Carbon-dioxide mitigation in the residential building sector: A household scale-based assessment

Energy Conversion and Management 198 (2019) 111915

Contents lists available at ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

Carbon-dioxide mitigation in the residential building sector: A household scale-based assessment Minda Maa,b,

⁎,1

, Xin Mac, Weiguang Caia,b,

⁎,2

T

, Wei Caid

a

International Research Center for Sustainable Built Environment, School of Management Science and Real Estate, Chongqing University, Chongqing, 400045, PR China Special Committee of Building Energy Consumption Statistics, China Association of Building Energy Efficiency, Beijing 100835, PR China c School of Science, Southwest University of Science and Technology, Mianyang 621010, PR China d College of Engineering and Technology, Southwest University, Chongqing 400715, PR China b

ARTICLE INFO

ABSTRACT

Keywords: CO2 mitigation CO2 intensity Residential building Decomposition analysis Ridge regression Emission mitigation strategy

Carbon-dioxide mitigation in residential building sector (CMRBS) has become critical for China in achieving its emission mitigation goal in the “Post Paris” period with the growing demand for household energy service in residential buildings. This is the first paper to investigate the factors that can mitigate carbon-dioxide (CO2) intensity and further assess CMRBS in China based on a household scale via decomposition analysis. The core findings of this study reveal that: (1) Three types of housing economic indicators and the final emission factor significantly contributed to the decrease in CO2 intensity in the residential building sector. In addition, the CMRBS from 2001 to 2016 was 1816.99 MtCO2, and the average mitigation intensity during this period was 266.12 kgCO2·(household·year)−1. (2) Ridge regression indicated that the robustness of the decomposition approach was sufficient for achieving reliable results for the decomposition analysis and CMRBS assessment. (3) The energy-conservation and emission-mitigation strategy caused CMRBS to effectively increase and is the key to promoting a more significant emission mitigation in the future. Overall, this paper covers the CMRBS assessment gap in China, and the proposed assessment model can be regarded as a reference for other countries and cities for measuring the retrospective CO2 mitigation effect in residential buildings.

1. Introduction IPCCa has declared that effective mitigation measures for the carbondioxide (CO2) emissions in the residential building sector are significant to suppress critical global warming trends [1] since residential buildings are responsible for nearly 20% of the final energy demand, which causes over 22% of CO2 emissions worldwide [2]. In developed countries (e.g., the United States and European Union countries), the emission feature shows that the residential building sector will achieve its emission peak soon [3,4]. On the contrary, regarding the developing country such as China, its residential building sector is facing a growing demand for household energy service. Therefore, large amounts of primary energy (e.g., coal and natural gas) and secondary energy (e.g., electrical power) have been consumed, which leads to large emission of CO2 in residential

buildings [5,6]. It has been reported that the CO2 which is released from the residential building sector has grown rapidly with a 6.57% increase per year over the past decade in China, and emissions measured at over 1.2 billion tons of CO2 in 2016 [7,8]. Reducing the large CO2 emissions from the residential building sector is critical for China in achieving its 2030 emission peak target, many scholars agree that the emission mitigation potential from residential buildings is considerable. As a typical case, McNeil et al. forecasted that effective mitigation measures for the CO2 emissions of residential buildings will contribute approximately 30% to the 2030 emission peak target in China [9], and this viewpoint has been further verified by their latest study [10]. Based on the efforts made by [9,11], Tan et al. assessed the contribution of the CO2 mitigation potential to emission peak goals in the Chinese residential building sector, and they believed that the

Abbreviations: C-BEED, China Building Energy and Emission Database; CBEM, China Building Energy Model; CMRBS, Carbon mitigation in residential building sector; CNY, Chinese Yuan; ECEM, Energy-conservation and emission-mitigation; EKC, Environmental Kuznets curve; FYP, Five-Year Plan; GDP, Gross domestic product; LMDI, Log-Mean Divisia index; Mtce, Million tons of standard coal equivalent; MtCO2, Million tons of carbon dioxide; VIF, Variance inflation factor ⁎ Corresponding authors at: School of Management Science and Real Estate, Chongqing University, Chongqing 400045, PR China. E-mail addresses: [email protected] (M. Ma), [email protected] (X. Ma), [email protected] (W. Cai), [email protected] (W. Cai). URL: https://scholar.google.com/citations?user=Gw9QUFIAAAAJhl=en (W. Cai). 1 https://scholar.google.com/citations?user=240qUyIAAAAJ&hl=en 2 https://scholar.google.com/citations?user=mcd-ggIAAAAJhl=en a Intergovernmental Panel on Climate Change (IPCC). https://doi.org/10.1016/j.enconman.2019.111915 Received 4 June 2019; Received in revised form 2 August 2019; Accepted 3 August 2019 0196-8904/ © 2019 Elsevier Ltd. All rights reserved.

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Nomenclature C c ch

cp d E e F H I i K P

CO2 released from residential buildings CO2 emission per floor space in residential buildings CO2 emission per household in residential buildings (i.e., CO2 intensity in residential buildings) CO2 emission per capita in residential buildings Housing purchasing power Energy consumption in residential buildings Energy consumption per floor space in residential buildings Floor spaces of residential buildings Amount of households Income of households Per capita income Final emission factor of residential buildings Population size

Greek letter

ch |0 chd che chi chK ch p chr chS

turning point of the emission peak will appear before 2030 (approximately in 2027) in the best emission mitigation scenario [12]. Nevertheless, regarding the retrospective CO2 mitigation volume, it has been barely investigated and assessed [13], especially concerning the mitigation indicator of CO2 intensityb, which is required to be preferentially analyzed in the emission peak scenario [14]. Therefore, three questions are proposed for the Chinese residential building sector, as shown below.

T

ch changes during Period T Effect of d onch Effect of e onch Effect of i onch Effect of K onch Effect of p onch Effect of r onch Effect of S onch Ridge parameter

CMRBS. Regarding Section 5, it covers three parts: Section 5.1 presents a robustness test to verify the reliability of the decomposition and assessment results; Section 5.2 reveals what leads to similar emission mitigation effects at different emission scales; Section 5.3 retrospects the ECEM strategy of the residential building sector and relevant policy implications are put forward. Section 6 focuses on conclusions, including core findings and upcoming studies.

• Q1: Are the intensity and total values of CO emissions mitigated in the retrospective phase ? • Q2: What leads to emission mitigation if it does exist? • Q3: How should the future mitigation effect be strengthened to c

Population size per household Housing price Housing price-to-income ratio Household age structure

p Pr r S

2

2. Literature review Currently, studies on CO2 mitigation in the building sector can be mainly summarized in two steps: a. the data collection on energy consumption and CO2 emissions, and b. the assessment model of CO2 mitigation. The first step is the basis for accessing the credible CO2 emission data and the second step is the key to assessing the specific volume of CO2 mitigation for evaluating the ECEM strategy. Regarding the case area of China, two representative groups assessed the timeseries data on energy and emission in the building sector (see Fig. 1) during the past decade to cover the gap of the unfinished data statistical work, which is officially conducted by MOHURDd [17,18]. As illustrated in Fig. 1, China Building Energy & Emission Database (C-BEED) and China Building Energy Model (CBEM) are two major databases used to describe the time-series data on energy and emission in the Chinese building sector [19]. Although the two databases assessed the energy and emission data via different approaches (C-BEED: up-bottom approach; CBEM: bottom-up approach), the annual results from C-BEED and CBEM are quite close (see each diagonal line of Fig. 1). For the major emission sectors (e.g., transportation and industry sectors), decomposition analysis has been extensively adopted to investigate the impacts of technical, economic, and social factors on CO2 emissions and to further assess the CO2 mitigation [20,21]. Decomposition analysis is mainly divided into two branches: structure decomposition analysis and index decomposition analysis. Compared with structure decomposition analysis, index decomposition analysis (see Fig. A2), with its typical version [e.g., Log-Mean Divisia index (LMDI)] has the advantage of an easy calculation process (e.g., no need to rely on input–output model) [22], reliable decomposition results (e.g., no residual exists in the decomposition process) [23]. Table 1 summarizes some of the typical literature on CO2 mitigation assessment via decomposition analysis. The approaches to calculating the CO2 emissions of the Chinese building sector and assessing CO2 mitigation in major emission sectors have been introduced above. Through the review, some issues should be addressed as follows:

achieve China’s 2030 emission peak?

