A bottom-up approach for estimating the economic potential of the rooftop solar photovoltaic system considering the spatial and temporal diversity

Applied Energy 232 (2018) 640–656

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

A bottom-up approach for estimating the economic potential of the rooftop solar photovoltaic system considering the spatial and temporal diversity

T

Minhyun Leea, Taehoon Honga, , Kwangbok Jeongb,a, Jimin Kima ⁎

a b

Department of Architecture and Architectural Engineering, Yonsei University, Seoul 03722, Republic of Korea Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, MI 48109-2125, United States

HIGHLIGHTS

economic potential of the rooftop solar PV system was estimated in South Korea. • The spatial and temporal diversity of the solar PV technology were considered. • The rooftop solar PV profitability was calculated based on a bottom-up approach. • The economic potential was 36-fold the installed capacity of the study area in 2016. • The • This study took one step farther towards the rooftop solar PV potential estimation. ARTICLE INFO

ABSTRACT

Keywords: Distributed solar generation Rooftop solar photovoltaic system Economic potential Rooftop solar photovoltaic profitability Hillshade analysis Life cycle cost

To successfully deploy distributed solar generation in urban environments, it is essential to investigate the potential to generate electricity from the rooftop solar photovoltaic (PV) system within a region. While various interpretations are possible for the rooftop solar PV potential, most of the previous studies focused on estimating the technical potential, not considering the economic viability and market dynamics. Therefore, it is necessary to estimate the economic potential of the rooftop solar PV system to quantify the amount of economically viable solar PV energy within a region and to evaluate the impact of the various factors affecting market access. Towards this end, this study proposed a bottom-up approach for estimating the economic potential of the rooftop solar PV system considering the market dynamics by adoption year. Accordingly, the economic potential of the rooftop solar PV system was estimated for the Gangnam district in Seoul, South Korea from 2008 to 2016. In terms of power capacity, it was analyzed that as of 2016, the actual installed capacity of the solar PV system in the Gangnam district was only 3% of the maximum economic potential of the rooftop solar PV system (i.e., economic potential for electricity business purposes), showing a high potential for additional rooftop solar PV adoption. In terms of electricity generation, it was shown that as of 2016, the annual economic potential of the rooftop solar PV system could supply up to 4.48% of the annual electricity consumption in the Gangnam district, while only 0.12% could be supplied from the annual electricity generation of the actual installed solar PV system. This study has significant contributions in that it took one step farther towards the rooftop solar PV potential estimation process, from the technical potential to the economic potential, considering the spatial and temporal diversity of the solar PV technology.

1. Introduction Recent concerns on climate change and resource depletion have brought about a demand for alternative energy sources and strategies, which highly contributed to the new energy paradigm shift from centralized to decentralized electricity generation with renewable energy, the so-called “distributed generation (DG)” [1]. DG allows onsite

⁎

electricity generation through renewable energy for nearby end users with small capacities, often with the solar photovoltaic (PV) system, instead of supplying electricity in a conventional manner, with large and centralized power plants [2]. In this regard, abundant rooftops and plenty of sunlight play a significant role in promoting distributed solar generation (DSG) in the building sector [3]. To successfully implement DSG by utilizing plenty of rooftops in

Corresponding author. E-mail address: [email protected] (T. Hong).

https://doi.org/10.1016/j.apenergy.2018.09.176 Received 14 June 2018; Received in revised form 17 September 2018; Accepted 22 September 2018 0306-2619/ © 2018 Elsevier Ltd. All rights reserved.

Applied Energy 232 (2018) 640–656

M. Lee et al.

urban environments, it is essential to investigate the potential of generating electricity from the rooftop solar PV system (i.e., rooftop solar PV potential) within an urban boundary [4]. In general, the rooftop solar PV potential can be categorized into five different hierarchical levels: (i) physical potential; (ii) geographic potential; (iii) technical potential; (iv) economic potential; and (v) market potential [5–7]. While various definitions and interpretations of the aforementioned rooftop solar PV potential are available, most of the previous studies focused on estimating the technical potential, the amount of energy that can be converted from solar radiation considering the geographic constraints (i.e., available rooftop area) and system performance (i.e., module efficiency) [7–10]. Early studies estimated the technical potential of the rooftop solar PV system of a region using relatively simplified methods, usually by approximating the available rooftop area with multipliers [3,11,12]. A study by Byrne et al. [3] used census data with certain factors that adopts an approximation of the available rooftop area, to estimate the rooftop solar PV potential in South Korea. Recent studies estimated the technical potential of the rooftop solar PV system of a region by integrating the geographic information system (GIS) technology [4,8,13–20]. Typically, studies by Buffat et al. [17] and Margolis et al. [18] utilized the satellite image data, Light Detection And Ranging (LiDAR) data in particular, to accurately estimate the rooftop solar PV potential in Switzerland and the United States (U.S.), respectively. A series of studies by Assouline et al. [19,20] used machine learning algorithms such as support vector machine and random forests, a relatively new approaches for estimating the rooftop solar PV potential, to estimate the rooftop solar PV potential in Switzerland. There is a dearth of studies, however, that attempted to investigate the economic and market potentials of the rooftop solar PV system worldwide. The economic potential is the subset of the technical potential, where the rooftop solar PV system is economically viable within a region. That is, the economic potential can explain how much energy is available with the economically viable rooftop solar PV system within a region. Meanwhile, the market potential refers to the amount of energy that can be expected from the market adoption of the rooftop solar PV system considering the market factors other than the economic aspects. Despite the significance of the economic and market potentials for promoting DSG, there have almost been no studies dealing with such potentials except for few researches done by National Renewable Energy Laboratory (NREL), a government-owned research center under the U.S. Department of Energy (DOE) specialized in renewable energy and energy efficiency. NREL estimated the economic potentials of various renewable energy sources in the U.S. considering the levelized cost of energy and levelized avoided cost of energy at the national level as of 2014 [6]. It was shown that the sum of the economic potentials of the different renewable technologies estimated in this study (i.e., wind energy, utility PV, DSG, hydropower, geothermal, and biopower) varied from one-third to over ten times the 2013 total U.S. generation from all energy sources, depending on the three distinct formulations defined in this study. NREL also estimated the market potentials of distributed energy resources for the residential, commercial, and industrial sectors in the U.S. through 2050 by developing the Distributed Generation Market Demand (dGen) model, a geospatially rich, bottom-up, marketpenetration model that simulates the potential adoption of various distributed energy resources (e.g. the dSolar module for DSG) [21,22]. Besides the U.S., the national level solar PV potential in South Korea has been estimated by Korea New & Renewable Energy Center (KNREC) under Korea Energy Agency (KEA), but only three hierarchical levels of the rooftop solar PV system (i.e., the physical, geographic, and technical potentials) were considered for analysis [5]. Therefore, to understand the economic viability of the rooftop solar PV system in a detailed way, and to evaluate the impact of various factors affecting market access (e.g., installation prices and policies), it is necessary to develop a method of estimating the economic potential and to conduct relevant studies in various regions. To address this challenge, this study aimed to develop a novel

bottom-up approach for estimating the economic potential of the rooftop solar PV system in a region. Using the developed method, the economic potential of the rooftop solar PV system was estimated for the Gangnam district, which consumes the most electricity (i.e., a total of 4,698,163 MWh in 2016) among the 25 local government districts of the city of Seoul in South Korea [23]. Towards this end, the following were mainly considered in this study, which differenciates it from other previous studies: (i) as a follow-up study, this study considered the technical potential of the rooftop solar PV system calculated for each building in the Gangnam district using the method developed by Hong et al. [8]; (ii) this study considered the rooftop solar PV profitability of each building in the Gangnam district based on the bottom-up approach to estimate the economic potential; (iii) this study considered the different installation purposes and subsidy payments to calculate the rooftop solar PV profitability and to estimate the economic potential; and (iv) this study considered the changes in the market conditions (i.e., installation cost, policies and support schemes, and electricity price) from 2008 to 2016 to calculate the rooftop solar PV profitability and to estimate the economic potential. First, this study calculated the technical potential of the rooftop solar PV system for each building in the Gangnam district using the method developed by Hong et al. [8]. As urban areas are not always suitable for installing the rooftop solar PV system due to the shadows on the rooftop from the surrounding buildings, Hong et al. [8] focused on calculating the available rooftop area by considering such shadows using Hillshade analysis to develop a method for estimating the technical potential. By using this method, each building in the Gangnam district would show a different technical potential, which would ultimately lead to different economic profitability for the rooftop solar PV system [24,25]. These differences in the technical and economic performance among the buildings were reflected to calculate the economic potential of the rooftop solar PV system within the Gangnam district. Second, this study calculated the rooftop solar PV profitability of each building in the Gangnam district based on the bottom-up approach for estimating the economic potential of the rooftop solar PV system. To calculate the rooftop solar PV profitability of each building in the Gangnam district, this study used life cycle cost (LCC) analysis based on the discounted cash inflows and outflows. Third, this study estimated the economic potential of the rooftop solar PV system for the different installation purposes and subsidy payments. In South Korea, the rooftop solar PV system can be installed for the following two purposes: (i) selfconsumption, and (ii) electricity business. When the rooftop solar PV system is installed for self-consumption purposes, the electricity generated from the system is primarily used for the building’s consumption. The surplus electricity generated from the rooftop solar PV system installed for self-consumption purposes can also sold to the grid depending on the selected business model. Meanwhile, when the rooftop solar PV system is installed for electricity business purposes, all the electricity generated from the system is sold to the grid. As the electricity generated from the installed rooftop solar PV system can be used for two different purposes, the profit structure also varies depending on the installation purpose. Therefore, this study considered the two different installation purposes, (i) self-consumption, and (ii) electricity business, to calculate the rooftop solar PV profitability, and ultimately, to estimate the economic potential. Furthermore, this study also considered the two different subsidy payments for self-consumption purposes, (i) without subsidy (i.e., baseline scenario), and (ii) with subsidy, to investigate the impact of subsidy payments, because there is always a possibility of not receiving such subsidy. Fourth, this study estimated the economic potential of the rooftop solar PV system from 2008 to 2016 to consider the changes in the market conditions (i.e., installation cost, policies and support schemes, and electricity price). As the market conditions of the solar PV industry, including the installation cost, policies and support schemes, and electricity price, vary significantly by year, the rooftop solar PV economic viability and profitability can be considerably different each year. Therefore, this study was conducted 641

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 1. Research framework.

for the target period from 2008 to 2016, the years with available data, to consider the market dynamics during such period. Towards the aforementioned end, this study was conducted in the five steps (refer to Fig. 1): (i) step 1: establishing the database; (ii) step 2: determining the system size and business model; (iii) step 3: calculating the electricity generation from the rooftop solar PV system; (iv) step 4: calculating the rooftop solar PV profitability; and (v) step 5: estimating the economic potential of the rooftop solar PV system.

