A meta-network-based risk evaluation and control method for industrialized building construction projects

Journal of Cleaner Production 205 (2018) 552e564

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

A meta-network-based risk evaluation and control method for industrialized building construction projects Tao Wang a, Shangde Gao a, Xiaodong Li b, *, Xin Ning c a

Department of Engineering Management, School of Management Science and Engineering, Central University of Finance and Economics, Tupei Building 403, South College Road 39, Haidian District, Beijing 100081, China b Department of Construction Management, School of Civil Engineering, Tsinghua University, Beijing 100084, China c School of Investment & Construction Management, Dongbei University of Finance and Economics, Dalian, Liaoning Province 116025, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 January 2018 Received in revised form 3 August 2018 Accepted 15 September 2018 Available online 17 September 2018

Construction methods in developing countries such as China are gradually industrializing due to rising labor costs, new techniques, tools, procedures and management methods. This makes it essential to develop new risk evaluation and control methods. Previous research seems to ignore the risk management of the application of new construction technologies and procedures. The identification of risk mechanisms is unclear, leading to a lack of clear guidance for risk avoidance and control. This paper describes a network model based on meta-network analysis of project objectives, risk events, risk factors and stakeholders in the construction process of building industrialization. According to ISO 31000, the main processes of risk management in the model has three parts: risk identification, risk analysis and evaluation, and risk treatment and control. Risk factors are identified from previous literature and site investigation, and the indirect impact on project objectives is analyzed and calculated with the networks in the meta-network. The order of importance is evaluated and used as the foundation of risk treatment and control. The analysis of the crucial risk factors makes it possible to identify the stakeholders who influence them, leading to suggestions for relevant control strategies. A residential building construction project in South China that uses the building industrialization construction system serves as a case study to verify the feasibility and applicability of the risk management system. Results show that critical risk factors are construction-related factors and design-related factors. The water and electricity engineers, the project manager, the project secretary, the field engineer, the project supervisor and the main contractor engineer have a significant influence on the risk factors. A stakeholder supervision system with targeted risk control strategies is proposed. © 2018 Published by Elsevier Ltd.

Keywords: Building industrialization Risk identification Risk evaluation Risk control Meta-network analysis Project stakeholders

1. Introduction Building industrialization has had various definitions and concepts in each period along with increasing demands and economic necessities. It is the integration of a wide range of new concepts and techniques, including e but not limited to e automation, robotics, reproduction, preassembly, standardization, mechanization, prefabrication and off-site construction (Zabihi, 2013). Industrialized construction has had a significant influence on improving efficiency, quality and the environmental impact of projects (Aye et al., 2012; Gibb, 1999; Wong and Yeh, 1985). The most significant

* Corresponding author. E-mail address: [email protected] (X. Li). https://doi.org/10.1016/j.jclepro.2018.09.127 0959-6526/© 2018 Published by Elsevier Ltd.

advantage of prefabrication is that better supervision on improving the quality of prefabricated products can be conducted (Tam et al., 2007). The movement of the construction industry toward industrialization in the fabrication system includes the trend of helping the building industry achieve benefits and share them with the end-user (Martinez et al., 2008). This trend gains considerable attention and support from the government in China (Ji et al., 2017). In the transformation process of building industrialization, new concepts and techniques create greater requirements for project management for several reasons. First, compared to the traditional construction process, construction workers typically need to learn new techniques and become more familiar with the process of prefabricated building components assembly. Second, prefabricated components and new construction equipment (e.g., selfclimb form, which is an intelligent wall formwork) greatly increase

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

the construction speed and enhance the flexibility of project planning (Jaillon and Poon, 2009). New management modes, such as interspersed construction e a new construction planning technology (Feng and Hu, 2016) e, have been applied to project management and have had significant influences on project duration and cost. Building industrialization brings positive changes in the construction process (Lovell and Smith, 2010) as well as complexity and uncertainty to project implementation. Industrialized construction projects are generally subjected to more risks compared to traditional projects (Karimiazari et al., 2011; Luo et al., 2015). Risk factors (e.g., the seam material is not securely fastened) increase with the new construction system of building industrialization (Luo et al., 2015), and they are more complex as more relationships exist among project objectives, risk events and risk factors. Some risk events can even lead to the failure of realization of project objectives, and risk factors can lead to risk events, so more strict requirements for risk management are necessary. Thus the relationships among project objectives, risk events and risk factors need to be considered (Bu-Qammaz et al., 2009; Fang and Marle, 2015; Fang et al., 2012; Guan and Guo, 2014; Luo et al., 2015; Tavakolan et al., 2017). However, there are still some limitations that need to be addressed. First, effective risk control methods are necessary for the successful risk management of industrialized building construction projects, while the risk management and control skills of managers of different levels (project managers, field engineers, supervision engineers) may fail to meet the requirements of a new construction management system (Liu, 2017). But very limited studies focus on proposing targeted risk control methods based on how the risk factor lead to the deviation of project objectives and affect each other (Kuo and Lu, 2013; Wang et al., 2016). Second, previous project risk management has analyzed and assessed risks from the perspective of the whole project (e.g. economics risks, market risks, risks from the external environment) (Luo et al., 2015; Xiahou et al., 2018), and there has been little attention focused on the potential risks associated with the implementation of building industrialization, which are listed in the case study of this paper. Third, a substantial proportion of researchers in this field have analyzed only the direct influences of risks (Luo et al., 2015), while overlooking the interaction between one risk event and other risk events (Han et al., 2008) (e.g. the interaction between “design changing” and “unreasonable construction planning”). Assumptions for some research methods of risk management limit certain kinds of associations. For example, the analytic hierarchy process (AHP) assumes that risk factors are isolated, which is the prerequisite of the comparison for the factors' weight judgment (Saaty, 2004). In the Bayesian network analysis process, closed loops among risk factors and events are not considered (Borsuk et al., 2004). In order to bridge the knowledge gaps of risk management and control of industrialized building construction projects, our research was based on meta-network analysis (MNA); thus, it is included a network analysis model composed of stakeholders, risk events, risk factors and project objectives. The meta-network was a complex network composed of various entities and connections among them (Li et al., 2015). MNA can conceptualize a project as multi-node with multi-link meta-networks composed of different node entities and their interdependencies (Zhu and Mostafavi, 2015). MNA has been applied in many fields, such as the social networks and supply chains in the e-commerce market (Wakolbinger and Nagurney, 2004), the integration of social networks with knowledge networks (Nagurney and Dong, 2005), a social media analysis (Carley et al., 2013a) and an assessment of public health systems (Lenz, 2012). The goal of our research was to systematically study the risk assessment and responding mechanisms for the deviation from project objectives in the application of

