Critical risk factors of public building green retrofit projects- an empirical study in Chongqing, China

ABSTRACT

In order to discuss the critical risk factors of green retrofit project in existing public buildings, this study identifies the risk factors from the whole life cycle firstly, and the experts are invited to assess the probability of the occurrence, the degree of the severity, and the follow-up effect of the risk occurrence, respectively. In addition, a Choquet integral-based FMEA (Failure mode and effect analysis) model is introduced to assess and analyze the risks in which the descriptive language results are transformed into triangle fuzzy number to analyze the risk assessment results quantitatively, and the relative preference relations are used to rank the risk factors, empirical study is conducted focusing on Chongqing. Based on the proposed risk assessment model, the top 10 high-risk factors in green retrofit projects of public buildings are extracted, the possible potential causes are analyzed, and the corresponding risk response measures are proposed, which provide references for the improvement of risk management in urban renewal.

KEYWORDS:

• Public buildings
• green retrofit
• critical risks, improved FMEA (failure mode and effect analysis)

1. Introduction

China has issued the “Action Plan for Carbon Dioxide Peaking Before 2030” in 2021, which calls for accelerating the transformation of urban and rural construction, leading to green and low-carbon development, promoting urban renewal actions, and improving the quality of green and low-carbon development. Furthermore, in 2022, the Ministry of Housing and Urban-Rural Development of the People’s Republic of China put forward the goal of “completing more than 350 million square meters in the energy-saving renovation of existing buildings by 2025”. As the challenges facing cities with regard to urban fabric and urban function are enormous, it is becoming increasingly problematic in this field (Liu, Chen, and Wang 2021; Huang et al. 2020). Compared to new buildings, the existing buildings were built earlier, lack of building energy or green awareness in the architectural design and construction processes at that time, which resulting in that over 90% of the existing builidngs are becoming high energy consumption. In addition, public building renovation is an essential part of urban renewal, while public building green renovation projects are complex, which facing with more risks than those of new buildings (Baek and Park 2012). The retrofits of existing public buildings are still being promoted gradually, from single energy-saving renovation to comprehensive green renovation. Therefore, adopting a scientific risk assessment approach to identify and evaluate critical risks, and to improve the risk management of green retrofit projects in public buildings is essential to further promote the successful implementation of urban renewal.

Most existing public buildings have the problems such as high energy consumption, imperfect use functions, and large negative environmental impacts, while demolition and reconstruction of these buildings will not only cause damage to the ecological environment but also a great waste of energy resources. Therefore, green retrofit for public buildings can bring an amount of benefits compared to demolition and reconstruction. While, due to a series of obstacles including the lack of design information for existing buildings in renovation projects, the lack of a mature green renovation market, and the lack of unified standards in green retrofit, the risks of green retrofit project for public buildings are prominent. Current studies on risk management of green renovation in public buildings focus on the retrofit financing risks (Chen, Zhang, and Zhao 2021), construction safety risks (Li et al. 2020), and government risks (Alam et al. 2019), and most of these studies only analyze a certain type of risk in existing building renovation. In addition, a few research analyzes risk impacts and probability of occurrence in the whole process of green renovation projects (Ranawaka and Mallawaarachchi 2018). As for the methods of risk assessment, the current research has adopted analytic hierarchy process (Abdelgawad and Fayek 2010), fuzzy comprehensive evaluation method (Bo et al. 2020). However, the above mentioned methods are more suitable for evaluating the overall risk level of the project, the key risk factors in specific stages are still lacking attention. Concerning about the uncertainty and complexity of green retrofit in public buildings, there is the lack of comprehensive risk assessment from multi-dimension and the whole life cycle.

Therefore, this study aims to assess the risks in green retrofit projects of public buildings comprehensively and to identify the critical risk factors. Firstly, the risk factors will be identified, then an improved FMEA model based on Choquet integral will be applied to analyze and rank the risk factors. Finally, an empirical study on public building green renovation will be conducted in Chongqing, discussing and analyzing the possible potential causes of the critical risk factors and the corresponding risk response measures. The main contribution of this study lies in that a risk assessment method based on Choquet integral improved FMEA (Failure mode and effect analysis) is introduced, which assessing risk factors of public buildings green retrofit projects from three dimensions, including the probability of occurrence (O), the degree of severity (S), and follow-up effect of the risk occurrence (E), which has practical significance for the successful implementation of green renovation in urban renewal.

The rest of this paper is thus organized as follows. Section 2 presents the comprehensive literature review of existing building renovation and risk management; Section 3 introduces a risk assessment method based on the Choquet integral improved FMEA for green renovation of public buildings; Section 4 presents the empirical study of risk assessment in the green retrofit process of existing public buildings in Chongqing; Section 5 discusses the possible potential causes of critical risk factors and corresponding risk countermeasures and recommendations, followed by the conclusion in Section 6.

2. Literature review

Existing research on retrofit buildings focuses on the areas of energy efficiency retrofit and green retrofit, including the policies and risks of existing public and residential buildings.

The first category of the existing literature focuses on energy efficiency retrofit for existing buildings, mainly on the positive aspects of energy efficiency retrofits, the barriers and risks encountered in energy efficiency retrofits, energy-saving renovation technologies, and the optimal renovation strategies. For example, Mikulić et al. (2021) studied the socio-economic impact of public energy efficiency retrofitting investments in Croatia, an open input-output model is developed, which shows that investing in public building retrofit has positive economic benefits. In the process of energy-saving renovation projects, there are also many obstacles in various aspects, such as Energy retrofit efficiency, retrofit strategies, and retrofit policies. Firstly, Lai et al. (2022) conducted empirical modeling on the financial reporting of energy renovation investments in residential and commercial buildings in New York, aiming to contribute to a knowledge base that addresses financial and information barriers to building energy efficiency. Secondly, Alam et al. (2019) analyzed the barriers to government energy efficiency retrofitting and related strategies to address the obstacles. The main obstacles of the projects were as follows: lack of political will, financing agreements, departmental\institutional capacity, industry capacity, quality assurance, and uncoordinated incentives. Lastly, Zhang et al. (2021) studied the implementation, success, and obstacles of energy-saving renovation policies for existing buildings in different countries, providing valuable references for the formulation of renovation strategies. Meanwhile, Lu et al. (2017) constructed a behavior-based decision-making model to study contract decision-making. The results of the study showed that the rebound effect of tenants prolongs the payback period of Energy Saving Performance Contracts, and tenants’ risk attitudes significantly influence the contract duration. Talebiyan and Mahsuli (2018) proposed a risk analysis approach to quantify the risk to portfolio of buildings, and quantified the reduction of risk on retrofitting buildings by computing the previously mentioned sensitivity measures. Finally, they built a probabilistic model that predicts the repair cost and the retrofit cost for masonry school buildings. Piccinini et al. (2021) Benzar et al. (2020) conducted comparative studies by combining examples to identify potential technologies for energy-saving renovation and determine the best measures for energy conservation in energy-saving renovation. In order to find the best strategy for building energy-saving renovation, Pazouki et al. (2021) used mathematical models to consider economic indicators such as profit, initial cost, and investment payback period, as well as environmental goals such as energy conservation, clean use, and renewable energy, and Chandrasekaran et al. (2022) used simulation methods to consider energy efficiency, visual comfort, and economy.