Table 1 Building information data on the building characteristics, energy consumption, and physical characteristics with spatial information. Classification Building characteristics

Detailed description

Reference

Building floor area Building type

m2

E-AIS [26]

Residential Commercial Educational Public

E-AIS [26]

Actual electricity consumption Average electricity consumption

kWh

E-AIS [26]

kWh/m2 by building type

Energy Consumption Survey 2014 [27]

m

SPACEN [28]

m2

SPACEN [28]

2. Material and methods Energy consumption

2.1. Step 1: Establishing the database 2.1.1. Step 1.1: Database on the building information The building information data were collected to calculate the electricity generation and profitability of the rooftop solar PV system in the study area (refer to Table 1). First, the data on the building characteristics and energy consumption (i.e., building floor area, building type, and electricity consumption) in the Gangnam district, Seoul, South Korea were collected from Electronic Architectural administration Information System (E-AIS) under the Ministry of Land, Infrastructure, and Transport (MOLIT) of the South Korean government [26]. These data were particularly used to determine the system size and business model in step 2. Second, the data on energy consumption (i.e., average electricity consumption) in South Korea were additionally collected from Energy Consumption Survey 2014 by Korea Energy Economics Institute (KEEI) under the Ministry of Trade, Industry, and Energy (MOTIE) of the

Physical characteristics with spatial information

Building elevation Toal rooftop area

South Korean government [27]. As E-AIS does not provide public access to the actual electricity consumption data for every building in South Korea, this study used the average electricity consumption per building floor area by building type to determine the system size and business model for buildings without actual electricity consumption data in step 2. Third, the data on the physical characteristics with spatial 642

Applied Energy 232 (2018) 640–656

M. Lee et al.

information (i.e., building elevation and total rooftop area) in the Gangnam district were collected from Spatial Information Industry Promotion Institute (SPACEN) under MOLIT [28]. These data were particularly used to calculate the available rooftop area for solar PV installation in step 3. To match the building information data from the different data sources, this study used the official building address because the buildings in South Korea are managed based on their official address. As a result, the building information data for the total of 21,681 buildings were collected to estimate the economic potential of the rooftop solar PV system. Among the 14 legal-status neighborhoods, Suseo-dong, Segok-dong, Jagok-dong, and Yulhyeon-dong, which are located in the Southern part of the Gangnam district, were excluded from the scope of this study because a large area of these neighborhoods were under development during the target period.

policies and support schemes can vary depending on these two different installation purposes.

• For self-consumption: The rooftop solar PV system installed for self-

2.1.2. Step 1.2: Database on the rooftop solar PV system The data on the solar radiation and module efficiency were collected to calculate the electricity generation of the rooftop solar PV system in the study area. First, the data on the hourly solar radiation from 2008 to 2016 in Seoul was collected from World Radiation Data Centre (WRDC), sponsored by the World Meteorological Organization, to calculate the technical potential of the rooftop solar PV system on an hourly basis [29]. Second, the data on the solar PV module efficiency were collected from various data sources to reflect the technology level of the solar PV market and industry from 2008 to 2016. The solar PV module efficiency has been continuously improved over time across the different types and manufacturers, due to the technology development [30,31]. Despite the variance in solar PV module efficiency by type and manufacturer, the efficiency of the commercial solar PV modules generally stayed between 13% and 16% within the target period (refer to Table 2) [32–34]. Accordingly, this study used the module efficiency of the commercial solar PV system which represents the typical efficiency of the most widely used solar PV modules in each year from 2008 to 2016.

•

2.1.3. Step 1.3: Database on the rooftop solar PV profitability As the market conditions of the solar PV industry, including the installation cost, policies and support schemes, and electricity price, vary significantly by adoption year, the rooftop solar PV profitability shows considerable differences. Therefore, the data on the various key factors affecting the rooftop solar PV profitability such as the installation cost, policies and support schemes, and electricity price for the target period in the study area were collected to calculate the rooftop solar PV profitability by adoption year. First, the data on the installation cost of the solar PV system from 2008 to 2016 in South Korea were collected from the National Survey Report of PV Power Applications in Korea 2015 by International Energy Agency (IEA)-Photovoltaic Power System Programme (PVPS) and HAEZOOM, the most popular solar PV consulting company in South Korea [35,36]. As shown in Table 3, the installation cost of the solar PV system in South Korea dropped dramatically from US$6235 (equivalent to KRW6,662,000 based on the KRW1068.50/USD exchange rate as of January 20, 2018) per kW in 2008 to US$1404 (equivalent to KRW1,500,000 based on the KRW1068.50/USD exchange rate as of January 20, 2018) per kW in 2016. Second, the data on the applicable policies and support schemes from 2008 to 2016 in South Korea were collected from various data sources. As the rooftop solar PV system can be installed for two purposes, (i) self-consumption, and (ii) electricity business, the applicable

consumption purposes can receive subsidies and grants from the federal or city government. According to the building type and system size, different types of subsidies and grants are available from the federal government: (i) residential buildings with the system size up to 3 kW (multi family houses with the system size up to 30 kW from 2015): home subsidy; (ii) non-residential buildings with the system size up to 50 kW: building subsidy; and (iii) public buildings: local grants. The data on the subsidies and grants from the federal government were collected from KNREC under KEA [37]. The city government of the study area (i.e., Seoul metropolitan government) also provides financial incentives through the Seoul mini power plant subsidy from 2011. The data on the Seoul mini power plant subsidy were collected from Energy White Paper by the Seoul metropolitan government [23,38–40]. As shown in Table 3, the amount of subsidies and grants offered by the government tended to decrease as the installation cost dropped from 2008 to 2016. For electricity business: The rooftop solar PV system installed for electricity business purposes can receive Feed in Tariff (FIT) from the federal government or sell renewable energy certificates (RECs) in the market, in addition to selling the electricity to Korea Electric Power Corporation (KEPCO). The South Korean government initially adopted FIT in 2002 by providing a long-term fixed tariff for 15 or 20 years for renewable energy generation, but it was replaced by Renewable Portfolio Standard (RPS) in 2012 [23]. Unlike FIT, which guarantees acceptable compensation for renewable energy generation, RPS creates an autonomous market environment through REC trading between renewable energy producers and power producers subject to RPS [37]. The data on the FIT rates and REC prices were collected from each of the following data sources, respectively: (i) KNREC under KEA [37]; and (ii) the REC Trading System under Korea Power Exchange (KPX) [41]. As shown in Table 3, FIT offered by the government slightly decreased as the installation cost dropped from 2008 to 2011, but still showed a high rate throughout its implementation (over US$0.5/kWh). Meanwhile, the REC price (1 REC = 1 MWh) depends on the market conditions (i.e., the supply and demand of RECs in the market), showing high volatility ranging from US$85.62 to 168.19 [42]. Overall, FIT was about four times more expensive than the REC price per kWh, bringing a financial burden on the government during its implementation. The city government of the study area (i.e., Seoul metropolitan government) also provides the financial support from 2013 through the Seoul FIT, which is duplicable with RPS, as well as the rooftop rental. The data on the Seoul FIT and rooftop rental were collected from Energy White Paper by the Seoul metropolitan government [23,38–40].

Third, the data on the electricity price from 2008 to 2016 in South Korea were collected from two different data sources depending on the installation purpose of the rooftop solar PV system, as follows.

• For self-consumption: The rooftop solar PV system installed for self-

consumption purposes profits by using the electricity generated by the system and lowering the electricity bill. Therefore, the data on the average retail price of electricity from 2008 to 2016 in the Gangnam district were collected from Power Big Data Center under

Table 2 Key factors affecting the technical potential of the rooftop solar PV system by year. Year

2008

2009

2010

2011

2012

2013

2014

2015

2016

Module efficiency (%)

13.60

13.70

14.00

14.45

14.90

15.35

15.68

16.00

16.20

643

Applied Energy 232 (2018) 640–656

M. Lee et al.

Table 3 Key factors affecting the rooftop solar PV profitability by year in South Korea. Classification

Eligible system size

2008

2009

2010

2011

2012

2013

2014

2015

2016

Note

Installation cost

Residential (US$/kW) Commercial (US$/kW)

∼10 kW ∼1 MW

6235 6235

5475 5475

4736 4736

3744 3744

2808 2808

2808 2808

2340 2620

1638 2106

1404 1872

Industry average

Applicable policies and support schemes (Federal)

Selfconsumption

∼3 kW

60%

60%

60%

50%

40%

1,076

880

786

627

3 kW ∼30 kW ∼50 kW

–

–

–

–

–

–

–

898

749

60%

60%

60%

50%

40%

1336

1207

1179

1020

per household for single family house per building for multi family house –

∼1 MW ∼30 kW 30 kW ∼200 kW 200 kW ∼1 MW ∼1 MW

50% 0.666 0.634

50% 0.605 0.581

50% 0.568 0.542

50% 0.568 0.542

50% – –

50% – –

50% – –

50% – –

50% – –

0.634

0.553

0.511

0.511

–

–

–

–

–

–

–

–

–

168

164

114

86

127

for 20 years

–

–

–

749

1217

1029

880

749

655

–

– – –

– – –

– – –

– – –

– – –

– 0.047 –

– 0.047 –

562 0.094 0.094

468 0.094 0.094

Duplicable with RPS

Rooftop rental (US$/kW/year)

1 kW ∼3 kW 3 kW∼ ∼50 kW 50 kW ∼100 kW ∼100 kW 100 kW∼

– –

– –

– –

– –

– –

23 23

23 23

19 23

19 23

–

Residential Commercia Educational SMP

– – – –

0.141 0.084 0.081 0.115

0.139 0.088 0.086 0.098

0.145 0.089 0.089 0.110

0.141 0.092 0.096 0.118

0.145 0.103 0.114 0.150

0.147 0.113 0.123 0.142

0.142 0.121 0.118 0.133

0.139 0.122 0.116 0.095

0.139 0.123 0.113 0.072

Average retail price

Electricity business

Home subsidy (US$/kW) Building subsidy (US$/kW) Local grants FIT (US$/kWh)

REC (US$/MWh) Applicable policies and support schemes (Seoul)

Electricity price (US$/kWh)

Selfconsumption Electricity business

Selfconsumption Electricity business

Seoul mini power plants subsidy (US$/kW) Seoul FIT (US$/kWh)

– for 15 years

Annual weighted average price

Note: % for home subsidy, building subsidy, and local grants indicates that a specified percentage of the installation cost is offered as subsidies or grants.