553

building industrialization. In this model, the association of risk events and factors that led to the deviation of project objectives was revealed, and targeted risk control strategies e based on the analysis of the relationship between stakeholders and risk factors e are proposed. The structure of this paper is as follows: In the next section, previous studies of risk management in building industrialization projects are reviewed, and the research gap is analyzed. In section 3, the establishment of the meta-network and our calculation methods are described in detail. To show how the risk evaluation and control model can be applied, a case study of a residential building construction project in South China, which applied a building industrialization construction system, is discussed in section 4. Related results and future work are described in section 5. 2. Literature review 2.1. Building industrialization and its influence on construction projects Previous research has shown that building industrialization has several notable characteristics. The first is standardization e including in the design phase, component production, actual construction and management. Standardization has a significant influence on the improvement of the quality of construction projects in the existing building system (Roy et al., 2005). Standardization and generalization of building components leads to the second characteristic: prefabrication in factories (Jaillon and Poon, 2009). Prefabrication refers to the mass production of building components in factories; these components are then delivered to a project site to be used in construction. The third characteristic is scientific management (Yan et al., 2004). Overall planning and new technology are the bases of scientific management in industrial construction. Building industrialization has been widely adopted around the world (Pan and Sidwell, 2011). According to Mao et al. (2013), in 1996, the levels of prefabrication in the construction industry in Germany, the Netherlands and Denmark already were 31%, 40% and 43%, and the size of the industrialized construction industry in the U.K. was £6 billion in 2006. Compared to these developed countries, building industrialization in China is still in its infancy (Luo et al., 2015), and developers in China lack the experience of industrialized construction approaches (Zhang and Skitmore, 2012). Building industrialization brings changes to the construction process and affects the achievement of project objectives. Technology associated with building industrialization e especially the use of prefabricated building components, such as prefabricated slabs, beams, precast concrete (PC) roofs, PC balconies and integrated bathrooms e has been gradually introduced in the residential building construction industry in China (Gan et al., 2017). This has had several effects on the construction process, including compressing construction duration (Goodier and Gibb, 2007), increasing building costs (Chiang et al., 2006; Lihong et al., 2013; Mao et al., 2013) and improving building quality (Gan et al., 2017). Assembled building provides an example to illustrate the changes and effects brought by building industrialization. In traditional construction processes, waste arises from design changes, design error, materials remaining, packaging and nonrecyclable consumables, and inclement weather (Faniran and Caban, 1998). Xu and Zhao (2010) proposed that assembled building could significantly reduce the disadvantages of traditional building processes while decreasing construction time, costs and environmental impact yet increasing quality. However, at this time, building industrialization is more costly than the conventional construction approach. By comparing assembled building projects

554

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

and cast-in-place projects in a case study, Lihong et al. (2013) discovered that the cost of the assembly construction process was 12.26% higher than it was for traditional projects. The main reason was related to the prefabricated components and installation, for there it was costlier to purchase PC components, and the salaries of workers who had experience with installation were higher than those without this specialized experience. In terms of construction safety, Rubio-Romero et al. (2014) concluded that depending on the project type, the safety of a prefabricated building might not be as good for a traditional building. In regard to environmental impact, by employing a life-cycle assessment (LCA), Hong et al. (2016) determined that building industrialization (assembled buildings) could reduce 4%e14% of the total life-cycle energy consumption, which covers the whole life cycle of the prefabricated components, including prefabrication manufacturing, transportation, on-site assembling and recycling in the demolition phase. With a similar method, Pons and Wadel (2011) analyzed a school project that adopted building industrialization in Catalonia, and the project made outstanding achievements in energy conservation and emission reduction. Overall, compared to traditional projects, building industrialization has improved efficiency and effectiveness, but it is still much more difficult for some projects to achieve their objectives (e.g. objectives about project cost, duration, quality, safety and environmental impact). A life-cycle assessment suggests that it is possible for the total cost to be lower, but in actual construction processes, the costs are much higher because of the use of precast concrete components and the lack of experienced professionals (Lihong et al., 2013). 2.2. Risk management methods of construction projects Changes in the achievement process of the project objectives leads to changes in the risk system. Therefore, successful risk management is critical. Previous research has described many risk management methods. Elsawah et al. (2016) adopted a risk matrix combining probability and influence from expert judgment. Zhang and Zou (2007) integrated fuzzy logic and the analytic hierarchy process in risk evaluation of critical infrastructure construction. Abdelgawad and Fayek (2011) combined fuzzy logic and fault tree analysis to reveal the sources of risks. Lin et al. (2015) built a multihierarchy risk analysis framework with risk matrices and applied it to risk examination and evaluation. Zhou and Zhang (2011) applied Fuzzy Bayesian Analysis (FBA) with experts' experience and a related construction database to deep basement projects and found that FBA had a significant influence on risk avoidance. Stergiopoulos et al. (2016) found that fuzzy logic and risk dependence charts could be used to effectively analyze the interdependence between different projects, and critical factors could be found by simulations of specific scenarios. To study building industrialization, Bari et al. (2012) analyzed and evaluated the types of critical factors in the construction process of building industrialization projects, including project factors (e.g. repeated use of design, molds or construction techniques from previous projects), economic factors and contractor factors (e.g. a contractor's staff and workers). Based on the related literature, Gan et al. (2017) proposed that in China, there were three kinds of risk factors related to project quality: design-related factors, production-related factors, and construction-related factors. Based on a Strengths, Weaknesses, Opportunities and Threats (SWOT) analysis, Jiang et al. (2018) analyzed 107 scientific papers and 85 government documents and determined the opportunities and challenges of building industrialization in the new urbanization in China. Jiang mentioned that the main challenges at this stage included high upfront costs, low collaboration between experts and project participants, incomplete regulations and standards, and low

market acceptance. Li et al. (2017) applied a system dynamics model to analyze the investment risk management of assembled buildings in China and classified the risk factors into five categories: economic risks, internal risks, technical risks, policy and legal risks, and market risks. They proposed related risk control strategies. From this brief literature review, we can observe that previous research has focused on total market risks, and there has been a lack of identification of the risk factors inherent in adopting the industrialized construction approach in the building construction process. Quantitative analysis of the interactions between risks has been insufficient, and risk response strategies after the risk assessment have not been given enough attention. We identified the following problems from our consideration of previous studies on the industrialization of buildings and the risk management studies of construction projects: (1) the level of achievement of project objectives in building industrialization projects was different from traditional projects, and the risk system should be changed; (2) most previous project risk research considered each risk as an independent event, and the correlation between risks has received limited attention; and (3) risk management research has usually stopped at the risk evaluation phrase with little attention given to risk response strategies. 2.3. Meta-network analysis Notably, meta-network analysis (MNA) e which can conceptualize construction projects as multi-node and multi-link metanetworks composed of different node entities and their interdependencies e can be applied to analyze the risk association and to propose risk control strategies (Zhu and Mostafavi, 2015). The structure of MNA includes the identification of entities and associations between these entities, and it proposes an analysis based on these associations and entities. The advantages of this method include the following: (1) risks are regarded as the relationships between different entities, and the MNA, which used the network in the analysis, is more intuitive than previous studies; (2) the mechanism of risk control by stakeholders' direct intervention in regard to the risk factors is revealed; (3) MNA provides an effective way to express the complex interactions of various factors involved in the project and extends the scope of analysis to multiple dimensions and forms an integrated project network that covers the relationships between the various influencing factors; and (4) MNA provides multi-dimensional perspectives and the ability to construct complex models that can accurately measure and diagnose the performance of project goals and tasks, helping to optimize task scheduling and distribution (Zhu and Mostafavi, 2016). MNA has been applied in some research. Li et al. (2015) compared construction projects of 11 auto dealers and built a meta-network with entities that included stakeholders, information and tasks. Zhu and Mostafavi (2015) put forward a comprehensive framework of organizational vulnerability assessment for complex construction projects to analyze the problem of cost management. In these studies, each project was broken down into different entities, such as stakeholders, information, resources and tasks, and the meta-network was formed with links between these entities. By analyzing the relationship between various types of project factors, MNA can conduct risk analysis and assessment under the condition of uncertain factors, so it is suitable for the risk management of these projects. In the MNA model proposed in this paper, the elements e the nodes of the meta-network e were divided into four categories: project objectives, risk events, risk factors and stakeholders. The standards for project success included the objectives of duration, cost, quality, influence on the environment and safety (health)