To answer the questions above, this paper first investigates the factors that can mitigate CO2 intensity and further assesses the CO2 mitigation in residential building sector (i.e., CMRBS, which includes both the intensity and total values) of China from 2000 to 2016 via the decomposition analysis. A robustness test is subsequently conducted via the ridge regression to verify the reliability of the decomposition and CMRBS results. In addition, the environmental Kuznets curve (EKC) on the CO2 emission feature is applied to explain similar emission mitigation effects for different emission scales in 2000–2016. Furthermore, the strategy for the energy-conservation and emission-mitigation (ECEM) of residential buildings is retrospected to explore policy patterns to achieve more emission mitigation effects in the future. The most significant contribution of this paper is to assess both the intensity and total values of CMRBS based on a household scale. To date, no studies on such a topic have been performed in China to the best of authors’ knowledge. Some of the recent literature have already reported cases on CO2 mitigation assessment of the building sector. However, their target is primarily focused on the CO2 mitigation of commercial buildings [15], or their assessment model follows the up-bottom approach which has yet to consider the effect of household scale on CMRBS and only focuses on the assessment level of energy savings instead of CO2 mitigation [16]. The above limitations will be further stated in the literature review. The remainder of this paper is conducted as follows: Section 2 states the literature review. Section 3 presents the materials and methods, which include the analytical approach of the CMRBS assessment model, variable definition, and data collection. Section 4 illustrates the decomposition result of the CO2 intensity and the assessment result of b Usually expressed as CO2 emission per household. Besides, CMRBS intensity shown in this paper represents CO2 mitigation per household in the residential building sector if not specified. c E.g., since the 21st century.

d

2

Ministry of Housing and Urban-Rural Development (MOHURD).

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. 1. Energy consumption and CO2 emissions of the Chinese building sector (2010–2016). Table 1 Typical literature on CO2 mitigation assessment in major emission sectors. Literature

Target

Scope

Period

Methods

Wang, Ang [24] Dong, Jiang [25] Achour and Belloumi [26] Wang, Zhang [27] Zhang, Zhao [28] Lin and Liu [29] Ma, Cai [15]

CO2 mitigation CO2 mitigation Energy savings CO2 mitigation CO2 mitigation CO2 mitigation CO2 mitigation

Global multiple emission sectors National emissions of 110 countries Transportation sector of Tunisia Transportation sector of China Industry sector of China Construction industry sector of China Commercial building sector of China

2000–2009 1980–2015 1985–2014 1985–2009 1993–2014 1991–2010 2000–2015

Structure decomposition analysis based on input–output model Index decomposition analysis based on LMDI approach Index decomposition analysis based on LMDI approach Index decomposition analysis based on LMDI approach Index decomposition analysis based on LMDI approach Index decomposition analysis based on LMDI approach Index decomposition analysis based on LMDI approach

Table 2 Definitions of variables. Symbol

Variable

Unit

Definition

C E K

CO2 released from residential buildings Energy consumption in residential buildings Final emission factor of residential buildings

Million tons of CO2 (MtCO2) Million tons of standard coal equivalent (Mtce) kgCO2·kgce-1

CO2 emission per household in residential buildings (i.e., CO2 intensity in residential buildings) CO2 emission per floor space in residential buildings Energy consumption per floor space in residential buildings Population size per household Per capita income Housing price-to-income ratio

kgCO2 per household kgCO2·m−2 kgce·m−2 persons per household CNY·person-1 %

– – K = C /E = Kl ) – – – – – S = Sj = 100%

ch = C / H c = C /F e = E /F p = P /H i = I /P

m2·CNY-1 kgCO2 per household kgCO2 per household

d = 1/Pr Eq. (4) Eq. (4)

Floor spaces of residential buildings Income of households Housing price Population size Amount of households Household age structure

F I Pr P H S ch c e p i r

d

ch |0 ch p chS chr chi chd che chK

T

Housing purchasing power ch changes during Period T Effect of p onch Effect Effect Effect Effect Effect Effect Ridge

Million square meters (m2) Billion Chinese Yuan (CNY) CNY·m−2 Million persons Million households %

kgCO2 kgCO2 kgCO2 kgCO2 kgCO2 kgCO2 –

of S onch of r onch of i onch of d onch of e onch of K onch parameter

per per per per per per

household household household household household household

Note: K = Kl (l = 1, 2, 3, 4, 5), where 1, 2, 3, 4, and 5 represent coal, oil, natural gas, electricity, and heating, respectively; S = 2, and 3 represent the age group of 0–14, 15–64, and over 65, respectively.

3

r=

Eq. Eq. Eq. Eq. Eq. Eq.

F Pr I

=

F H P H

Pr I P

(4) (4) (4) (4) (4) (4) (0, 1)

Sj = 100% ( j = 1, 2, 3 ), where 1,

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Regarding the topic of emission mitigation assessment, although CO2 mitigation assessment has been widely discussed in the major emission sectors such as transportation and industry, such a study has not been proposed for the residential building sector to the best of the authors’ knowledge (see Fig. A3). Although the typical literature has reported the CO2 mitigation of the building sector [15], the studies actually focused on emission mitigation assessment in commercial buildings, and the assessment model has yet to investigate the intensity feature of CO2 mitigation. Therefore, it is valuable to conduct such an individual assessment for residential buildings to close the gap. Regarding the approach of emission mitigation assessment, Table 1 proves that decomposition analysis has been extensively applied to assess CO2 mitigation across different emission sectors and the decomposition results are reliable enough for CO2 mitigation assessment. Since the residential building sector is a typical nonproductive sector, its energy flow is hard to express by the input–output model (i.e., structure decomposition analysis is not suitable). However, complete time-series data on the energy and emissions of the residential building sector can be accessed and verified from C-BEED in China [30,31]. Therefore, index decomposition analysis, with its commonly used version (LMDI), is the preferred approach to assess CO2 mitigation rather than structure decomposition analysis. The findings above reveal that an individual study on CO2 mitigation assessment for the residential building sector needs to be urgently presented. Thus, this paper aims to investigate the factors that can mitigate CO2 intensity and further assesses CMRBS in China to fill the research gap mentioned above. Contributions cover the following two aspects:

decomposition analysis and CO2 mitigation assessment.

• Both the intensity and total values of CMRBS are considered in the assessment.

