market have also exhibited volatility, as is illustrated by the
graph below.

Graph 2: Price of ST Transactions of Electricity on Indian Energy Exchange

(IEX): FY 2009-10 to FY 2012-13
2009-10
Average - INR 5.171/kwh

2011-12
Average - INR 3.498/kwh

Source: IEX Data

PACE-D Technical Assistance Program

2010-11
Average - INR 3.564/kwh

2012-13
Average - INR 2.794/kwh

For a long time in India, electricity has been considered as social commodity, to be provided at a low
cost to the citizens. Despite significant increases in load, electricity tariffs have changed very little.
This has led to inefficient use of energy and electricity distribution utilities incurring significant
financial losses, which in turn has slowed down investments in the necessary infrastructure. This
has further affected industries where reliability of electricity supply is of paramount importance.
Additionally, the industrial and commercial sectors have been impacted by high electricity tariffs
while the residential consumer segment has been less impacted. With the introduction of open
access regulations, industrial consumers are increasingly purchasing power from the market, which
is at times less expensive than the tariff set by the utility and is more reliable. Thus, there is an
automatic price signal sent to the participating industrial customers with the evolution of the
wholesale markets. As observed from the following graph, 40-45 percent of overall trades on the
day-ahead markets are from direct customers availing themselves of open access. While relatively
low in overall terms, the volumes are significant (annually about 6-7 BU), constituting close to 0.8
percent of the overall electricity supply in the country.

Volume (In Million Units)

Graph 3: Volume of Participation of Open Access (OA) Consumers in

Day Ahead Market
1400.00
1200.00
1000.00
800.00
600.00
400.00
200.00
0.00
Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar
09 09 09 10 10 10 10 10 10 11 11 11 11 11 11 12 12 12 12 12 12 13 13

Source: IEX (till March 2013)

With increasing price volatility and higher costs for peak power, it becomes important to
communicate price signals to consumers who may be otherwise used to them. Such signals
through dynamic pricing are likely to create benefits in the form of avoided generation, transmission
and distribution costs. These consumers are both commercial and residential in nature. Graph 4
indicates the growing stock of residential buildings across key towns in India.

Smart Grids: An Approach to

Dynamic Pricing in India

Graph 4: Estimated Growth in Residential Stock

Bangalore
Chennai
Mumbai
Pune
Ahmedabad
Noida
Gurgaon
Delhi
0

5000

10000

15000

2016-17

2015-16

20000

25000

2014-15

30000

2013-14

35000

40000

2012-13

Number of Residential Buildings

Source: Data collated from www.magicbricks.com, www.makaan.com, www.99acres.com and

www.magicbricks.com

The pricing design has to be differentiated from conventional TOD pricing since the peaks are much
more volatile than in the past and the prices can vary radically between days of the month and
between seasons.
Demand response is typically reliability-based (customers reduce their consumption in response to
system conditions), or price responsive (customers change consumption behavior in response to a
financial incentive). The key difference between reliability-based demand response and dynamic
pricing is that demand response targets relatively fewer but larger customers, while dynamic pricing
targets more customers however with less consumption. For price-responsive demand response,
dynamic pricing and demand response are complementary to each other.
In this regard, it is essential that an in-depth review of the current pricing programs that have been
piloted or implemented in different countries be conducted prior to developing an approach for
dynamic pricing that would be applicable to Indian smart grid pilot projects. It is expected that a
review of past pilot programs on dynamic pricing will provide the following insight:
i.

Key factors to the success of dynamic pricing programs

ii. Understanding of why customers do or do not accept or change their usage behavior in
response to various treatments
iii. Correlation of program success to enabling technology
iv. What are the most effective incentives to the customers?

PACE-D Technical Assistance Program

v. What are the key considerations for a successful incentive design?

vi. How important is market segmentation for the pilot implementation?
vii. Is there a trend worldwide in the popularity of the dynamic pricing options?
viii.What are the key contributing factors affecting the shift in demand?
ix. Customer engagement options that significantly impact the success of dynamic
pricing programs.
To help answer these questions, this report includes a review of current dynamic pricing methods
and options worldwide, lessons learned from current dynamic pricing programs, and a discussion
that will lead to the development of an approach to dynamic pricing experiments in India.

