You are on page 1of 9

Energy Policy 136 (2020) 111050

Contents lists available at ScienceDirect

Energy Policy
journal homepage: http://www.elsevier.com/locate/enpol

Estimation of the usage fee for peer-to-peer electricity trading platform: The
case of South Korea
Juyong Lee, Youngsang Cho *
Department of Industrial Engineering, College of Engineering, Yonsei University, 50 Yonsei-Ro, Seodaemun-gu, Seoul, 03722, South Korea

A R T I C L E I N F O A B S T R A C T

Keywords: The peer-to-peer (P2P) electricity trading platform (ETP) is an electricity-sharing system by which individuals
P2P electricity trading platform can sell surplus electricity or purchase electricity to meet a deficit in a regional power network. South Korea has
Energy prosumer allowed energy prosumers to engage in electricity trading from 2016 and is seeking ways to vitalize the elec­
Willingness-to-pay
tricity trading market. This study aims to estimate the willingness to pay (WTP) of respondents for the monthly
Contingent valuation method
Spike model
usage of P2P ETP by using the contingent valuation method. Of the 1000 Korean respondents, 639 were willing
to pay a fee for using the platform. The estimated WTP is KRW 5958.86 (USD 5.50) per month. Moreover, the
estimated WTP is higher in respondents who have more background knowledge of the P2P ETP and who pay
higher electricity fees. The estimated WTP is lower than the usage fees of existing ETPs in other countries, which
is primarily due to Korea’s low electricity prices. Furthermore, the estimation results show that enhancing
consumer awareness of P2P trading and introducing market strategies based on electricity usage should be
considered by electricity trading companies and regulators to enhance P2P ETP usage.

1. Introduction Today, which argues that consumers can become producers through the
development of power generation technology (Nevitt and McLuhan,
Reducing greenhouse gas emissions has emerged as one of the most 1972). Energy prosumers participate in the electricity market through
prominent global environmental issues. Therefore, the development and generation aggregators or platforms and sell surplus electricity through
introduction of renewable energy are actively pursued globally. In an electricity trading platform. Similarly, neighbors who have higher
response to the growing interest in renewable energy, new business electricity bills or a higher electricity demand can buy the electricity
models in the electricity industry have been launched worldwide, with that the prosumers sell on the platform. This intermediary platform is
the gradual expansion of distributed generation (also known as micro­ called a peer-to-peer (P2P) electricity trading platform (ETP). P2P
grid) based on the energy storage system (ESS) and photovoltaic power electricity trading refers to the end consumer playing the role of an
generation. These small-scale energy systems have enabled energy energy prosumer who produces electricity and trades their surplus
consumers to become energy producers or sellers. With energy con­ electricity with other consumers in the power grid.
sumers starting to generate renewable energy at their homes, new forms There are established P2P ETPs—such as Piclo in the United
of energy products and energy markets—such as solar panels, batteries, Kingdom, Sonnencommunity in Germany, and Vandebron in the Neth­
smart home applications, and energy trading services—are emerging erlands—that are commercialized for mediating the electricity trading
(Parag and Sovacool, 2016). between prosumers and consumers. The United States launched a pilot
Energy prosumers—who have emerged due to the improvement in ETP commercialization business through the Brooklyn Microgrid proj­
the economic feasibility of distributed resources and changes in con­ ect. South Korea has also been promoting the electricity trading business
sumer perceptions due to the technological advances—are causing a between neighbors in a similar microgrid since March 2016 and is
paradigm shift in the electricity industry (Shandurkova et al., 2012; seeking ways to energize the electricity trading market. Korea is still in
Marques and Nixon, 2013). The word “prosumer” is a compound word the process of reviewing the feasibility of commercializing P2P ETPs and
derived from “producer” and “consumer.” The concept of energy pro­ the existing studies on the subject focus mainly on technical issues, such
sumers was first introduced by McLuhan and Nevitt in their book Take as the application of blockchain or big data technology in the platform

* Corresponding author.
E-mail addresses: dlwndyd@yonsei.ac.kr (J. Lee), y.cho@yonsei.ac.kr (Y. Cho).

https://doi.org/10.1016/j.enpol.2019.111050
Received 4 June 2019; Received in revised form 19 September 2019; Accepted 12 October 2019
Available online 22 October 2019
0301-4215/© 2019 Elsevier Ltd. All rights reserved.
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