•

KEPCO by building type (i.e., residential, commercial, and educational) [43]. For electricity business: The rooftop solar PV system installed for electricity business purposes profits by selling the electricity generated by the system. Therefore, the data on the system marginal price (SMP) from 2008 to 2016 in South Korea were collected from the Electric Power Statistics Information System (EPSIS) under KPX [44].

owners to use the electricity generated from their rooftop solar PV system and sell the surplus electricity back to the grid through KEPCO, the largest electric utility in South Korea, responsible for the majority of South Korea’s electricity supply [46]. The main difference between the two is their profit structure; net metering credits the building owners for the surplus electricity that they send to the grid so that they are billed only for their net electricity use while PPA for self-consumption lets building owners literally sell their surplus electricity back to the grid. That is, when a building is net-metered, the building owner needs to pay only for the net electricity used based on the retail price of electricity whereas when a building adopts PPA for self-consumption, the building owner has to fully pay for the electricity used from the grid with the retail price of electricity and sell the surplus electricity separately back to the grid based on the SMP, which is the wholesale price of electricity. Net metering is the most common and preferred business model for self-consumption purposes as it allows building owners to sell their surplus electricity at the retail price, which is usually higher than the wholesale price. Meanwhile, building owners in South Korea who decide to install a rooftop solar PV system with host ownership for electricity business purposes can sell electricity through the following two business models: (i) electricity market trading; and (ii) PPA. Traditionally, power producers, usually large utility-scale power plants, trade electricity directly from the electricity market through KPX, the organization that operates the electricity market and power system in South Korea [47]. This electricity market trading system, however, involves high additional costs (e.g., the cost of installing the meters required for the electricity trading process) for power producers as it was originally planned for the electricity trade across large utility-scale power plants. Therefore, small renewable energy producers such as rooftop solar PV system adoptors had difficulty participating in the electricity market to sell electricity. To overcome this limitation, the South Korean government allowed small renewable energy producers to sell their electricity directly to KEPCO via PPA beginning in 2005 rather than trade electricity

2.2. Step 2: Determining the system size and business model 2.2.1. Step 2.1: Determining the business models to be considered As the solar PV industry showed remarkable growth in the past decade, various business models and support schemes for it constantly emerged and evolved. Therefore, the business models and their financing mechanisms available during the target period in the study area were identified and analyzed based on the two aforementioned installation purposes to determine the business models to be considered in this study. Tables 4 and 5 summarized the business models and their financing mechanisms according to the installation purpose of the rooftop solar PV system in South Korea, and how they changed over the years. As shown in Tables 4 and 5, the business models and financing mechanisms of the rooftop solar PV adoption for both the self-consumption and electricity business purposes can be mainly classified into three categories: (i) host ownership; (ii) third-party ownership; and (iii) community-based financing [45]. First, host ownership is the most traditional way to install a rooftop solar PV system where the building owner keeps the ownership of the installed rooftop solar PV system. In South Korea, building owners who decide to install a rooftop solar PV system with host ownership for selfconsumption purposes can choose between two business models: (i) net metering; and (ii) Power Purchase Agreement (PPA) for self-consumption. Both net metering and PPA for self-consumption allow building 644

Applied Energy 232 (2018) 640–656

M. Lee et al.

Table 4 Business models of the rooftop solar PV system for self-consumption purposes in South Korea. Purpose

Self-consumption

Business model

Host ownership

Applicable sector Implementing year Eligible system size

Profit structure Applicable policies and support schemes (Federal)

Applicable policies and support schemes (Seoul) Applicability

Third-party ownership

Community-based financing

Net metering

PPAa for self-consumption

Solar lease

Prosumer electricity trading

All sectors (Mainly residential) 2005 less than 3 kW (2005–2011) less than 10 kW (2012–2015) less than 50 kW (2016) less than 1000 kW (2017∼) ① Electricity bill savings

All sectors

Residential

2005 less than 200 kW (2005–2009) less than 1000 kW (2010∼)

2013 Single family: 3–9 kW per household Multi family: 10–30 kW per building ① Pay for solar lease ② Electricity bill savings

Residential (has potential to be extended to other sectors in the future) 2017 less than 1,000 kW

Residential: Home subsidy (2004∼) Non-residential: Building subsidy (1993∼) Public: Local grants (2006∼) Seoul mini power plants (2011∼) Most common

① Electricity bill savings ② Selling surplus electricity ③ Selling REC (2016∼) Residential: Home subsidy (2004∼) Non-residential: Building subsidy (1993∼) Public: Local grants (2006∼) Seoul mini power plants (2011∼) Alternative to net metering

Undefined

n/a

Residential: Home subsidy (2004∼)

n/a

Seoul mini power plants (2011∼)

Rapidly growing

Emerging

a

Note: PPA refers to Power Purchase Agreement.

from the electricity market through KPX. Accordingly, PPA has become the most common business model for installing rooftop solar PV systems for electricity business purposes. Second, third-party ownership and community-based financing are relatively new and recently developed methods of installing a rooftop solar PV system introduced after 2012. In third-party ownership, the ownership of the installed rooftop solar PV system is retained by a third-party, while in community-based financing, a community shares the electricity generated from a rooftop solar PV system. The concept of third-party ownership for self-consumption purposes has emerged to allow consumers to install a solar PV system on their rooftops without a high upfront installation cost. In South Korea, consumers who desire to install a rooftop solar PV system with third-party ownership for selfconsumption purposes can adopt the solar lease. In the solar lease model, rather than buy the system, residents pay for the use of the rooftop solar PV system over a period of time to a solar leasing company, who installs and owns the solar PV system on the resident’s rooftop. This solar lease model was introduced in South Korea in 2013 and is a rapidly growing business model as it does not involve an upfront installation cost. Community-based financing for self-

consumption purposes is the most recently developed business model providing a new energy trade platform with the concept of peer-to-peer (P2P) energy trading. Under this P2P energy trading concept, prosumer electricity trading was introduced in South Korea in 2017, where prosumers, who can both produce and consume energy, can sell their surplus electricity directly to their neighbors. Meanwhile, the concept of third-party ownership and communitybased financing for electricity business purposes has emerged to promote the installation of a solar PV system on the unused rooftops of public buildings, including educational facilities (i.e., schools). In South Korea, people who want to operate an electricity business but who do not own a rooftop where they can install the solar PV system can choose between the following two business models depending on the number of participants: (i) PPA with rooftop rental; and (ii) solar power cooperative. Basically, both PPA with rooftop rental and solar power cooperative allow electricity business operators to sell the electricity generated from solar energy by renting the unused rooftop of a public building and installing a solar PV system on it. The only difference between the two is the number of participants in a single electricity business; an individual or a single entity participates in the electricity

Table 5 Business models of the rooftop solar PV system for electricity business purposes in South Korea. Purpose

Electricity business

Business model

Host ownership

Applicable sector Implementing year Eligible system size Profit structure

Applicable policies and support schemes (Federal) Applicable policies and support schemes (Seoul) Applicability

Third-party ownership

Electricity market trading

PPAa

PPAa with rooftop rental

Community-based financing Solar power cooperative

All sectors 1962 more than 200 kW (2005–2009) more than 1000 kW (2010∼) ① Selling electricity (SMP) ② Receiving FIT (2002–2011) or selling REC (2012∼)

All sectors 2005 less than 200 kW (2005–2009) less than 1000 kW (2010∼) ① Selling electricity (SMP) ② Receiving FIT (2002–2011) or selling REC (2012∼)

Educational/Public 2013 n/a

Educational/Public 2012 n/a ① Pay for rooftop rental ② Selling electricity (SMP) ③ selling REC

FIT (2002–2011) RPS (2012∼) Seoul FIT (2013∼)

FIT (2002–2011) RPS (2012∼) Seoul FIT (2013∼)

① Pay for rooftop rental ② Selling electricity (SMP) ③ selling REC RPS (2012∼)

RPS (2012∼)

Seoul FIT (2013∼)

Seoul FIT (2013∼)

Not recommended

Most common

Emerging

Emerging

Note: PPAa refers to Power Purchase Agreement. 645

Applied Energy 232 (2018) 640–656

M. Lee et al.

business through PPA with rooftop rental, whereas a group of people participates in the electricity business through a solar power cooperative and shares the benefit from the electricity generated from a rooftop solar PV system among themselves. Therefore, the basic profit structures of these two business models are identical, besides the profit sharing structure of the solar power cooperative among the members. Based on the aforementioned business models by installation purpose, this study considered the following business models for calculating the rooftop solar PV profitability: (i) self-consumption: net metering and PPA for self-consumption; and (ii) electricity business: PPA and PPA with rooftop rental. Other business models were excluded for calculating the rooftop solar PV profitability for the following reasons. First, solar lease was excluded for calculating the rooftop solar PV profitability in this study because it is hard to compare its LCC analysis result with those of other business models as it does not have an upfront cost. Moreover, the economic performance of the solar lease model highly depends on the monthly electricity consumption of an individual household and the South Korean government has already analyzed and officially announced the profitability of the solar lease model according to the monthly electricity consumption level. Second, prosumer electricity trading was excluded for calculating the rooftop solar PV profitability in this study because it was implemented in 2017 and its profit structure is currently undefined. Third, electricity market trading was excluded for calculating the rooftop solar PV profitability in this study because it entails high additional costs for power producers and is rarely adopted in the rooftop solar PV system. Fourth, solar power cooperative was excluded for calculating the rooftop solar PV profitability in this study because the basic profit structure of solar power cooperative is the same as that of PPA with rooftop rental. Therefore, by including PPA with rooftop rental as one of the business models for calculating the rooftop solar PV profitability, the overall economic performance of solar power cooperative can also be analyzed.

the actual electricity consumption data. System size D of the buildings without actual electricity consumption data was calculated based on the average electricity consumption per building floor area by building type collected from Energy Consumption Survey 2014 by KEEI [27]. Since this electricity consumption data is used only to determine system size D, the effect on the final results (i.e., the economic potential of the rooftop solar PV system) due to the substitution of average values in place of the actual values would be insignificant. System size D of each building can be calculated using Eq. (2). The data on the technical potential of the rooftop solar PV system, required module surface area, and performance ratio, which are necessary for calculating system size D, are explained in detail in step 3.