Dynamic Pricing Options

There are four widely used types of dynamic pricing options:
TOU: Prices vary by rate period and day of week, but do not change based on system
conditions (technically, not a dynamic rate option). TOU prices and time periods are generally
fixed at least one year in advance. TOU offers consumers the lowest reward-risk profile.
RTP: Prices change on an hourly or sub-hourly basis to reflect the true cost of supply in the
wholesale market. The volatile nature of RTP can harm consumers and may discourage
participation in voluntary dynamic pricing programs, although RTP offers consumers the
highest reward compared to traditional flat-rate pricing, but at the highest risk.
CPP: CPP captures the true cost of power generation during peak demand periods (i.e., top
100 to 200 hours a year). In exchange for paying very high prices during those peak hours,
customers receive a discounted rate for all remaining hours of the year. The actual times in
which the CPP will be in effect are identified on a day-ahead (and sometimes day-of) basis,
depending on the demandsupply balance.
Peak Time Rebate (PTR): PTR is similar to CPP, but instead of higher prices during peak
periods on selected days, customers are paid to reduce load (technically, not a rate, but a
pay-for-performance program). PTR allows customers to remain on their current flat rate
while receiving a cash rebate for each kilowatt hour (kWh) of energy usage they reduce from
their baseline usage during the peak hours. Thus, PTR provides an opportunity to customers
to save money on their monthly bill.

Smart Grids: An Approach to

Dynamic Pricing in India

Lessons Learned from Dynamic Pricing Programs

Time of Use Rates
TOU pricing of electricity is the most widely used dynamic pricing method in power systems today.
In the U.S. and Canada, it has gone well beyond the experimental pilot project stage and has been
widely adopted. TOU pricing was instituted in Ontario, Canada, in 2006 by the Ontario Electric Board,
and today covers 90 percent of residential customers. In California, at the direction of the California
Public Utilities Commission, businesses are moving to a full time-based TOU electric rate structure
and while TOU rates are optional for residential customers.
TOU rates require the functionality of interval meters or TOU meters that has existed for years in
older technologies and is included today within the features of smart meters. Interval and TOU
meters that have been installed historically to measure commercial and industrial customers with
time-based rates often lacked automatic meter reading. In California, since 2006, nine million smart
meters have been installed for Pacific Gas and Electric Company customers and five million smart
meters have been installed for Southern California Edison customers.
A smart meter is an electrical meter that records consumption of electric energy in intervals of an
hour or less and communicates that information at least daily back to the utility for monitoring and
billing purposes. Smart meters extend the functionality of interval meters or TOU meters in addition
to automatic meter reading. Smart meters usually involve real-time or near real-time sensors, power
outage notification, power quality monitoring, and generally can support two way communications
as part of AMI in a smart grid implementation.
These additional features of smart meters enable more than simple automated meter reading.
Smart meters enable two-way communication between the meter and the central system. An AMI
with smart meters differs from traditional automated meter reading in that it includes increased
functionalities enabled by communications with the smart meters. Unlike home energy monitors,
smart meters can gather data for remote reporting. While home energy monitors and smart meters
can be installed together, these are distinct devices with different functions that should not
be confused.

Smart Meters and Time of Use

The American Council for an Energy-Efficient Economy reviewed more than 36 different residential
smart metering and feedback programs internationally. They published their results in 2010
(reference #3). Their conclusion was: To realize potential feedback-induced savings, advanced
meters [smart meters] must be used in conjunction with in-home (or on-line) displays and welldesigned programs that successfully inform, engage, empower and motivate people. Their results
call for a national social marketing campaign to help raise awareness of smart metering and give
customers the information and support they need to become more energy efficient, and raise
customer awareness regarding changes they must make to realize the potential of smart meters.

10 PACE-D Technical Assistance Program

While TOU rates and smart meters are widespread, their implementation has often been
controversial. For example, while in Ontario, Canada, 90 percent of residential customers pay TOU
rates and almost every household and small business now has a smart meter, in British Columbia,
Canada, the situation is more complicated. In British Columbia, TOU electric rates are opposed by
the provincial government and do not accompany smart meters.
BC Hydro had set a goal of implementing smart meters to all customers by the end of 2012.
However, after smart meter installations were associated with several fires, the Union of British
Columbia Municipalities voted in favor of a moratorium on smart meter installations in British
Columbia, and many BC municipalities passed motions opposing the installation of smart meters.
BC Hydro is not obliged to abide by these municipal decisions, and the provincial government insists
that installations will proceed based on global standards. In January 2013, when smart meters were
installed in 90 percent of customer homes, BC Hydro announced that smart meters will not be
installed at any home unless the company has the customer's permission, and headlines read BC
Hydro Smart Meter Installation Not Mandatory.
Italy and the Netherlands present two somewhat contrasting European examples of smart meter
installation. In Italy, the world's largest smart meter deployment was undertaken by the utility Enel
SpA with more than 30 million customers. Between 2000 and 2005, Enel deployed smart meters to
its entire customer base. These smart meters are fully electronic, all solid-state, with integrated bidirectional communications, advanced power measurement and management capabilities, and an
integrated, software-controllable disconnect switch. They communicate over low voltage power line
using standards-based power line technology to data concentrators, at which point they
communicate via IP to Enel's enterprise servers, demonstrating that smart grids do not require
wireless devices. The system provides a wide range of advanced features, including: the ability to
remotely turn power on and off; read usage information from a meter; detect a service outage;
change the maximum amount of electricity that a customer may demand at any time; detect
"unauthorized" use of electricity and remotely shut it off; and remotely change the meter's billing
plan from credit to prepay, as well as, from flat-rate to multi-tariff.
In the Netherlands, as part of a national energy reduction plan, the government proposed in 2007
that all seven million households in the country should have a smart meter by 2013. The roll out of
these meters was delayed in August 2008 for several reasons including difficulties registering smallscale local energy production (such as by solar panels), and then again in April 2009 after consumer
groups raised privacy concerns. As a result, the government decided to make the use of smart
meters voluntary.