(Yoo et al., 2017; Kang et al., 2018; Lee et al., 2018). However, to the The Brooklyn Microgrid is a photovoltaic energy trading platform
best of our knowledge, there have been no studies that investigate the that was launched as a demonstration project in Brooklyn, New York, in
financial aspects of P2P electricity trading such as usage fees of the the United States in April 2017 (Mengelkamp et al., 2018). Fifty resi­
trading platform, even though a large investment is essential for dents of Brooklyn took part in the project, installing solar (or photo­
installing the platform. Based on this, the main purpose of this study is voltaic) panels and selling surplus energy. Prosumers can choose
two-fold; (1) to estimate the usage fee1 for a P2P ETP by determining the whether to trade the energy generated by the photovoltaic panels on a
willingness-to-pay (WTP) of potential users and (2) to derive future peer-to-peer basis, store the surplus electricity in an on/offline storage
government policies and company strategies for activating and growing device, or use it in the home. The Brooklyn Microgrid gains profit from
P2P ETP businesses. the sales of energy storage units and commission from electricity
The remainder of this study is organized as follows: Section 2 de­ trading. Table 1 summarizes the characteristics of the existing P2P ETPs
scribes the concept and background of P2P ETPs and reviews previous discussed in this section.
studies related to the subject. Section 3 explains the data and method­ Unlike the above countries with established and commercialized P2P
ology used in this study, while Section 4 presents the estimation results ETPs, Korea’s P2P ETP market is still in its early stages. In 2016, Korea
and the discussion. Section 5 provides the main findings, implications, launched a demonstration project of P2P electricity trading with phys­
and limitations of the study. ical infrastructure such as solar panels and ESS in two small towns to
foster energy prosumers’ business and activate the P2P electricity trade
2. Background through new trading platforms. However, this project is still in the early
stage of examining the feasibility of P2P ETP, and has therefore not been
2.1. Peer-to-peer electricity trading platform (P2P ETP) promoted actively in Korea.
Currently, Korean energy prosumers who have registered electricity
In P2P electricity trading, energy prosumers’ trading partners are trading businesses are trading mostly in electricity generated by
neighbors who want to buy electricity from someone other than the photovoltaic panels. The transactions are with the Korea Electric Power
electricity sales company. Prosumers, therefore, get sales revenue by Corporation (KEPCO). There are several other possible energy sources
selling surplus electricity to their neighbors, and this transaction is for P2P electricity trading, such as wind power and biomass; however,
performed via a trading platform. P2P ETPs are being developed glob­ photovoltaic-based electricity trading is the main source that is currently
ally based on technological improvements such as the application of being promoted in Korea. However, photovoltaic power generation has
blockchain and big data. In several countries—such as the United characteristics of intermittent availability of time for generation and
Kingdom, the Netherlands, Germany, and the United States—pilot high generation cost, compared with traditional generation methods.
businesses are already in operation or being commercialized. Vandebron This is one of the reasons the P2P power trading market is not actively
is a Dutch web-based electricity trading platform launched in April 2014 operated in Korea. The prosumers and consumers are less likely to
(Vandebron, 2019). Prosumers introduce their power generation modes, participate in P2P electricity trading as the average electricity charge is
residential areas, and trading conditions on the platform and once the lower than that of photovoltaic generation cost.2 To achieve grid parity
consumer has selected the prosumer they want to trade with, the by lowering the generation cost of renewable energy to the level of the
transaction takes place within 2 min. In December 2018, Vandebron had average electricity charge, the Korean government is planning to invest
approximately 100,000 members and prosumers on the platform selling KRW 140 trillion (USD 129 billion) in renewable energy R&D and
electricity generated by wind, biomass, and solar energy. Vandebron has related facilities by 2030 and to invest KRW 1.2 trillion (USD 1.11 bil­
a profit structure based on platform usage fees, charging a membership lion)—approximately 62.1% of the Electrical Industry Foundation
fee of USD 12 per month for both consumers and prosumers.
In the United Kingdom, Piclo was launched in October 2015 under Table 1
sponsorship of the United Kingdom Department of Energy and Climate Existing cases of P2P ETPs.
Change (DECC) and the energy fund of Nominet Trust (Piclo, 2019).
P2P ETP (Country, Characteristics Monthly usage fee
Piclo matches energy consumers and prosumers every 30 min using Launch year)
electric meter data, power generation costs, and consumer preference
Vandebron - Consumers select prosumers - USD 12
information. Both consumers and prosumers can use Piclo online ser­ (Netherlands, 2014) directly and deal at a price
vices through computers and smartphones, and electricity trading takes set by the prosumers.
place when the consumers’ desired prices and the prosumers’ trans­ - Transaction takes place in
action conditions are both met. Piclo is managed and supported by the 2 min.
Piclo (UK, 2015) - Web and App based platform - Free
government and venture capital investment and there is no fee for using
- Platform matches energy - government and
the Piclo platform. consumers and prosumers venture capital
Sonnencommunity is Germany’s renewable energy trading platform every 30 min. investment
and was established in 2016 (KEMRI, 2016; Sonnenbatterie, 2019). Sonnencommunity - Fixed-price system (23 - EUR 19.99
Sonnencommunity combines energy trading and battery storage tech­ (Germany, 2016) cents/kWh)
- Stable power supply with
nology, providing surplus power and services that can be stored in battery storage technology
batteries and for emergencies. This means that prosumers can save their Brooklyn Microgrid - Able to sell, store, and use - Free
surplus electricity to the battery and use it when needed or sell it again at (USA, 2017) photovoltaic energy that - commissions of
a future date. Unlike other platforms, where the prosumer presents prosumers generate electricity trading
- Provide solar panels and on/
prices directly, Sonnencommunity adopts a fixed-price system and
offline energy storage device
consumers pay 23 cents per kWh to the prosumers. Platform users pay a
fee of EUR 19.99 per month and additional costs—such as transaction (Source: Vandebron, 2019; Piclo, 2019; Sonnencommunity, 2019; Mengelkamp
fees—are not incurred. et al., 2018)

1
In this study, “usage fee” is the amount payable for the right to access the
2
grid and trade electricity on the platform. The payment for using P2P ETP in­ In Korea, the average electricity charge of households in 2017 was KRW
cludes transaction fees and platform maintenance costs and does not include 110/kwh (EPSIS, 2019). Conversely, the average cost of photovoltaic genera­
power generation facilities and ESS construction costs. tion in 2017 was KRW 182/kwh (Song et al., 2018).

2
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Fund—in renewable energy businesses in 20193 (Park and Koo, 2018). initial bid amount that has a significant influence on the WTP estimation
is difficult (Ready and Hu, 1995). Despite these drawbacks, most recent
2.2. Literature review concerning P2P ETPs studies utilized the DC format because of its evident merits: high
response rates, low biases, and a low likelihood of yielding unreliable
Technical feasibility studies concerning P2P ETPs have been actively WTP measures (Hanemann, 1984; Oerlemans et al., 2016).
conducted. Werth et al. (2015) conducted a conceptual study applying The DC format can be categorized into a single-bounded dichoto­
nanogrid—a technologically smaller and simpler version of micro­ mous choice (SBDC) proposed by Bishop and Heberlein (1979) and
grid—in the distributed energy network and the application was proven double-bounded dichotomous choice (DBDC) proposed by Hanemann
to be valid on a conceptual level. K. Zhang et al. (2018) proposed a P2P (1985). The DBDC format asks a respondent a yes/no question twice,
ETP model in a microgrid by applying the game theory and simulated whereas the SBDC format asks just one yes/no question. When the
electricity trading with their model. The simulation results showed that respondent answers “yes” to an initial bid, a follow-up question with a
P2P electricity trading improves the balance between power generation second bid—higher than the initial bid—is presented. If the respondent
and consumption and the more peers, the easier it is to reach that responds with “no” to the initial bid, the second bid—lower than the
equilibrium. Di Silvestre et al. (2018) proposed an experimental sce­ initial bid—is presented. The DBDC format is preferred in many studies,
nario of a blockchain-based microgrid ETP. They discovered that the due to the SBDC format having only one question and therefore being
superposition of electricity trading in the microgrid causes power losses statistically inefficient compared to the DBDC format. However, the
in every branch and expressed the need for technological alternatives to DBDC format can potentially have compliance problems or rejection
address this issue. Alam et al. (2019) proposed an optimal trading model problems4 (Cameron and Quiggin, 1994). To overcome the disadvan­
integrating a demand-side management (DSM) system with the P2P tages of both the SBDC and DBDC formats, Cooper et al. (2002) proposed
ETP. The energy cost is increased under unrestrained trading conditions. the one and one-half bounded dichotomous choice (OOHBDC) format, in
However, when the proposed demand size adjusting algorithm is which ranges are presented to respondents for both the lower bid and
applied, the total energy cost of the peers can be minimized. the upper bid. In the OOHBDC format, one of the two bids is randomly
There are few studies on the economic and social aspects of P2P presented to the respondents as an initial bid. If the respondent answers
ETP—such as charging schemes or consumer preferences—compared to “yes” to the upper bid or “no” to the lower bid, the survey ends. Whereas,
the number of studies on technical aspects. There are some studies, if a respondent presents answer of “yes” (“no”) to the lower (upper) bid,
however, on consumer preference for renewable energy-based policy or a follow-up question with the upper (lower) bid is presented. If the
market expansion. Bertsch et al. (2016) analyzed public acceptance of respondent answers “no” (“yes”) to the follow-up question, the survey
the renewable energy grid expansion policy in Germany using survey ends. This study applied the OOHBDC format to estimate WTP, as this
questionnaires. They found that public acceptance improves as the format has the advantages of reducing both inefficiency and bias
amount of renewable energy sources and the level of education on en­ (Cooper et al., 2002).
ergy increases. There are also studies concerning the activation of the
microgrid or P2P electricity market, but they were all conducted
through literature and case review (Franke et al., 2005; Sousa et al., 3.2. WTP estimation model
2019). To our knowledge, there are no studies that carried out empirical
analyses on the subject. To successfully introduce and operate a P2P In this study, we use the utility difference model proposed by
ETP, a WTP estimation concerning the platform usage from the Hanemann (1984). The utility of the respondent with income (m) and
perspective of consumer preference is needed. characteristics (S) for the state of the non-market goods (J) is expressed
by the indirect utility function (u). Here, J ¼ 0 means that P2P ETP is not
3. Methodology introduced and J ¼ 1 means that P2P ETP is introduced and available in
the contingent market. The indirect utility function can then be
3.1. The one and one-half bounded dichotomous choice contingent expressed by the observable deterministic part (VðJ; m; SÞ) and the un­
valuation method observable stochastic part (εj ) as follows:

uðJ; m; SÞ ¼ VðJ; m; SÞ þ εj (1)


This study uses a contingent valuation method (CVM) to estimate the
monthly usage fee for a P2P ETP. CVM is a survey-based method and one
where εj is independent and identically distributed with a zero mean. If
of the most widely-used approaches for measuring the value of non-
the respondent replies “yes” to the question “Are you willing to pay A for
market goods. In the electricity industry, CVM has in the past been
using P2P ETP?” they maximize the utility by willingly paying amount
utilized to estimate the WTP of consumers concerning the acceptance of
A. The probability that the respondent answers “yes” can be expressed
green energy sources, measuring the social value of nuclear energy, and
as:
avoiding blackouts (Nomura and Akai, 2004; Stigka et al., 2014; Kim
et al., 2015). Prf}yes}g ¼ PrfΔVðAÞ � ηg ¼ Fη ½ΔVðAÞ� (2)
In a CVM survey, a specially designed questionnaire sets up a hy­
pothetical situation representing changes in non-market goods and re­ where η ¼ ε0 ε1 and Fη ð ⋅ Þ represents the cumulative distribution
spondents answer how much they are willing to pay for the market function (CDF) of η. For a respondent whose WTP (denoted as C) exceeds
change. There are several survey formats suitable for eliciting the WTP or equals the bid presented, the probability of answering “yes” can be
of respondents in CVM, including open-ended (OE), payment card (PC), expressed as:
and dichotomous choice (DC) formats (Boyle et al., 1996; Welsh and Prf}yes}g ¼ PrfC � Ag � 1 GC ðAÞ (3)
Poe, 1998). The DC format is an approach that gives respondents a “yes”
or “no” choice for the amounts of bids presented by an interviewer and where Gc ðAÞ is defined as the CDF of C. Therefore, we can derive the
the WTP is estimated through econometric analyses such as logit or following relationship from equations (2) and (3):
probit model. The DC format is costly since a large number of obser­
vations must be obtained to ensure reliability. Moreover, setting an
4
The compliance problem is when respondents answer “yes” to satisfy in­
terviewers’ expectations although they actually wanted to answer “no.” The
3
The yearly average exchange rate in February 2018 was 1083.29 KRW/USD rejection problem is when respondents answer “no’ unconditionally (or without
(IMF, 2019). thinking) because the repeated questions are troublesome to them.

3
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Fη ½ΔVðAÞ� � 1 GC ðAÞ (4) benefits of using the platform: (1) additional revenue from sales of
surplus electricity; (2) a reduction in energy wastage by purchasing only
When the upper and lower bids are presented to respondents, there
lacking electricity; (3) getting information concerning power genera­
are six possible outcomes: “yes,” “yes-yes,” “yes-no,” “no,” “no-yes,” and
tion, transactions, and usage from the platform provider. We then
“no-no.” For convenience in the formula, we use the binary value indi­
explained to the respondents that financial investments such as the
cator variables (I) for each outcome as follows:
installation of an ESS and power generation facilities are necessary for
8 9
> Y > using P2P ETP and these investments can be satisfied by electricity
< I i ¼ 1ðrespondent i answers "yes"Þ
> >
= trading through the platform. In the survey, the payment vehicle for
NY
Iðupper bid as an initial bidÞ ¼ I i ¼ 1ðrespondent i answers "no yes"Þ
>
> > using the P2P ETP was the monthly platform usage fee per household.
: I NN ¼ 1ðrespondent i answers "no no"Þ ​ >
;
i Generally, the payment vehicle in CVM studies on public or environ­
(5) mental goods is an additional payment of taxes, such as income tax.
However, the P2P ETP is a platform for electricity trading, which is
and different from general public or environmental goods. Therefore, using
8 9 income tax as a payment vehicle is not appropriate for this study. This is
> > because the direct consumer benefits from the P2P ETP are limited to
YY
< I i ¼ 1ðrespondent i answers "yes yes"Þ >
> =
YN
Iðlower bid as an initial bidÞ ¼ I i ¼ 1ðrespondent i answers "yes no"Þ platform users and do not apply to the general public on a national level.
> >
>
: I N ¼ 1ðrespondent i answers "no"Þ
i
>
; We therefore set the usage fee for the P2P ETP as a payment vehicle.
Several existing CVM studies on the electricity industry have used usage
(6)
fees (bills) as payment vehicles in a similar way (Yoo and Kwak, 2009;
where 1 (…) is an indicator function with a value of 1 if the argument is Kim et al., 2015). Moreover, as the direct consumer benefits are not
true, and 0 otherwise. derived when the transaction is stopped, we assumed that the payments
When the respondent answers “no” to a lower bid or “no-no” to an are made monthly for the period of the platform’s use. Furthermore, we
upper bid, the WTP of the respondent may be less than the lower bid or explained to the respondents that their incomes are limited and they
zero. To determine whether the respondent’s WTP is zero or at least not should be used for a variety of purposes.
zero, we adopt the spike model proposed by Kristro €m (1997). For re­ Prior to the actual survey, we conducted an online pilot survey with
spondents who answer “no” to a lower bid or “no-no” to an upper bid, an 500 respondents to determine the initial bid sets for the actual survey.
additional question was presented: “Are you willing to pay just KRW 1 or The upper and lower 5% outliers from the pilot test were excluded to
more for using the P2P ETP?” Based on this, the value indicator variables reduce bias and we determined five sets of values for the initial bids –
can be added as follows: KRW (1000; 5000), (5000; 10,000), (10,000; 15,000), (15,000; 20,000),
( ) (20,000; 25,000). The actual survey was conducted by a specialized
I NNY’
i ¼ 1ðrespondentianswers "no no yes"Þ market survey company, Gallop Korea, by utilizing one-on-one face-to-
Iðupper bid asaninitial bidÞ¼ NNN’ face interviews in March 2018. For the actual survey, 1000 respondents
Ii ¼ 1ðrespondent ianswers "no no no"Þ
were selected by applying stratified sampling to reflect the population of
(7)
South Korea.5 The demographic characteristics of the survey re­
and spondents are described in Table 2.
! ( ) The ten bid sets were allocated to 1000 respondents at a ratio similar
IiNY’ ¼ 1 respondent i answers }no yes}Þ to each other. Table 3 presents the distribution of the responses corre­
I lower bid as an initial bid ¼
IiNN’ ¼ 1 respondent i answers }no no}Þ sponding to each bid. When a lower bid was given as an initial bid, 177
out of 495 respondents (35.8%) answered “no-no.” When the upper bid
(8)
was given as the initial bid, 184 out of 505 (36.4%) answered “no-no-
Possible answers for the additional question are denoted as Y’ and no.” The respondents who answered “no-no” or “no-no-no” represent a
N’. The WTP is zero when the respondent answers “no-no-no” or “no-no” zero WTP. This indicates that 361 out of all respondents (36.1%) have
and the WTP is positive but less than the lower bid when the respondent zero WTP for using a P2P ETP.
answers “no-no-yes” or “no-yes.” From the eight indicator variables, the The zero WTP respondents’ intention can be categorized in two
log likelihood function for the OOHBDC spike model is as follows: ways: (1) their WTP is actually zero; (2) they disagree with or are dis­
( � � �� � � � �� ) pleased with the survey topic and the presented bid. Concerning re­
XN
I Yi þ I YY ln 1 GC AUi þ I YN i þ Ii
NY
ln GC AUi GC ALi
ln L¼ i
� � � � � spondents who are dissatisfied with the questionnaire and whose
i¼1 þ I NNY’
i þ I NY’
i ln GC ALi GC ð0Þ þ I NNN’
i þ I NN’
i ln½GC ð0Þ� answers are zero, their actual WTP may not be zero. These answers are
(9) called protest WTP (Carson et al., 2003; Ramajo-Herna �ndez and del
Saz-Salazar, 2012). In the survey, we presented an additional question to
Here, assuming that the respondent’s WTP follows the logistic CDF,
the zero WTP respondents, asking why the WTP is zero to distinguish the
GC ðAÞ has the following functional form with parameter estimates a and
true zero responses from the protest WTP responses. Table 4 shows the
b:
responses concerning the reasons for their zero WTPs. We considered the
8
< ½1 þ expða bAÞ� 1 if A > 0 respondents who selected “Not enough information” or “Against the
GC ðAÞ ¼ ½1 þ expðaÞ� 1 if A ¼ 0 (10) electricity trading platform” as the protest WTP respondents. Of the 361
:
0 if A < 0 zero WTP respondents, 142 were the protest-WTP respondents who are
1 39.4% of the zero-WTP respondents. The protest-WTP respondents have
The spike is defined as ½1 þ expðaÞ� and the mean WTP is calculated
as ð1 =bÞln½1 þ expðaÞ�.