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

(Kerzner, 1989). These standards formed the category of “project objective.” The failure to achieve an objective and the deviation from the project objectives were defined as project risks (Purdy, 2010), which are described below. Risk events were the direct reflection of a loss in a project and had a direct impact on project objectives. Risk factors referred to factors that lead to a risk event. Risk factors may be related to individual actions or come from various broader conditions (including societal conditions, legislation, politics, economics and nature). Stakeholders were the project participants who had a direct or indirect influence on the project. We introduced these elements to the meta-network because (1) the purpose of risk management is to reduce the deviation from the objectives in the implementation process of the project, while the deviation is reflected as the consequence of the occurrence of risk events; (2) to conduct risk control, the risk factors which lead to the occurrence of risk events need to be discovered by analyzing how risk events happen; and (3) operative risk control strategies should be proposed from the responsibility assigned to the stakeholders. The process of MNA is described in detail in the following sections. 3. Risk system identification and MNA in building industrialization Fig. 1 shows the process of MNA for risk management. The first step is risk identification and collection of risk levels. The second step is the establishment of a meta-network, including transforming entities into nodes and weighting the links between the nodes. Stakeholders, risk factors, risk events and project objectives are transformed into nodes and combined with weighted links to form several networks, which are integrated into a meta-network. The final step is risk analysis and calculation, including risk evaluation and proposing risk control strategies. Risk levels of the risk factors are transferred to the hierarchy of objectives through the meta-network, and relationships between stakeholders and

555

different risk factors are identified through the networks. Major factors are evaluated based on the model's networks between risk factors and project objectives. The strategy for risk response is based on the analysis of the networks that relate stakeholders and risk factors, and specific measures are designed for specific stakeholders. The results can provide an objective and comprehensive reference for the risk management of building industrialization projects. 3.1. Risk identification Because the new risk system for building industrialization is different from the system for traditional projects, the risk events and risk factors should be reidentified, and a new risk library for building industrialization construction needs to be established. The unique risks for industrialization construction processes include the risks arising from the introduction of new construction technology, building materials, construction equipment and management methods (Li-zi et al., 2015). Risk identification for the process of introducing building industrialization construction methods involves identifying the potential risks and creating a risk database that can be used to construct the meta-network. There are two sources for the risk library: the summary of previous research on building industrialization and interviews with professionals and project managers who have related experience. Analysis of the risk management methods and conclusions in previous studies helps to identify potential risks, and interviews of the professionals and project managers help to revise the risk list to yield the risk library. The risk library was the basis of the following analysis. 3.2. Establishment of a meta-network based on risk identification Based on the risk identification, connections can be made between the identified project objectives, risk events and risk factors

Fig. 1. The process of a meta-network analysis (MNA).

556

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

to form a risk evaluation system with three categories. There are interactions among these entities. The occurrence of one risk event may result in a chain reaction leading to a series of other risk events, and there also may be interactions among risk factors. Risk factors may not directly affect project objectives, but they may have an indirect influence due to the occurrence of risk events. The dividing line between risk factors and risk events is not strict. The main basis for the division is related to how directly they affect the project objectives. Based on the above relationships among project objectives, risk events and risk factors, the risk level of risk factors will ultimately influence the achievement of objectives through risk events. To achieve targeted risk control, nodes and related networks of stakeholders are added to the meta-network. Stakeholders are the main participants in the project, and they include project managers, developers, designers, construction workers and suppliers. In this research, the responsibilities of stakeholders for specific risk factors are analyzed, and stakeholders are linked to the risk factors with logic connections. If one risk factor has to be controlled, related stakeholders need to be found for targeting responses. The nodes of stakeholders are the ends of the risk control system. Fig. 2 shows the relationships between the four kinds of nodes. As displayed in Fig. 2, the four types of nodes can connect the following seven networks: (1) the project objective network formed by the association between the project objective nodes, (2) the risk impact network formed by the association between the project objective nodes and the risk event nodes, (3) the risk event network formed by the association between risk event nodes, (4) the risk decomposition network formed by the association between risk event nodes and risk factor nodes, (5) the risk factor network formed by the association between risk factor nodes, (6) the social network formed by the association between stakeholder nodes and (7) the risk distribution network formed by the association between risk factor nodes and stakeholder nodes. 3.3. Calculation processes in a meta-network analysis To quantify the meta-network, risk levels of risk factors and

weights of links in the networks need to be assigned values. The network conveys the level of risk through the connections between the nodes, where the sources of risk level are the risk factors because they are the smallest units that contain the risk content. This is because the risk factors lead to the occurrence of risk events and the deviation of project objectives. The connections between nodes for the level of risk transfer efficiency may be different; their practical significance is that the influence from risk factors to risk events or from risk events to objectives is different, which needs to be judged. The following algorithm is proposed for the calculations: First, from the experts' judgment, we assigned values of risk level to different risk factors, leading to the vector risk factor:

risk factor ¼ ðf1 ; f2 ; …; fn Þ

(1)

The vectors risk events and project objective were established without values:

risk event ¼ ðe1 ; e2 ; …; en Þ

(2)

project objective ¼ ðo1 ; o2 ; …; on Þ

(3)

Next, the links between the six networks that formed the metanetwork were assigned values that represented the different transfer efficiency of risk level from the risk factors. From a mathematical perspective, networks can be regarded as matrices, and the elements in the matrices represent the weight of the links. The relationships among vectors in the networks are shown below:

risk event ¼ risk factor  risk factor network  risk decomposition network  risk event network (4) project objective ¼ risk event  risk impact network  project objective network

(5)

By assigning values to risk levels for risk factors and transferring the risk levels through the networks, the risk level of the project

Fig. 2. Relationships among different nodes in the meta-network.