Dynamic Pricing Pilot Projects

Smart meters enable and support various time-based dynamic pricing methods in addition to TOU
electricity rates. Most current smart grid experiments with dynamic pricing study time-based pricing
options that go beyond TOU rates.
Most analyses of pilot show that customers do respond to dynamic pricing rates by lowering peak
usage, and the response varies depending on the intensity of the price signal. The response
improves when enabling technologies are added. Pilot projects in the U.S. have demonstrated that
the dynamic pricing does not hurt low-income customers; on the contrary, it provides benefit to

Smart Grids: An Approach to

Dynamic Pricing in India

11

many low-income customers. In the U.S., the PowerCentsDC program has been highly successful,
as demonstrated by the fact that 89 percent of the participants would recommend it to their friends
and neighbors, and 93 percent preferred dynamic prices to flat rates (see Graph 3, also references #
10 and 11).

Categories for Defining Dynamic Pricing Research Objectives

In its Dynamic Pricing and Consumer Behavior Studies Webinar (April 20, 2010), the U.S. Department
of Energy addresses the following four categories which are important for defining dynamic pricing
research objectives:
1. What to test?
a. Pricing option Each pricing option can have several attributes such as:
i. Price levels by rate period will higher peak prices generate more impact?
ii. Number of rate periods how many rate periods would have more impact?
iii. Length of peak period will shorter period have more impact?
iv. Timing of rate period will seasonal rate periods be more relevant?
v. Combination of options as overlays on existing pricing tiers.
b. Enabling technologies - Enabling technologies such as direct load control, home area
network, programmable communicating thermostats, message alerts, in-home displays
(IHD), and web portals are anticipated to have an impact on the end users.
The following questions will be relevant in the development of a dynamic pricing
experiment:
i. What technologies are to be tested and what selection criteria should be used?
ii. Customer eligibility criteria for using the technology, if any.
iii. Cost impact.
c. Enrollment options Enrollment options could be mandatory, opt-in, or opt-out. It is
important to know how enrollment varies across different customer segments, rates,
enrollment options, and marketing strategies. Issues that need consideration include:
i. Desired enrollment model for the full-scale roll out.
ii. Attrition mitigation measures (e.g., bill protection) for a full-scale role out.
iii. Mitigation models for the trial/pilot phase.
d. Marketing strategy It is important to decide which features (e.g., message, package,
incentive, mode of communication, etc.) are to be included in the marketing offer. Issues
that need consideration include:
i. Relative importance of marketing features.
ii. Offering different incentives across different customer segments.
iii. Consideration of unconventional marketing methods.

12 PACE-D Technical Assistance Program

2. What population to reach? There are several target markets to choose from:
a. Residential customers (low/median/high income, etc.).
b. Commercial customers (size and type of business).
c. Industrial customers (timing of processes, size).
3. What impact to measure?
a. Changes in peak demand and energy use by time period and resultant savings.
b. Differential acceptance/enrollment/attrition rates for each option and market segment.
c. Changes in consumer behavior underlying the changes in energy use.

Summary of Dynamic Pricing Pilot Projects

Based on the study of 17 pilots listed in Graph 5, the following inferences can be drawn:

Based on the fact that more than 70 percent of the pilots reviewed were experimented on
residential customers, it may make more sense to start the pilot with residential test
market.

Out of five most successful pilots (reference #, 2, 11, 12, 16 Graph 5), four were tested on
residential customers.