5
In this study, we used proportionate probability stratified sampling, which
3.3. Survey design and data descriptions
is the most widely-used method for national-level surveys. The proportions of
respondents (households) were allocated according to the national population
We designed the OOHBDC questionnaire to estimate the monthly composition ratio by “region.” The sample was then extracted according to
WTP for using P2P ETP. We started the questionnaire by describing the demographic composition ratios such as “age” and “gender.” By utilizing this
backgrounds and basic concepts of the energy prosumers and the stratified sampling process, we were able to obtain a representative sample of
concept of P2P ETP to the respondents. We also presented the possible Korea.

4
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Table 2 priori, the coefficient of the bid amount is negative. We can therefore
Characteristics of survey respondents. interpret that a higher bid amount is less likely to elicit a “yes” response.
Characteristic Group Number of Ratio The estimated WTP to use the P2P ETP is KRW 5958.86 (USD 5.50)
respondents (%) per month and statistically significant at the 1% level. Additionally, to
Gender Male 507 50.7 calculate the confidence intervals of the estimated WTP, we conduct the
Female 493 49.3 Monte Carlo simulation, a parametric bootstrap approach proposed by
Age 20–29 184 18.4 Krinsky and Robb (1986). First, we repeatedly generate 5000 individual
30–39 201 20.1 parameter values using the estimated coefficients and the
40–49 233 23.2
50–59 229 22.9
variance-covariance matrix and then we calculate 5000 mean WTPs.
60–69 153 15.3 After listing the calculated WTPs in order of magnitude, we obtain the
Regiona Metropolitan Area 715 71.5 95% and 99% confidence intervals by omitting 2.5% and 0.5% at both
Non-metropolitan 285 28.5 ends. The 95% and 99% confidence intervals are presented in the last
Area
two rows of Table 5.
Average monthly electricity Under 9 2 0.2
bill per household (KRW 10–29 100 10.0 To analyze the effects of the respondents’ socio-economic charac­
1000) 30–49 488 48.8 teristics on the WTP for using a P2P ETP, we estimated the model with
50–69 348 34.8 covariates. The estimation results are presented in Table 6. Among the
70–89 41 4.1 covariates, gender, age, level of education, number of family members,
90–109 14 1.4
Over 110 7 0.7
and income do not have statistically significant effects on WTP. In other
Average monthly income per Under 299 165 16.5 words, most of demographic characteristics have no statistically signif­
household (KRW 10,000) 300–399 190 19.0 icant relations with the WTP for using P2P ETP. Several studies have
400–499 232 23.2 analyzed that electricity demand has income elasticity (Fatai et al.,
500–699 277 27.7
2003; Zachariadis and Pashourtidou, 2007; Kim et al., 2015), but our
Over 700 136 13.6
Education level (graduation) Lower than high- 452 45.2 result show that WTP for using P2P ETP is not income elastic. On the
school other hands, region, electricity fee, and background knowledge about
Higher than 548 54.8 P2P ETPs are statistically significant at the 1% level. The estimated
University/college coefficients of the variables “fee” and “knowledge” are positive, which
Number of family members 1 76 7.6
indicates that respondents who pay higher electricity fees or have more
2 165 16.5
3 264 26.4 background knowledge concerning P2P ETP are willing to pay more for
4 437 43.7
More than 5 58 5.8
Table 4
Notes:
a
Reasons for zero-WTP.
In South Korea, cities with populations of more than one million are referred
to as “metropolitan cities” (Huh and Kim, 2003). We applied this criterion to Reasons mentioned for zero-WTP Number of respondents Ratio (%)
organize the survey data concerning the different regions. Cannot afford to pay the additional fee 33 9.1
Should be provided free of charge 96 26.6
Not enough information 106 29.4
been generally excluded from the analysis data set because their re­
Do not care 90 24.9
sponses do not represent their actual WTPs (Boyle and Bishop, 1988; Against the electricity trading platform 36 10.0
Chien et al., 2005; Meyerhoff and Liebe, 2010). However, when the Total of reasons mentioned 361 100
responses of the protest WTP respondents are excluded from the data set,
the overall WTP can be overestimated (Strazzera et al., 2003). We
therefore considered the WTP values of the protest WTP respondents as Table 5
actual zero and included their responses in the data set. Estimation results (without covariates).
Variables Coefficient estimates
4. Results and discussion
Constant 0.4687 (7.02)a
Bid 0.1602 ( 27.41)a
The estimation results of the OOHBDC spike model without cova­ Spike 0.3849 (24.34)a
riates are shown in Table 5. According to the result of the Wald test Log-likelihood 1298.484
performed, the null hypothesis—which states that values of all co­ Wald statistic (p-value) 592.51 (0.000)
efficients in the estimated equation are zero—is statistically rejected at WTP (KRW) Mean 5958.86 (24.34)a
95% Confidence interval 5483.40–6464.08
the 1% significance level, meaning that the estimated equation is sta­ 99% Confidence interval 5335.22–6639.39
tistically appropriate. All estimates are statistically significant at the 1%
level. The number of zero WTP responses is 361 out of the 1000 re­ Notes:
The unit of bid is KRW 1000 for convenience.
spondents and the estimated value of the spike is 0.3849, which virtually
Values in parenthesis except Wald statistic are t-statistics.
coincides with the actual proportion of zero WTP responses. We can a
Significant at the 1% level.
therefore say that the spike is estimated well. Additionally, as expected a