D=

TRAGB × r MRA

(2)

where D stands for the system size for the 100% self-sufficiency of each building (kW), ECGB stands for the annual electricity consumption of a given building (kWh), TechnicalGB stands for the annual technical potential of the rooftop solar PV system for a given building per unit area (kWh/m2), MSA stands for the required module surface area per kW (m2/kW), and PR stands for the performance ratio (89%). Second, the system size and business model were determined for the rooftop solar PV system of the 21,681 buildings in the Gangnam district by considering the actual installation trend in the study area. Basically, for both the self-consumption purposes and electricity business purposes, the installation size of the rooftop solar PV system for each building cannot exceed the maximum installable system size on each rooftop area due to the physical constraint of the rooftop area (i.e., less than system size A). Based on this maximum installable system size (i.e., system size A), the appropriate system size and business model for installing the rooftop solar PV system were determined according to the installation purpose, as follows.

• For self-consumption: As shown in Fig. 2, the four system size alter-

2.2.2. Step 2.2: Determining the system size and business model The system size and business model of the rooftop solar PV system was determined for each building to calculate the electricity generation and profitability of the rooftop solar PV system for all the buildings in the study area. Based on the applicable policies and support schemes from 2008 to 2016, the system size and business model of the rooftop solar PV system was determined for each building considering its total rooftop area, building type, adoption year, and annual electricity consumption (refer to Fig. 2). First, five possible system size alternatives (refer to system sizes A ∼ E in Fig. 2) were calculated for the rooftop solar PV system of all the 21,681 buildings in the Gangnam district, as follows: (i) maximum installable system size on each rooftop area (system size A); (ii) eligible system size for subsidy (system size B) considering its building type; (iii) eligible system size for net metering (system size C) considering its adoption year; (iv) system size for the 100% self-sufficiency of each building (system size D) considering its annual electricity consumption; and (v) eligible system size for PPA (system size E) considering its adoption year. System sizes B, C, and E of each building were simply determined based on the building type or adoption year, whereas system size A of each building was calculated based on the total rooftop area, using Eq. (1).

A=

ECGB TechnicalGB × MSA × PR

•

(1)

natives (i.e., system sizes A ∼ D) were compared to one another to determine the appropriate system size and business model for installing the rooftop solar PV system according to the following criteria: (i) the system size was determined to receive the subsidy from the federal government for all buildings (i.e., less than system size B); (ii) the system size was determined to not exceed the system size for the 100% self-sufficiency of each building considering its annual electricity consumption (i.e., less than system size D); (iii) net metering has priority over PPA as it is the more common and preferred business model for self-consumption purposes; and (iv) when the system size exceeds the eligible size for net metering, PPA can be selected as an alternative. For electricity business: As shown in Fig. 2, the two system size alternatives (i.e., system sizes A and E) were compared to each other to determine the appropriate system size for PPA for all buildings (i.e., less than system size E). Meanwhile, the appropriate business model for installing the rooftop solar PV system was determined according to the building type and adoption year, as follows: (i) PPA was selected for residential and commercial buildings; (ii) PPA was selected for educational and public buildings with an adoption year before 2013; and (iii) PPA with rooftop rental was selected for educational and public buildings with an adoption year after 2013.

2.3. Step 3: Calculating the electricity generation of the rooftop solar PV system

where A stands for the maximum installable system size on each rooftop area (kW), TRAGB stands for the total rooftop area of a given building (m2), r stands for the ratio of the total rooftop area allowable for installing the solar PV system (70% according to Building Solar PV System Installation Guideline by Seoul metropolitan government [48]), and MRA stands for the minimum rooftop area required for installing a 1 kW solar PV system (16.5 m2 according to KNREC [37]). Meanwhile, system size D of each building was calculated based on

The electricity generation of the rooftop solar PV system was calculated for each building to evaluate the expected technical performance of the rooftop solar PV system in the study area in the following two steps: (i) calculating the available rooftop area for solar PV installation; and (ii) calculating the technical potential of the rooftop solar PV system. 646

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 2. Algorithm for determining the system size and business model.

The available rooftop area for solar PV installation was calculated to estimate the technical potential of the rooftop solar PV system for each building in the Gangnam district. First, the shaded rooftop area was calculated using Hillshade analysis to quantify the rooftop area where the solar PV system cannot perform at its optimal level based on the actual location of the sun. There are two types of input data required to calculate the shaded area using Hillshade analysis: (i) the building data on the physical characteristics with spatial information; and (ii) the location data of the sun at the time of the analysis. This study used the building physical characteristics (i.e., building elevation and total rooftop area) with the spatial information of the Gangnam district collected from SPACEN in step 1.1 [28]. The location data (i.e., altitude and azimuth) of the sun was calculated using the Sun Altitude and Azimuth Calculation tool from Korea Astronomy & Space Science Institute (KASI) [49]. By using these input data, this study conducted Hillshade analysis for 12 days (on the 15th of each month from January to December) at hourly intervals (from 6 a.m. to 7 p.m.) using ArcMap 10.1, a widely used GIS software developed by Environmental Systems Research Institute (ESRI) [50]. As a result, a total of 156 simulations were performed to calculate the shaded area in the Gangnam district. Second, the available rooftop area of each building for solar PV installation was calculated by removing the shaded rooftop area based on Hillshade analysis results. By removing the shaded rooftop area from the total rooftop area of an individual building, the available rooftop area for every building in the study area where the solar PV system can perform at its optimal level without any disturbance of the building shadow was quantified. For a more detailed description of the calculation process of

the available rooftop area for solar PV installation, please refer to the study by Hong et al. [8]. The technical potential of the rooftop solar PV system was calculated considering the available rooftop area and hourly solar radiation to estimate the practical and realistic electricity generation of the rooftop solar PV system for each building in the Gangnam district. First, the total solar radiation on the available rooftop area of a building considering the solar PV module efficiency was normalized with its total rooftop area to calculate the technical potential of the rooftop solar PV system for each building in the Gangnam district. Second, the expected electricity generation of the rooftop solar PV system for each building in the Gangnam district was then calculated considering the technical potential and size of the rooftop solar PV system. Accordingly, the monthly technical potential and expected electricity generation of the rooftop solar PV system for a given building were calculated using Eqs. (3) and (4), by integrating various key factors including the available rooftop area, hourly solar radiation, solar PV module efficiency, system size, required module surface area, and performance ratio. Among these key factors, the required module surface area per kW can be calculated using Eq. (5), based on the previous studies [12,51,52]. The performance ratio was also determined to consider the overall effect of the losses on the actual solar PV system output due to various factors (e.g., array temperature, soiling, shading, mismatch) [53,54]. Among these various loss factors, the shading losses were excluded for determining the performance ratio, as this study already considered the loss due to shading for calculating the technical potential of the rooftop solar PV system [52,55,56]. Apart from these key 647

Applied Energy 232 (2018) 640–656

M. Lee et al.

factors, the orientation and tilt angle of the rooftop solar PV system were determined to fully consider the hourly calculated results of the available rooftop area. As the available rooftop area without any disturbance from the building shadow changes with time due to the hourly calculation method, this study assumed that the solar PV panels are installed horizontally on rooftops. For a more detailed description of the calculation process of the technical potential of the rooftop solar PV system, please refer to the study by Hong et al. [8]. 18

TechnicalGBi =

j=6

ARAGBij ×

n

SRijk × ePV

k=1

TRAGB

EGGBi = TechnicalGBi × SSGB × MSA × PR MSA =

The expected ROI and PP of the rooftop solar PV adoption for each building were calculated based on the discounted cash inflows (i.e., benefits) and outflows (i.e., costs) using Eqs. (6)–(13). The benefits and costs of the rooftop solar PV adoption for each building were calculated using Eqs. (8)–(13) depending on the installation purpose (i.e., selfconsumption or electricity business) and business model (i.e., net metering, PPA for self-consumption, PPA, or PPA with rooftop rental), as the profit structure differs by installation purpose and business model. As the rooftop solar PV system installed for self-consumption purposes profits by receiving subsidies from the government and saving the electricity bill, the benefit from electricity generation was calculated based on the amount of subsidy and monthly average retail price of electricity by building type. As the rooftop solar PV system installed for electricity business purposes profits by selling the electricity and either receiving FIT (2008–2011) or selling RECs (2012–2016; along with the Seoul FIT from 2013), the benefit from electricity generation was calculated based on the monthly weighted average SMP, FIT, REC price, and Seoul FIT.

PC SISTC × ePV

(3) (4) (5)

where TechnicalGBi stands for the technical potential of the rooftop solar PV system for a given building per unit area in month i (kWh/m2), EGGBi stands for the expected electricity generation of the rooftop solar PV system for a given building in month i (kWh), ARAGBij stands for the available rooftop area of a given building on the 15th of month i at time j to j + 1 (m2), SRijk stands for the solar radiation on day k of month i at time j to j + 1 (kWh/m2), ePV stands for the solar PV module efficiency in adoption year (%), TRAGB stands for the total rooftop area of a given building (m2), SSGB stands for the system size of a given building (kW), MSA stands for the required module surface area per kW (m2/kW), PR stands for the performance ratio (89%), PC stands for the power capacity (1 kW), SISTC stands for the solar irradiance under standard test condition (STC) (1,000 W/m2), i stands for the month (i = 1, 2, 3, …, 12), j stands for the time in 24-hour format (j = 6, 7, 8, …, 18), k stands for the day of a month (k = 1, 2, 3, …, 31), and n stands for the total number of days in month i. By multiplying (i) the total solar radiation at a certain time frame (e.g., 12 to 1 p.m.) for an entire month, (ii) the available rooftop area of a given building at a certain time frame in month i, and (iii) the solar PV module efficiency, the technical potential of the rooftop solar PV system for a given building in month i at a certain time frame was calculated. Then, the technical potential of the rooftop solar PV system in month i for a given building was normalized with its total rooftop area. This process was carried out for every time frame and every month to calculate the monthly technical potential of the rooftop solar PV system for a given building. The calculated technical potential of the rooftop solar PV system for a certain building per unit area can be now used to estimate the monthly electricity generation of the rooftop solar PV system with the desired system size considering the various system losses.

n

ROIGB =

t=1 n t=1

PPGB = T ,

Bt (1 + r )t Ct (1 + r )t

when

(6) T t=1

Bt (1 + r )t

T

Ct (1 + r )t

t=1 12

Bt for SC (NM ) = St × SSGB + i=1

Bt for SC (PPA) = St × SSGB + +

12 i=1

(

EGGBi ECGBi (1 + d )t

)

× SMPi +

12

=0

(7)