The second category of existing study investigates the green retrofit of existing buildings, including residential buildings and public buildings. Green building projects advocate targeted sustainable development goals, and some specific risks in green building projects may lead to some adverse results (Koc, Kunkcu, and Gurgun 2023). The number of existing non-green buildings still far exceeds the number of green buildings and significantly negatively impacts the environment (Nguyen and Macchion 2022). Liu et al. (2021) took the Huazhong University of Science and Technology library as an example to establish an evaluation model based on the cloud model-TOPSIS method for evaluating public green retrofit buildings. They obtained the green ranking of renovation projects through engineering conditions, environmental protection requirements, and expert opinions to explore the implications of green retrofit technologies and policies on retrofit projects. Jun et al. (2021) analyzed the energy consumption of buildings actually used by public institutions against the background of green retrofit projects under the Green New Deal policy in Korea, and determined the relationship between the main policy direction and actual energy use to improve the energy efficiency of public buildings. Liu et al. (2020) researched existing retrofit policies and barriers in China and proposed relevant policies for addressing these barriers, as well as Galatioto et al. (2019) examined the effectiveness of national policies related to building renovation proposed for the refurbishment of historic public buildings in Italy, providing valuable reference for the government to formulate new policies to address obstacles and for other countries to initiate green renovation policies. With the promotion of green technologies and green policies, green retrofit projects in public buildings have gradually increased. In addition, compared with new construction, it is more likely to lead to deterioration of environmental adaptation and stability during actual project modifications (Alba-Rodríguez et al. 2017; Jun, Ahn, and Park 2021) and to generate other types of risks. Examples include energy efficiency uncertainty (Ranawaka and Mallawaarachchi 2018), procurement delay warranty cost risks affecting green retrofits (Ranawaka and Mallawaarachchi 2018), lack of new products to meet green retrofit requirements (Wang et al. 2018), and lack of experience with green retrofit design (Wang et al. 2018). Among the current studies on risk management for the green retrofit of public buildings, Chen et al. (2021) focused on the financing dilemma of green renovation, and developed the perceived payoff matrix and evolutionary game model of the government, Energy Service Companies (ESCOs), banks, and owners, Li et al. (2020) based on typical construction safety problems in public buildings renovation establish a construction safety risk assessment model of renovation project based on entropy-unascertained measure theory. Ranawaka and Mallawaarachchi (2018) combined the questionnaire data with the risk assessment matrix to evaluate the risks associated with the green renovation project in Sri Lanka, and formulated a risk response framework accordingly.

The above discussion shows that existing studies paid attention to the renovation technologies, retrofit correlation policies, and impacts of retrofit. Furthermore, there are few studies on the risks of green retrofit projects in existing public buildings (Jun, Ahn, and Park 2021). Most of the research that has been carried out focuses on certain types of risks in green retrofit projects of public buildings, only a few comprehensive risk studies have adopted subjective risk assessment methods to evaluate the probability and severity of the risk. Therefore, it is vitally important for considering the risks of green retrofit projects in public buildings comprehensively and multi-dimensionally. Thus, the significance of this research is to analyze the characteristics of risks from a multidimensional perspective and to assess the whole life cycle risks of public building green retrofits. By combining the subjective and objective methods, an improved FMEA with Choquet integral is developed to establish the risk assessment model, and the empirical study in Chongqing will contribute to pointing out the critical risk factors in the whole life cycle of public building green retrofits to derive the corresponding risk measures and recommendations.

3. Research methods

FMEA is a commonly used risk assessment method in previous studies, which taking the function of the occurrence rate (O), severity (S), and difficulty of detection (D) of failure or risk modes as the risk priority number (RPN) and ranking potential risks based on the estimation of RPN that can help in identifying all failure modes within a system, assessing impact, and planning for corrective actions, has been extensively adopted in various fields such as aerospace, machinery, automotive, and medical devices (Karamoozian and Wu 2020; Liu et al. 2018; Zheng, Liu, and Wang 2021). However, in the traditional FMEA method, there are three primary defects, one is that the three parameters (O, S and D) included in the RPN should have the same importance weight, whereas in practice, the importance of each parameter potentially is different; Second, the risk factor rating process may cause the same RPN value for different combinations of evaluation results for the three parameters, but actually the risk role of various failure modes may be completely different; Third, the three parameters are assessed with specific numerical values, and it is hard to express the factor assessment values as exact values in practical applications. To overcome the shortcomings of FMEA, scholars have focused on the improvement of this method. Zhou et al. (2021) proposed a new method based on FMEA, which aimed at prioritizing the potential failure modes and identifying the optimal design scheme of a substrate under grey environments. Abdelgawad and Fayek (2010) integrated FMEA with fuzzy logic and fuzzy analytical hierarchy process to risk management in the construction industry. To solve the problem of indicator interactions in multi-index evaluation, the Choquet integral theory with the characteristics of non-additive and nonlinear has attracted more attention (Marichal 2000). Choquet integrals are used to aggregate the relevant decision criteria, representing decision-maker preferences as well. They can be measured with the interaction between indicators, and are widely used in many fields such as software quality assessment, multi-criteria decision-making, and fuzzy measures (Li et al. 2013).

In order to investigate the risks in green retrofit projects of public buildings, this research adopts an improved FMEA model, in which the risk assessment is conducted by applying the FMEA model with the Choquet integral, and relative preference relationship for the priority number ranking is adopted to improve the effectiveness of the traditional FMEA model and reasonably avoid the above mentioned three major defects. A flowchart is presented in Figure 1 to illustrate the application of the proposed research methods.

Figure 1. The flowchart of the proposed Choquet integral FMEA method.

The main steps are as follows.

Stage I: Identify the risk factors. Identifying risk factors from a whole life cycle perspective and building a risk list for green retrofit projects in public buildings.