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

objectives can be obtained. For risk control, the risk distribution network and social network were established after introducing stakeholders to the meta-network. Unlike the risk evaluation system, links in these networks were not weighted. Because they only represented the interactions between stakeholders and the relationships of stakeholders to risk factors, their influence was not relevant. The risk factor network also was considered to express the association among risk factors. The actual influence between stakeholders and risk factors was obtained from the nodes of the stakeholders: 0

risk distribution network ¼ social network  risk distribution network  risk factor network

(6)

These calculations were used as a comparison of the correlations between stakeholders and risk factors, and critical stakeholders for one specific risk factor were identified. Meanwhile, in the metanetwork, the changes of nodes and links led to perturbations of the project objectives, and changes in the nodes or links and related vectors and networks changed accordingly. Therefore, with an effective risk control measurement, controlling some risk factors or the links related to these factors will ultimately reduce the risk level of the project objectives. 4. Case analysis To verify the feasibility and applicability of the meta-network analysis of risk evaluation and to provide direction for risk control, a residential building construction project in South China that applied building industrialization was analyzed. The developer of this project is a famous Chinese residential real estate developer who aims to improve construction efficiency and quality by introducing industrialized construction approaches. Therefore, our research focused on the risk evaluation and control in the application of the new construction system of building industrialization. The basic information for this developer and the project is shown in Table 1. The process of applying the proposed risk evaluation and control method included the following steps: (1) formulate the specific risk factor list based on the literature related to building industrialization and interviews with project managers and determine the relationship between these risk factors and the new construction system; (2) establish hierarchies of project objectives, risk events, risk factors and stakeholders and form the framework of the metanetwork based on the results of interviews and the scientific literature; (3) conduct a survey of project managers and collect the judgment of risk level and weight of links from the experts and then assign values to the meta-network; and (4) carry out risk evaluation based on the calculation methods and propose a risk control strategy. 4.1. Risk identification and establishment of the meta-network

557

risk factors list were mainly based on interviews with project managers and related information of the new construction system of building industrialization in that project. Interviews were conducted with the project manager, the project assistant manager, the field engineers, the water and electricity engineers and the project supervisor. Because the project was a normal residential building project, our research focused on the risk evaluation and control in the application of the new construction system of building industrialization. In this case, the project objectives were similar to traditional projects, and the risk events were related to the application of construction technologies of the new system. The stakeholders were the major participants in this project. The foundation of the establishment of the meta-network included two parts: the actual experience of project managers from the construction process, which was determined by interviews and the relevant information obtained from technicians involved in this project. We established a preliminary list of risk factors by analyzing and summarizing technique information. We then interview project managers to get to revise the preliminary list into the final risk factor list (see Table 2). Fig. 3 shows the process of the establishment of the meta-network risk database: The risk list shown in Table 2 includes all the nodes in the metanetwork, which can be divided into four categories: project objectives, risk events, risk factors and stakeholders. According to the demonstration of networks in section 3.2, there were seven networks in the meta-network: (1) the project objective network to disclose the interactions among the project objectives; (2) the risk impact network to present how risk events lead to the bias of the project objectives; (3) the risk event network to show the interactions among the risk events; (4) the risk decomposition network to disclose how risk factors lead to risk events; (5) the risk factor network to show the interactions among the risk factors; (6) the social network to show the relationships among stakeholders; and (7) the risk distribution network to show how the stakeholders can prevent the risk factors. To determine the existence and weight of the links in the meta-network, surveys were conducted with the project managers. The questionnaire directly collected their judgments of the existence and weights of links. Fig. 4 summarizes the results of the questionnaires and the construction of the network by ORA-NetScenes (Carley et al., 2013b). The meta-network was the foundation of the MNA. In Fig. 4, the meta-network has round nodes which represent the project stakeholders; square nodes represent the project objectives; diamond nodes represent the risk events and triangle nodes represent the risk factors. The links in the meta-network appeared to show that risk factors had an indirect influence on project objectives by causing risk events, and stakeholders could bring down the risk levels of the project objectives by controlling risk factors. Therefore, critical risk factors which had the most significant impact on project objectives could be determined by comparing the influence levels on objectives of risk factors. Meanwhile, by calculating the influence of the stakeholders on the risk factors, the targeted risk control strategies could be proposed and assigned to the project stakeholders who had a great impact on the critical risk factors.

In our research, the risk identification and development of the

Table 1 Basic information of the project and the developer in this case. Type of company

One of the top five real estate developers in China

Structure type Floor number Greening ratio Total project investment

Reinforcement concrete shear wall 28 with two underground floors 60%

Location Land area Plot ratio 1.5 billion RMB (about $237 million

Guangdong Province 20,234 m2 2.78 U.S.)

558

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

Table 2 Nodes and their contents for the meta-network in this case. Project objectives 1 Time 4 Safety Risk events 1 Prefabricated wall panel technology 4 Structural seam technology 7 Ceramic insulation board technology 10 Self-healing waterproofing technology 13 PVC wallpaper technology Risk factors 1 The supply of wall panels is unstable. 4 The mounting of the flexible connector is not horizontal or vertical. 7

The openings in the dry section are not blocked off on time. 10 Drainage in the wet section is not installed on time. 13 Poor construction quality of basement. 16 Cannot implement correction after the installation finish. 19 Low quality due to lack of adequate supervision. 22 Conservation work is not timely.

25 Drawings of water, electricity and the garden are not completed before construction. 28 The flatness of structural wall panels and stability of supports do not meet requirements. 31 The pipes crossing the beams deviate from the design direction. 34 Concrete leakage into the water stop. Stakeholders 1 Project manager 4 Project secretary 7 Subcontractor's engineers (decoration) 10 Supplier

2 5

Cost Environment

3

Quality

2 5 8 11 14

Full cast-in-place exterior craft Aluminum formwork High-pressure water roughen technology Integrally curing floor technology Interspersed construction technology

3 6 9 12 15

Prefabricated components Intelligent climbing formwork technology Precise positioning of water and electricity equipment Toilet cubicle technology Water interception between floors technology

2 5

The quality of wall panels is poor. The seam material is not securely fastened.

3 6

8

Drainage in the dry section is not installed on time.

11 Inefficient costs because of a failure to intercept water from other floors. 14 Large differences between design and site conditions. 17 Filling wall and structural shear wall displacement and deformation of vertical flexible joints during concrete pouring. 20 Inaccurate positioning of the starting point causes the assembly to fail or causes the stairs to be distorted. 23 The flatness of the board and the base are not up to standards. 26 Worker- and goods-related equipment installation failed to be completed on time.

Cracks exist at wall joints. Horizontal seam materials are not installed after pouring floor concrete and before the concrete's first solidification. 9 The openings in the wet section are not blocked off on time. 12 Workers get hurt by high pressure water.

15 Faults at the end of the wall panel. 18 The supply of aluminum formwork delayed by design changes. 21 The concrete is damaged due to uncompleted concealed construction. 24 Conflicts between design drawings of the wall panel parts and the construction site. 27 Decoration work is impacted by low-quality interception of water from other floors.

29 Fastening of the model is loose, and the shape of support 30 Mortar leakage in the embedded lamp box. changes during structural board construction. 32 Oblique cracking exists while installing the wire box.