Table 3
Distribution of responses in the survey.
Bid amount (KRW) Lower bid is given as an initial bid Upper bid is given as an initial bid

Lower Upper “yes-yes” “yes-no” “no-yes” “no-no” Total “yes” “no-yes” “no-no-yes” “no-no-no” Total
1000 5000 26 32 17 24 99 36 22 13 29 100
5000 10,000 15 20 36 29 100 26 16 25 36 103
10,000 15,000 8 11 46 36 101 12 6 42 39 99
15,000 20,000 12 9 37 42 100 7 11 42 45 105
20,000 25,000 8 4 37 46 95 10 8 45 35 98
Total (%) 69 (13.9) 76 (15.4) 173 (34.9) 177 (35.8) 495 (100) 91 (18.0) 63 (12.5) 167 (33.1) 184 (36.4) 505 (100)

5
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Table 6 monthly WTP for consumers using the CVM. The estimated WTP is USD
Estimation results (with covariates). 5.50 per household per month, and the national-level WTP is USD 1.46
Variables Coefficient estimates billion per year. Based in the estimation results, we derive some impli­
cations concerning the introduction and vitalization of a P2P ETP system
Constant 0.5782 ( 0.93)
Bid 0.1663 ( 27.03)a in Korea. First, the estimated WTP of the P2P ETP in this study (USD
Gender 0.1580 (1.28) 5.50) is less than half of the monthly ETP usage fees in the Netherlands
Age 0.0058 ( 1.10) (USD 12) and Germany (EUR 19.99; USD 22.53). This result may be due
Region 0.4007 ( 2.94)a to low consumers’ incentives for electricity trading through the P2P ETP
Fee 0.1174 (2.73)a
Knowledge 0.4380 (5.83)a
due to uncertainties concerning the infrastructure construction costs and
Education 0.0117 (0.36) the risks related to implementing new electricity trading methods.
Family 0.0418 (0.60) However, the low incentives and high risks for P2P ETP trading are also
Income 0.0443 ( 0.94) attributed to the current electricity price in Korea. As the current elec­
Spike 0.3790 (23.71)a
tricity price does not appropriately reflect the various external and so­
Log-likelihood 1271.476
Wald statistic (p-value) 561.61 (0.000) cial costs, consumers are not motivated to adopt new technologies and
WTP (KRW) Mean 5835.97 (23.64)a systems despite their advantages and potential benefits. We therefore
95% Confidence interval 5362.21–6332.23 propose that the main underlying cause of the low WTP for P2P ETP is
99% Confidence interval 5204.48–6439.49 the low electricity price in Korea. Unlike the other countries, the elec­
Notes: tricity price in Korea is set not by the market, but by the government6
The unit of bid is KRW 1000 for convenience. (Park and Hong, 2014) and the Korean government has maintained
Values in parenthesis except Wald statistic are t-statistics. relatively low electricity price as a priority to create a stable electricity
a
Significant at the 1% level. supply and demand. Korea’s electricity supply is therefore based on
providing lower-cost energy sources such as coal and nuclear power
using the P2P ETP. On the other hand, the estimated coefficient of the preferentially.7 Due to the government’s decades-long policy of main­
variable “region” is negative, indicating that households not in metro­ taining low electricity prices, Korean consumers are accustomed to the
politan cities are willing to pay more for using P2P ETP than those who low price and thus have less interest in improving energy efficiency or
live in metropolitan cities. The average monthly WTP is estimated to be enhancing renewable and eco-friendly energy generation (Kim and Shin,
KRW 5835.97 (USD 5.39) per month and statistically significant at the 2016; Kim and Cho, 2017). Furthermore, the price of secondary energy,
1% level. To sum up, respondents living in non-metropolitan areas, electricity, has been lower than that of primary energy in Korea until
those paying more for electricity, and those that have more prior recently. This means that the electricity price system of Korea is “dis­
knowledge about ETPs are willing to pay a higher usage fee than those torted.” The lower WTP of Koreans (compared to other countries) can
who do not. therefore be understood based on the historical context and current
Expanding the household-level estimates to the entire population circumstances in Korea. To implement strategies and policies for acti­
could be helpful in estimating the economic benefit of P2P ETPs at a vating P2P ETP, normalization of the current electricity price scheme of
national level. Since this study’s survey respondents were selected by Korea should be premised preferentially. For the normalization of the
stratified sampling to reflect the entire Korean population, the sample distorted electricity price scheme, we propose that the Korean govern­
provides representative data for the total population. The estimated ment needs to comprehensively reflect the energy calories and social
WTP for using P2P ETP is KRW 5958.86 (USD 5.50) per month. costs of electricity—such as GHG emissions and particulate mat­
Therefore, the annual WTP is KRW 71,506.32 (USD 66.01) per house­ ters—and secure fairness in the tax regime across all energy sources.
hold. The number of households paying their own electricity fees in Second, the results show that respondents living in non-metropolitan
South Korea is 22.08 million as of January 2019 (KOSIS, 2019). cities are willing to pay more for using a P2P ETP. Renewable energy
Therefore, the annual and national-level WTP is KRW 1.58 trillion (USD facilities such as photovoltaic panels and ESSs and microgrid networks
1.46 billion). This was determined by multiplying the number of are relatively easier to construct in non-metropolitan cities than in
households with the annual WTP per household. metropolitan cities. A higher WTP for non-metropolitan residents means
higher acceptance of P2P ETP. Therefore, the high acceptance of po­
5. Conclusion and policy implications tential consumers and the relative ease of infrastructure establish­
ment—which refers to the cost-benefit efficiency of the non-
The reduction of greenhouse gas emissions to cope with climate metropolitan prosumers—can have positive impacts on the activation
change is emerging as one of the most critical environmental issues in of P2P ETP. Therefore, introducing a P2P electricity trading system in
the world and P2P ETP has become one of the most significant new
forms of the energy market solutions in several countries such as the
United Kingdom, Germany, Netherlands, and the United States. The 6
The electricity price adjustment process in Korea is as follows. When the
South Korean government announced the “2030 New Energy Industry KEPCO board of directors submits the electricity price adjustment plan to the
Diffusion Strategy” in 2015 to preemptively address the future energy Ministry of Trade, Industry and Energy (MOTIE), this plan is first discussed
system. The strategy included the proliferation of energy prosumers as between MOTIE, the Ministry of Economy and Finance (MOEF), and the Korea
part of the relaxation of the regulations as well as new business creation Electricity Regulatory Commission (KOREC). After an agreement is reached and
in the energy sector. Moreover, the Korean government has revised the the price adjustment is sent to KEPCO, the result of the electricity price
“Guidelines for the transmission of small-scale renewable energy power adjustment is implemented. KOREC is affiliated with MOTIE, and the members
generation” on March 2018, thereby laying the foundation for the entry of KOREC consists of experts from various fields and institutions, such as uni­
of energy prosumers into the electric power trading market. To versities, research institutes, private organizations, and legal entities. However,
the final approval of the electricity price adjustment is provided by government.
encourage and assist energy prosumers through new trading platforms
This procedure is completely different from the general market pricing
and to activate the P2P electricity trade, in 2016, Korea has launched a
mechanism.
demonstration project of P2P ETP targeting two small towns. However, 7
In Korea, coal and nuclear power plants were built in large numbers during
the P2P ETP market in Korea is still in its early stages. the process of achieving rapid industrialization and economic growth, which
This study estimates the usage fee—or WTP—of consumers for P2P started in the late 1900s (Kim et al., 2011). In 2018, the share of electricity
ETP in South Korea. As Korea’s P2P ETP market is not yet activated and generation by energy source was 42.4% for coal power and 23.4% for nuclear
commercialized, we assume a virtual ETP market and estimate the power (KEEI, 2019).