Since this paper assesses CMRBS, the total volume of CO2 mitigation is required for analysis. Furthermore, the intensity feature of CMRBS needs to be assessed as CO2 intensity is the key indicator in controlling the total CO2 emissions, which is usually characterized as CO2 emission per household in residential buildings [5]. Thus, the CO2 intensity feature should first be analyzed through conducting the investigation on factors affecting CO2 mitigation and the assessment of CO2 mitigation. In this paper, both intensity and total values of CMRBS are assessed through the decomposition analysis for emission peak confirmation and policymaking. 3. Materials and methods Section 3.1 presents an assessment model to investigate the factors that can mitigate CO2 intensity and further assesses the retrospective CMRBS of China. Moreover, the materials of this study, including both the variable definition and original data collection are introduced in Sections 3.2 and 3.3. To present a clear and logic flow, Fig. 2 expresses the study framework. 3.1. CMRBS assessment model

• This paper is the first to assess CMRBS based on a household scale.

It is widely accepted that the smallest unit of energy survey in a residential building is the household scale, and energy/CO2 intensity is usually characterized as the energy consumption/CO2 emission per household [32–34]. Thus, the CO2 intensity feature should first be analyzed for conducting the investigation of the factors affecting CO2 mitigation and the assessment of CO2 mitigation [35]. As introduced in Table 1 and Fig. A2, decomposition analysis has been widely adopted to investigate the factors that can mitigate CO2 intensity and to assess the CO2 mitigation in emission sectors. Regarding the residential building sector of China, this paper presented the CMRBS assessment via a decomposition

Only several literature has reported the CO2 mitigation assessment in the building sector and their cases focused on emission mitigation in commercial buildings rather than residential buildings. In addition, few studies have considered the effect of household scale to CMRBS. Therefore, this paper first assesses CMRBS based on a household scale via the decomposition analysis. Moreover, a robustness test via ridge regression is conducted to ensure the reliability of results in the

Fig. 2. Schematic of this paper. 4

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. 3. Decomposition flow of Kaya identity on CO2 intensity in the residential building sector.

analysis. To achieve this goal, a series of impact factors were considered to reflect the CO2 intensity feature which were identified by the Kaya identity as follows: Kaya identity [36], optimized on the basis of Impact of Population, Affluence, and Technology identity [37], indicates the coupling of CO2 emission and its impact factors, including three aspects: 1) population, 2) affluence [e.g., the gross domestic product (GDP) per capita], and 3) technology [e.g., CO2 intensity (the product of energy intensity and the emission factor)] [38], as expressed in Eq. (1).

CO2 = Population ×

Energy GDP CO2 × × Population GDP Energy

Let ch =

(1)

ch |0

Factor VI – Per capita income (i = I/P ) Factor VII – Housing purchasing power (d = 1/Pr ) a

Sj , r =

F H P H

Pr I P

, i = P, d = I

1 , Pr

e=

E , F

and (3)

= ch |T

T

ch |0

= ch p + chS + chr + chi + chd + che + chK

(4)

Specifically, every parameter (e.g., ch p ) on the right side of Eq. (4) can be further expressed as:

ch p = L (ch |T , ch |0 ) ln a lna

L (a , b ) =

b , lnb

a

p|T P|T H|0 = L (ch |T , ch |0 ) ln p|0 P|0 H|T

(5)

b (a > 0, b > 0) (6)

0, a = b (a > 0, b > 0)

Furthermore, CMRBS is expressed as follows:

CMRBSintensity|0

Sj )

=

S=

Thereafter, this paper used the LMDI approach to decompose Eq. (3) to confirm the effects of seven factors on CO2 intensity. LMDIe [40] has been widely applied with Kaya identity to evaluate the effects of various factors on energy consumption or CO2 emissions [41,42]. Through the LMDI application handbook [43], the changes of CO2 intensity in residential buildings during Period T ( ch |0 T ) are decomposed as follows:

Box 1 | Parameters shown in Fig. 3. Factor I – Energy consumption per floor space in residential building sector (e = E / F ) Factor II – Final emission factor of residential building sector (K = C / E = K l ) Factor III – Population size per household (p = P/H )

F Pr I

P , H

ch = p S r i d e K

a

Factor V – Housing price-to-income ratio (r =

p=

Kl ; Eq. (2) can be converted to Eq. (3).

K=

To decompose the CO2 intensity in the residential building sector and obtain its relevant factors, Kaya identity on CO2 intensity is required to be built as the first step in deploying the decomposition analysis [39]. To present a reasonable decomposition process and ensure the reliability of the decomposition results, this paper referred to the latest highly related study to build the specific Kaya identity on CO2 intensity for the residential building sector: Liang et al. reported a case of Kaya identity on CO2 intensity in the residential building sector, which considered social, economic and technical features of residential buildings [5], as illustrated in Fig. 3 and Box 1.

Factor IV – Household age structure (S =

C , H

F Pr H P I ) H P

CMRBS|0

T

= H|0

where chm |0 |0

The parameters of Box 1 are introduced in Table 2.

T

<0

T

T

T

= ×(

| chm |0 | chm |0

T| T |)

(7) (8)

{ ch p, chS , chr , chi, chd, che, chK }, chm (9)

Mathematical expression of Fig. 3 is as follows.

C P = ch = H H

Sj

F P H r P I H P

I 1 E P Pr F

Kl

e Technically, LMDI includes two versions (i.e., LMDI-I and LMDI-II). Since this paper chooses the additive form to conduct the decomposition analysis (see Eq. (4)), the LMDI-I approach is preferred [43]. LMDI shown in this paper represents the LMDI-I approach if not specified.

(2)

5

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. 4. Schematic of CMRBS assessment model.

Fig. 4 presents a schematic of the decomposition analysis on ch to make the assessment model clearly understood.

contributed positively to the growth of CO2 intensity during 2000–2016, as expressed by the blue blocks shown in Fig. 6. This finding reveals the significant coupling of energy and emission intensities, which is linked by the final emission factor of the residential building sector (the purple blocks shown in Fig. 6). Regarding the negative factors, the housing purchasing power contributed the most in decreasing the CO2 intensity, as demonstrated by the light blue blocks shown in Fig. 6 (e.g., c hd |2000 2004 = 37.06%, c hd |2004 2008 = 33.35% , c hd |2008 2012 = 44.36%, and c hd |2012 2016 = 29.69%). Compared to the housing purchasing power, the housing price-to-income ratio also contributed negatively in terms of increasing the CO2 intensity from 2000 to 2008, as expressed by the green blocks shown in Fig. 6. However, the impact of the housing price-to-income ratio on CO2 intensity shifted from a negative status to a positive status from 2008 to 2016 (e.g., c hr |2004 2008 = 17.29% and c hr |2008 2012 = 0.87% ). As indicated in Eq. (2), the household age structure and the final emission factor in the residential building sector can be further extended into a series of subfactors (i.e., S = Sj and K = Kl ), respectively. Hence, the red and purple blocks shown in Fig. 6 were further decomposed. Regarding the energy structure changes, an optimal energy structure can lead to the maximum potential of CO2 mitigation. From 2000 to 2016, the proportion of coal consumption in the residential building sector decreased significantly (from 44.33% to 26.18%), and the proportions of electricity and heating increased during the same period (from 40.57% to 50.73%). Let the 2012–2016 period serve an example,

3.2. Variables 3.3. Data The original data used in this paper are at the national scale, which cover the data on nine variables of Eq. (4) over the period of 2000–2016. The data on E , C , F , and K were collected from C-BEED [7]. Meanwhile, the data on P , S , Pr , i , and p were accessed from the National Bureau of Statistics of PR China (http://data.stats.gov.cn/ english/). Fig. 5 and Table B1 summarize the data above. 4. Results 4.1. Decomposition results on CO2 intensity in the residential building sector Fig. 6 presents the LMDI decomposition results of CO2 intensity in the Chinese residential building sector during the 2000–2016 period via Eqs. (4) to (6). For the positive factors promoting the CO2 intensity growth in the residential building sector, per capita income played the most sigc hi |2000 2004 = 50.26% , c hi |2004 2008 = 61.63% , nificant role (e.g., c hi |2008 2012 = 56.09%, and c hi |2012 2016 = 37.03% ), as illustrated by the light green blocks shown in Fig. 6. Compared to the per capita income, energy consumption per floor space in the residential building sector also 6

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. 5. Data on core variables in the decomposition analysis (Eq. (4)) from 2000 to 2016. Note: E , C , and F in Fig. 5a represent energy consumption, CO2 emissions, and floor spaces of residential buildings, respectively; P and S = Sj in Fig. 5b represent population size and household age structure ( j = 1, 0–14 age group; j = 2 , 15–64 age group; j = 3 , over 65 age group), respectively; Pr and i in Fig. 5c represent housing price and per capita income, respectively; K and p in Fig. 5d represent the final emission factor of residential buildings and population size per household, respectively.