EGGBi × RPi (1 + d )t 12 i=1

(

ECGBi (1 + d)t

× RPi

(8)

)

(NRECi × RECi × wREC )

i=1

(10)

Ct for SC = ICt × SSGB + OMCt 12

EGGBi × SMPi + (1 + d )t

Bt for EB (FIT ) = i=1

Bt for EB (RPS ) =

12 i=1

+

(

EGGBi (1 + d )t 12

i=1

(

)

× SMPi +

EGGBi (1 + d )t

(9)

12 i=1

12 i=1

EGGBi × FITt (1 + d )t

(11)

(NRECi × RECi × wREC )

) × SeoulFIT

t

(12)

Ct for EB = (ICt + RRCt ) × SSGB + OMCt

2.4. Step 4: Calculating the rooftop solar PV profitability

(13)

where ROIGB stands for the expected ROI of the rooftop solar PV system for a given building, PPGB stands for the expected PP of the rooftop solar PV system for a given building (years), Bt stands for the benefit from electricity generation in year t (US$), Ct stands for the cost from rooftop solar PV installation in year t (US$), SC stands for self-consumption purposes, SC(NM) stands for net metering for self-consumption purposes, SC(PPA) stands for PPA for self-consumption purposes, EB stands for electricity business purposes, EB(FIT) stands for electricity business purposes with FIT (2008–2011), EB(RPS) stands for electricity business purposes with RPS (2012–2016), n stands for the analysis period, r stands for the real discount rate, St stands for the amount of subsidy per kW in year t (US$), SSGB stands for the system size of a given building (kW), EGGBi stands for the expected electricity generation of the rooftop solar PV system for a given building in month i (kWh), ECGBi stands for the electricity consumption of a given building in month i (kWh), d stands for the annual degradation rate of the rooftop solar PV system, RPi stands for the average retail price of electricity per kWh in month i (US$/kWh), SMPi stands for the SMP per kWh in month i (US$/kWh), FITt stands for the FIT in year t (US$/kWh), NRECi stands for the number

The rooftop solar PV profitability was calculated for each building to evaluate the expected economic performance of the rooftop solar PV system in the study area for each target year. Over this period, the solar PV market has gone through major changes, resulting in different rooftop solar PV profitability by adoption year. To calculate the rooftop solar PV profitability of each individual building, this study conducted LCC analysis using the return on investment (ROI) and payback period (PP), the most widely used relative valuation metrics for deciding on whether to adopt the solar PV system [57,58]. ROI represents the ratio of the discounted cash inflows and outflows, as shown in Eq. (6). When ROI exceeds “1,” it can be inferred that the economic viability of the rooftop solar PV system has been achieved, and a decision on rooftop solar PV adoption can be made [59,60]. Meanwhile, PP represents the length of time (e.g., year) required for recovering the investment cost and reaching a break-even point, as shown in Eq. (7), making it simple and easy to determine the economic feasibility of rooftop solar PV adoption [60–62]. 648

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 3. Analysis results of the rooftop solar PV profitability by adoption year.

of RECs issued in month i, RECi stands for the REC price in month i (US $), wREC stands for the REC weight for installing the solar PV system on buildings (1.0 for self-consumption and 1.5 for electricity business on rooftops), SeoulFITt stands for the Seoul FIT in year t (US$/kWh), ICt stands for the installation cost of the rooftop solar PV system per kW in year t (US$), OMCt stands for the operation and maintenance (O&M) cost in year t (US$), and RRCt stands for the rooftop rental cost per kW in year t (US$). To calculate the expected ROI and PP of the rooftop solar PV system for each building, the following assumptions for LCC analysis were defined: (i) analysis approach; (ii) analysis period; (iii) real discount rate; (iv) inflation rate; (v) operation and maintenance (O&M) cost; and (vi) system degradation rate [59].

degradation rate for conducting LCC analysis. 2.5. Step 5: Estimating the economic potential of the rooftop solar PV system The economic potential of the rooftop solar PV system in the study area was estimated to quantify the amount of economically viable rooftop solar PV energy within the study area during the target period. Accordingly, the economic potential of the rooftop solar PV system in the Gangnam district was evaluated based on the analysis results of the rooftop solar PV profitability from two perspectives: (i) power capacity: the economic potential of the rooftop solar PV system was estimated by calculating the amount of economically viable rooftop solar PV system in terms of power capacity within the study area during the target period; and (ii) electricity generation: the economic potential of the rooftop solar PV system was estimated by calculating the amount of electricity that can be generated from the economically viable rooftop solar PV systems within the study area during the target period. It was assumed that when ROI exceeds “1” and PP falls within the analysis period (i.e., 25 years), the economic viability of the rooftop solar PV system has been reached.

• Analysis approach and period: This study used the present-worth •

i=

method for the analysis period of 25 years based on the useful life and warranty period of the solar PV panel [59,60,63]. Real discount rate: This study calculated the real discount rate using Eq. (14), based on the following data: (i) nominal interest rate: base rate from 2008 to 2016 provided by the Economic Statistics System (ECOS) of Bank of Korea [64]; (ii) inflation rate: inflation rate from 2008 to 2016 provided by ECOS of Bank of Korea [64]; (iii) electricity price growth rate: retail price of electricity and SMP from 2008 to 2016 provided by Power Big Data Center and EPSIS, respectively [43,44]; and (iv) REC price growth rate: REC price from 2012 to 2016 provided by the REC trading system [41].

(1 + in ) 1 (1 + f )

3. Results and discussion 3.1. Analysis of the rooftop solar PV profitability The rooftop solar PV profitability in the Gangnam district was analyzed by adoption year from 2008 to 2016, based on the different installation purposes and subsidy payments: (i) self-consumption (without subsidy); (ii) self-consumption (with subsidy); and (iii) electricity business. Towards this end, the number and percentage of buildings with rooftop solar PV profitability was calculated to determine how many buildings would be economically viable for rooftop solar PV adoption. In addition, the expected economic performance in terms of ROI and PP was evaluated for the buildings with profitability to determine the extent to which such buildings could profit from the installation of the rooftop solar PV system. Fig. 3 shows the number and percentage of buildings with rooftop solar PV profitability by adoption year based on the different installation purposes and subsidy payments. Figs. 4 and 5 show the box plot of the expected economic performance in terms of ROI and PP, respectively, for the buildings with profitability, by adoption year. For a better understanding of the analysis results, Figs. 6 and 7 were presented, showing the changes in the key factors affecting the rooftop solar PV profitability during the target period in Seoul. As shown in Fig. 3, the rooftop solar PV profitability in the Gangnam district differed

(14)

where i stands for the real discount rate; f stands for one of the followings: the inflation rate, electricity price growth rate, or REC price growth rate; and in stands for nominal interest rate.

• O&M cost: This study assumed the annual O&M cost to be 1% of the •

installation cost of the rooftop solar PV system based on the previous studies [59,60,63]. The O&M cost includes various costs for safety management, insurance, tax, cleaning, repair, and replacement that can occur during the operation of the rooftop solar PV system. System degradation rate: This study assumed that the technical performance of the rooftop solar PV system would be degraded by 20% during its useful life of 25 years, based on the previous studies and the actual PV panel data [63,65,66]. Since solar PV systems experience the decline in their technical performance over time due to various environment conditions exposed to moisture, oxidation, and dust, it is important to determine and assume the system 649

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 4. Box plot of the expected economic performance in terms of ROI by adoption year.

not only by adoption year but also by installation purpose and subsidy payment. The details are discussed below. First, the rooftop solar PV profitability for self-consumption without subsidy was the lowest compared to that for self-consumption with subsidy and for electricity business. If the solar PV system were installed on all the building rooftops in the Gangnam district for selfconsumption without subsidy, there would have been no building with profitability until 2013, but the percentage of buildings with profitability gradually would have gradually increased from 0.1% in 2014 to 57.1% in 2016. With a slight decrease in the installation cost of the rooftop solar PV system from US$2808 in 2013 to US$2340–2620 in 2014 per kW (refer to Fig. 6), only 0.1% of the buildings in the Gangnam district could have profited from rooftop solar PV adoption in 2014. The percentage of buildings with profitability substantially increased in 2015; as such, almost half of the buildings in the Gangnam district (i.e., 46.2% of the buildings) were able to profit from rooftop solar PV adoption for self-consumption without subsidy. Even though almost half of the buildings in the Gangnam district could profit from rooftop solar PV adoption for self-consumption without subsidy after 2015, their expected economic performance in terms of ROI and PP was low overall (refer to the box plot colored orange in Figs. 4 and 5). If the rooftop solar PV system were installed in

2015, the median values of the expected ROI and PP would have been 1.23 and 18.50 years, respectively, requiring a relatively long period to recover the investment cost. If the rooftop solar PV system were installed in 2016, the median value of the expected PP would have been shortened to 15.64 years, but no building would have been able to expect a PP of within 10 years. Therefore, it can be said that the subsidy from the government still plays an important role in promoting the rooftop solar PV system in the Gangnam district even if the majority of the buildings can profit from rooftop solar PV adoption without subsidy as of 2016. Second, the rooftop solar PV profitability for self-consumption with subsidy was improved compared to that for self-consumption without subsidy, showing steady growth from 2010 to 2016. If the solar PV system were installed on all the building rooftops in the Gangnam district for self-consumption with subsidy, there would have been no building with profitability until 2009 despite subsidy payment, but the percentage of buildings with profitability would have gradually increased from 0.1% in 2010 to 97.2% in 2016. As the installation cost of the rooftop solar PV system after subsidy payment significantly decreased from US$1460 in 2014 to US$852–927 in 2015 per kW (refer to Fig. 6), almost all the buildings in the Gangnam district (i.e., as many as 95.8% of buildings) could have profited from rooftop solar PV adoption

Fig. 5. Box plot of the expected economic performance in terms of PP by adoption year. 650

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 6. Key factors affecting the rooftop solar PV profitability for self-consumption purposes.