Based on the data in Graph 5, the size of the test market does not seem to have a strong
correlation with the success of the pilot. However, three out of five most successful pilots
were conducted across a smaller cross-section of customers (1,000 or lower). It seems a
smaller size of test market will be more manageable from resource- and cost- containment
perspective.

A homogenous market mix would help establish a definite trend in the outcome. It may also
make sense to test the pilot in more than one test market for any differentiation purposes.

Pilots must provide enabling technologies (e.g., IHD, smart thermostat) to the customers in
the chosen test market.

Based on the pilots reviewed, CPP and TOU pricing with a combination of CPR was
more successful.

Considerations for Dynamic Pricing Design

Below is an outline of important considerations for the program design process adapted from the
American Electric Power, Texas experience (reference #4).

Begin with a soft launch: A soft launch is a limited release of a program which is essentially a
feedback process that provides a preview of the response and pre-validation of assumptions
before incurring significant resources on the program. A soft launch enhances the probability
of success of the program. A typical soft launch may be conducted with a goal of reaching
about 10 percent of the target population.
Smart Grids: An Approach to
Dynamic Pricing in India

13

Define the population and align the sample to maximize external validity: The population in
this context is the electricity consumers/end users (also known as the treatment group) that
would benefit most from the program design. The external validation ensures that the
results of the program can be extrapolated from the target to larger population.

Include a control group to maximize internal validity: A control group is similar to the
treatment group in electricity consumption, but without dynamic pricing option. Internal
validity will determine the impact of dynamic pricing on the electricity consumption of the
target population.

Design treatment groups to isolate impacts of research variables: The evaluation will be
more meaningful if the dynamic pricing program is designed in a way not to combine with
any other program.

Recruit sufficient sample to achieve analysis goals: Size of the treatment group or sample is
important for making the outcome of the program meaningful. However, this could be a key
cost driver. Appropriate balance has to be made by weighing benefit versus cost in
determining the sample size. Statistical calculations can be employed in determining the
optimum sample size.

Over-recruit to account for dropouts: In order to ensure the availability of adequate data for
the analysis, it is important to recruit more participants than required to mitigate participant
drop outs. Over recruitment and drop outs will vary depending on several factors such as:
o Ease of the program interface for the participant
o Length of the evaluation period Some participants may not want to be tied up for
too long
o Participants' particulars Renters/home owners etc.

Plan for an assessment period sufficient for evaluation of long-term impacts: Short-term
programs may not adequately reflect long-term impacts. Also, response may be initially
slower due to the learning curve.

Combine various data sources to improve understanding of impacts: Provision should be

made to collect additional data that may impact energy consumption (e.g., weather data,
power outages, etc. will be helpful for the analysis).

Designing Future Dynamic Pricing Pilots

The following observations drawn from the completed dynamic pricing pilots are useful in designing
future dynamic pricing program:

Customers respond better to dynamic pricing with planned communication and education.

Enabling technologies (e.g., smart meter, smart thermostat, IHD) play a key role in the
success of a pilot.

14 PACE-D Technical Assistance Program

Pilot design and roll-out must mirror a utility's full deployment approach as much
as possible.

Low-income customers with a flat load profile do benefit from dynamic pricing.

A variety of choice offerings (CPP, RTP, TOU, etc.) may help determine the most effective
option for the selected treatment group.

In some situations, RTP may not be popular with some customers, due to exposure to high
risk. Customers may accept higher risks if there a corresponding reward to go with it.

In order for the results to be demonstrative, both the treatment and control groups should
have a similar consumption pattern.

Summary of Results from Dynamic Pricing Pilots

The 2013 ISGAN Issue Brief (reference #5) discusses a TOU tariff simulation study and concludes
that the net impact on consumer bill on account of aforementioned TOU charges will be almost
negligible if the consumer does not make any change in consumption pattern.
As indicated in this report, results from international pilots exhibit varying impacts from the tariff
schemes and a degree of sensitivity to the way in which customers interact with the tariff (for
example, via in-home devices). Graph 5 displays the varying impacts experienced by the U.S. and
Canada utilities due to different tariff schemes and consumer interaction methods. Further
information on the U.S. and Canada time-based tariffs can be found in reference #6.