33 Deviations exist during distribution box installation.

2 5 8 11

3 6 9

Field engineers Project supervision Subcontractor's engineers (garden) Designers

4.2. Risk evaluation with the meta-network The next step in the risk evaluation system was to analyze the risk factors to determine their relative impact on the project objectives. After establishing the meta-network that connected the risk factors, risk events and project objectives, the risk levels for the risk factors were determined. Risk levels referred to the probability of risk factors. Questionnaires were used to collect the judgments of experts who also were the participants in the risk identification phase. The survey results are shown in Fig. 5. The influence of risk factors on project objectives that were derived from the results of the questionnaires were put into the equations listed in section 3. This involved transferring experts' judgments on risk factors to the hierarchy of project objectives through weighted links in the meta-networks. As demonstrated in section 3, the essence of the meta-network is the matrix, and the transfer of risk levels from risk factors to risk events and project objectives is essentially a multiplication of vectors of risk levels and matrices in different networks to get the influence on the risk events and project objectives. The resulting matrix of these calculations is shown in Fig. 6. In this matrix, red cells indicate high risk level, while green signify a low risk. On the whole, for the five objectives, safety received the least influence from the risk factors, which showed that the application of the new building industrialization construction system had less impact on safety. To select the critical risk factors, the Pareto principle was introduced to the analysis. We

Water and electricity engineers Main contractor's engineers Workers

considered the first 20% of the risk factors as the major source of risk and formulated a risk control strategy based on them. Table 3 shows the risk factors at the first 20% risk level for the 34 risk factors. In the calculation of risk evaluation, these risk factors had the largest impact on the project objectives when combining the risk level and the transfer effect of the network in the matrix of Fig. 6. There are several reasons for this. First, the most important influence on the project from building industrialization construction methods is on the construction of the main buildings. Assembly construction, prefabricated components and module building all affect the process of the main building construction. Previous studies (Bari et al., 2012; Gan et al., 2017) have suggested that the critical factors in the application of building industrialization in China that influence project quality can be divided into three categories: design-related factors, production-related factors, and construction-related factors. For the critical risk factors listed in Table 3, Risk Factor_18 represented a design-related and production-related factors, including overall design issues and the template production problems caused by the use of aluminum templates. Risk Factor_3, 4 and 15 were construction-related factors, including the issues of wall panel installation, the problems in the construction process of the structure and the matter of quality control. In essence, during the process of change from the traditional construction methods to a new type of industrial construction, the equipment, technology and materials were new to the works of design, management and construction. If the design phase

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

Fig. 3. The process for the establishment of the meta-network risk database.

559

560

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

Fig. 4. Meta-network of the risk management for the new construction system of building industrialization in this case.

Fig. 5. The probability of the risk factors in the meta-network.

RF_1

RF_2

RF_3

RF_4

Time

23.1 51.975 329.175

Cost

23.4

RF_5 16.8

RF_6

RF_7

RF_8

RF_9

RF_10 RF_11 RF_12

58.8

105.3

89.7

89.7

89.7

91.8

1.95

333.45

16

16

56

99.9

85.1

85.1

85.1

89.7

1.85

Quality

21.9 49.275 312.075

15.6

15.6

54.6

97.2

82.8

82.8

82.8

85.5

1.8

Safety

10.4

148.2

8.6

8.6

30.1

59.4

50.6

50.6

50.6

51

1.1

23.1 51.975 329.175

16.8

16.8

58.8

105.3

89.7

89.7

89.7

91.8

1.95

Environment

52.65

16.8

23.4

RF_13 RF_14 RF_15

RF_16 RF_17 RF_18 RF_19 RF_20 RF_21 RF_22 RF_23 RF_24

74.3

32.1

107

21.4

102.4

214

192.6

10.7

16.4

4.1

31.2

39.9

70.2

31.35

104.5

20.9

99.6

209

188.1

10.45

14.4

3.6

29.6

38.75

68.5

30.6

102

20.4

97.2

204

183.6

10.2

14.8

3.7

28.8

37.8

42

16.8

56

11.2

53.4

112

100.8

5.6

9.6

2.4

17.6

21.2

74.3

32.1

107

21.4

102.4

214

192.6

10.7

16.4

4.1

31.2

39.9

RF_25 RF_26 RF_27

RF_28 RF_29 RF_30 RF_31 RF_32 RF_33 RF_34

25.9

14.8

33.3

4.1

4.1

31.2

7.8

7.8

31.2

7.8

26.6

15.2

34.2

3.6

3.6

29.6

7.4

7.4

29.6

7.4

24.5

14

31.5

3.7

3.7

28.8

7.2

7.2

28.8

7.2

14

8

18

2.4

2.4

17.6

4.4

4.4

17.6

4.4

25.9

14.8

33.3

4.1

4.1

31.2

7.8

7.8

31.2

7.8

Fig. 6. Influence matrix between risk factors (RF) and project objectives.

was not perfect, substantial design changes might be necessary (Risk Factor_18), which could lead to faulty construction processes, such as casting issues in the panel and connection installation and shear walls (Risk Factor_3, 4 and 15). Second, in the industrialized construction process, the new project planning method e interspersed construction technology (Feng and Hu, 2016) e was used to improve project efficiency. Interspersed construction technology means that a sub-project was interspersed during the construction of another sub-project; therefore, it was necessary to provide a clean and dry workplace for the next sub-project after the previous sub-project. Thus, floor water interception could meet the requirement of interspersed construction. Floor water interception means intercepting the wasted water from other floors e which were doing wet construction works e to provide a clean and dry workplace for the following works (e.g., decoration work). Low quality of floor water interception will pollute the workplace of

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

561

Table 3 Critical risk factors and related analysis of the risk evaluation. Risk Factor ID

Risk Factor

Analysis

Risk Factor_3

Cracks exist at wall joints.

Risk Factor_18

The supply of aluminum formwork delayed by design changes.

Risk Factor_11

Inefficient costs because of a failure to intercept water from other floors.

Risk Factor_15

Faults of the ends of wall panels.

Risk Factor_4

The mounting of the flexible connector is not horizontal or vertical.

Risk Factor_27

Decoration work is impacted by low- quality interception of water from other floors.

Wall joints are critically important components in the technology of prefabricated interior walls. If cracks exist, they will have a significant influence on the quality of the prefabricated walls and perhaps the entire building. Unlike the timber formwork, the aluminum formwork cannot be changed at the last minute when there are design changes. This will lead to a delay of the formwork supply and have a great impact on the project duration. If there is a failure to intercept the wasted water from other floors, the decoration work area will be polluted, and the quality of decoration work will be affected. This will lead to additional costs (e.g., rework and cleaning). The structure walls are the components that bear the load, so if the faults exist, the load-bearing capacity of the entire building will be weakened. It is necessary for the mounting of the flexible connector to meet the design requirements, which normally requires the connector to be mounted horizontally or vertically. If not meeting these requirements, the related construction will be negatively affected. Similar to Risk Factor_11, the decoration work area will be polluted, and the efficiency and quality of decoration work will be negatively impacted.