in 2015. Accordingly, the percentage of buildings with rooftop solar PV profitability for self-consumption with subsidy exceeded that of the buildings with profitability for electricity business after 2015 (i.e., 83.2% in 2015 and 91.0% in 2016). Even though the percentage of buildings with profitability gradually increased from 2010, their expected economic performance in terms of ROI and PP was relatively low until 2014 (refer to the box plot colored yellow in Figs. 4 and 5). Until 2014, more than half of the buildings with rooftop solar PV profitability could expect an ROI of only under 1.5 and a PP of over 15 years, requiring a relatively long period to recover the investment cost. From 2015, however, more than half of the buildings with profitability could expect an ROI of over 1.5 and a PP of within 15 years, showing a relatively high economic performance through the installation of the rooftop solar PV system. In 2016, more than half of the buildings with profitability could even expect a PP of within 10 years. Therefore, it can be said that the subsidy from the South Korean government was not sufficient to financially encourage the people in the Gangnam district to install the rooftop solar PV system until 2014, but the expected economic performance of the rooftop solar PV system improved after 2015 due to the installation cost reduction and well-structured subsidy scheme. Third, the rooftop solar PV profitability for electricity business

showed a pattern different from that for self-consumption, both with and without subsidy, showing a high economic potential throughout the target period. If the solar PV system were installed on all the building rooftops in the Gangnam district for electricity business purposes, more than 80% of the buildings in the Gangnam district could have profited from rooftop solar PV adoption during the whole target period (i.e., 2008–2016). During the period when the South Korean government adopted FIT (i.e., until 2011), the percentage of buildings with profitability steadily increased from 85.7 to 97.6% due to the high FIT from the South Korean government. Particularly, almost all the buildings in the Gangnam district (i.e., up to 97.6% of the buildings) could have profited from rooftop solar PV adoption in 2011 because the high FIT remained unchanged despite the considerable decline in the installation cost, as shown in Fig. 7 (i.e., from US$4736 in 2010 to US $3744 in 2011). During this whole period, more than half of the buildings in the Gangnam district could have expected an ROI of over 1.2 and a PP of within 12 years, showing a relatively high economic performance through the installation of the rooftop solar PV system (refer to the box plot colored gray in Figs. 4 and 5). Meanwhile, during the period when FIT was replaced by RPS (i.e., from 2012), the percentage of buildings with profitability decreased from 88.3% in 2012 to 80.01% in 2014 and then increased to 91.0% in

Fig. 7. Key factors affecting the rooftop solar PV profitability for electricity business purposes. 651

Applied Energy 232 (2018) 640–656

M. Lee et al.

rooftop solar PV system in terms of power capacity was compared to the power capacities of the maximum installable and actual installed solar PV systems in the Gangnam district. First, the maximum installable capacity of the rooftop solar PV system was estimated to be 256.19 MW in the Gangnam district. This indicates that if the solar PV system is installed on all the rooftops in the Gangnam district with the maximum installable system size considering the total rooftop area, it is possible to adopt the rooftop solar PV system of at most 256.19 MW. Second, the economic potential of the rooftop solar PV system in terms of power capacity was estimated to range from 0 to 249.26 MW in the Gangnam district depending on the installation purpose, subsidy payment, and adoption year. Among the three installation options (i.e., self-consumption without subsidy, self-consumption with subsidy, and electricity business), electricity business showed the highest economic potential throughout the target period (i.e., ranging from 215.22 to 249.26 MW), reaching 84–97% of the maximum installable capacity. Self-consumption with subsidy also showed a relatively high economic potential after 2015, with a power capacity of over 200 MW, but the economic potential was somewhat low, with a power capacity of under 100 MW before 2014. Meanwhile, self-consumption without subsidy showed a relatively low economic potential throughout the target period, with the highest power capacity of 123.61 MW in 2016. There was even no economic potential for installing the rooftop solar PV system for self-consumption without subsidy from 2008 to 2013. Third, the actual installed capacity of the solar PV system in the Gangnam district was far less than the economic potential estimated in this study, indicating that there remains a significant possibility of additional rooftop solar PV adoption. Regardless of the high economic potential of the rooftop solar PV system up to 235.30 MW, the actual installed capacity of the solar PV system in the Gangnam district was 6.46 MW as of 2016. This indicates that it is possible to install an additional rooftop solar PV system of up to 228.84 MW for electricity business purposes (refer to C in Fig. 8) in the Gangnam district when the economic profitability is considered (117.15 MW for self-consumption without subsidy (refer to A in Fig. 8) and 196.70 MW for self-consumption with subsidy (refer to B in Fig. 8)). As shown in Table 6 and Fig. 9, the economic potential of the rooftop solar PV system in terms of electricity generation was compared to the calculated electricity generation of the maximum installable and actual installed solar PV systems in the Gangnam district. First, the annual electricity generation of the maximum installable rooftop solar PV systems in the Gangnam district was estimated to be 223,042 MWh. This indicates that if the solar PV system is installed on all the rooftops in the Gangnam district with the maximum installable system size considering the total rooftop area, it is possible to produce at most 223,042 MWh of electricity in a year. Second, the annual economic potential of the rooftop solar PV system in terms of electricity generation was estimated to range from 0 to 219,356 MWh in the Gangnam district depending on the installation

2016 due to the change in the SMP and the REC price, as shown in Fig. 7. From 2012 to 2014, the percentage of buildings with profitability constantly decreased along with the decline in the SMP and the REC price as the installation cost remained almost constant during this period. In 2015, however, the percentage of buildings with profitability increased despite the decline in the SMP and the REC price due to the considerable drop in the installation cost, as shown in Fig. 7 (i.e., from US$2340 ∼ 2620 in 2014 to US$1638–2106 in 2015). Even though more than 80% of the buildings in the Gangnam district could have profited from rooftop solar PV adoption for electricity business with RPS, their expected economic performance in terms of ROI and PP until 2014 was relatively lower than that of rooftop solar PV adoption for electricity business with FIT (refer to the box plot colored gray in Figs. 4 and 5). Until 2014, the majority of the buildings with rooftop solar PV profitability could have expected an ROI of only under 1.3 and a PP of over 14 years, requiring a relatively long period to recover the investment cost. Moreover, no buildings with profitability could have expected a PP of within 10 years until 2014 through the electricity business. More than half of the buildings with profitability, however, could have expected a ROI of over 1.3 and a PP of within 14 years in 2015, and more than 25% of the buildings with profitability could have expected a PP of within 10 years in 2016 due to the steady decline in the installation cost and the increase in the REC price, as shown in Fig. 7 (i.e., from US$4736 in 2010 to US$3744 in 2011). Therefore, it can be said that the electricity business with FIT from 2008 to 2011 was designed to guarantee profitability from the installation of the solar PV system with a high FIT rate whereas the profitability of the solar PV system for electricity business with RPS from 2012 to 2016 was highly dependent on the market conditions (i.e., installation cost, SMP, and REC price) in South Korea. 3.2. Analysis of the economic potential of the rooftop solar PV system The economic potential of the rooftop solar PV system in the Gangnam district was analyzed by adoption year from 2008 to 2016, based on the different installation purposes and subsidy payments: (i) self-consumption (without subsidy); (ii) self-consumption (with subsidy); and (iii) electricity business. Towards this end, the economic potential of the rooftop solar PV system estimated in this study was compared to the maximum installable and actual installed solar PV systems in the Gangnam district in terms of two aspects: (i) power capacity; and (ii) electricity generation. The data on the actual installed capacity of the solar PV systems in the Gangnam district were obtained from the Seoul metropolitan government and KNREC. Table 6 and Figs. 8 and 9 show the comparative analysis results of the economic potential of the rooftop solar PV system with the maximum installable and actual installed solar PV systems in the Gangnam district by adoption year in terms of power capacity and electricity generation. As shown in Table 6 and Fig. 8, the economic potential of the

Table 6 Comparative analysis results of the economic potential of the rooftop solar PV system by adoption year. Classification Power capacity (MW)

Maximum installable capacity Economic Self-consumption potential Electricity business Actual installed capacity

Electricity generation (MWh)

without subsidy with subsidy

Electricity from maximum installable system Economic Self-consumption without potential subsidy with subsidy Electricity business Electricity from actual installed system

2008

2009

2010

2011

2012

2013

2014

2015

2016

256.19 –

256.19 –

256.19 –

256.19 –

256.19 –

256.19 –

256.19 0.09

256.19 82.66

256.19 123.61

– 226.28 –

– 228.94 –

0.06 241.57 –

2.40 249.26 –

28.56 233.76 –

65.52 230.64 –

97.04 215.22 –

200.47 215.63 –

203.16 235.30 6.46

223,042 –

223,042 –

223,042 –

223,042 –

223,042 –

223,042 –

223,042 89

223,042 74,884

223,042 113,688

– 204,251 –

– 206,077 –

59 215,151 –

2,334 219,356 –

25,729 210,189 –

61,797 208,042 –

89,316 196,173 –

176,109 195,212 –

177,556 210,589 5,624

652

Applied Energy 232 (2018) 640–656

M. Lee et al.

Fig. 8. Comparative analysis results of the economic potential of the rooftop solar PV system in terms of power capacity.

purpose, subsidy payment, and adoption year. Among the three installation options (i.e., self-consumption without subsidy, self-consumption with subsidy, and electricity business), electricity business showed the highest economic potential throughout the target period (i.e., ranging from 196,173 to 219,356 MWh), reaching 88–98% of the maximum installable capacity. Self-consumption with subsidy also showed a relatively high economic potential after 2015, being able to produce over 150,000 MWh of electricity, but the economic potential was somewhat low with the expected electricity generation of under 100,000 MWh before 2014. Meanwhile, self-consumption without subsidy showed a relatively low economic potential throughout the target period being able to produce at most 113,688 MWh of electricity in 2016. There was even no economic potential for installing the rooftop solar PV system for self-consumption without subsidy from 2008 to 2013. Third, the electricity generation of the actual installed solar PV system in the Gangnam district as of 2016 was calculated to be 5624 MWh based on the average technical potential of the rooftop solar PV system for a building per unit area in the Gangnam district (i.e., 158.47 kWh/m2). It was analyzed to be far less than the economic potential estimated in this study, indicating that there remains a significant possibility of electricity supply from additional rooftop solar PV adoption. At most additional 204,965 MWh of electricity can be supplied to the Gangnam district by the rooftop solar PV systems for

electricity business (refer to C in Fig. 9) when the economic profitability is considered (171,931 MWh for self-consumption without subsidy (refer to A in Fig. 9) and 108,063 MWh for self-consumption with subsidy (refer to B in Fig. 9)). Fourth, the economic potential of the rooftop solar PV system and the electricity generation of the maximum installable and actual installed solar PV systems were compared with the total electricity consumption and the average electricity consumption per household in the Gangnam district as of 2016 (refer to Table 7). As shown in Table 7, with the annual electricity generation of the actual installed solar PV systems (i.e., 5,624 MWh), it was analyzed that it could supply only 0.12% of the annual total electricity consumption in the Gangnam district (i.e., 4,698,163 MWh as of 2016) [23]. This amount of electricity can cover only 1821 households (i.e., 0.77% of the total households in the Gangnam district) according to the annual average electricity consumption per household in the Gangnam district (i.e., 3088.32 kWh) [43]. The self-sufficiency rate can be increased up to 3.78% based on the economic potential of the rooftop solar PV system for self-consumption with subsidy (i.e., 177,556 MWh) (2.42% for selfconsumption without subsidy (i.e., 113,688 MWh)). This amount of electricity can cover 57,493 households (i.e., 24.19% of all the households) in the Gangnam district (36,812 households (i.e., 15.49% of all the households) for self-consumption without subsidy). The self-sufficiency rate can be further increased to 4.48% based on the economic