Stage II: Establish an improved FMEA model with Choquet integral. Firstly, evaluating the importance of the three parameters by experts. Secondly, using natural language to evaluate risk factors, transforming the descriptive language of risk decision-makers into Triangle Fuzzy Number for quantitative analysis, and using Choquet integrals to establish the interactions between parameters. Thirdly, using the Choquet integral formula, the fuzzy value of risk priority for each risk factor is calculated by combining the assessment results of steps first and second.

Stage III: Obtain the top 10 high-risk factors. The relative preference relationship between fuzzy numbers is selected to rank the priority of each risk, and the top 10 high-risk factors are listed and analyzed in depth.

3.1. Identification of risk factors of green retrofit project in public buildings

Green retrofit project of public buildings inevitably faces various risks from the whole life cycle, not only the common risks of public buildings but also the unique risks of green renovation. The Assessment Standard for Green Retrofitting of Existing Building GB/T 51141-2015 defines the green renovation of buildings as maintaining, updating, and strengthening existing buildings with the aim of saving energy, improving living environment, and enhancing usage functions, which provides a theoretical basis for the evaluation criteria of green retrofit project of public buildings. Thus, by analyzing and summarizing existing literature and case studies of green retrofit projects and existing building renovation, as well as referring to relevant standards such as Assessment Standard for Green Retrofitting of Existing Building GB/T 51141-2015, and General Code for Maintenance and Renovation of Existing Building GB 55022-2021, the risk factors are identified and categorized from the whole life cycle. As shown in Table 1, a total of 30 risk factors of public building green retrofit projects have been identified in four phases, namely, the decision stage, the design phase, the construction phase, and the operation phase.

Table 1. Risk factors of green retrofit projects in public buildings.

Phase	Code	Risk Factors	References
Decision Stage Risks	FM1	Risk of uncompleted policies	(Alam et al. 2019; Berghorn and Syal 2016; Lee, Lam, and Lee 2015; Wang et al. 2018)
	FM2	Risk of delays in government approvals	(Lee, Lam, and Lee 2015; Lu, Zhang, and Chen 2017; Wang et al. 2018)
	FM3	Interest rate exchange rate change risk	(Berghorn and Syal 2016)
	FM4	Risk of policy or standard adjustments	(Berghorn and Syal 2016; Lee, Lam, and Lee 2015; Lu, Zhang, and Chen 2017; Wang et al. 2018)
	FM5	Financing risks for renovation projects	(Alam et al. 2019; Berghorn and Syal 2016; Lee, Lam, and Lee 2015; Lu, Zhang, and Chen 2017)
	FM6	Risk of inadequate contractual agreements	(Liu, Wang, and Bai 2016; Berghorn and Syal 2016)
	FM7	Risk of increased consultation and retrofitting costs	(Liu, Wang, and Bai 2016)
	FM8	Risk of inaccurate measurement of existing building status	(Liu, Wang, and Bai 2016; Alam et al. 2019; Berghorn and Syal 2016; Lee, Lam, and Lee 2015)
	FM9	Fierce competition in the green retrofit market	(Liu, Wang, and Bai 2016)
	FM10	Risk of inaccurate projection of retrofit benefits	(Wang et al. 2018)
	FM11	Risk of inaccurate targeting of transformation	(Wang et al. 2018)
Design phase risk	FM12	Risk of deviation from the retrofit design plan	(Berghorn and Syal 2016)
	FM13	Risk of missing original design information for existing buildings	(Wang et al. 2018)
	FM14	Risk of the inexperience of design units for public building renovation	(Jafari, Valentin, and Bogus 2019; Wang et al. 2018)
	FM15	Risk of changing standards for public building renovation	(Wang et al. 2018)
	FM16	Lack of green awareness in public building renovation	(Alam et al. 2019; Ferreira and Almeida 2015; Jafari, Valentin, and Bogus 2019; Wang et al. 2018)
Construction phase risk	FM17	Force majeure risk	(Liu, Wang, and Bai 2016; Berghorn and Syal 2016; Lee, Lam, and Lee 2015)
	FM18	Insufficient experience in green renovation construction management	(Wang et al. 2018)
	FM19	The impact of renovation construction on the original building and surrounding buildings	(Chen 2019; Marique and Rossi 2018)
	FM20	Liquidity risks	(Berghorn and Syal 2016; Liu, Wang, and Bai 2016)
	FM21	Risk of improper selection of green retrofit technology for public buildings	(Liu, Wang, and Bai 2016; Berghorn and Syal 2016; Ferreira and Almeida 2015; Lee, Lam, and Lee 2015; Li et al. 2015; Wang et al. 2018)
	FM22	Risk of unreasonable selection of new materials or equipment	(Liu, Wang, and Bai 2016; Chen 2019; Lee, Lam, and Lee 2015; Wang et al. 2018)
	FM23	Lack of insurance related to green retrofitting of public buildings	(Lee, Lam, and Lee 2015; Wang et al. 2018)
Operational phase risks	FM24	The risk that the renovation does not meet the relevant norms and standards	(Liu, Wang, and Bai 2016; Beijing Municipal Commission of Housing and Urban-Rural Development, Beijing Major Projects Construction Headquarters Office 2020; Berghorn and Syal 2016; Chen 2019; Lee, Lam, and Lee 2015)
	FM25	Risk of improper green technology mastery by managers	(Beijing Municipal Commission of Housing and Urban-Rural Development, Beijing Major Projects Construction Headquarters Office 2020; Berghorn and Syal 2016; Li et al. 2015; Nguyen, Do, and Macchion 2021; Wang et al. 2018)
	FM26	Risk of material and equipment obsolescence	(Wang et al. 2018)
	FM27	Subject of interest responsibility for sharing unknown risks	(Lee, Lam, and Lee 2015; Wang et al. 2018)
	FM28	Risk of inadequate operation and maintenance	(Berghorn and Syal 2016; Lee, Lam, and Lee 2015; Wang et al. 2018)
	FM29	Risk of inappropriate green awareness and behavior of property owners	(Berghorn and Syal 2016)
	FM30	Risk of inaccurate measurement of operational energy consumption	(Liu, Wang, and Bai 2016; Wang et al. 2018)

3.2. Risk assessment model

In the process of FMEA analysis, RPN is mainly utilized for the quantitative analysis of the complex system, and each failure mode is ranked according to the RPN value. The magnitude of the RPN value is the risk level of failure mode (Fan et al. 2020).