decoration works, negatively impacting quality and efficiency. For the critical risk factors listed in Table 3, Risk Factor_11 and 27 were the risk factors related to floor water interception. In summary, because of the critical risk factors listed in Table 3, the project quality may not reach the expected outcome. Serious mistakes may lead to the need for much rework, which significantly increases the total project costs and duration. These risk factors were not completely avoided in this test case, as illustrated by Risk Factor_3. According to the experience of field engineers and experts in the case study, there were a large number of issues in the installation of wall panels, and cracks appeared in the connecting parts. However, with interspersed construction technology, the effectiveness of the project was so high that there was not enough time for workers to totally solve these issues. The result was that even after the decoration phase, many cracks were still obvious, which had a significant influence on the project quality. A comparison of the results of MNA and the actual conditions of this case showed that the risk factors identified from MNA were nearly the same as the potential risks. This means the process of risk identification and evaluation was effective. Risk control strategies are analyzed in the next phase. 4.3. Risk control based on MNA

influence on the risk factors. These results were consistent with the field study. One possible reason is that Stakeholder_1 and 4 were the highest-level managers of the project; they were responsible for the organization and coordination of the entire project. Stakeholders_2, 3, 5 and 6 were the project managers lowest in the hierarchy; they were directly responsible for the construction activities. These workers were from the project team, construction units and supervision unit. It can be concluded that for projects applying building industrialization, the major risk stakeholders were the top-level managers and basic managers from these three categories. The water and electricity engineers, project manager and secretary had more

RF_1

Project manager Field engineers Water and electricity engineers Project secretary Project supervision Main contractor's engineers Subcontractor's engineers (decoration) Subcontractor's engineers (garden) Workers Supplier Designers

RF_2

4 3 4 4 4 4 4 4 2 3 2

RF_3

3 2 2 2 2 2 2 2 1 2 1

RF_4

4 4 4 4 4 3 3 3 4 2 4

RF_5

3 3 4 3 3 3 3 3 3 1 3

RF_6

3 3 4 3 3 3 3 3 3 1 3

RF_7

3 3 4 3 3 3 3 3 3 1 3

RF_8

4 4 4 4 4 4 3 3 4 1 4

RF_9

4 4 4 4 4 4 3 3 4 1 4

RF_10 RF_11 RF_12

4 4 4 4 4 4 3 3 4 1 4

4 4 4 4 4 4 3 3 4 1 4

3 3 3 3 4 3 2 2 3 1 3

2 1 2 2 1 2 2 2 2 0 2

RF_13 RF_14 RF_15 RF_16 RF_17 RF_18 RF_19 RF_20 RF_21 RF_22 RF_23 RF_24

Based on the above process, stakeholders with major responsibilities for the critical risk factors were identified by the calculation of social network, risk factor network and risk distribution network, and the calculation results are shown in Fig. 7. In this matrix, red cells indicate a high level of influence level, while green means there was a low level of influence. Fig. 7 shows that Stakeholder_3 (the water and electricity engineers), Stakeholder_1 (the project manager) and Stakeholder_4 (the project secretary) were the critical stakeholders with the highest levels of influence, and the influence levels of Stakeholder_2 (the field engineer), Stakeholder_5 (the project supervisor) and Stakeholder_6 (the main contractor engineer) were near the influence levels of the critical stakeholders. Therefore, all six stakeholders were considered to propose risk control strategies. The basic responsibilities of these stakeholders are listed in Table 4 in the order of the level of

3 3 4 3 3 3 3 3 3 1 3

3 2 3 3 2 2 2 2 3 1 3

3 3 4 3 3 3 3 3 3 1 3

2 1 1 1 1 2 2 2 0 1 0

2 1 2 2 1 2 2 2 2 0 2

2 2 3 2 2 3 3 3 2 0 2

4 4 4 4 4 3 3 3 4 2 4

2 2 3 2 3 2 2 2 2 1 2

3 3 4 3 3 3 3 3 3 1 3

2 1 2 2 1 2 2 2 2 0 2

RF_25 RF_26 RF_27 RF_28 RF_29 RF_30 RF_31 RF_32 RF_33 RF_34

5 5 4 5 5 4 3 3 5 2 5

3 3 4 3 3 3 3 3 3 1 3

3 3 4 3 3 3 3 3 3 1 3

3 4 3 3 3 3 2 2 3 1 3

3 4 3 3 3 3 2 2 3 1 3

3 4 3 3 3 3 2 2 3 1 3

3 4 3 3 3 3 2 2 3 1 3

3 4 3 3 3 3 2 2 3 1 3

1 1 0 1 1 1 0 0 1 0 1

1 1 0 1 1 1 0 0 1 0 1

Fig. 7. Matrix of influence between stakeholders and risk factors (RF).

2 2 0 2 2 1 0 0 2 1 2

3 2 2 3 3 4 3 3 2 0 2

562

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

Table 4 Critical stakeholders and related responsibilities. Stakeholder ID

Stakeholder

Responsibility

Stakeholder_3 Stakeholder_1 Stakeholder_4 Stakeholder_2 Stakeholder_5 Stakeholder_6

Water and electricity engineers Project manager Project secretary Field engineer Project supervisor Main contractor engineer

Construction and installation of water and electricity equipment (e.g. water pipe, electricity wire). The top manager of this project; responsible for planning, organizing, leading and controlling the entire project. The support staff for the project manager; responsible for coordination and communication. The management of detailed construction activities. Supervision of the project activities to ensure that the project is implemented according to plan. The management of detailed construction activities and the arrangement of workers.

impact on the risk factors, while the field engineers from the development units, main contractor engineers from the construction units and project supervisors from the supervision units played supporting roles. After agreeing on the critical stakeholders, risk control strategies were proposed based on the results of the risk evaluation. For construction-related factors and water interception related factors e such as Risk Factor_3 and 11ecritical stakeholders should have adopted several measures. As the highest-level managers of the project, the project manager and project secretary should have strengthened the communication of the engineers and proposed a flexible construction plan based on the real-time conditions of the project. As members of the construction units, the main contractor engineers bear the brunt of risk events and should implement strict control methods for the workers and provide correct instruction about construction technology. This can reduce the probability of the risk factors. Water and electricity engineers and field engineers from the project team should be more familiar with the new construction methods, related technologies and design, so they should guide people higher in the hierarchy. They need real-time monitors at all project venues to ensure that design ideas are implemented and that the building industrialization construction methods are implemented correctly. As support for the project team, supervision engineers should monitor the project from the perspective of project quality and preventing accidents by reducing risks, such as Risk Factor_3, 4 and 15. They should provide for careful examination during and after the construction process. For design-related risk factors, such as Risk Factor_18, the major risk stakeholders are the members of the project team. For example, in the implementation of the “full cast-in-place exterior craft,” changing water and electricity lines directly leads to changes in the aluminum formwork; thus, the water and electricity engineers and field engineers should revise the requirements for the aluminum formwork as soon as possible to reduce the impact on project duration and total cost. Also, the project manager and secretary should give assistance and coordinate the communication between the project and the design teams. Based on these strategies, a strict responsibility system for the industrialized construction process should be established among these six critical stakeholders e namely the water and electricity engineers, the project manager, the project secretary, the field engineer, the project supervisor and the main contractor engineer. First, as the highest-level managers of the project, the project manager and project secretary are responsible for the communication and coordination of the stakeholders inside and outside the project, such as the design team and the basic managers of the project. Second, to achieve the greatest degree of risk aversion, a supervision mechanism is needed among the basic managers. Engineers from the main contractor, who are directly responsible for implementation of the industrialized construction approach, should receive supervision from the water and electricity engineers, field engineers and project supervisors, who are more familiar with these approaches. The quality control of the project supervisors should be examined by the engineers from the project team.