Fig. 9. Comparative analysis results of the economic potential of the rooftop solar PV system in terms of electricity generation. 653

Applied Energy 232 (2018) 640–656

M. Lee et al.

Table 7 Comparative analysis results of the electricity supply capability in 2016. Classification Electricity from maximum installable system Economic potential

Self-consumption Electricity business

without subsidy with subsidy

Electricity from actual installed system

potential of the rooftop solar PV system for electricity business (i.e., 210,589 MWh), and if the solar PV system is installed on all the rooftops in the Gangnam district with the maximum installable system size, it can supply 4.75% of the annual total electricity consumption in the Gangnam district, with 223,042 MWh of electricity. Accordingly, the annual economic potential of the rooftop solar PV system for electricity business can cover 68,189 households (i.e., 28.70% of all the households) while the expected electricity generation of the maximum installable rooftop solar PV systems can cover 72,221 households (i.e., 30.39% of all the households) in the Gangnam district.

% of households

4.75%

72,221

30.39%

2.42% 3.78% 4.48%

36,812 57,493 68,189

15.49% 24.19% 28.70%

0.13%

1821

0.77%

Based on the calculated rooftop solar PV profitability, the economic potential of the rooftop solar PV system in the Gangnam district was estimated for the different installation purposes and subsidy payments by adoption year in terms of two aspects: (i) power capacity; and (ii) electricity generation.

• Power capacity: The economic potential of the rooftop solar PV

This study developed a method for estimating the economic potential of the rooftop solar PV system in a region based on a novel bottom-up approach. Towards this end, the rooftop solar PV profitability was calculated using various methods, such as Hillshade analysis and LCC analysis, along with 21,681 actual building data in the Gangnam district. Most importantly, the market conditions (i.e., installation cost, policies and support schemes, and electricity prices), which vary over time, were taken into account in estimating the economic potential of the rooftop solar PV system to consider the market dynamics by adoption year from 2008 to 2016. As a result, the rooftop solar PV profitability of the total of 21,681 buildings in the Gangnam district were calculated for the different installation purposes and subsidy payments by adoption year, as follows.

•

• For self-consumption (without subsidy): No buildings were economic-

•

No. of households

conditions (i.e., installation cost, SMP, and REC price).

4. Conclusion

•

Self-sufficiency rate

ally viable for the installation of the rooftop solar PV system until 2013 whereas almost half of the buildings could profit from rooftop solar PV adoption after 2015 even without subsidy. It was analyzed, however, that the subsidy from the government would still play an important role in promoting the rooftop solar PV system because the expected economic performance of the buildings with rooftop solar PV profitability was relatively low (i.e., could not expect a PP of within 10 years) even with the installation cost reduction. For self-consumption (with subsidy): No buildings were economically viable for the installation of the rooftop solar PV system until 2009, whereas more than half of the buildings could profit from rooftop solar PV adoption after 2014. It was analyzed, however, that until 2014, the subsidy from the South Korean government was not sufficient to financially encourage people to install the rooftop solar PV system because the expected economic performance of the buildings with rooftop solar PV profitability was relatively low (i.e., could not expect a PP of within 10 years). After 2015, the expected economic performance of the rooftop solar PV system improved due to the installation cost reduction and well-structured subsidy scheme. For electricity business: More than 80% of the buildings in the Gangnam district could have profited from rooftop solar PV adoption from 2008 to 2016. The expected economic performance turned out to be different, however, depending on the applicable policy: (i) FIT (2008–2011); and (ii) RPS (2012–2016). It was analyzed that the electricity business with FIT was designed to guarantee rooftop solar PV profitability with a high FIT rate, whereas the profitability for electricity business with RPS highly depends on the market

system for electricity business, self-consumption with subsidy, and self-consumption without subsidy was analyzed to reach up to 97% (i.e., 249.26 MW), 79% (i.e., 203.16 MW), and 48% (i.e., 123.61 MW) of the maximum installable capacity (i.e., 256.19 MW), respectively. Meanwhile, the economic potential for electricity business, self-consumption with subsidy, and self-consumption without subsidy was analyzed to be 36, 31, and 19-fold compared to the actual installed capacity (i.e., 6.46 MW), respectively, as of 2016. Electricity generation: It was analyzed that only 0.12% of the annual total electricity consumption in the Gangnam district (i.e., 4698,163 MWh) could be supplied from the annual electricity generation of the actual installed solar PV system (i.e., 5624 MWh) as of 2016. If all the economically viable rooftop solar PV systems are installed in the Gangnam district for electricity business, self-consumption with subsidy, and self-consumption without subsidy, however, they were analyzed to be capable of supplying 4.48% (i.e., 210,589 MWh), 3.78% (i.e., 177,556 MWh), and 2.42% (i.e., 113,688 MWh) of the annual total electricity consumption in the Gangnam district, respectively.

This study has significant implications and contributions in that it (i) took one step farther towards the rooftop solar PV potential estimation process, from the technical potential to the economic potential; (ii) calculated the technical and economic performance of the rooftop solar PV system for each building within the regional scope based on the bottom-up approach (i.e., spatial diversity); and (iii) considered the changes in the market conditions of the solar PV industry by adoption year (i.e., temporal diversity). Using the methodology proposed in this study, it is possible to provide an initial estimate of the economic viability of the rooftop solar PV generation at a specific location. It can also be useful for screening the possibility for future applications and in assessing the impact of various market conditions, including installation cost, policies and support schemes, and electricity price, on the economic viability of the rooftop solar PV system. This type of estimate is highly different from just conducting an economic assessment of a single or group of rooftop solar PV system, as it provides an important insight into the urban energy planning from the macroscopic perspective. Even with the aforementioned points of superiority, this study has some space left for future improvements, as follows: (i) as this study was conducted on a macro scale for an entire district, not all the details of each building required for installing the rooftop solar PV system (e.g., detailed rooftop characteristics and installation conditions) could be considered; (ii) the economic potential of the rooftop solar PV system estimated in this study was based on the historical data from 654

Applied Energy 232 (2018) 640–656

M. Lee et al.

2008 to 2016, which leaves a potential for extending the applicability to the latest data available; and (iii) the proposed method was applied only to the Gangnam district, which leaves a potential for extending the regional boundary to the city (i.e., Seoul) or national level (i.e., South Korea), and even to other countries. To address these issues, some efforts can be made in the future researches to generalize or extend the methods developed and proposed in this study. In particular, the algorithm for determining the system size and business model of the rooftop solar PV system depending on the building characteristics (i.e., building type, total rooftop area, and electricity consumption) can be made into a generalized model by considering the global emerging market trends and applicable policies, instead of the local or old ones.