3.2.1. Weights of the parameters in improved RPN

In this study, the risk assessment of public buildings green retrofit projects is conducted by considering the risk characteristics of the project and defining the parameters of improved RPN as the probability of occurrence of risk factors (O), the severity degree of risk factors (S) and the follow-up effect of the risk occurrence (E), where O is the likelihood that the risk factor will occur, S is the level of the negative effects of the risk factor on the project and the difficulty of its resolution, and E refers to the probability of other potential risks caused by risk, and the extent of its impact on the subsequent life cycle. In traditional FMEA methods, D is commonly used to evaluate the degree of fault detection before an aircraft fault occurs, nevertheless, in green renovation projects of public buildings, the risks are easily affected by human factors and policies, making it difficult to predict in advance the extent of all possible risks. Therefore, it is as an alternative to D in the traditional FMEA approach (Wang et al. 2018).

To determine the weights of the three parameters (O, S and E) in the RPN, the experts are invited to grade them, respectively, based on their own work experiences, with a score interval of [0,1]. The function of improved RPN is positively correlated with the importance of the parameters.

3.2.2. Risk assessment

Complied with the identified 30 risk factors, the experts in the building industry and the research field are invited to evaluate the risk level of each risk factor in terms of the probability of occurrence of the risks, the degree of severity of the risks, and follow-up effect of the risk occurrence, the above-mentioned parameters O, S and E.

In the risk assessment process, the complexity of the risk system, the ambiguity of the risk decision maker’s perception and experiences make it tough to quantify the risks using exact value. Therefore, natural language is devoted to assessing the magnitude of the risk factors. The evaluation includes five levels, very low, low, medium, high, and very high. Experts select five assessment levels with natural language levels for each risk factor, then the descriptive language is converted into triangular fuzzy number for quantitative analysis. Table 2 shows the definition of the risk assessment level and the triangular fuzzy number transformation.

Table 2. Risk assessment level and triangular fuzzy number.

Risk assessment level	Triangular fuzzy number
Very high (VH)	(0.75,1.00,1.00)
High (H)	(0.50,0.75,1.00)
Medium (M)	(0.25,0.50,0.75)
Low (L)	(0.00,0.25,0.50)
Very low (VL)	(0.00,0.00,0.25)

3.2.3. Calculation of RPN

In this study, fuzzy measure and Choquet integral are used to establish the interaction between the risk factors (Wang et al. 2018). Firstly, experts rate the three parameters between 0 and 1 to indicate their judgment on the weight of each parameter in the function of RPN. Subsequently, the Choquet integral value is used to indicate the magnitude of each risk factor in the RPN calculation.

Fuzzy measurement. In 1974, the Japanese scholar Sugeno first proposed to replace the additive class of set functions with relatively weak monotonicity, which is known as the fuzzy measure (Sugeno 1974).

Definition 1

(Tan 2011). Suppose the set of risk factors is X = (X₁, X₂, … , X_n), and X is a function of the power set of F(X) onto [0, 1], satisfying the condition,

1. φ(∅) = 0，φ(X) = 1;

2. There exist sets A and B when$\mathrm{\forall }$A, B ∈ φ(X), with A$\subseteq$B, then φ(A) ≤ φ(B).

φ is a λ-fuzzy measure defined on F(X) if the following equation is satisfied (Joshi and Kumar 2016; Sugeno 1974):

If $\mathrm{\forall }$A, B, A$\cap$B=∅, and the following equation is satisfied, φ is an additive measure on X. The expression is

(1) $\phi (A\cup B)=\phi \left(A\right)\hspace{0.17em}+\hspace{0.17em}\phi \left(B\right)\hspace{0.17em}+\hspace{0.17em}\lambda \phi \left(A\right)\phi \left(B\right)$(1)

λ∈[−1,$\mathrm{\infty }$), $\mathrm{\forall }$A, B∈φ(X)and A∩B=∅

The λ− fuzzy measure indicates the interaction between each set A and B. This interaction is determined by the parameter, which is provided as follows:

1. If λ>0, Equation (1)(1) $\phi (A\cup B)=\phi \left(A\right)\hspace{0.17em}+\hspace{0.17em}\phi \left(B\right)\hspace{0.17em}+\hspace{0.17em}\lambda \phi \left(A\right)\phi \left(B\right)$(1) is denoted as φ(A∪B) > φ(A) +φ(B), which indicates that there exists a multiplicative effect.

2. If λ = 0, Equation (1)(1) $\phi (A\cup B)=\phi \left(A\right)\hspace{0.17em}+\hspace{0.17em}\phi \left(B\right)\hspace{0.17em}+\hspace{0.17em}\lambda \phi \left(A\right)\phi \left(B\right)$(1) is denoted as φ(A∪B) = φ(A) +φ(B), which means that there is no connection between set A and B.

3. If λ< 0, Equation (1)(1) $\phi (A\cup B)=\phi \left(A\right)\hspace{0.17em}+\hspace{0.17em}\phi \left(B\right)\hspace{0.17em}+\hspace{0.17em}\lambda \phi \left(A\right)\phi \left(B\right)$(1) is denoted as φ(A∪B) < φ(A) +φ(B), which shows that the set {A, B}has a substitutive effect.

Definition 2

(Kojadinovic, Marichal, and Roubens 2005). Suppose X = (X₁, X₂, … , X_n) is a finite set, and the λ-fuzzy measure φ defined on F(X) can be expressed as the following equation.

(2) $\phi \left(\chi \right)=\{\begin{array}{cc}\frac{1}{\lambda }\left[{\displaystyle {\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)-1}\right]& \text{if}\lambda \ne 0\\ {\displaystyle {\sum }_{i=1}^n\phi \left(x_i\right)}& \text{if}\lambda =0\end{array}$(2)

Definition 3

(Kojadinovic, Marichal, and Roubens 2005). According to Equation (2)(2) $\phi \left(\chi \right)=\{\begin{array}{cc}\frac{1}{\lambda }\left[{\displaystyle {\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)-1}\right]& \text{if}\lambda \ne 0\\ {\displaystyle {\sum }_{i=1}^n\phi \left(x_i\right)}& \text{if}\lambda =0\end{array}$(2) , considering the boundary condition φ(χ) = 1, the unique λ value can be found in Equation (3)(3) $\lambda +1={\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)$(3) .

(3) $\lambda +1={\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)$(3)

Calculation of RPN based on Choquet integral. The magnitude of the fuzzy RPN is expressed by the Choquet integral value, and the mean value method is adopted to summarize the fuzzy risk priority number of each expert.

Definition 4

(Wang et al., 2018). Suppose φ is a fuzzy measure on χ, and the set of risk factors is X = (X₁, X₂, …, X_n), the discrete Choquet integral can be defined as the following equation.