6
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

non-metropolitan cities and gradually expanding it to other regions and commercialize P2P ETP in the future. Second, many of the de­
should be considered. Feldhiem, a small agricultural town in Germany, mographic characteristics have no significant association with WTP for
is Germany’s first energy self-sufficient village, generating 100% of its using P2P ETP. As we derived implications with only statistically sig­
own electricity and contributing to regional economic revitalization nificant covariates, we did not interpret the other non-significant de­
through electricity sales (Young and Brans, 2017). In Korea, small-scale mographic variables further. Further analyses are therefore necessary to
cities such as Chungnam and Jeonbuk are piloting energy self-sufficient discover why most of the demographic variables are not statistically
village programs, but they are not yet generating 100% of their own significant and to determine whether this result is specific to Korea or
electricity and electricity trading remains in the early stages of whether it can be commonly observed in other countries. Third, we set
introduction. up the contingent market of the P2P ETP based on the current status of
Third, the results showed that respondents who have more prior Korea with its distorted electricity price scheme. Since the estimated
knowledge about P2P ETPs are willing to pay more for using the plat­ WTP in this study reflects the current status of the Korean electricity
form. This means that enhancing consumer awareness can raise social market, the estimation result can vary as the market changes into the
acceptance of energy prosumers and P2P electricity trading. It is renewable energy base or the electricity price is determined by the
therefore necessary to gradually expand P2P ETP related policies and market not by the government. Therefore, by setting up several sce­
systems along with appropriate education and publicity, so that existing narios for the future electricity market of Korea and performing WTP
electricity consumers or energy prosumers who have traded only estimation for each scenario, more detailed information and different
through KEPCO can enter the ETP as early adopters. In previous studies implications for introducing P2P ETP can be obtained. Lastly, for the
concerning new energy market (or technology) introduction, the WTP of introduction of the P2P ETP, the establishment of a reasonable pricing
consumers that have more prior knowledge about the market were scheme is important in conjunction with the estimation of the consumer
higher and it has been suggested that enhancing consumers’ prior WTP. The usage fee can be set as a flat price, but it can also be set
knowledge is a significant factor for raising social acceptance and differently according to the users or the services offered. For example,
preference for the market (Roche et al., 2010; Kim et al., 2018; Xie and the pricing schemes of the platform can be established based on (1) the
Zhao, 2018). These existing studies analyzed various electricity-related amount of electricity usage, (2) the volume of electricity sales or pur­
industries and markets, such as green electricity and electric vehicles; chases, or (3) the quantity and quality of the functions or services pro­
our study, however, is the first—to our knowledge—to analyze the vided. However, an appropriate pricing scheme for P2P ETP will be
relationship between prior knowledge and WTP specifically concerning somewhat limited if derived only from our analysis results, as further
P2P ETP. studies are needed. There are several studies on the quality (or services)
Lastly, respondents paying higher electricity fees have higher WTP differentiated charging scheme in the electricity industry (Fumagalli
for using the P2P ETP. This indicates that the consumers with high et al., 2001; Woo et al., 2014; Kim et al., 2015; C. Zhang et al., 2018), but
electricity consumption have high acceptance and preference for ETPs, to the best of our knowledge, there is no study proposing charging
and their utilities for using the ETP are also high. This means that schemes for P2P ETPs from the platform providers’ perspective.
considering a strategy introducing P2P ETP to consumers with high
electricity consumption preferentially and then gradually expanding the Declaration of interest
scope of the target users may be effective. This phased introduction of
P2P ETP based on electricity consumption may mean that ETP com­ None.
panies can secure financial resources effectively.
The contributions of this study are twofold. First, from a consumer Acknowledgements
point of view, we successfully estimate the usage fee of P2P ETPs in
Korea, even though this platform has not been activated yet. Previous This research did not receive any specific grant from funding
studies on P2P ETPs have focused on the technical development and agencies in the public, commercial, or not-for-profit sectors.
application of technologies for improving the efficiency of the platform
but have not—to our knowledge—considered the consumer’s perspec­ Appendix. OOHBDC CVM Questionnaire
tive. This study further proposes strategies for activating the P2P ETP
market by including various socio-economic variables in the analysis. As a result of the pilot survey, the expected usage fee is between
Second, we compare the estimated usage fee of this study with that of (Lower) to (Upper) KRW per month.
other countries where the P2P ETP market has been introduced and Type A [Type A questions were posed to approximately half of the
commercialized, and discover that consumer acceptance of the platform respondents.]
in Korea is lower than those of countries. The main cause of the low Q1. Would you be willing to pay (Lower) KRW per month for using
consumer acceptance in Korea is the distorted electricity market. To P2P ETP?
address this, we propose that the electricity price structure should be
normalized to enable P2P ETP market vitalization. (1) Yes → Go to Q2
Despite these clear contributions, this study is subject to certain (2) No → Go to Q3
limitations. First, we estimated the usage fee of the P2P ETP by deter­
mining respondents’ willingness to pay for the presented bids based on Q2. If so, would you be willing to pay (Upper) KRW per month for
the CVM questionnaire format. However, when determining whether to using P2P ETP?
use the P2P ETP, consumers may consider the various technical and
social attributes of the platform, such as functionality, convenience, and (1) Yes → Survey ends
network externality. This means that the consumers’ WTP can vary (2) No → Survey ends
depending on the quality or quantity of each attribute. Therefore, it is
necessary for future studies to consider subdividing the attributes of the Type B [Type B questions are posed to the other half of the
P2P ETP when designing WTP questionnaires and conducting a conjoint respondents.]
analysis. Additionally, micro- and macro-level cost-benefit studies on Q1. Would you be willing to pay (Upper) KRW per month for using
prosumers and the electricity market are necessary to estimate the P2P ETP?
appropriate usage fee that will enable optimal operation of the P2P ETP
in the electricity market. A sensitivity analysis of the P2P ETP sub­ (1) Yes → Survey ends
scription rate considering the usage fee level is necessary to introduce (2) No → Go to Q2