Fig. 6. Changes of CO2 intensity in the Chinese residential building sector ( ch ) via a decomposition analysis (2000–2016). Note: ch p , chs , chr , chi , chd , che , and chK represent the impacts of population size per household, household age structure, housing price-to-income ratio, per capita income, housing purchasing power, energy consumption per floor space, and the final emission factor to CO2 intensity in residential building sector, respectively.

7

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

where the CO2 mitigation contributions of the five main energy sources are summarized as follows: 69.60 (coal), 28.64 (oil), 32.77 (natural gas), 65.93 (electricity), and 68.96 (heating) kgCO2 per household. Regarding the impacts of population size per household and the household age structure on CO2 intensity, the residents within the 15–64 age group were the main force in promoting the CO2 intensity decrease from 2000 to 2012. However, the impact of the population size per household on the CO2 intensity shifted from a negative status to a positive status from 2012 to 2016 ( c hp |2008 2012 = 5.21% and c hp |2012 2016 = 3.27%). Thus, the contribution of residents within the 15–64 age group to the CO2 intensity increase changed from a negative status to a positive status from 2012 to 2016. 4.2. Retrospective CO2 mitigation in the residential building sector Fig. 7a reflects the trends on total and intensity values of CMRBS from 2001 to 2016 in China via the calculation based on Eqs. (7) to (9). To express the uncertainty level of CMRBS, two error bands were added in Fig. 7a (error band value of CMRBS intensity: 89.45 kgCO2 · (household · year) −1, and error band value of CMRBS: 40.19 MtCO2 per year). Furthermore, the average intensity of CMRBS during different Five-year Plan (FYP) periods in China is summarized as follows: 199.75 (10th FYP Period: 2001–2005), 307.19 (11th FYP Period: 2006–2010), and 284.45 kgCO2 · (household · year)-1 (12th FYP Period: 2011–2015). Moreover, CMRBS values during the three periods above are: 393.68 (10th FYP Period), 648.10 (11th FYP Period), and 641.40 MtCO2 (12th FYP Period). Besides, Fig. 7c assessed CMRBS intensity at another two scales (i.e., CO2 mitigation per capita and CO2 mitigation per floor space). The fitting estimations shown in Fig. 7b and d reflect that the

Fig. 8. Assessed and official expected values of energy savings in the Chinese residential building sector. Note: official expected values are . accessed from [46–48]

continuous growth of CMRBS at four different scales is obvious. After assessing CMRBS values, a comparative analysis of the official expected and assessed values of energy savings in the residential building sector was presented, as illustrated in Fig. 8. To provide a comparable condition, the CMRBS values of Fig. 7 were converted to the energy-saving values via the final emission factor in the residential building sector introduced in Eq. (2). Fig. 8 reveals that the assessed

Fig. 7. a and b. Total and intensity values of CMRBS in China during 2001–2016; c and d. CMRBS intensity at another two scales during 2001–2016 (CO2 mitigation per capita and CO2 mitigation per floor space). 8

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

values were much higher than the official expected values during the 11th and 12th FYP Periods. Furthermore, although the 13th FYP Period (2016–2020) is still in process, the assessed energy-saving value in 2016 reached 60.80 Mtce, constituting over 60% of the official expected value from 2016 to 2020. It is believed that the residential building sector is able to meet its 13th FYP energy-saving target. It should be noted that the Chinese government has officially conducted the nationwide effort to reach the ECEM target since 2006 [44,45], which means that the official expected value of energy savings in the residential building sector during the 10th FYP Period (2001–2005) was lacking and the relevant comparative analysis of energy savings is difficult to represent, as is shown in Fig. 8. However, the results of the comparative analysis focusing on 2006–2016 are enough to prove that the Chinese residential building sector has achieved obvious energysaving benefits over the past decade.

5. Discussion 5.1. Robustness test of the CMRBS assessment model As declared in Section 1, this paper investigates the factors that mitigate CO2 intensity and to further assess the intensity and total values of CMRBS in China via decomposition analysis. Although the Kaya identity at the CO2 scale for LMDI decomposition is built based on the latest highly related study [5] whose analytical result is credible, the reliability of decomposition results in this paper should be further evaluated; in short, whether the LMDI decomposition result is reliable for CMRBS assessment. The robustness test is the final step in completely answering questions 1 and 2 in Section 1. To respond to the issues above, Section 5.1 presents a robustness test to conduct a comparative analysis on impact factor elasticity in LMDI and ridge regression scenarios. Compared to LMDI, the regression analysis is also a widely adopted approach to achieve the goal of investigating the effects of different factors on CO2 intensity [49]. To deploy the robustness test, this

Fig. 9. Ridge regression results with the optimal ridge parameter ( = 1). 9

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

paper designed the identity linked with the explained and the explaining variables based on Eq. (3) via the Stochastic Impacts by Regression on Population, Affluence, and Technology approach, as shown below.

ch = p S r i d e K =

+

p lnp

+

lnch r lnr

+

i lni

+

d lnd

+

e lne

+

K lnK

(10)

where variable S of Eq. (3) was removed from Eq. (10) since the value of Sj always equals one, which shows that regression estimation on variable S is meaningless. As the basis from which to choose the type of regression experiment (ordinary least squares or its modified version such as the ridge regression), the multicollinearity among the explaining variables needs to be tested [50,51]. The variance inflation factor (VIF) is the indicator used to characterize the multicollinearity level as expressed below:

VIFv = (1

Rv2 ) 1 (v = 1, 2, 3,

, n)

(11)

If a serious multicollinearity exists, the ordinary least squares will no longer be stable for the regression estimation. Nevertheless, ridge regression is able to overcome the serious multicollinearity, as shown in Eq. (12).

X = y where ( ) = (X T X + I ) 1X T y

Fig. 10. Comparative analysis on impact factor elasticity in LMDI and ridge regression scenarios.