5. Conclusions Our research evaluated the risk levels in an industrialized building construction project and suggested targeted risk control measures based on the proposed integrated risk evaluation and control approach. We used meta-network analysis to incorporate the experts' knowledge and data from our on-the-spot investigation for the final model of risk management, supplementing the body of knowledge in the area of industrialized construction projects. First, we identified risks by using the scientific literature and previous construction experience. Next, the results were converted into a meta-network that consisted of many associated networks. Finally, risk evaluation and control were carried out with reference to the meta-network. From a theoretical perspective, this framework can thoroughly analyze risk factors in a building industrialization construction project, study the influence of the associations among risks and offer effective risk control strategies based on risk evaluation. Compared to previous risk management methods, the advantages of MNA are that it can identify the critical risk factors more visibly (by graphing the meta-network) and provide effective guidance for risk control by analyzing the participation of stakeholders for controlling risk factors. The traditional methods usually stopped at the risk evaluation phase. Meanwhile, the use of this method is simple. For building industrialization projects, project managers can somewhat easily handle the risk management related to the risk factors identified by MNA. We applied our proposed MNA method to a residential building construction project in South China that introduced building industrialization to the construction process. During the construction process in this project, because of new construction and management approaches of industrialized construction, risk factors in the construction process were quite different than those in a traditional project. Therefore, risk identification for the industrialized construction process was conducted by interviewing related experts and experienced engineers, and the results of these interviews were used to create a meta-network of risk. As shown in the results of the risk evaluation process, six risk factors in three categories were considered critical risk factors: (1) constructionrelated factors, such as Risk Factor_3 (cracks exist on wall panel joints and other unqualified low-quality factors); (2) design and production-related factors, such as Risk Factor_18 (the supply of aluminum model was delayed due to design changes); and (3) water interception-related risk factors, such as Risk Factor_27 (decoration work was impacted by low-quality interception of water from other floors). The results were consistent with the field study results. The risk factors were not completely avoided during the project. A risk control analysis was based on the results of the risk evaluation to find the critical stakeholders who had a significant influence on the risk factors, and these stakeholders were Stakeholder_3 (the water and electricity engineers), Stakeholder_1 (the project manager), Stakeholder_4 (the project secretary), Stakeholder_2 (the field engineer), Stakeholder_5 (the project supervisor) and Stakeholder_6 (the main contractor engineer). To avoid risk factors and events, risk control strategies were proposed.

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

In addition, a responsibility system among these stakeholders should be established. First, the project manager and project secretary should be responsible for the communication and coordination of the stakeholders inside and outside the project, such as the design team and the basic managers of the project. Second, a supervision mechanism is needed among the basic managers. Engineers from the main contractor should receive supervision from the water and electricity engineers, field engineers and project supervisors. The quality control of project supervisors should be examined by the engineers from the project team. In this risk control strategy, all stakeholders should minimize the possibility of risk factors occurring within their area of responsibility. In actual practice, risk managers should build a hierarchy of project objectives, risk events, risk factors, stakeholders and related networks based on their actual situation. The meta-network proposed for this project is just an example; the establishment of a network should be based on each project's unique nature. These results render novel and practical ideas for risk management, particularly to evaluate and control the risks of industrialized building construction projects. Future work should use MNA in other scenarios and add more entities and relationships to the meta-network to make the method more effective and feasible. Acknowledgement The study is financially supported by the National Science Foundation of China (No. 71871235). The authors would like to thank the managers and workers participated in this study. References Abdelgawad, M., Fayek, A.R., 2011. Fuzzy reliability analyzer: quantitative assessment of risk events in the construction industry using fuzzy fault-tree analysis. J. Construct. Eng. Manag. 137, 294e302. Aye, L., Ngo, T., Crawford, R., Gammampila, R., Mendis, P., 2012. Life cycle greenhouse gas emissions and energy analysis of prefabricated reusable building modules. Energy Build. 47, 159e168. Bari, N.A.A., Yusuff, R., Ismail, N., Jaapar, A., Ahmad, R., 2012. Factors influencing the construction cost of industrialised building system (IBS) projects. Procedia Soc. Behav. Sci. 35, 689e696. Borsuk, M.E., Stow, C.A., Reckhow, K.H., 2004. A Bayesian network of eutrophication models for synthesis, prediction, and uncertainty analysis. Ecol. Model. 173, 219e239. Bu-Qammaz, A.S., Dikmen, I., Birgonul, M.T., 2009. Risk assessment of international construction projects using the analytic network process. Can. J. Civ. Eng. 36, 1170e1181. Carley, K.M., Pfeffer, J., rgen, Liu, H., Morstatter, F., Goolsby, R., 2013a. Near real time assessment of social media using geo-temporal network analytics. In: Ieee/acm International Conference on Advances in Social Networks Analysis and Mining, pp. 517e524. Carley, K.M., Reminga, J., Storrick, J., Pfeffer, J., Columbus, D., 2013b. ORA User's Guide 2013. Ora Users Guide. Chiang, Y.-H., Chan, E.H.-W., Lok, L.K.-L., 2006. Prefabrication and barriers to entryda case study of public housing and institutional buildings in Hong Kong. Habitat Int. 30, 482e499. Elsawah, H., Bakry, I., Moselhi, O., 2016. Decision support model for integrated risk assessment and prioritization of intervention plans of municipal infrastructure. J. Pipeline Syst. Eng. Pract. 7, 04016010. Fang, C., Marle, F., 2015. A Framework for the Modeling and Management of Project Risks and Risk Interactions. Fang, C., Marle, F., Zio, E., Bocquet, J.C., 2012. Network theory-based analysis of risk interactions in large engineering projects. Reliab. Eng. Syst. Saf. 106, 1e10. Faniran, O., Caban, G., 1998. Minimizing waste on construction project sites. Eng. Construct. Architect. Manag. 5, 182e188. Feng, W., Hu, J., 2016. Application of BIM Technology in Interspersed Construction of High-rise Residences. Construction Technology. Gan, Y., Shen, L., Chen, J., Tam, V.W., Tan, Y., Illankoon, I., 2017. Critical factors affecting the quality of industrialized building system projects in China. Sustainability 9, 216. Gibb, A.G., 1999. Off-site Fabrication: Prefabrication, Pre-assembly and Modularisation. John Wiley & Sons. Goodier, C., Gibb, A., 2007. Future opportunities for offsite in the UK. Construct. Manag. Econ. 25, 585e595. Guan, D.J., Guo, P., 2014. Constructing Interdependent Risks Network of Project Portfolio Based on Bayesian Network. International Conference on Management