(4) $fd\phi ={\displaystyle {\sum }_{i=1}^{n}f\left(x_{\left(i\right)}\right)\left[\phi \left(A_{\left(i\right)}\right)-\phi \left(A_{\left(i-1\right)}\right)\right]}$(4)

Among them,

The $\int fd\phi$ -integral value indicates the magnitude of the risk priority number (RPN);

$f\left(x_{\left(i\right)}\right)$ is a monotone non-decreasing function, i. e. $f\left(x_{\left(1\right)}\right)\ge f\left(x_{\left(2\right)}\right)\cdots \ge f\left(x_{\left(n\right)}\right)\ge 0$,

$A_{\left(1\right)}=\left\{x_{\left(1\right)}\right\},A_{\left(2\right)}=\left(x_{\left(1\right)},x_{\left(2\right)}\right),\cdots ,\text{break}{A}_{\left(n\right)}=\left(x_{\left(1\right)},x_{\left(2\right)},\cdots ,x_{\left(n\right)}\right)$

$f\left(x_{\left(ij\right)}\right)$ denotes the risk level assessment result of the ith risk factor influenced by the jth risk factor,

$\phi \left(A\right)$=$\{\phi$(O), $\phi$(S), $\phi$(E)}, where φ(O), φ(S), and φ(E) denote the importance of each of the three types of risk factors, respectively.

3.2.4. Ranking of the risk factors

The RPN obtained by Equation (4)(4) $fd\phi ={\displaystyle {\sum }_{i=1}^{n}f\left(x_{\left(i\right)}\right)\left[\phi \left(A_{\left(i\right)}\right)-\phi \left(A_{\left(i-1\right)}\right)\right]}$(4) is a fuzzy number, and the procedure of ranking fuzzy numbers should be conducted to calculate the final risk priority and rank the risk factors.

The relative preference relationship between fuzzy numbers is selected to rank the risk priority of each risk, and the relative preference relationship indicates the degree of preference of several fuzzy numbers to the mean value (Wang 2015).

Suppose X = (${\xi }_1$, ${\xi }_2$, … ${\xi }_n)$ is a group of $\bar{n}$ triangular fuzzy numbers, where ${\xi }_{\bar{l}}$ = (${\xi }_{\bar{l}}^l$, ${\xi }_{\bar{l}}^m$, ${\xi }_{\bar{l}}^u$), $\bar{l}$ = 1, 2, … $\bar{n}$. Assume $\bar{X}$ = (${\bar{\xi }}^l$, ${\bar{\xi }}^m$, ${\bar{\xi }}^u$) be the average value of the set X. The function ${\eta }_{P_{\bar{l}}}\left({\xi }_{\bar{l}},\bar{X}\right)$ indicates the relative preference relationship between $\bar{X}$ and ${\xi }_{\bar{l}}$. So,

(5) ${\eta }_{P_{\bar{l}}}\left({\xi }_{\bar{l}},\bar{X}\right)=\frac{1}{2}\left(\frac{\left({\xi }_{\bar{l}}^l-{\bar{\xi }}^l\right)+2\left({\xi }_{\bar{l}}^m-{\bar{\xi }}^m\right)+\left({\xi }_{\bar{l}}^u-{\bar{\xi }}^u\right)}{2\parallel T_{\mathrm{X}}\parallel }+1\right)$(5)

(6) $\Vert T_X\Vert =\{\begin{array}{cc}\frac{\left(T_{Xl}^{+}-T_{Xu}^{-}\right)+2\left(T_{Xm}^{+}-T_X^X{\overline{X}}^{-}\right)+\left(T_{Xu}^{+}-T_{Xl}^{-}\right)}{2}& \text{if}{T}_{Xl}^{+}\ge T_{Xu}^{-}\\ \frac{\left(T_{Xl}^{+}-T_{Xu}^{-}\right)+2T_{Xm}^{+}-T_X\overline{X}\text{+}\left({\text{T}}_{\text{Xu}}^{\text{+}}{\text{-T}}_{\text{Xl}}^{\text{-}}\right)}{\text{2}}\text{+2}\left({\text{T}}_{\text{Xu}}^{\text{-}}{\text{-T}}_{\text{Xl}}^{\text{+}}\right)& {\text{ifT}}_{\text{Xl}}^{\text{+}}{\text{T}}_{\text{Xu}}^{\text{-}}\end{array}$(6)

where $T_{Xl}^{+}$ = $ma{x}_{\bar{i}}\left\{{\xi }_{\bar{i}}^l\right\}$, $T_{Xm}^{+}$ = $ma{x}_{\bar{i}}\left\{{\xi }_{\bar{i}}^m\right\}$, $T_{Xu}^{+}$ = $ma{x}_{\bar{i}}\left\{{\xi }_{\bar{i}}^u\right\}$, $T_{Xl}^{-}$ = $mi{n}_{\bar{i}}\left\{{\xi }_{\bar{i}}^l\right\}$, $T_{Xm}^{-}$ = $mi{n}_{\bar{i}}\left\{{\xi }_{\bar{i}}^m\right\}$, $T_{Xu}^{-}$ = $mi{n}_{\bar{i}}\left\{{\xi }_{\bar{i}}^u\right\}$, $\bar{i}$ = 1, 2, … $\bar{n}$.

Overall, this research establishes a list of risk factors from the whole-life perspective of green retrofit in public buildings, and develops an improved FMEA model with the Choquet integral to assess the risk factors. RPN is used to calculate and rank the risk factors, in which three parameters O, S and E are involved. Risk assessment consists of two dimensions, the weights of the parameters and the grade of the risk factors. According to the transformation rules of fuzzy triangular number, the results of risk factors assessed at the natural language level are quantified. Utilizing the Choquet integral with the fuzzy measure, the fuzzy RPN is obtained. Finally, by selecting the preference relationship between fuzzy numbers, the priority of risk factors is ranked, and the ranking is the results of risk assessment.

4. Empirical study

By the end of 2020, Chongqing had completed a total of 11.74 million square meters of energy-saving renovation of public buildings. Whereas, with the rapid development of the economy, land use and building energy consumption in the urban area cannot meet the needs of social development obviously. The subsequent contradictions in development, land use inefficiency, poor living environment and other phenomena, will affect urban planning obviously. Therefore, in line with the established research methodology, an empirical study will be conducted in Chonqing, focusing on the risk factors in public building green retrofit projects. The experts in the fields of construction industry including real estate developers, construction units, supervision and consulting unit are invited to participate in a questionnaire survey to analyze and assess the risks in green renovation.