7
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Q2. If so, would you be willing to pay (Lower) KRW per month for Kim, C., Shin, J., 2016. Electricity pricing policy alternatives to control rapid
electrification in Korea. J. Electr. Eng. Technol. 11 (2), 285–299.
using P2P ETP?
Kim, D., Kim, N., Kim, W., 2018. The effect of patent protection on firms’ market value:
the case of the renewable energy sector. Renew. Sustain. Energy Rev. 82,
(1) Yes → Survey ends 4309–4319.
(2) No → Go to Q3 Kim, H., Shin, E.S., Chung, W.J., 2011. Energy demand and supply, energy policies, and
energy security in the Republic of Korea. Energy Policy 39 (11), 6882–6897.
Kim, K., Nam, H., Cho, Y., 2015. Estimation of the inconvenience cost of a rolling
Q3. If so, would you willing to pay just 1 KRW or more per month for blackout in the residential sector: the case of South Korea. Energy Policy 76, 76–86.
using P2P ETP? Kim, K., Cho, Y., 2017. Estimation of power outage costs in the industrial sector of South
Korea. Energy Policy 101, 236–245.
KOSIS (Korean Statistical Information Service), 2019. Number of Resident Registration
(1) Yes → Survey ends Households by Administrative District [accessed 16.05.19]; Available from: htt
(2) No → Go to 4 p://kosis.kr/statHtml/statHtml.do?orgId¼101&tblId¼DT_1B040B3.
Krinsky, I., Robb, A.L., 1986. On approximating the statistical properties of elasticities.
Rev. Econ. Stat. 715–719.
Q4. What is your main reason for not being willing to pay? Please Kristr€om, B., 1997. Spike models in contingent valuation. Am. J. Agric. Econ. 79 (3),
mark the most appropriate reason below. 1013–1023.
Lee, M., Hong, T., Jeong, K., Kim, J., 2018. A bottom-up approach for estimating the
economic potential of the rooftop solar photovoltaic system considering the spatial
(1) I cannot afford to pay for using P2P ETP. and temporal diversity. Appl. Energy 232, 640–656.
(2) It should be provided free of charge. Marques, B., Nixon, K., 2013, September. The gamified grid: possibilities for utilising
game-based motivational psychology to empower the Smart Social Grid. In: 2013
(3) There is not enough information.
Africon. IEEE, pp. 1–5.
(4) I do not care about P2P ETP. Mengelkamp, E., G€ arttner, J., Rock, K., Kessler, S., Orsini, L., Weinhardt, C., 2018.
(5) I am against P2P ETP. Designing microgrid energy markets: a case study: the Brooklyn Microgrid. Appl.
Energy 210, 870–880.
Meyerhoff, J., Liebe, U., 2010. Determinants of protest responses in environmental
valuation: a meta-study. Ecol. Econ. 70 (2), 366–374.
References Nevitt, B., McLuhan, M., 1972. Take Today: the Executive as Dropout. Harcourt Brace
Jovanovich, New York.
Alam, M.R., St-Hilaire, M., Kunz, T., 2019. Peer-to-peer energy trading among smart Nomura, N., Akai, M., 2004. Willingness to pay for green electricity in Japan as
homes. Appl. Energy 238, 1434–1443. estimated through contingent valuation method. Appl. Energy 78 (4), 453–463.
Bertsch, V., Hall, M., Weinhardt, C., Fichtner, W., 2016. Public acceptance and Oerlemans, L.A., Chan, K.Y., Volschenk, J., 2016. Willingness to pay for green electricity:
preferences related to renewable energy and grid expansion policy: empirical a review of the contingent valuation literature and its sources of error. Renew.
insights for Germany. Energy 114, 465–477. Sustain. Energy Rev. 66, 875–885.
Bishop, R.C., Heberlein, T.A., 1979. Measuring values of extramarket goods: are indirect Parag, Y., Sovacool, B.K., 2016. Electricity market design for the prosumer era. Nat.
measures biased? Am. J. Agric. Econ. 61 (5), 926–930. energy 1 (4), 16032.
Boyle, K.J., Bishop, R.C., 1988. Welfare measurements using contingent valuation: a Park, H., Hong, W.K., 2014. Korea‫ ׳‬s emission trading scheme and policy design issues to
comparison of techniques. Am. J. Agric. Econ. 70 (1), 20–28. achieve market-efficiency and abatement targets. Energy Policy 75, 73–83.
Boyle, K.J., Johnson, F.R., McCollum, D.W., Desvousges, W.H., Dunford, R.W., Park, R., Koo, M.G., 2018. South Korea’s renewable energy policy: coming together or
Hudson, S.P., 1996. Valuing public goods: discrete versus continuous contingent- drifting apart? J. Int. Area Stud. 25 (2), 61–85.
valuation responses. Land Econ. 381–396. Piclo, 2019. Building Software for a Smarter Energy Future [accessed 16.05.19];
Cameron, T.A., Quiggin, J., 1994. Estimation using contingent valuation data from a" Available from: https://piclo.energy/.
dichotomous choice with follow-up" questionnaire. J. Environ. Econ. Manag. 27 (3), Ramajo-Hern� andez, J., del Saz-Salazar, S., 2012. Estimating the non-market benefits of
218–234. water quality improvement for a case study in Spain: a contingent valuation
Carson, R.T., Mitchell, R.C., Hanemann, M., Kopp, R.J., Presser, S., Ruud, P.A., 2003. approach. Environ. Sci. Policy 22, 47–59.
Contingent valuation and lost passive use: damages from the Exxon Valdez oil spill. Ready, R.C., Hu, D., 1995. Statistical approaches to the fat tail problem for dichotomous
Environ. Resour. Econ. 25 (3), 257–286. choice contingent valuation. Land Econ. 491–499.
Chien, Y.L., Huang, C.J., Shaw, D., 2005. A general model of starting point bias in Roche, M.Y., Mourato, S., Fischedick, M., Pietzner, K., Viebahn, P., 2010. Public attitudes
double-bounded dichotomous contingent valuation surveys. J. Environ. Econ. towards and demand for hydrogen and fuel cell vehicles: a review of the evidence