563

Science & Engineering, pp. 1587e1592. Han, S.H., Du, Y.K., Kim, H., Jang, W.S., 2008. A web-based integrated system for international project risk management. Autom. ConStruct. 17, 342e3568. Hong, J., Shen, G.Q., Mao, C., Li, Z., Li, K., 2016. Life-cycle energy analysis of prefabricated building components: an inputeoutput-based hybrid model. J. Clean. Prod. 112, 2198e2207. Jaillon, L., Poon, C.S., 2009. The evolution of prefabricated residential building systems in Hong Kong: a review of the public and the private sector. Autom. ConStruct. 18, 239e248. Ji, Y., Zhu, F., Li, H.X., Al-Hussein, M., 2017. Construction industrialization in China: current profile and the prediction. Appl. Sci. 6. Jiang, R., Mao, C., Hou, L., Wu, C., Tan, J., 1 February 2018. A SWOT analysis for promoting off-site construction under the backdrop of China's new urbanisation. J. Clean. Prod. 173, 225e234. Karimiazari, A., Mousavi, N., Mousavi, S.F., Hosseini, S., 2011. Risk assessment model selection in construction industry. Expet. Syst. Appl. Int. J. 38, 9105e9111. Kerzner, H., 1989. Project Management: a Systems Approach to Planning, Scheduling, and Controlling. Wiley. Kuo, Y.C., Lu, S.T., 2013. Using fuzzy multiple criteria decision making approach to enhance risk assessment for metropolitan construction projects. Int. J. Proj. Manag. 31, 602e614. Lenz, R., 2012. Topology of local health officials' advice networks: mind the gaps. J. Publ. Health Manag. Pract. 18, 602e608. Li-Zi, L., Chao, M., Li-Yin, S., Zheng-Dao, L., 2015. Risk factors affecting practitioners' attitudes toward the implementation of an industrialized building system: a case study from China. Eng. Construct. Architect. Manag. 22, 622e643. Li, M., Li, G., Huang, Y., Deng, L., 2017. Research on investment risk management of Chinese prefabricated construction projects based on a system dynamics model. Buildings 7, 83. Li, Y., Lu, Y., Li, D., Ma, L., 2015. Metanetwork analysis for project task assignment. J. Construct. Eng. Manag. 141, 04015044. Lihong, L.I., Geng, B., Baoku, Q.I., Lei, Y., Luan, L., 2013. Empirical Analysis on the Cost of Prefabricated Construction Contrast with Cast-in-place Building. Construction Economy. Lin, Y.C., Paul, A., Corotis, R.B., Liel, A.B., 2015. Framework methodology for riskbased decision making for transportation agencies. ASCE-ASME J. Risk Uncertain. Eng. Syst. Part A Civ. Eng., 04015006 Liu, H., 2017. A new perspective of building industrialization to promote rapid development of China's construction. World Construct. 6. Lovell, H., Smith, S.J., 2010. Agencement in housing markets: the case of the UK construction industry. Geoforum 41, 457e468. Luo, L.Z., Mao, C., Shen, L.Y., Li, Z.D., 2015. Risk factors affecting practitioners' attitudes toward the implementation of an industrialized building system: a case study from China. Eng. Construct. Architect. Manag. 22, 622e643. Mao, C., Shen, Q., Pan, W., Ye, K., 2013. Major barriers to off-site construction: the developer's perspective in China. J. Manag. Eng. 31, 04014043. Martinez, S., Jardon, A., Navarro, J.M., Gonzalez, P., 2008. Building industrialization: robotized assembly of modular products. Assemb. Autom. 28, 134e142. Nagurney, A., Dong, J., 2005. Management of knowledge intensive systems as supernetworks: modeling, analysis, computations, and applications. Math. Comput. Model. 42, 397e417. Pan, W., Sidwell, R., 2011. Demystifying the cost barriers to offsite construction in the UK. Construct. Manag. Econ. 29, 1081e1099. Pons, O., Wadel, G., 2011. Environmental impacts of prefabricated school buildings in Catalonia. Habitat Int. 35, 553e563. Purdy, G., 2010. ISO 31000: 2009dsetting a new standard for risk management. Risk Anal. 30, 881e886. Roy, R., Low, M., Waller, J., 2005. Documentation, standardization and improvement of the construction process in house building. Construct. Manag. Econ. 23, 57e67. Rubio-Romero, J.C., Su~ a, r.-C.M., Abad, J., 2014. Modeling injury rates as a function of industrialized versus on-site construction techniques. Accid. Anal. Prev. 66, 8e14. Saaty, T.L., 2004. Fundamentals of the analytic network process d dependence and feedback in decision-making with a single network. J. Syst. Sci. Syst. Eng. 13, 129e157. Stergiopoulos, G., Kotzanikolaou, P., Theocharidou, M., Lykou, G., Gritzalis, D., 2016. Time-based critical infrastructure dependency analysis for large-scale and cross-sectoral failures. Int. J. Crit. Infrastruct. Protect. 12, 46e60. Tam, V.W.Y., Tam, C.M., Zeng, S.X., Ng, W.C.Y., 2007. Towards adoption of prefabrication in construction. Build. Environ. 42, 3642e3654. Tavakolan, M., Etemadinia, H., Tavakolan, M., Etemadinia, H., 2017. Fuzzy weighted interpretive structural modeling: improved method for identification of risk interactions in construction projects. J. Construct. Eng. Manag. 143. Wakolbinger, T., Nagurney, A., 2004. Dynamic supernetworks for the integration of social networks and supply chains with electronic commerce: modeling and analysis of buyer-seller relationships with computations. Netnomics 6, 153e185. Wang, T., Wang, S., Zhang, L., Huang, Z., Li, Y., 2016. A major infrastructure riskassessment framework: application to a cross-sea route project in China. Int. J. Proj. Manag. 34, 1403e1415. Wong, A.K., Yeh, S.H.K., 1985. Housing a Nation : 25 Years of Public Housing in Singapore. Xiahou, X., Yuan, J., Liu, Y., Tang, Y., Li, Q., 2018. Exploring the driving factors of construction industrialization development in China. Int. J. Environ. Res. Publ.

564

T. Wang et al. / Journal of Cleaner Production 205 (2018) 552e564

Health 15. Xu, X., Zhao, Y., 2010. Some Economic Facts of the Prefabricated Housing. Industry Report, vol. 3. Rutgers Business School, Newark, NJ. Yan, W., CAO, Y.-h., LI, G.-r., 2004. Development of assembly-type RC structure and building industrialization. J. Chongqing Jianzhu Univ. 26, 131e136. Zabihi, H., 2013. Definitions, concepts and new directions in industrialized building systems (IBS). Ksce J. Civ. Eng. 17, 1199e1205. Zhang, G., Zou, P.X.W., 2007. Fuzzy analytical hierarchy process risk assessment approach for joint venture construction projects in China. J. Construct. Eng. Manag. 133, 771e779.

Zhang, X., Skitmore, M., 2012. Industrialized housing in China: a coin with two sides. Int. J. Strat. Property Manag. 16, 143e157. Zhou, H.B., Zhang, H., 2011. Risk assessment methodology for a deep foundation pit construction project in shanghai, China. J. Construct. Eng. Manag. 137, 1185e1194. Zhu, J., Mostafavi, A., 2015. Ex-ante assessment of vulnerability to uncertainty in complex construction project organizations. Zhu, J., Mostafavi, A., 2016. Metanetwork framework for integrated performance assessment under uncertainty in construction projects. J. Comput. Civ. Eng. 31, 04016042.