Graph 5: Varying Impact of Tariff Scheme

IHD and Prepayment
Impacts

IHD-Only Impacts
20%

IHD and Time-Varying

Rates Impact

16%
14%
12%
10%
8%
6%
4%
2%

Hydro One TOU 2

Country Energy

SRP

Woodstock Hydro

BC Hydro

Hydro One RTM

Hydro One TOU 1

0
Newfoundland
Power

Conservation Impact (%)

18%

Smart Grids: An Approach to

Dynamic Pricing in India

15

Table 1 includes data and observations from 17 dynamic pricing pilots. The data used in the table are
based on information provided in reference # 7.
Additional sources of information on international smart grid pilot projects and dynamic pricing are
found in references # 5 and 6.
The observations listed in Table 1 indicate that the consumption pattern will change if:

The ratio of peak/off peak rate is high it is important to note that the energy savings may
not translate into a proportional direct cost savings.

There is a price limit hedge (or announced maximum price) for the program participants.

Enabling technologies are employed so that:

o

The consumers can monitor their electricity usage from IHD.

Smart meters and smart thermostats are in place to automate data collection and
control of appliances - ONCOR, Inc.'s Smart Texas program applied a monthly surcharge
for program participants towards installation of the smart meters.

The program produces consistent results with minimum error Pepco, who conducted one
of the most successful dynamic pricing pilots in the U.S.(PowerCents), conducted a sample
test on each production run of smart meters before installation to ensure accuracy.

The end users have adequate training and knowledge of the program.

The program is limited to a homogeneous mix of participants.

Questions that the various programs tried to address in designing their dynamic pricing
program include:

Would people be willing to participate in the program?

How much money would consumers with central air-conditioning save, and how would this
compare to consumers without central air-conditioning?

How comfortable is the treatment group in handling technologies such as IHD, smart meter,
and smart thermostat?

16 PACE-D Technical Assistance Program

Smart Grids: An Approach to

Dynamic Pricing in India

17

Southern
Company/
Georgia
Power/Power
Rewards,
Georgia,

CenterPoint
Energy, /In
Home Display
Pilot, Houston,
Texas, 2010,
Program
budget N/A.

ISO-NE/DR
109
Reserve Pilot,
Commercial
New England,
customers.
2006-10,
Program budget
N/A.

500
Residential
customers
given smart
meters, IHD.

1,000
Residential
customers
already
having AMI.

280,000
Residential
with smart
meter.

Test Market

2008-2009,
Program
budget N/A.

Potomac
Electric power
Co. (PEPCO)/
Power Cents
DC,
Washington,
DC, 20082009,
20102011;
Program
Budget - N/A.

Company/
Program

Sl. #

RTP, VPP, TOU,

Flat rate.

TOU

CPP - Rewards
based difference
between actual and
projected usage
(required 30 hours
of admin time per
event). It was
difficult to predict
consumption
behavior of
residential
customers.

CPP- prices 7 times

the normal price for
60 peak hours/year;
CPR - rebates
earned for lower
consumption in
peak hours; HP hourly change in
price based on
wholesale prices.

Dynamic
Pricing Option

35% load
reduction of
enrolled
participants.

NA

Program reduced
peak energy
demand.

CPP - 30% -13%

(summer
/winter); CPR 13% - 5%
(summer/winter;
HP - 4% - 2%
(summer/winter).

Changes in
Peak Demand

NA

NA

35/ kWh
of energy
saved, with
a maximum
CPP period
of 50
hours/ year.

CPP - 2%;
CPR - 5%;
HP - 39%.

Dollar
savings

NA

NA

Overall
8% bill
savings
over 2
years.

Consumer
Savings

Table 1 - Dynamic Pricing Pilots

NA

Advertising,
direct
communication,
internet, media
relations, and
collateral
materials.

Customers
were notified
at least one
day before
CPP event
via phone/email.

Program
brochure, and
explanation
in bill.

Customer
Communication/
Education

Base line adjustments in

calculations negatively
impacted load reduction
performance. Performance
of load reduction assets
always less than the
demand response reserve
contract amount.

Periodic meter checking

and automating routing
transactions are
important for accuracy
and reduction in routine
service visits.

Customer education and

energy management
advice, and enabling
technology are important
for the acceptance and
satisfaction of end users
as well as program
reliability; Payout to
customers higher than the
highest RTP (88/kWh vs.
29/kWh) indicating flaw
in the incentive design.

Smart thermostats
were used in houses
with electric heating;
Low income
participant's demand
reduction was
smaller; Billing
system testing was
essential.

Other Issues/
Observations

18 PACE-D Technical Assistance Program

71
Residential
customers.

85
Residential
customers.

Anaheim
Critical Peak
Pricing
Experiment,
Anaheim,
California,
2005, Program
budget - N/A.

Idaho Power
Residential
Pilot Program,
Idaho, 20052006, Program
budget - N/A.

2009-11, USD
Budget - 572
million.

3.4 million
customers
from all
sectors, with
smart meter,
IHD.