(12)

where denotes the ridge parameter (changing from 0 to 1) to shape the ridge and VIF traces (see Fig. 9) for seeking the optimal regression result of Eq. (10). It is observed that when = 0 , Eq. (12) is simplified as the ordinary least squares and VIF values of the explaining variables are much high than the maximum value of 10 (see Fig. 9b: the max VIF value: 2216.28 and the average VIF value: 678.93), which means that serious multicollinearity does exist [52]. Therefore, it is suitable to use the ridge regression to build the regression scenario for the robustness test. Fig. 9a shows the optimal ridge regression estimation with the optimal ridge parameter value of 1. It is observed that the VIFs have effectively reduced from the average value of 678.93 to 0.1264, which reveals that the ridge regression has successfully overcome the serious multicollinearity. Furthermore, the fit goodness shown in Fig. 9c and d indicates the regression estimation is reliable at the signification level of 95%. After estimating the regression results, this study conducted a comparative analysis on the impact factor elasticity in LMDI and ridge regression scenarios, as shown in Fig. 10. It is observed that the elasticity values of each factor in different scenarios are dissimilar [e.g., every 1% increase of ch will cause a 2.19% increase of i (LMDI scenario) and every 1% increase of lnch will cause a 18.89% increase of lni (ridge regression scenario)]; however, the impact of elasticity traces in the two scenarios have the same trend (the M−shaped curve) from 2000 to 2016. This finding reveals that the two elasticity traces are consistent, which proves that the robustness of the LMDI approach is sufficient for achieving reliable results in decomposition analysis and CMRBS assessment, as illustrated in Figs. 6 and 7. Furthermore, the robustness test used for the CMRBS assessment model is the final step in answering questions 1 and 2 in Section 1, as expressed in Fig. 2.

only one turning point exists in a normal EKC, and this hypothesis has been tested in Fig. 11. It should be noted that regarding the existing shortterm CO2 emission feature in the residential building sector, its EKC follows the inverted U-shaped pattern and the extended EKC (e.g., the Nshaped pattern [55]) has yet been observed via the fitting analysis. Following the arrow direction, it is observed that the order of occurrence time of turning points lists, as follows: EKC (c), 2013 EKC (cp ), 2019 (predictive value) EKC (ch ), 2020 (predictive value) EKC (C), 2024 (predictive value). This finding further verifies the hypothesis that the order of achieving emission peak is correct in the residential building sector of China, which has been previously verified by [5]. Through analyzing the EKC feature shown in Fig. 11, the cause leading to a similar emission mitigation effect at different emission scales (see Fig. 7) can be explained. Let Fig. 7 a serve an example: As illustrated in Fig. 11, only EKC at CO2 emission per floor space scale achieved its turning point in 2001–2016. However, the EKCs at scales of CO2 intensity and total CO2 emissions have yet to meet their turning points during this period, which means that CO2 emission changes in EKC (ch ) and EKC (C) faced monotonical increases with different decreasing growth levels, and their emission features were linked by the household size (see Eq. (8)). This phenomenon can explain the similar CO2 mitigation effects for CMRBS and its intensity scales. 5.3. Retrospection and implication of ECEM strategy in the residential building sector To answer question 3 from Section 1, ECEM strategy needs to be reviewed. ECEM strategy of the building sector is defined as the set of code, act, policy document, energy conservation standard, energy efficiency label, energy conservation technology, and economic intensive approach to achieve the ECEM target in the building sector [56,57]. Specifically, ECEM strategy can be mainly summarized into three parts: a. mandatory strategy (e.g., energy conservation standard), b. information strategy (e.g., energy efficiency label, ladder-type price of electricity), and c. economic intensive strategy (e.g., special funding, financial subsidy) [58]. Regarding the residential building sector of China, its official ECEM strategy was fully deployed at the beginning of 11th FYP Period (2006), which includes over 10 relevant codes and acts, more than 80 policy documents, and at least 50 mandatory standards by the end of 2015f. For example, to fully promote the ECEM strategy in the building sector, the Chinese government issued China

5.2. What leads to similar emission mitigation effects at different emission scales? Section 4 summarizes the decomposition results of CO2 intensity in the residential building sector and CMRBS values from 2001 to 2016 in China. Through the analyses, one question, shown in Fig. 7, is raised as follows: what leads to the similar mitigation effects in the residential building sector? In short, what causes the similar trends in the intensity and total values of CMRBS from 2001 to 2016? To answer this question, the CO2 emission features of the residential building sector should be further discussed. Thus, EKC theory was utilized in this section to reveal the timeseries relationship between CO2 emissions and per capita income in residential buildings. To completely reveal the emission features, the EKCs at four emission scales were illustrated in Fig. 11, which included: the EKC at CO2 emission per floor space scale [EKC (c)], the EKC at CO2 emission per capita scale [EKC (cp )], the EKC at CO2 intensity scale [EKC (ch )], and the EKC at total CO2 emission scale [EKC (C)]. As demonstrated in [53,54],

f

10

Source: http://www.mohurd.gov.cn/jzjnykj/jzjnykjzcfb/index.html

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

further assessed the intensity and total values of CMRBS in China from 2000 to 2016 via decomposition analysis. A robustness test was subsequently conducted via the ridge regression to verify the reliability of the decomposition and CMRBS results. Moreover, the EKC on CO2 emission feature was applied to reveal similar emission mitigation effects at different emission scales in the residential building sector. Furthermore, the ECEM strategy of residential buildings was retrospected for exploring policy patterns to achieve more emission mitigation effects in the future. Section 6 introduces core findings and upcoming studies, as follows: 6.1. Core findings

• CMRBS from 2001 to 2016 is 1816.99 MtCO

2 and the average mitigation intensity during this period is 266.12 kgCO2 · (household · year)-1.

The CO2 mitigation assessment model proposed in this paper was built via decomposition analysis. Through observing the decomposition results on CO2 intensity conducted using the LMDI approach, three housing economic indicators (housing purchasing power, housing price-to-income ratio, and population size per household) contributed significantly to decrease CO2 intensity. As for the technical factor effect, the impact of the final emission factor on CO2 intensity was also negative, which means the “coal to electricity” effect of the residential building sector is significant over the past decade. Regarding CMRBS, the average intensity of CMRBS during different periods is summarized as follows: 199.75 (2001–2005), 307.19 (2006–2010), and 284.45 kgCO2 · (household · year)-1 (2011–2015). In addition, CMRBS values during the above three periods are: 393.68, 648.10, and 641.40 MtCO2. Compared to the official expected energy-saving target, the assessed value of energy savings based on emission mitigation has already exceeded the expected value during the abovementioned three periods.

Fig. 11. EKCs at different emission scales in the Chinese residential building sector (2000–2025).

Energy Conservation Code (1997, 2007, 2016, 2018 versions) and China Act on Energy Conservation of Civil Buildings (2008). Following the two codes’ guidance above, a series of specific approaches such as the energy conservation standards of newly built buildings, economic intensive approaches on the energy conservation retrofit of existing buildings, the evaluation system of green buildings and the ECEM technology of civil buildings have been proposed for the public, as is mainly illustrated in Fig. 12. According to Fig. 12, it is obviously observed that the Chinese government has expended much effort to develop the ECEM cause, and this significant achievement directly leads to the CMRBS increase (e.g., see Figs. 5 and 6, the final emission factor has been affected by the “coal to electricity” strategy, which directly promoted the CMRBS increase [59,60]). In addition, it is believed that more significant emission mitigation will be realized via the further deployment and implementation of ECEM strategy in the residential building sector in the upcoming phase. Considering the space limitation, the implications for ECEM strategy in the residential building sector of China are conducted briefly as follows: a. implementing the official policy assessment system for ECEM which is built based on the ECEM effect [61,62]; b. establishing the multidimensional data statistical system for energy and emission which includes the large sample of building microdata and the sustainable monthly/quarterly/annual data on energy and emission at national and provincial scales [63,64]; c. issuing the design/construction mandatory standard for residential buildings guided by the ECEM effect [65,66]; d. increasing the financial expenditure on the formulation and implementation of the ECEM strategy [67,68]; e. spreading a series of high energy-efficiency technologies and productions such as the microgrid and distributed energy system [69,70], and the nearly zero energy building technology [71,72]. Overall, the above policy implication answers question 3 in Section 1, as illustrated in Fig. 2.