4.1. Background information

This study identifies a total of 30 risk factors in public building retrofit projects, so FM_t (t = 1, 2, 3, … 30) denotes the tth risk factor to be evaluated. A total of 25 experts are invited to participate in the risk evaluation, and DM_t (t = 1, 2, 3, … 25) denotes the tth expert involved in the evaluation. The practitioners and researchers related to the retrofitting public building projects are the potential respondents of the questionnaires, including the organizations of real estate developers, construction units, supervision and consulting units, design units, government sectors, and research institution. More than 50% of experts have over 6 years of working experience in renovation projects. Table 3 shows the specific number and proportion of experts in various organizations, as well as the distribution of their working experiences.

Table 3. Background information of experts.

	Number	Proportion
Type of Organization
Real estate developers	2	8%
Construction units	8	32%
Supervision and Consulting units	5	20%
Design Units	2	8%
Government Sectors	2	8%
Research Institution	6	24%
Years of Working Experience
3–5 years	12	48%
6–10 years	8	32%
10 years or more	5	20%

4.2. Results of the parameter weights

The fuzzy measures of the three parameters in RPN are obtained from the experts, as shown in Table 4. To illustrate the fuzzy measures of the parameters, the results of DM₆ assessment are taken as an example. The fuzzy measure of DM₆ can be expressed as

Table 4. Weights of the parameters in the case study.

Experts	Weights of the parameters			λ	Experts	Weights of the parameters			λ
	O	S	E			O	S	E
DM1	1	0.5	0.5	−1	DM14	0.2	0.4	0.4	0
DM2	0.4	0.5	0.4	−0.5845	DM15	0.3	0.4	0.2	0.3719
DM3	0.7	0.8	0.8	−0.9862	DM16	0.3	0.4	0.4	−0.2580
DM4	0.7	0.5	0.3	−0.7985	DM17	0.4	0.3	0.3	0
DM5	0.9	0.9	0.8	−0.9979	DM18	0.4	0.4	0.3	−0.2580
DM6	0.4	0.5	0.8	−0.9025	DM19	0.5	0.9	1	−1
DM7	0.3	0.4	0.3	0	DM20	0.9	0.6	0.7	−0.9857
DM8	0.2	0.5	0.3	0	DM21	0.5	0.2	0.3	0
DM9	0.8	0.9	0.6	−0.9909	DM22	0.4	0.5	0.3	−0.4516
DM10	0.3	0.4	0.2	0.3719	DM23	0.6	0.4	0.2	−0.4796
DM11	0.5	0.8	0.7	−0.9606	DM24	0.5	0.8	1	−1
DM12	0.9	0.6	0.6	−0.9800	DM25	0.9	0.5	0.6	−0.9736
DM13	1	0.5	0.5	−1

$\phi \left(A\right)$=$\{\phi$(O), $\phi$(S), $\phi$(E)}= $\left\{0.4,0.5,0.8\right\}$

where the parameter λ = −0.9025 is calculated from Equation (3)(3) $\lambda +1={\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)$(3) .

4.3. Results of RPN calculation

According to the calculation results in section 4.2, the fuzzy RPN can be calculated using Equation (3)(3) $\lambda +1={\prod }_{i=1}^n\left(1+\lambda \phi \left(x_i\right)\right)$(3) . The mean value method is used to summarize each expert’s fuzzy RPN. At last, the parameters $\bar{X}$ are calculated and the parameter $\parallel T_X\parallel$ is figured out with Equation (6)(6) $\Vert T_X\Vert =\{\begin{array}{cc}\frac{\left(T_{Xl}^{+}-T_{Xu}^{-}\right)+2\left(T_{Xm}^{+}-T_X^X{\overline{X}}^{-}\right)+\left(T_{Xu}^{+}-T_{Xl}^{-}\right)}{2}& \text{if}{T}_{Xl}^{+}\ge T_{Xu}^{-}\\ \frac{\left(T_{Xl}^{+}-T_{Xu}^{-}\right)+2T_{Xm}^{+}-T_X\overline{X}\text{+}\left({\text{T}}_{\text{Xu}}^{\text{+}}{\text{-T}}_{\text{Xl}}^{\text{-}}\right)}{\text{2}}\text{+2}\left({\text{T}}_{\text{Xu}}^{\text{-}}{\text{-T}}_{\text{Xl}}^{\text{+}}\right)& {\text{ifT}}_{\text{Xl}}^{\text{+}}{\text{T}}_{\text{Xu}}^{\text{-}}\end{array}$(6) . According to Equation (5)(5) ${\eta }_{P_{\bar{l}}}\left({\xi }_{\bar{l}},\bar{X}\right)=\frac{1}{2}\left(\frac{\left({\xi }_{\bar{l}}^l-{\bar{\xi }}^l\right)+2\left({\xi }_{\bar{l}}^m-{\bar{\xi }}^m\right)+\left({\xi }_{\bar{l}}^u-{\bar{\xi }}^u\right)}{2\parallel T_{\mathrm{X}}\parallel }+1\right)$(5) , the RPN of each risk factor is obtained and ranked.

$\bar{X}$= (0.2366, 0.4604, 0.7013), $\parallel T_X\parallel$ = 0.9519.

${\eta }_{P_1}\left(FM1,\bar{X}\right)=\frac{1}{2}\left(\frac{\left(0.2996-0.2366\right)+2\ast \left(0.5189-0.4604\right)+\left(0.7499-0.7013\right)}{2\ast 0.9519}+1\right)=0.57350$

The rest of the calculation and ranking results are shown in Table 5.

Table 5. Results of RPN calculation.