Test Market

1.4 million
customers
from all
sectors with
smart meter,
smart
thermostat.

2009-12, USD
Budget - 7.3
million.

ONCOR,
Inc./Smart
Texas
Program,
Dallas, Texas,

Company/
Program

SDG&E Smart
Meters, San
Diego,
California,

Sl. #

Time of Day
(TOD)

CPP (Rebates
are calculated
on usage
reductions
noon - 6 p.m)

CPP

TOU

Dynamic
Pricing Option

50% reduction.

12% (reduction
is larger on
higher
temperature
CPP

NA

NA

Changes in
Peak Demand

NA

NA

NA

Dollar
savings

NA

NA

NA

Consumer
Savings

High-price day
notification via phone
or email when the
price of electricity
was over USD 0.10
per kWh.

TOD rates with very low

peak/off peak rate ratio will
have no effect on shifting usage
A higher peak/off-peak rates
ratio is needed to induce
customers to shift usage from
peak to off-peak periods.

APU paid an average of 7

times more/ KWh for the
load reductions achieved
during CPP days than the
rebate amount.

Utility's focus has been on

security/integrity first;
Variable-pricing programs
employing smart meters have
not been utilized, except for
business customers; used 9
customer/smart meter
metrics.

Web site devoted to

informing consumers
about saving energy;
SDG&E has met with
more than 25
stakeholder groups in
academia, business,
customer advocacy,
and government since
late 2010 in order to
understand their smart
grid preferences.

NA

Monthly surcharge for smart

meter installation (USD
2.19/month); Benefit not fully
realized since TOU pricing is
not wide spread; DLR pilot
using smart meters will
measure congestion relief and
extrapolate potential economic
effects in the overall Oncor
service area.

Other Issues/
Observations

Web site devoted to

informing consumers
about saving energy
(includes links to
compliance reports);
Advertisements;
Door hangers;
Hosted 8 Mobile
Experience Center
events.

Customer
Communication/
Education

Smart Grids: An Approach to

Dynamic Pricing in India

19

Ontario
Energy Board
Smart Price
Pilot, Canada,
2006-2007,
Program
budget - N/A.

12
373
Residential
customers.

379
Residential
customers
(my Power
Sense); 319
Residential
customers
(my Power
Connection).

112
Residential
and
Industrial,
Commercial.

Bonneville
Power
Administration
(BPA)/Olympic
Peninsula
Project,
Washington,
2006-07,
Program
budget - N/A.

PSEG/my
Power Sense
and myPower
Connection,
New Jersey,
2006-2007,
Program
budget - N/A.

1,500
Residential
customers.

Test Market

Community
Energy
Cooperative's
Energy-Smart
Pricing Plan,
Illinois, 20032005, Program
budget - N/A.

Company/
Program

11

10

Sl. #

Regulated
Price Plan
(RPP) TOU,
RPP TOU with
CPP (TOU
CPP), RPP
TOU with CPR
(TOU CPR).

TOU/CPP
(includes off
peak, peak,
base rates summer
months).

Fixed;
TOU/CPP; RTP.

RTP

Dynamic
Pricing Option

Load shift 6% (TOU

only); 4.7%
(TOU CPP);
7.4% (TOU
CPR) for full
pilot duration.

my Power
Sense - 17% ;
my Power
Connection 47%; for
combined
impact of TOU
and CPP.

29.7% for 500kW (fall), 19%

for 750-kW
(winter); No
measurement
provided for
1500-kW
(summer)
period, which
did not
need peak
management.

3-4% of
summer
electricity
usage.

Changes in
Peak Demand

NA

NA

Mean
savings:
TOU/CPP
-30%;
RTP
27%;
Fixed
2%.

Dollar
savings

NA

NA

10% from
previous
year's bill.

Consumer
Savings

Monitoring device
allows customers to
view an estimate of
the CO2 emissions
produced as a result
of their electricity
consumption.

Customers were
educated on TOU
tariff and notified on
CPP day ahead.

NA

Energy usage
education provided to
participants.

Customer
Communication/
Education

Large average peak reductions

under all three of the rate
designs, with the TOU CPP
impact being the highest,
followed by the TOU CPR, and
then the pure TOU; Results were
achieved without any
accompanying incentives or
price schemes.

my Power Connection
customers had enabling
technology.

Much of the decline in peak

demand can be attributed to
automated responses from the
smart appliances installed.
Default settings, which would
shut off appliance usage at
critical high price times, were
rarely overridden; Users were
given an average of USD 150
(which could be more or less
depending on energy savings)
for participation.