• The robustness of the LMDI approach is sufficient for achieving reliable results in decomposition analysis and CMRBS assessment.

To answer whether the LMDI decomposition result is reliable for CMRBS assessment, regression analysis was applied to verify the robustness of the LMDI approach. Due to the serious multicollinearity existing among the explaining variables of LMDI decomposition, ridge regression is preferred to ordinary least squares. The comparative analytical result on the impact factor elasticity in LMDI and ridge regression scenarios shows that impact elasticity traces in both scenarios are consistent as both elasticity trace trends are M−shaped curves from 2000 to 2016. Thus, the robustness of the LMDI approach has been verified, and this approach can be trusted to achieve reliable results in decomposition analysis and CMRBS assessment.

• The ECEM strategy promotes CMRBS increase effectively and it is the key to promoting more significant emission mitigation in the future.

Using the retrospection on ECEM strategy in the residential building sector, it is observed that the official ECEM strategy has been fully deployed since 2006 and the Chinese government has expended much effort to develop the ECEM cause. This significant achievement directly leads to CMRBS increase. In addition, more significant emission mitigation will be realized via different policy patterns in the future, such as implementing the official policy assessment system for ECEM which is built based on the ECEM effect; establishing the multidimensional data statistical system of energy and emission; issuing the design/construction mandatory standard guided by the ECEM effect; increasing financial expenditure on the formulation and implementation of the ECEM strategy; and spreading a series of high energy-efficiency technologies and productions.

6. Conclusions This paper investigated the factors for mitigating CO2 intensity and 11

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. 12. Development pattern of ECEM strategy in the Chinese residential building sector (2006–2019). Note: a. principal line of ECEM strategy; b. mandatory design standards for energy conservation of residential buildings across different building climate zones.

6.2. Upcoming studies

regions differs significantly [77], individual studies on provincial-level/ city-level CO2 mitigation assessments in residential buildings in urban and rural China are worthy to be conducted, respectively, which will help the central and local governments implement more effective targets and strategies for ECEM. At last, to conduct the sustainable development of built environment, the scope of emission mitigation assessment in built environment needs to be further extended, such as greenhouse gas mitigation [78], particulate matter mitigation [79], etc.

Upcoming studies need to be deployed to fill in several gaps in the current study. Regarding the factors affecting CMRBS, the impact of climate change on energy and emission in residential buildings is significant [73]. Specifically, the temperature change does affect the energy and emission intensities of residential buildings since residents have different requirements on heating, ventilating and air conditioning system operation in different temperatures [74,75]. Hence, the future study should attempt to consider the temperature impact on CMRBS. For the unusual emission mitigation peak which occurred in 2009 (see Fig. 7), the internal cause needs to be investigated. Since the time point on the peak is close to the time point of the 2008 financial crisis [76], the econometrics approach (regression discontinuity, difference-in-difference analysis, etc.) may be adopted to investigate the cause leading to the peak effect. Furthermore, regarding the case area, since the CO2 emission feature of the residential building sector in urban and rural

Acknowledgements This study was supported by the Fundamental Research Funds for the Central Universities of PR China (2018CDYJSY0055 and 2019CDJSK03XK04), the National Planning Office of Philosophy and Social Science Foundation of China (19BJY065), and the National Key R &D Program of China (2018YFD1100201).

Appendix A. Extra figures and tables The range of CO2 released from residential building sector is expressed by red circle of Fig. A1.

12

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Fig. A1. Range of CO2 released from the residential building sector (red circle) (). Source: https://www.iea.org/topics/energyefficiency/buildings/

Fig. A2. Index decomposition analysis versus structure decomposition analysis in the energy and emission analysis.

Fig. A3. Distribution of the literature on CO2 mitigation assessment in the top three emission sectors. 13

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

Appendix B. Raw dataset Table B1 Raw data on core variables in the decomposition analysis (Eq. (4)) from 2000 to 2016. Year

E Mtce

C MtCO2

K kgCO2·kgce-1

K1 Coal %

K2 Oil %

K3Natural gas %

K4 Electricity %

K5 Heating %

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

189.00 204.00 227.00 255.00 284.00 312.00 333.00 358.00 372.00 385.00 399.00 420.00 445.00 479.00 493.00 517.00 553.00

442.00 481.00 529.00 600.00 657.00 717.00 761.00 810.00 847.00 895.00 939.00 988.00 1090.00 1139.00 1119.00 1145.00 1217.00

2.339 2.358 2.330 2.353 2.313 2.298 2.285 2.263 2.277 2.325 2.353 2.352 2.449 2.378 2.270 2.215 2.201

0.53 0.49 0.46 0.46 0.46 0.44 0.42 0.38 0.35 0.33 0.31 0.31 0.30 0.27 0.26 0.25 0.24

0.11 0.10 0.10 0.10 0.11 0.10 0.11 0.11 0.09 0.10 0.09 0.09 0.09 0.09 0.11 0.12 0.13

0.03 0.04 0.03 0.03 0.04 0.04 0.05 0.07 0.08 0.09 0.10 0.11 0.12 0.12 0.12 0.12 0.12

0.12 0.13 0.12 0.13 0.14 0.15 0.17 0.19 0.20 0.22 0.21 0.23 0.24 0.25 0.25 0.25 0.26

0.21 0.25 0.29 0.28 0.26 0.26 0.25 0.24 0.27 0.27 0.28 0.26 0.26 0.26 0.26 0.26 0.24

Year

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

F

P

H

S215–64 age group %

S3Over 65 age group %

I

Pr

104 households

S10–14 age group %

108 square meters

104 persons

Billion CNY

CNY per square meter

306.04 311.23 325.89 334.26 349.47 369.15 377.50 386.14 395.98 406.76 419.60 436.52 453.31 468.80 484.34 500.43 519.81

126,743 127,627 128,453 129,227 129,988 130,756 131,448 132,129 132,802 133,450 134,091 134,735 135,404 136,072 136,782 137,462 138,271

36843.90 37335.96 37900.23 38277.42 38685.25 41735.52 41431.90 41687.74 41973.27 42425.78 43255.16 44505.93 44917.85 45690.18 46007.39 44342.58 44460.13

22.90 22.50 22.40 22.10 21.50 20.30 19.80 19.40 19.00 18.50 16.60 16.50 16.50 16.40 16.50 16.52 16.64

70.10 70.40 70.30 70.40 70.90 72.00 72.30 72.50 72.70 73.00 74.50 74.40 74.10 73.90 73.40 73.01 66.66

7.00 7.10 7.30 7.50 7.60 7.70 7.90 8.10 8.30 8.50 8.90 9.10 9.40 9.70 10.10 10.50 16.70

4704.49 5179.79 5804.68 6452.57 7337.33 8324.66 9478.30 11319.06 13199.18 14631.98 16771.57 19646.95 22569.82 24915.87 27584.96 30195.18 32937.53

1948.00 2017.00 2092.00 2197.00 2608.00 2936.96 3119.25 3645.18 3576.00 4459.00 4725.00 4993.17 5429.93 5850.00 5933.00 6473.00 7203.00