Manag. 50 (2), 362–377. and methodological implications. Energy Policy 38 (10), 5301–5310.
Cooper, J.C., Hanemann, M., Signorello, G., 2002. One-and-one-half-bound dichotomous Shandurkova, I., Bremdal, B.A., Bacher, R., Ottesen, S., Nilsen, A., 2012. A Prosumer
choice contingent valuation. Rev. Econ. Stat. 84 (4), 742–750. Oriented Energy Market. Developments and Future Outlooks for Smart Grid Oriented
Di Silvestre, M.L., Gallo, P., Ippolito, M.G., Sanseverino, E.R., Zizzo, G., 2018. A technical Energy Markets. A State-Of-The-Art Perspective, vol. 3. IMPROSUME Publication
approach to the energy blockchain in microgrids. IEEE Transactions on Industrial Series, Halden.
Informatics 14 (11), 4792–4803. Song, J., Seo, K., Kang, Y., 2018. Introduction of capacity charge system to expand PV
EPSIS (Electric Power Statistics Information System) [accessed 16.05.19]; Available market. Bull. Korea Photovolt. Soc. 4 (1), 35–41 (in Korean).
from: http://epsis.kpx.or.kr/epsisnew/selectEksaScfGrid.ajax?menuId¼060700. Sonnenbatterie, 2019. What Is the SonnenCommunity [accessed 16.05.19]; Available
Fatai, K., Oxley, L., Scrimgeour, F.G., 2003. Modeling and forecasting the demand for from: https://sonnengroup.com/sonnencommunity/.
electricity in New Zealand: a comparison of alternative approaches. Energy J. Sousa, T., Soares, T., Pinson, P., Moret, F., Baroche, T., Sorin, E., 2019. Peer-to-peer and
75–102. community-based markets: a comprehensive review. Renew. Sustain. Energy Rev.
Franke, M., Rolli, D., Kamper, A., Dietrich, A., Geyer-Schulz, A., Lockemann, P., 104, 367–378.
Schmeck, H., Weinhardt, C., 2005, June. Impacts of distributed generation from Stigka, E.K., Paravantis, J.A., Mihalakakou, G.K., 2014. Social acceptance of renewable
virtual power plants. In: International Sustainable Development Research energy sources: a review of contingent valuation applications. Renew. Sustain.
Conference, vol. 11. Energy Rev. 32, 100–106.
Fumagalli, E., Black, J.W., Ilic, M., Vogelsang, I., 2001, July. A reliability insurance Strazzera, E., Scarpa, R., Calia, P., Garrod, G.D., Willis, K.G., 2003. Modelling zero values
scheme for the electricity distribution grid. In 2001 Power Engineering Society and protest responses in contingent valuation surveys. Appl. Econ. 35 (2), 133–138.
Summer Meeting. In: Conference Proceedings (Cat. No. 01CH37262), vol. 1. IEEE, Vandebron, 2019. Duurzame Energie Van Nederlandse Bodem [accessed 16.05.19];
pp. 261–266. Available from: https://vandebron.nl.
Hanemann, W.M., 1984. Welfare evaluations in contingent valuation experiments with Welsh, M.P., Poe, G.L., 1998. Elicitation effects in contingent valuation: comparisons to a
discrete responses. Am. J. Agric. Econ. 66 (3), 332–341. multiple bounded discrete choice approach. J. Environ. Econ. Manag. 36 (2),
Hanemann, W.M., 1985. Some issues in continuous-and discrete-response contingent 170–185.
valuation studies. Northeast. J. Agric. Resour. Econ. 14(1204-2016-69798), 5. Werth, A., Kitamura, N., Tanaka, K., 2015. Conceptual study for open energy systems:
Huh, W.K., Kim, H., 2003. Information flows on the internet of Korea. J. Urban Technol. distributed energy network using interconnected DC nanogrids. IEEE Trans. Smart
10 (1), 61–87. Grid 6 (4), 1621–1630.
IMF (International Monetary Fund), 2019 [accessed 16.05.19]; Available from: https Woo, C.K., Ho, T., Shiu, A., Cheng, Y.S., Horowitz, I., Wang, J., 2014. Residential outage
://www.imf.org/external/np/-fin/data/param_rms_mth.aspx. cost estimation: Hong Kong. Energy Policy 72, 204–210.
Kang, E.S., Pee, S.J., Song, J.G., Jang, J.W., 2018. April). A blockchain-based energy Xie, B.C., Zhao, W., 2018. Willingness to pay for green electricity in Tianjin, China: based
trading platform for smart homes in a microgrid. In: 2018 3rd International on the contingent valuation method. Energy Policy 114, 98–107.
Conference on Computer and Communication Systems (ICCCS). IEEE, pp. 472–476. Yoo, S.H., Kwak, S.Y., 2009. Willingness to pay for green electricity in Korea: a
KEEI (Korea Energy Economics Institute), 2019. Energy Supply and Demand Brief contingent valuation study. Energy Policy 37 (12), 5408–5416.
2019.02 (in Korean). Yoo, Y.S., Hwang, T., Kang, S., Newaz, S.S., Lee, I.W., Choi, J.K., 2017. October). Peer-to-
KEMRI (KEPCO Management Research Institute), 2016. Overview of E-Prosumer and peer based energy trading system for heterogeneous small-scale DERs. In: 2017
domestic/foreign trends. KEMRI Power Economics Review 2016–2017 (in Korean). International Conference on Information and Communication Technology
Convergence (ICTC). IEEE, pp. 813–816.

8
J. Lee and Y. Cho Energy Policy 136 (2020) 111050

Young, J., Brans, M., 2017. Analysis of factors affecting a shift in a local energy system Zhang, C., Wu, J., Zhou, Y., Cheng, M., Long, C., 2018. Peer-to-Peer energy trading in a
towards 100% renewable energy community. J. Clean. Prod. 169, 117–124. Microgrid. Appl. Energy 220, 1–12.
Zachariadis, T., Pashourtidou, N., 2007. An empirical analysis of electricity consumption Zhang, K., Mao, Y., Leng, S., He, Y., Maharjan, S., Gjessing, S., Zhang, Y., Tsang, D.H.,
in Cyprus. Energy Econ. 29 (2), 183–198. 2018. Optimal charging schemes for electric vehicles in smart grid: a contract
theoretic approach. IEEE Trans. Intell. Transp. Syst. (99), 1–13.

You might also like