Day-ahead (DA) announcement

of the hourly electricity prices
for the next day and a price
limit hedge of USD 0.50 per
kWh for participants, meaning
that the maximum hourly price
set at USD 0.50 per kWh for
the participants in the program.

Other Issues/
Observations

20 PACE-D Technical Assistance Program

89
Residential
customers.

Commercial
& Industrial
(C&I).

100,000
(Touraine)
and 200,000
(Lyons)
Residential
customers.
Residential.

AmerenUE
Critical Peak
Pricing Pilot,
Missouri,
2004-2005,
Program
budget - N/A.

Impact
Evaluation of the
California
Statewide
Pricing Pilot
(SPP), California,
2003-2004,
Program
budget - N/A.

ERDF/Pilot Linky,
Touraine and
Lyon, France,
2009-2011,
budget - ~4.3
billion Euro.

Load
Management
Pilot Project,
ESKOM, South
Africa, 20092012, Program
budget - N/A.

14

15

16

17

50,000
(Residential/
customers
50/50).

Test Market

Energy
Australia/TOU
Tariff Program,
New South
Wales,
Australia,
2005, Program
budget - N/A.

Company/
Program

13

Sl. #

TOU

TOU, CPP, CPR

TOU, CPP-F
(Fixed period),
CPP-V (variable
length of peak
w/ enabling
tech).

TOU-CPP, TOUCPP-Tech
(enabling
technology).

TOU/DPP

Dynamic
Pricing Option

11.2 kW (pre
winter) &
2.63.0 kW
(Winter) DR
per house for
1.5kW limit.

CPP 16%,
CPR 12%,
TOU 5%.

CPP-F:
13.1%; CPPV:~20%.

TOU-CPP:
12%; TOUCPP-Tech:
35%.

~20%

Changes in
Peak Demand

NA

NA

NA

NA

Dollar
savings

NA

NA

NA

NA

~20%

Consumer
Savings
Other Issues/
Observations

Provides each
customer an
electricity demand
display instrument
(EDDI) that shows
the real-time
demand of various
electrical appliances
at work.

In-Home Displays
(IHD), websites and
informative billing.

NA

NA

Enabling technology limits the

amount of electricity supplied
to households during highconstraint periods; When a
household does not comply
with the load limit that was set
during the pilot, that individual
house is subjected to load
shed.

55% Reduction Non technical

losses; Installed low loss
transformers; installed Linky
meters for customers to be
able to monitor energy
consumption.

Comparing the CPP-F and the

CPP-V results suggests that
usage impacts are significantly
larger with an enabling
technology than without it.

Enabling technology was a key

component for the success of
the program for TOU-CPP-Tech
group.

Price signals through Required in house display (IHD)

Short Message
and on line access to data;
Service (SMS),
Large estimation error from
telephone, email, or heterogeneity of business
the display unit.
customers; Day ahead
notification was effective,
impacts of load shift was
seasonal.

Customer
Communication/
Education

Conclusions
Several conclusions can be drawn from the review of the time-based dynamic pricing methods
that have been implemented internationally as also the available data on pilot project dynamic
pricing experiments.
There is a scarcity of smart grid dynamic pricing pilot projects outside of North America that go
beyond TOU tariffs, compared to the widespread efforts at implementing AMI in smart grid
projects internationally.
Successful introduction of smart meters and time-based dynamic electricity pricing requires a
well-planned social marketing campaign to help raise awareness and give customers the
information and support they need to become more energy efficient. It is also necessary to raise
customer awareness on what changes they must make to realize the potential benefits.
At the current stage of power system commercial development in developing countries for which
data is available in Asia and Africa:
1.

These systems are ripe for AMI to support 100 percent metering of loads.

2. With progress in metering of loads, AMI can be supplemented by implementing relatively

straight-forward and transparent TOU pricing.
Dynamic pricing can be an effective complement to wholesale market access and demand
response initiatives that are already underway in India. More sophisticated dynamic pricing
experiments will require the same levels, if not more, of infrastructure and customer education.
The planned smart grid pilot projects in India can provide the required level of infrastructure. The
benefits of such pricing programs may be larger than TOU pricing. In the long run, it may be more
efficient to move directly to such a program rather than first moving to TOU pricing and then to a
more sophisticated dynamic pricing program later on.
For any such experiment, a cost-benefit analysis for the experiment, and full-scale rollout should
be performed and updated as better information becomes available. If the cost-benefit analysis
does not show an overall benefit for full-scale implementation, it may not make sense to spend
the effort on the smaller scale experiment.