[10] Zhou N, Khanna N, Feng W, Ke J, Levine M. Scenarios of energy efficiency and CO 2 emissions reduction potential in the buildings sector in China to year 2050. Nat Energy 2018;3:978–84. [11] Delmastro C, Lavagno E, Mutani G. Chinese residential energy demand: Scenarios to 2030 and policies implication. Energy Build 2015;89:49–60. [12] Tan X, Lai H, Gu B, Zeng Y, Li H. Carbon emission and abatement potential outlook in China's building sector through 2050. Energy Policy. 2018;118:429–39. [13] Qian D, Li Y, Niu F, O'Neill Z. Nationwide savings analysis of energy conservation measures in buildings. Energy Convers Manage 2019;188:1–18. [14] Heidarinejad M, Cedeño-Laurent JG, Wentz JR, Rekstad NM, Spengler JD, Srebric J. Actual building energy use patterns and their implications for predictive modeling. Energy Convers Manage 2017;144:164–80. [15] Ma M, Cai W, Cai W. Carbon abatement in China's commercial building sector: a bottom-up measurement model based on Kaya-LMDI methods. Energy 2018;165:350–68. [16] Yan R, Ma M, Pan T. Estimating energy savings in Chinese residential buildings from 2001 to 2015: a decomposition analysis. J Eng Sci Technol Rev 2001;10(2017):107–13. [17] MOHURD, China. Statistical Statement System of Energy Consumption in Civil Buildings. 2015. Available at http://www.mohurd.gov.cn/wjfb/201512/ t20151218_226035.html (in Chinese). [18] MOHURD, China. Statistical Statement System of Energy & Resource Consumption in Civil Buildings. 2018. Available at http://www.mohurd.gov.cn/wjfb/201803/ t20180314_235394.html (in Chinese). [19] Huo T, Ren H, Zhang X, Cai W, Feng W, Zhou N, et al. China's energy consumption in the building sector: a Statistical Yearbook-Energy Balance Sheet based splitting method. J Cleaner Prod 2018;185:665–79. [20] Feng C, Wang M. Journey for green development transformation of China’s metal

References [1] IPCC. Global Warming of 1.5 °C - an IPCC special report on the impacts of global warming of 1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty. IPCC2018. Available at http://www.ipcc.ch/report/sr15/. [2] International Energy Agency. World Energy Outlook 2018. IEA2019. Available at https://www.iea.org/weo2018/. [3] Berardi U. A cross-country comparison of the building energy consumptions and their trends. Resour Conserv Recycl 2017;123:230–41. [4] Wang H, Chen W, Shi J. Low carbon transition of global building sector under 2- and 1.5-degree targets. Appl Energy 2018;222:148–57. [5] Liang Y, Cai W, Ma M. Carbon dioxide intensity and income level in the Chinese megacities' residential building sector: Decomposition and decoupling analyses. Sci Total Environ 2019;677:315–27. [6] Su C, Madani H, Palm B. Heating solutions for residential buildings in China: Current status and future outlook. Energy Convers Manage 2018;177:493–510. [7] China Association of Building Energy Efficiency. China Building Energy Consumption Report (2018). in: W.G. Cai, (Ed.). Beijing, PR China, 2018. Available at http://www.cabee.org/site/content/22960.html (in Chinese). [8] Ma M, Cai W. What drives the carbon mitigation in Chinese commercial building sector? Evidence from decomposing an extended Kaya identity. Sci Total Environ 2018;634:884–99. [9] McNeil MA, Feng W, du Can SdlR, Khanna NZ, Ke J, Zhou N. Energy efficiency outlook in China’s urban buildings sector through 2030. Energy Policy 2016;97:532–9.

14

Energy Conversion and Management 198 (2019) 111915

M. Ma, et al.

[50] Farrar DE, Glauber RR. Multicollinearity in regression analysis: the problem revisited. Rev Econ Stat 1967:92–107. [51] Ma M, Yan R, Cai W. An extended STIRPAT model-based methodology for evaluating the driving forces affecting carbon emissions in existing public building sector: evidence from China in 2000–2015. Nat Hazards 2017;89:741–56. [52] Marquardt DW, Snee RD. Ridge regression in practice. The American Statistician 1975;29:3–20. [53] Qiao H, Zheng F, Jiang H, Dong K. The greenhouse effect of the agriculture-economic growth-renewable energy nexus: evidence from G20 countries. Sci Total Environ 2019;671:722–31. [54] Wang Z, Bu C, Li H, Wei W. Seawater environmental Kuznets curve: evidence from seawater quality in China's coastal waters. J Cleaner Prod 2019;219:925–35. [55] Allard A, Takman J, Uddin GS, Ahmed A. The N-shaped environmental Kuznets curve: an empirical evaluation using a panel quantile regression approach. Environ Sci Pollut Res 2018;25:5848–61. [56] Su C, Madani H, Palm B. Building heating solutions in China: A spatial technoeconomic and environmental analysis. Energy Convers Manage 2019;179:201–18. [57] Cai W, Liu C, Lai K-H, Li L, Cunha J, Hu L. Energy performance certification in mechanical manufacturing industry: a review and analysis. Energy Convers Manage 2019;186:415–32. [58] Ma M, Cai W, Wu Y. China Act on the Energy Efficiency of Civil Buildings, A decade review. Sci Total Environ 2008;651(2019):42–60. [59] Chen H, Chen W. Potential impact of shifting coal to gas and electricity for building sectors in 28 major northern cities of China. Appl Energy 2019;236:1049–61. [60] Ma X, Mei X, Wu W, Wu X, Zeng B. A novel fractional time delayed grey model with Grey Wolf Optimizer and its applications in forecasting the natural gas and coal consumption in Chongqing China. Energy. 2019;178:487–507. [61] Tian W, de Wilde P, Li Z, Song J, Yin B. Uncertainty and sensitivity analysis of energy assessment for office buildings based on Dempster-Shafer theory. Energy Convers Manage 2018;174:705–18. [62] Tan Y, Liu G, Zhang Y, Shuai C, Shen GQ. Green retrofit of aged residential buildings in Hong Kong: A preliminary study. Build Environ 2018;143:89–98. [63] Amasyali K, El-Gohary NM. A review of data-driven building energy consumption prediction studies. Renew Sustain Energy Rev 2018;81:1192–205. [64] Wu W, Ma X, Zeng B, Wang Y, Cai W. Forecasting short-term renewable energy consumption of China using a novel fractional nonlinear grey Bernoulli model. Renewable Energy 2019;140:70–87. [65] Zuo J, Pullen S, Rameezdeen R, Bennetts H, Wang Y, Mao G, et al. Green building evaluation from a life-cycle perspective in Australia: a critical review. Renew Sustain Energy Rev 2017;70:358–68. [66] Jing R, Wang M, Zhang Z, Wang X, Li N, Shah N, et al. Distributed or centralized? Designing district-level urban energy systems by a hierarchical approach considering demand uncertainties. Appl Energy 2019;252:113424. [67] Zhao X, Zuo J, Wu G, Huang CJASR. A bibliometric review of green building research 2000–2016 2019(62):74–88. [68] Dong K, Sun R, Li H, Liao H. Does natural gas consumption mitigate CO 2 emissions: testing the environmental Kuznets curve hypothesis for 14 Asia-Pacific countries. Renew Sustain Energy Rev 2018;94:419–29. [69] Jing R, Wang M, Wang W, Brandon N, Li N, Chen J, et al. Economic and environmental multi-optimal design and dispatch of solid oxide fuel cell based CCHP system. Energy Convers Manage 2017;154:365–79. [70] Jing R, Zhu X, Zhu Z, Wang W, Meng C, Shah N, et al. A multi-objective optimization and multi-criteria evaluation integrated framework for distributed energy system optimal planning. Energy Convers Manage 2018;166:445–62. [71] Wu W, Skye HM. Net-zero nation: HVAC and PV systems for residential net-zero energy buildings across the United States. Energy Convers Manage 2018;177:605–28. [72] Chen X, Shuai C, Chen Z, Zhang Y. What are the root causes hindering the implementation of green roofs in urban China? Sci Total Environ 2018;654:742–50. [73] Li Y, Pizer WA, Wu L. Climate change and residential electricity consumption in the Yangtze River Delta, China. Proc Natl Acad Sci 2019;116:472. [74] Fan J, Wu L, Zhang F, Cai H, Ma X, Bai H. Evaluation and development of empirical models for estimating daily and monthly mean daily diffuse horizontal solar radiation for different climatic regions of China. Renew Sustain Energy Rev 2019;105:168–86. [75] Wu L, Huang G, Fan J, Zhang F, Wang X, Zeng W. Potential of kernel-based nonlinear extension of Arps decline model and gradient boosting with categorical features support for predicting daily global solar radiation in humid regions. Energy Convers Manage 2019;183:280–95. [76] Ma M, Cai W. Do commercial building sector-derived carbon emissions decouple from the economic growth in Tertiary Industry? A case study of four municipalities in China. Sci Total Environ 2019;650:822–34. [77] Guo J, Du L, Wei C. Equity-efficiency trade-off in China's energy capping policy. Energy Policy. 2019;126:57–65. [78] Zhang Y, Shen L, Shuai C, Bian J, Zhu M, Tan Y, et al. How is the environmental efficiency in the process of dramatic economic development in the Chinese cities? Ecol Ind 2019;98:349–62. [79] Zhang Y, Shuai C, Bian J, Chen X, Wu Y, Shen L. Socioeconomic factors of PM2.5 concentrations in 152 Chinese cities: Decomposition analysis using LMDI. J Cleaner Prod 2019;218:96–107.