Risk Factors	FMEA model with the Choquet integral
	O	S	E	Fuzzy RPN	Ranking
FM1	0.2996	0.5189	0.7499	0.57350	6
FM2	0.2942	0.5260	0.7600	0.57705	5
FM3	0.0968	0.2972	0.5472	0.30807	30
FM4	0.2671	0.4886	0.7291	0.53664	11
FM5	0.3361	0.5407	0.7602	0.60537	3
FM6	0.3348	0.5765	0.8085	0.63538	1
FM7	0.2807	0.5282	0.7637	0.57235	7
FM8	0.2181	0.4389	0.6889	0.47672	19
FM9	0.2866	0.5249	0.7274	0.56414	10
FM10	0.2480	0.4636	0.7051	0.50833	15
FM11	0.2319	0.4439	0.6889	0.48619	17
FM12	0.2318	0.4498	0.6980	0.49149	16
FM13	0.2873	0.5143	0.7463	0.56419	9
FM14	0.2073	0.4403	0.6793	0.46956	21
FM15	0.2194	0.4324	0.6639	0.46766	22
FM16	0.1431	0.3271	0.5751	0.35345	29
FM17	0.3083	0.5311	0.7597	0.58655	4
FM18	0.2566	0.4901	0.7356	0.53384	12
FM19	0.2510	0.4854	0.7304	0.52742	14
FM20	0.3087	0.5520	0.7915	0.60557	2
FM21	0.1745	0.3973	0.6473	0.42325	25
FM22	0.1551	0.3800	0.6270	0.39955	27
FM23	0.1415	0.3508	0.6008	0.37128	28
FM24	0.2220	0.4395	0.6895	0.47917	18
FM25	0.1886	0.4271	0.6771	0.45304	23
FM26	0.2165	0.4390	0.6890	0.47606	20
FM27	0.2773	0.5203	0.7703	0.56842	8
FM28	0.2564	0.4879	0.7304	0.53137	13
FM29	0.1971	0.4201	0.6676	0.45129	24
FM30	0.1605	0.3800	0.6300	0.40310	26

5. Discussions

This section discusses the critical risk factors in public buildings green retrofit projects and their possible potential causes, and corresponding risk countermeasures and suggestions are proposed.

In the case study, the top 10 risk factors are identified as risk of imperfect contract agreement (FM₆), risk of capital turnover (FM₂₀), risk of renovation project financing (FM₅), risk of force majeure (FM₁₇), risk of delay in government approval (FM₂), risk of uncompleted policy (FM₁), risk of negotiation and renovation cost increase (FM₇), risk of unclear sharing of responsibility among interest parties (FM₂₇), risk of missing original design information of existing buildings (FM₁₃), and risk of fierce competition in the green renovation market (FM₉).

5.1. Potential causes of risk

Among them, there are six risk factors belong to the decision-making stage, risk management in this stage directly affects the overall quality and the construction level of the project. Avoiding high-risk factors as early as possible plays an indirect role in facilitating the subsequent risk control, reducing the probability of other risks in the subsequent process of the project as well. In the design stage, 1 risk factor “The lack of original architectural design information” is involved, which will directly impact the renovation design process, the schedule of the project, and the cost. The green retrofit project of public buildings is complex and has a long construction period, 2 risk factors are related to this phase. In the construction phase, it inevitably faces risks from the internal and the external environment, which will directly influence the construction efficiency and quality of the renovation project (Braulio-Gonzalo, Jorge-Ortiz, and Bovea 2022). The operation stage is to maintain and manage the daily operation of the renovation project after the completed project, in which 1 risk factor is involved. Due to the large number of participants and complex responsibilities, the renovation effect and operation income of public building green retrofits is prone to be affected severely.

Contractual imperfection risk (FM₆) is the top one risk factor in the renovation project. When signing the contract, the insufficient binding force or the incomplete contents can lead to the failure to protect the rights and interests of all participants involved in the renovation project (Berghorn and Syal 2016). The imperfection of the contract content includes a series of specific unclear clauses such as the scope of the renovation project and the return mechanism, as well as the unclear contract system including the project contract and procurement contract. Inadequate contract agreements that are not supplemented timely may lead to deep-rooted risks. In addition, if the risk responsibilities of all the participants failed to be clearly agreed on, it may lead to unclear or uneven risk-taking and the disputes among the participants.

Project materials and equipment procurement require large amounts of capital, which susceptible to policy and interest rate changes. At present, the EPC is the primary way to implement public building retrofits in China, while most of the energy-saving service companies are small firms with a unsound financial system and creditable assets, low-rated bank credit rating, and complicated procedures for loans and guarantees application (Wang et al. 2019). As the public building green retrofit projects generally have a long payback period, most of the energy-saving companies have high pressure financing and capital turnover, which leading to the projects facing huge capital turnover risk (FM₂₀) and renovation project financing risk (FM₅).

Due to climate changes or natural disasters, the delays of the construction period in green renovation projects will generate, which inevitably resulting in increasing costs, and the projects are suffering from force majeure risks (FM₁₇).

As there is lack of clear approval regulations for urban renewal in China, the responsible departments for project construction varies, and the requirements and processes of the renovation scheme are different (Arslan et al. 2020; Mastrucci et al. 2020). Since the government sectors have no exact references for approving such projects, it faces a high risk of delay in government approval (FM₂). As a result, the project approval may be delayed, which indirectly causes the schedule is fell behind.

Policy changes during the whole life cycle of public building green retrofit projects will impact the project comprehensively. As an emerging industry, the relevant legal policies, regulations, and supervision systems need further improvement (Alam et al. 2019). Simultaneously, the lack of government-related subsidy policies prevents the market investment from being incentivized. In addition, the current public building renovation is exposed to the pressure of carbon reduction under the target of “carbon peak and carbon neutralization”. To achieve the target of green development, innovations on urban-rural construction paths should be implemented in urban planning, and relevant legal policies and technical standards should be modified based on the requirements of carbon reduction (Huang et al. 2020). Hence, the public building renovation is also hindered by the risk of uncompleted policies (FM₁) significantly.

There are a variety of transaction costs in the green retrofit process. Except for the equipment rents and the renovation costs, it also includes expenses arising from multi-party negotiation and multi department approval. Due to the decentralized property rights of some public green renovation projects, the negotiation and renovation costs increase accordingly. The project faces the risk of increased negotiation and renovation costs (FM₇), which may bring the consequence of increased investment costs.

The focus of the stakeholders in the responsibility sharing and the behavior selection varies in different stages of the green retrofit projects. Consequently, the aim of responsibility sharing among participants is to reach the benefit maximization in renovation projects. If the responsibility sharing is not explicit in the project, it is difficult for the government, the owners, and other stakeholders to carry out their duties effectively, which will cause the risk of unclear sharing of responsibility among stakeholders (FM₂₇) and hindering the development of the public building green retrofit market.

Since some existing public buildings have been operated for a long time, the original design drawings, equipment operation records, and measurement data of energy-using systems may loss or incomplete (Wang et al. 2018). Therefore, there is the risk of missing the original design information of existing buildings (FM₁₃). Once such information misses, it will greatly increase the difficulty of public building green renovation work.

The green renovation market, as an emerging industry, has tremendous potential for development to form a large market gradually, which generating the risk of intense competition in the green renovation market (FM₉), resulting in the squeeze of the profit margins (Liu, Wang, and Bai 2016), and leading to a viciously competitive environment where the market order is disturbed.