[19] Assouline D, Mohajeri N, Scartezzini J-L. Quantifying rooftop photovoltaic solar energy potential: A machine learning approach. Sol Energy 2017;141:278–96. https://doi.org/10.1016/j.solener.2016.11.045. [20] Assouline D, Mohajeri N, Scartezzini JL. Large-scale rooftop solar photovoltaic technical potential estimation using Random Forests. Appl Energy 2018;217:189–211. https://doi.org/10.1016/j.apenergy.2018.02.118. [21] Sigrin B, Gleason M, Preus R, Baring-gould I, Margolis R. The Distributed Generation Market Demand Model (dGen): documentation. Colorado (CO); 2016. [22] Cole W, Lewis H, Sigrin B, Margolis R. Interactions of rooftop PV deployment with the capacity expansion of the bulk power system. Appl Energy 2016;168:473–81. https://doi.org/10.1016/j.apenergy.2016.02.004. [23] Seoul Metropolitan Government. Energy White Paper. Seoul; 2016. [24] Muaafa M, Adjali I, Bean P, Fuentes R, Kimbrough SO, Murphy FH. Can adoption of rooftop solar panels trigger a utility death spiral? A tale of two U.S. cities. Energy Res Soc Sci 2017;34:154–62. https://doi.org/10.1016/j.erss.2017.06.041. [25] Palmer J, Sorda G, Madlener R. Modeling the diffusion of residential photovoltaic systems in Italy: An agent-based simulation. Technol Forecast Soc Change 2015;99:106–31. https://doi.org/10.1016/j.techfore.2015.06.011. [26] MOLIT. Electronic Architectural administration Information System (E-AIS). Minist Land, Infrastructure, Transp. < http://www.eais.go.kr/ > ; n.d. [accessed 22. 10.17]. [27] Korea Energy Economics Institute (KEEI). Energy Consumption Survey 2014; 2015. [28] Spatial Information Industry Promotion Institute (SPACEN). < http://www.spacen. or.kr/main.do > ; n.d. [accessed 05.10.17]. [29] World Radiation Data Centre (WRDC). < http://wrdc.mgo.rssi.ru/ > ; n.d. [accessed 05.10.17]. [30] Green MA. Developments in crystalline silicon solar cells. Sol. Cell Mater., John Wiley & Sons, Ltd; 2014, p. 65–84. < https://doi.org/10.1002/9781118695784. ch4 > . [31] Green MA, Hishikawa Y, Warta W, Dunlop ED, Levi DH, Hohl-Ebinger J, et al. Solar cell efficiency tables (version 50). Prog Photovoltaics Res Appl 2017;25:668–76. https://doi.org/10.1002/pip.2909. [32] Ardani K, Margolis R. 2010 Solar Technologies Market Report. Colorado (CO); 2011. doi: DOE/GO-102011-3318. [33] Woodhouse M, Jones-Albertus R, Feldman D, Fu R, Horowitz K, Chung D, et al. On the Path to SunShot: The Role of Advancements in Solar Photovoltaic Efficiency, Reliability, and Costs. Colorado (CO); 2016. doi: NREL/TP-6A20-65872. [34] Barbose G, Darghouth N. Tracking the Sun IX; 2016. [35] Park C, Kim KW, Seo J, Song J, Jeong D. National Survey Report of PV Power Applications in Korea 2015. St. Ursen; 2015. [36] HAEZOOM. < http://haezoom.com/ > ; n.d. [accessed 05.10.17]. [37] Korea New & Renewable Energy Center (KNREC). < http://www.knrec.or.kr/ knrec/13/KNREC130110.asp?idx=554 > ; n.d. [accessed 05.10.17]. [38] Seoul Metropolitan Government. Energy White Paper. Seoul; 2013. [39] Seoul Metropolitan Government. Energy White Paper. Seoul; 2014. [40] Seoul Metropolitan Government. Energy White Paper. Seoul; 2015. [41] Renewable Energy Certificate (REC) Trading System. < http://rec.kmos.kr/index. do > ; n.d. [accessed 08.01.18]. [42] Lee M, Hong T, Yoo H, Koo C, Kim J, Jeong K, et al. Establishment of a base price for the Solar Renewable Energy Credit (SREC) from the perspective of residents and state governments in the United States. Renew Sustain Energy Rev 2017;75:1066–80. https://doi.org/10.1016/j.rser.2016.11.086. [43] Power Big data Center. < http://home.kepco.co.kr/kepco/BD/bigData/main/ bigDataMain.do > ; n.d. [accessed 15.01.18]. [44] Electric Power Statistics Information System (EPSIS). < http://epsis.kpx.or.kr/ epsisnew/ > ; n.d. [accessed 05.10.17]. [45] Hamwi M, Lizarralde I. A review of business models towards service-oriented electricity systems. Procedia CIRP Elsevier; 2017. p. 109–14. https://doi.org/10. 1016/j.procir.2017.03.032. [46] Korea Electric Power Corporation (KEPCO). < http://home.kepco.co.kr/kepco/ main.do > ; n.d. [accessed 11.01.18]. [47] Information on distributed generation. < http://home.kepco.co.kr/kepco/CO/H/ htmlView/COHAHP002.do?menuCd=FN040603 > ; n.d. [accessed 26.01.18]. [48] Seoul Solarmap. < http://solarmap.seoul.go.kr/index.do > ; n.d. [accessed 20. 01.18]. [49] Korea Astronomy & Space science Institute (KASI). < https://astro.kasi.re.kr:444/ index > ; n.d. [accessed 05.10.17]. [50] Environmental Systems Research Institute (ESRI). < https://www.esri.com/en-us/ home > ; n.d. [accessed 05.10.17]. [51] Schallenberg-Rodríguez J. Photovoltaic techno-economical potential on roofs in regions and islands: The case of the Canary Islands. Methodological review and methodology proposal. Renew Sustain Energy Rev 2013;20:219–39. https://doi. org/10.1016/j.rser.2012.11.078. [52] Fouad MM, Shihata LA, Morgan EI. An integrated review of factors influencing the performance of photovoltaic panels. Renew Sustain Energy Rev 2017;80:1499–511. https://doi.org/10.1016/j.rser.2017.05.141. [53] Reich NH, Mueller B, Armbruster A, van Sark WGJHM, Kiefer K, Reise C. Performance ratio revisited: is PR > 90% realistic? Prog Photovoltaics Res Appl 2012;20:717–26. https://doi.org/10.1002/pip.1219. [54] Khalid AM, Mitra I, Warmuth W, Schacht V. Performance ratio – Crucial parameter for grid connected PV plants. Renew Sustain Energy Rev 2016;65:1139–58. https:// doi.org/10.1016/j.rser.2016.07.066. [55] PVWatts Calculator. < http://pvwatts.nrel.gov/index.php > ; n.d. [accessed 05. 10.17]. [56] Dobos AP. PVWatts Version 5 Manual (NREL/TP-6A20-62641). Colorado (CO); 2014.

Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT; Ministry of Science and ICT) (NRF-2018R1A2A1A19020868). References [1] Pesaran HAM, Huy PD, Ramachandaramurthy VK. A review of the optimal allocation of distributed generation: Objectives, constraints, methods, and algorithms. Renew Sustain Energy Rev 2017;75:293–312. https://doi.org/10.1016/j.rser.2016. 10.071. [2] Camilo HF, Udaeta MEM, Veiga Gimenes AL, Grimoni JAB. Assessment of photovoltaic distributed generation – Issues of grid connected systems through the consumer side applied to a case study of Brazil. Renew Sustain Energy Rev 2017;71:712–9. https://doi.org/10.1016/j.rser.2016.12.099. [3] Byrne J, Taminiau J, Kurdgelashvili L, Kim KN. A review of the solar city concept and methods to assess rooftop solar electric potential, with an illustrative application to the city of Seoul. Renew Sustain Energy Rev 2015;41:830–44. https://doi. org/10.1016/j.rser.2014.08.023. [4] Suomalainen K, Wang V, Sharp B. Rooftop solar potential based on LiDAR data: Bottom-up assessment at neighbourhood level. Renew Energy 2017;111:463–75. https://doi.org/10.1016/j.renene.2017.04.025. [5] Korea New and Renewable Energy Center (KNREC). 2016 New & Renewable Energy White Paper; 2016. [6] Brown A, Beiter P, Heimiller D, Davidson C, Denholm P, Melius J, et al. Estimating renewable energy economic potential in the United States: methodology and initial results. Colorado (CO); 2015. [7] Gagnon P, Margolis R, Melius J, Phillips C, Elmore R. Rooftop solar photovolatic technical potential in the united states: A detailed assessment. Colorado (CO); 2016. [8] Hong T, Lee M, Koo C, Jeong K, Kim J. Development of a method for estimating the rooftop solar photovoltaic (PV) potential by analyzing the available rooftop area using Hillshade analysis. Appl Energy 2017;194:320–32. https://doi.org/10.1016/j. apenergy.2016.07.001. [9] Freitas S, Catita C, Redweik P, Brito MC. Modelling solar potential in the urban environment: State-of-the-art review. Renew Sustain Energy Rev 2015;41:915–31. https://doi.org/10.1016/j.rser.2014.08.060. [10] Izquierdo S, Rodrigues M, Fueyo N. A method for estimating the geographical distribution of the available roof surface area for large-scale photovoltaic energypotential evaluations. Sol Energy 2008;82:929–39. https://doi.org/10.1016/j. solener.2008.03.007. [11] Peng J, Lu L. Investigation on the development potential of rooftop PV system in Hong Kong and its environmental benefits. Renew Sustain Energy Rev 2013;27:149–62. https://doi.org/10.1016/j.rser.2013.06.030. [12] Kurdgelashvili L, Li J, Shih CH, Attia B. Estimating technical potential for rooftop photovoltaics in California, Arizona and New Jersey. Renew Energy 2016;95:286–302. https://doi.org/10.1016/j.renene.2016.03.105. [13] Santos T, Gomes N, Freire S, Brito MC, Santos L, Tenedório JA. Applications of solar mapping in the urban environment. Appl Geogr 2014;51:48–57. https://doi.org/10. 1016/j.apgeog.2014.03.008. [14] Jakubiec JA, Reinhart CF. A method for predicting city-wide electricity gains from photovoltaic panels based on LiDAR and GIS data combined with hourly Daysim simulations. Sol Energy 2013;93:127–43. https://doi.org/10.1016/j.solener.2013. 03.022. [15] Ko L, Wang JC, Chen CY, Tsai HY. Evaluation of the development potential of rooftop solar photovoltaic in Taiwan. Renew Energy 2015;76:582–95. https://doi. org/10.1016/j.renene.2014.11.077. [16] Strzalka A, Alam N, Duminil E, Coors V, Eicker U. Large scale integration of photovoltaics in cities. Appl Energy 2012;93:413–21. https://doi.org/10.1016/j. apenergy.2011.12.033. [17] Buffat R, Grassi S, Raubal M. A scalable method for estimating rooftop solar irradiation potential over large regions. Appl Energy 2018;216:389–401. https://doi. org/10.1016/j.apenergy.2018.02.008. [18] Margolis R, Gagnon P, Melius J, Phillips C, Elmore R. Using GIS-based methods and lidar data to estimate rooftop solar technical potential in US cities. Environ Res Lett 2017;12:074013https://doi.org/10.1088/1748-9326/aa7225.

655

Applied Energy 232 (2018) 640–656

M. Lee et al. [57] Sommerfeldt N, Madani H. Revisiting the techno-economic analysis process for building-mounted, grid-connected solar photovoltaic systems: Part one – Review. Renew Sustain Energy Rev 2017;74:1379–93. https://doi.org/10.1016/j.rser.2016. 11.232. [58] Sommerfeldt N, Madani H. Revisiting the techno-economic analysis process for building-mounted, grid-connected solar photovoltaic systems: Part two Application. Renew Sustain Energy Rev 2017;74:1394–404. https://doi.org/10. 1016/j.rser.2017.03.010. [59] Lee M, Hong T, Koo C, Kim C-J. A break-even analysis and impact analysis of residential solar photovoltaic systems considering state solar incentives. Technol Econ Dev Econ 2017;4913:1–25. https://doi.org/10.3846/20294913.2016.1212745. [60] Lee M, Hong T, Koo C. An economic impact analysis of state solar incentives for improving financial performance of residential solar photovoltaic systems in the United States. Renew Sustain Energy Rev 2016;58:590–607. https://doi.org/10. 1016/j.rser.2015.12.297. [61] Rai V, Robinson SA. Agent-based modeling of energy technology adoption: Empirical integration of social, behavioral, economic, and environmental factors.

[62] [63] [64] [65] [66]

656

Environ Model Softw 2015;70:163–77. https://doi.org/10.1016/j.envsoft.2015.04. 014. Robinson SA, Rai V. Determinants of spatio-temporal patterns of energy technology adoption: An agent-based modeling approach. Appl Energy 2015;151:273–84. https://doi.org/10.1016/j.apenergy.2015.04.071. Branker K, Pathak MJM, Pearce JM. A review of solar photovoltaic levelized cost of electricity. Renew Sustain Energy Rev 2011;15:4470–82. https://doi.org/10.1016/ j.rser.2011.07.104. Economic Statistics System (ECOS) of the Bank of Korea. < http://ecos.bok.or.kr/ > ; n.d. [accessed 05.10.17]. Lee M, Hong T, Kang H, Koo C. Development of an integrated multi-objective optimization model for determining the optimal solar incentive design. Int J Energy Res 2017;41:1749–66. https://doi.org/10.1002/er.3744. López Prol J, Steininger KW. Photovoltaic self-consumption regulation in Spain: Profitability analysis and alternative regulation schemes. Energy Policy 2017;108:742–54. https://doi.org/10.1016/j.enpol.2017.06.019.