industry: a spatial econometric analysis. J Cleaner Prod 2019;225:1105–17. [21] Shao S, Yang L, Gan C, Cao J, Geng Y, Guan D. Using an extended LMDI model to explore techno-economic drivers of energy-related industrial CO 2 emission changes: a case study for Shanghai (China). Renew Sustain Energy Rev 2016;55:516–36. [22] Wang H, Ang B, Su B. Assessing drivers of economy-wide energy use and emissions: IDA versus SDA. Energy Policy. 2017;107:585–99. [23] Ang BW. The LMDI approach to decomposition analysis: a practical guide. Energy Policy. 2005;33:867–71. [24] Wang H, Ang BW, Su B. A Multi-region Structural Decomposition Analysis of Global CO2 Emission Intensity. Ecol Econ 2017;142:163–76. [25] Dong K, Jiang H, Sun R, Dong X. Driving forces and mitigation potential of global CO2 emissions from 1980 through 2030: evidence from countries with different income levels. Sci Total Environ 1980;649(2019):335–43. [26] Achour H, Belloumi M. Decomposing the influencing factors of energy consumption in Tunisian transportation sector using the LMDI method. Transp Policy 2016;52:64–71. [27] Wang WW, Zhang M, Zhou M. Using LMDI method to analyze transport sector CO2 emissions in China. Energy. 2011;36:5909–15. [28] Zhang X, Zhao X, Jiang Z, Shao S. How to achieve the 2030 CO 2 emission-reduction targets for China's industrial sector: Retrospective decomposition and prospective trajectories. Global Environ Change 2017;44:83–97. [29] Lin B, Liu H. CO2 mitigation potential in China's building construction industry: a comparison of energy performance. Build Environ 2015;94:239–51. [30] Jing R, Wang M, Zhang Z, Liu J, Liang H, Meng C, et al. Comparative study of posteriori decision-making methods when designing building integrated energy systems with multi-objectives. Energy Build 2019;194:123–39. [31] Ma M, Cai W, Cai W, Dong L. Whether carbon intensity in the commercial building sector decouples from economic development in the service industry? Empirical evidence from the top five urban agglomerations in China. J Cleaner Prod 2019;222:193–205. [32] Energy Information Administration, US. Residential Energy Consumption Survey (RECS) 2015. EIA2015. Available at https://www.eia.gov/consumption/ residential/. [33] Department for Business, Energy & Industrial Strategy, UK. Domestic National Energy Efficiency Data-Framework (NEED) methodology. 2019. Available at https://www.gov.uk/government/publications/domestic-national-energyefficiency-data-framework-need-methodology. [34] MOHURD, China. Energy & Resource Survey for Civil Buildings. 2018. Available at http://www.mohurd.gov.cn/wjfb/201803/W020180314052735.doc (in Chinese). [35] Liu Y, Yang L, Zheng W, Liu T, Zhang X, Liu J. A novel building energy efficiency evaluation index: establishment of calculation model and application. Energy Convers Manage 2018;166:522–33. [36] Kaya Y. Impact of carbon dioxide emission control on GNP growth: interpretation of proposed scenarios. Intergovernmental Panel on Climate Change/Response Strategies Working Group, May; 1989. [37] Ehrlich PR, Holdren JP. Impact of population growth. Science 1971;171:1212–7. [38] Shuai C, Shen L, Jiao L, Wu Y, Tan Y. Identifying key impact factors on carbon emission: evidences from panel and time-series data of 125 countries from 1990 to 2011. Appl Energy 1990;187(2017):310–25. [39] Wang M, Feng C. Using an extended logarithmic mean Divisia index approach to assess the roles of economic factors on industrial CO2 emissions of China. Energy Econ 2018;76:101–14. [40] Ang BW, Liu FL. A new energy decomposition method: perfect in decomposition and consistent in aggregation. Energy 2001;26:537–48. [41] Feng C, Huang J-B, Wang M. The driving forces and potential mitigation of energyrelated CO2 emissions in China's metal industry. Resour Policy 2018;59:487–94. [42] Ma M, Yan R, Du Y, Ma X, Cai W, Xu P. A methodology to assess China's building energy savings at the national level: An IPAT–LMDI model approach. J Cleaner Prod 2017;143:784–93. [43] Ang B. LMDI decomposition approach: A guide for implementation. Energy Policy 2015;86:233–8. [44] State Council, China. Comprehensive Work Plan for Energy Efficiency and Emission Reduction. 2007. Available at http://www.gov.cn/xxgk/pub/govpublic/mrlm/ 200803/t20080328_32749.html (in Chinese). [45] Cai W, Lai K-H, Liu C, Wei F, Ma M, Jia S, et al. Promoting sustainability of manufacturing industry through the lean energy-saving and emission-reduction strategy. Sci Total Environ 2019;665:23–32. [46] MOHURD, China. 12th FYP Building Energy Efficiency Special Plan. 2012. Available at http://www.mohurd.gov.cn/zxydt/201206/t20120605_210123.html (in Chinese). [47] MOHURD, China. 13th FYP Building Energy Efficiency and Green Building Development Plan. 2017. Available at http://www.mohurd.gov.cn/wjfb/201703/ t20170314_230978.html (in Chinese). [48] State Council, China. China Act on the Energy Efficiency of Civil Buildings (2008). Beijing, PR China, 2008. Available at http://www.gov.cn/flfg/2008-08/07/ content_1067062.htm (in Chinese). [49] Ma M, Pan T, Ma Z. Examining the Driving Factors of Chinese Commercial Building Energy Consumption from 2000 to 2015: A STIRPAT Model Approach. J Eng Sci Technol Rev 2000;10(2017):28–34.

15