5.2. Risk response measures and recommendations

Based on the risk analysis and assessment results, the following risk response measures and recommendations are summarized.

To avoid the risk of contractual imperfection (FM₆), the contract should clearly specify the explicit clauses, containing project scope, return mechanism, price mechanism, performance standards, and evaluation methods. Meanwhile, it requires to improve the contract system, such as project contracts, financing contracts, operation contracts, and procurement contracts (Ranawaka and Mallawaarachchi 2018). In the contract, it is recommended to specify the time limit for the fulfillment of obligations to avoid the risk of force majeure (FM₁₇), and reasonably delineate the boundaries of participants’ responsibilities. The participants should uphold the spirit of contract and bear the risks and liabilities after signing the contract.

There is a proposal to build the public building green retrofit of management system comprehen-sively, and promote the transformation of the public buildings from energy-saving renovation to green renovation. The government can develop the intelligent upgrading of urban infrastructure, and further implement the construction of the city information model (CIM) platform. The application of “CIM+” can enhance the level of urban public building renovation and operational efficiency, as well as achieve effective project capital control, and avoid the risk of capital turnover (FM₂₀) and renovation project financing risk (FM₅).

To achieve multi-channel financing, it needs the renovation project of finance by introducing social capital. In order to reduce the repayment of financing loans to companies, banks offer low-interest loan operations for green renovation projects (Chen, Zhang, and Zhao 2021). The contractual agreements can diversify the risk of financing renovation projects, and the insurance companies can transfer the risk (Baek and Park 2012). Meanwhile, in advance to avert the risk of financing renovation projects (FM₅), it is certainly paying close attention to the national financial policy dynamics and preparing a financing feasibility analysis report.

In response to the risk of uncompleted policies (FM₁), the government could formulate relevant policies, and develop the overall standards and regulations for green retrofit programs, which promote the successful implementation of retrofit projects (Liu et al. 2020; Tan et al. 2018). The government formulates a set of management and control rules, so that can guide the implementation and avoid the risk of delay in government approval (FM₂). The government, as the policy maker, urges participants to reach an agreement and maintain a fair environment. The government implements a series of preferential policies including tax incentives, loan incentives, and financial subsidies, which encourage enterprises to participate in and further promote the market-oriented operation of green renovation.

The following measures can be taken to deal with the risk of unclear responsibility sharing among stakeholders (FM₂₇). Based on different stages of the responsibility sharing characteristics, all stakeholders have a dynamic sharing mechanism that comes into operation during different stages of green renovation projects. In addition, they should reasonably share the responsibility, which can realize revenue and liability sharing, and promote the benign development of the cooperative relationship of green renovation bodies (Wade, Bush, and Webb 2020).

In response to the risk of negotiation and renovation cost increase (FM₇), Big Data technology can be used to collect information, and properly promote the concept of green building projects to the public. It can raise awareness of green renovation to the public by popularizing relevant knowledge and the concept of green building correctly, which can achieve the purpose of reducing consultation costs. For the risk of missing original design information of existing buildings (FM₁₃), the original design data of the existing part can be properly archived; With the missing design data, construction unit should enhance the communication efficiency with the original design unit based on fully understanding the original design information, and reduce the project cost.

The retrofit projects face the risk of fierce competition in green renovation market (FM₉). To avoid excessive competition and waste of resources, it is recommended to build a market order for healthy competition and win-win cooperation among green renovation enterprises. Besides, there are some measures to enhance the enthusiasm of enterprises to participate in. It can be proposed to prevent enterprises from monopolizing competition, to enhance the division of labor and collaboration within enterprises, and to maintain the order of the green renovation market.

6. Conclusions

With the increasing number of aged buildings, the green retrofit of public buildings is an effective way in the urban renewal process. Many efforts have been made to promote green retrofit, and a part of the study carried out risk management research. Moreover, studying on risks of the green retrofit project of existing public buildings is conducive to improving the risk management level. However, few studies have been conducted on comprehensive risk assessment from multi-dimension and the whole life cycle. Therefore, this study aims to assess the risks in green retrofit projects of public buildings comprehensively and to identify the critical risk factors.

This research identifies 30 risk factors in green retrofit projects of public buildings, and assesses the level of risk factors from three dimensions, including the probability of risk occurrence (O), severity degree of risk (S), and follow-up effect of the risk occurrence (E), and the risk factors are quantified based on triangular fuzzy numbers. To assess the risk factors, an improved FMEA assessment model with Choquet integral is developed, and the preference relationship between fuzzy numbers is selected to rank the RPN. Based on the background of existing public building green retrofit projects in Chongqing, a case study is conducted, the risk assessment is carried out by the improved FMEA method to identify the top10 risk factors, and the possible potential causes of critical risk factors are discussed, and the risk countermeasures and recommendations are proposed. For instance, building market orders for healthy competition among green renovation enterprises, making a series of relevant green retrofit policies by the governments, and offering low-interest loan operations by the banks, all of these measures have great practical significance for public building green retrofit, which promoting the green renovation behaviors in public building retrofit. The results of this work can provide suggestions for the participants in green retrofits to carry out efficient risk control measures, and provide references for the government sectors to issue relevant policies and sound specification standards. However, there are still some limitations in this study that can be investigated in future works, the coupling mechanism between the risks factors in green retrofit projects of public buildings is not considered. Therefore, for future research, dynamic risk assessment can be carried out on the risk of green retrofit of public buildings to further explore its risk influence mechanism.

Acknowledgements

The authors are grateful to all the participants in the questionnaire survey.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Chongqing Municipal Education Commission Project [KJQN202300732]; Ministry of Education Humanities and Social Sciences General Project [21YJCZH048]; Chongqing Natural Science Foundation project [CSTB2023NSCQ-MSX0724].

Notes on contributors

Xiaosen Huo

Xiaosen Huo was born in Xinxiang City, Henan Province, P. R. China in 1990. She is the associate professor in School of Economics and Management of Chongqing Jiaotong University. Her current research interests include risk management in green retrofit projects and infrastructures. E-mail: [email protected]

Tong Hao

Tong Hao was born in Chongqing City, P. R. China in 1998. She is a postgraduate student and is currently pursuing the master degree in Management Science and Engineering in School of Economics and Management of Chongqing Jiaotong University. E-mail: [email protected]

Liudan Jiao

Liudan Jiao was born in Chongqing City, P. R. China in 1989. He is the associate professor in School of Economics and Management of Chongqing Jiaotong University. His current research interests include urban resilience and rail transit resilience. E-mail: [email protected]