Smart Grids: An Approach to

Dynamic Pricing in India

21

Acronyms
AMI

advanced metering infrastructure

AT&C

aggregate technical and commercial loss

BU

billion unit

CPP

critical peak pricing

CPR

critical peak rebate

DP

dynamic pricing

DPP

dynamic peak pricing

DR

demand reduction

EPAct

Energy Policy Act

HP

hourly pricing

IHD

in-home display

kWh

kilowatt hour

MOP

Ministry of Power

NE

New England

PHEV

plug-in hybrid electric vehicle

PSEG

Public Service Electric and Gas Company

PTR

peak time rebate

RTP

real time pricing

SDG&E

San Diego Gas and Electric Company

TOD

time of day

TOU

time of use

VPP

variable peak pricing

22 PACE-D Technical Assistance Program

References
No.

Reference

Potential Impacts of Plug-in Hybrid Electric Vehicles on Regional Power Generation Stanton W. Hadley, Alexandra Tsvetkova, OAK RIDGE NATIONAL LABORATORY, January
2008 (http://web.ornl.gov/info/ornlreview/v41_1_08/regional_phev_analysis.pdf)

Assessment of Demand Response and Advanced Metering Staff Report, Docket Number
AD-06-2-000, August 2006/2008 Revised (http://www.ferc.gov/legal/staffreports/demand-response.pdf)

Advanced Metering Initiatives and Residential Feedback Programs: A Meta-Review for

Household Electricity-Saving Opportunities, Karen Ehrhardt-Martinez, Kat A. Donnelly,
& John A. Skip Laitner, Report Number E105, June 2010 (http://sedc-coalition.eu/wpcontent/uploads/2011/06/ACEEE-08-06-01-Energy-Information-Feedback-Studies1.pdf)

Effective Research Methodology for Smart grid Enabled Consumer Behavior Impact
Assessment, 2012 ACEEE Summer Study on Energy Efficiency in Buildings - Jordan
Michel, Geavista Group, and Pam Osterloh, AEP Texas
(http://www.aceee.org/files/proceedings/2012/data/papers/0193-000041.pdf)

Dynamic Tariff Structures for Demand Side Management and Demand Response - An
Approach Paper from India, 2013 ISGAN - Sanjeev Kumar, Director, MoP, GoI, India
N.S. Sodha, Executive Director, POWERGRID, India, Kumud Wadhwa, Dy. General
Manager, POWERGRID, India

The Impact of Informational Feedback on Energy Consumption Survey of the

Experimental Evidence, Sanem Sergici, Ph.D., Ahmad Faruqui, Ph.D.,
http://www.brattle.com/_documents/uploadlibrary/upload772.pdf

Comparison of Results Across Dynamic Pricing and Time-Based Rate Pilot Programs:
Quantifying the Benefits of Dynamic Pricing In the Mass Market, Edison Electric
Institute, January 2008

Smart Grid Legislative and Regulatory Policies and Case Studies U.S. Energy
Information Administration (EIA) December 2011

Dynamic Pricing: What Have We Learned? - Sanem Sergici, Ph.D., Ahmad Faruqui, Ph.D.,
The Brattle Group, 5/19/2011

10

Appendix E: Comparison of Results Across Dynamic Pricing and Time-Based

Rate Pilot Programs, Edison Electric Institute (EEI), Sanem Sergici, The Brattle Group,
January 2008

11

Dynamic Pricing and Its Discontents, Regulation, Ahmad Faruqui, Ph.D., The Brattle
Group, Fall 2011

Smart Grids: An Approach to

Dynamic Pricing in India

23

No.

Reference

12

U.S. Department of Energy's Smart grid Investment Grant Program: Dynamic Pricing and
Consumer Behavior Studies Webinar, April 20, 2010

13

Dynamic Pricing and Low-Income Customers, Lisa Wood and Ahmad Faruqui, Public
Utilities Fortnightly, November 2010

14

Architecting the Future of Dynamic Pricing, Ahmad Faruqui, Ph.D., The Brattle Group,
July 24, 2012

15

Smart grid Legislative and Regulatory Policies and Case Studies, U.S. Energy Information
Administration (EIA), December 2011

16

PowerCentsDC Program, Final Report, September 2010

(http://www.powercentsdc.org/ESC%2010-09-08%20PCDC%20Final%20Report%20%20FINAL.pdf)

24 PACE-D Technical Assistance Program

NOTES

NOTES

U.S. Agency for International Development

1300 Pennsylvania Avenue, NW
Washington, DC 20523
Tel: (202) 712-0000
Fax: (202) 216-3524
www.usaid.gov