Life cycle sustainability performance assessment

Building and Environment 138 (2018) 21–41

Contents lists available at ScienceDirect

Building and Environment journal homepage: www.elsevier.com/locate/buildenv

Life cycle sustainability performance assessment framework for residential modular buildings: Aggregated sustainability indices

T

Mohammad Kamali, Kasun Hewage∗, Abbas S. Milani School of Engineering, University of British Columbia, Kelowna, BC V1V 1V7, Canada

A R T I C LE I N FO

A B S T R A C T

Keywords: Modular construction Sustainability criteria Sustainability indicators Life cycle performance Benchmarking MCDA ELECTRE AHP TOPSIS

Construction of residential buildings using off-site methods, in particular modular construction, is receiving considerable attention. However, the sustainability performance of modular buildings has rarely been investigated through a life cycle perspective. In this paper, a life cycle sustainability performance assessment framework is developed for modular buildings and its application is examined. In the first part of the paper, suitable life cycle sustainability performance criteria (SPCs) for modular buildings were developed and ranked. In this regard, potential SPCs were identified through a comprehensive literature review and expert interviews. These SPCs were then evaluated by construction experts through two questionnaire surveys against three evaluation criteria: applicability, data availability, and data accuracy. The evaluation criteria's weights were determined through a group decision making process using the Analytic Hierarchy Process (AHP) multi-criteria decision analysis (MCDA) method. Consequently, the experts' feedback was analyzed with the help of the Elimination and Choice Translating Reality (ELECTRE) outranking MCDA method and all the SPCs were ranked within their associated sustainability categories (i.e., environmental, economic, and social). In the second part of the paper, application of the proposed framework has been discussed and validated through a case study of a modular building in British Columbia, Canada. Sustainability performance of modular buildings in the proposed framework were assessed by developing aggregated sustainability indices for the selected SPCs and comparing them with corresponding benchmarks. In this regard, appropriate sustainability performance indicators (SPIs) under each selected SPC have been developed and calculated. Consequently, through an aggregation process, sustainability indices are developed using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) MCDA method. In this paper, the environmental life cycle performance of the case study building has been benchmarked and recommendations have been made for performance improvements. This research is deemed useful for the construction practitioners since it provides a methodical framework for life cycle sustainability performance assessment of modular buildings and assists with the selection of sustainable methods of construction.

1. Introduction 1.1. Background Similar to ‘conventional’ ‘stick-built’ (on-site) buildings, buildings constructed using modular construction method are permanent structures. These two types of buildings differ in their respective life cycle phases. The main difference is the design and construction phase. In the case of modular construction, the building is designed based on a number of modules, in which they are built in a modular construction facility for majority of the construction work and then transported to the building site and placed on a permanent foundation [1,2]. According to the published literature, modular construction as one

∗

of the principal methods of off-site construction offers various advantages. Speed of construction, safety, productivity, product quality, and less environmental impacts, are among the advantages of using modular construction [3–10]. Conversely, transportation restraints, increased coordination and communication, and public's negative perception are among the disadvantages of this method of construction [11–14]. Despite many reported advantages of modular construction, its application is still limited when compared to the conventional construction approach [15–18]. This is mainly because the various advantages of using modular construction are not well understood by different stakeholders [19–21]. Many studies claimed that certain buildings are ‘sustainable’ only

Corresponding author. E-mail addresses: [email protected] (M. Kamali), [email protected] (K. Hewage), [email protected] (A.S. Milani).

https://doi.org/10.1016/j.buildenv.2018.04.019 Received 24 November 2017; Received in revised form 20 March 2018; Accepted 13 April 2018 Available online 17 April 2018 0360-1323/ © 2018 Elsevier Ltd. All rights reserved.


M. Kamali et al.

construction) to gain a deeper understanding of the sustainability performance of modular buildings. The main objective of this research is to propose a methodical framework that can be used to benchmark the life cycle sustainability performance of residential modular buildings. To attain the main objective, the following specific sub-objectives have been accomplished in this paper:

because they perform well in some aspects, such as energy consumption. In other words, sustainability has often been considered and evaluated exclusively according to the environmental dimension and thus ‘sustainable building’, ‘environmentally sustainable building’, and ‘green building’ are known terms that are interchangeably used. However, this approach is not appropriate for determining a sustainable building [22–25]. The concept of sustainability has been categorized in the environmental, social and economic dimensions (namely triple bottom line or TBL) [26]. The environmental dimension indicates minimizing the environmental impacts over the life cycle of a building [27]. The economic dimension implies the affordability to support the direct and indirect costs of a building, without neglecting other essential needs [28]. However, this requisite depends on the context and people and also recalls the time uncertainty of economic sustainability. In fact, a change in what is an economically sustainable choice in buildings is possible according to economic cycles and markets developments. A sustainable building should deliver economic value over time, taking into account future life-cycle costs of operation, maintenance, refurbishment and disposal [27]. The social dimension of a sustainable building is the most ignored one as it was rarely investigated. However, some studies mentioned the characteristics of a building that encourages social sustainability [29–31]. It is not expected from a building to simultaneously show the best environmental, economic, and social performance, since in some cases they are contradictory. For example, construction of an energy efficient building requires more costs. Therefore, balancing the impacts of a building on these three dimensions (not individually maximizing/minimizing) over the entire life cycle is a key factor towards sustainable buildings. Sustainability assessments are intended to gather and provide information to ease decision-making processes [32]. Several methodologies and systems have been developed and published to assess the level of sustainability in buildings. One of the widely used sustainability assessment methods includes the rating systems (also called green building rating systems). Several rating tools that are used to assess environmental sustainability of buildings include, LEED (International), Green Globes (US and Canada), LBC (International), BREEAM (International), CASBEE (Japan), among others. Rating systems deal with mainly environmental sustainability performance of buildings by providing a set of performance criteria and scoring each building project based on those criteria. These systems examine the current performance or the expected performance of a “whole building” and allow comparison of different buildings [33]. Despite many advantages, a number of shortcomings were reported with the use of some of the rating systems, such as the complexity and diversity of criteria (e.g., energy modelling), the bureaucratic process of assessment, and cost factors (e.g., assessment collation and certification fees), and so forth [34]. Another important category for conducting (environmental) sustainability assessment of buildings consists of the life cycle assessment systems or LCA-based tools, such as ATHENA (US and Canada), BEES (US), Eco-Quantum (Netherlands), EcoEffect (Sweden), ENVEST (UK), among others. LCA-based tools were mainly developed to evaluate the life cycle environmental impacts of a building as a whole. They usually follow a bottom-up approach, meaning that the impacts of the building's materials and components are combined to determine the environmental impacts of the whole building [35].

- Identification and selection of appropriate sustainability criteria for modular buildings; - Determination of suitable sustainability indicators under each sustainability criterion as well as their weights, measurement methods, and benchmarks; - Development of sustainability indices for benchmarking the performance of modular buildings; and - Validation of the proposed framework with a case study modular building. The framework presented in this paper is intended to assist with making informed decisions on selection of the best method of construction (modular vs. conventional). Furthermore, it can be used to explore and improve the low sustainability performing areas over the life cycle of a new modular building design, even if the decision on the construction method has already been made. The methodology outlined in this paper can also be adopted for sustainability assessment benchmarking in other construction practices or by researchers in other fields to evaluate a process or product's performance metrics with respect to their benchmarks. This paper is structured as follows: Section 2 outlines the methodology of the proposed framework and explains how the detailed analyses are conducted. Results of the conducted surveys and analyses followed by discussions are presented in Section 3. Within the same section, for the proof-of-concept, the proposed framework has been validated with a case study of a modular building in the Okanagan, British Columbia, Canada. The last section (Section 4) briefly summarizes the main conclusions and recommendations for future study. 2. Methodology The conceptual framework proposed in this research for sustainability assessment of modular buildings is presented in Fig. 1. The framework consists of two main parts that have been separated by a dashed line in the figure. In the first part, primary potential sustainability performance criteria (SPCs) were compiled and grouped into three main sustainability categories through screening the existing criteria available in the literature. Three SPC evaluation criteria were defined and their importance weights were assigned with the help of the Analytic Hierarchy Process (AHP) multi-criteria decision analysis (MCDA) method. Then, two questionnaire surveys were designed and conducted to evaluate the SPC categories against the evaluation criteria. The construction experts' feedback collected through the surveys was analyzed using the Elimination and Choice Translating Reality (ELECTRE) MCDA method to rank the SPCs within the sustainability categories. In the second part of the framework, the application of the developed SPC categories in the sustainability performance assessment of modular buildings was shown. First, by considering the rank orders, maximum possible number of SPCs within the intended sustainability category is selected depending on circumstances of the building project. To develop sustainability indices for the selected SPCs, suitable sustainability performance indicators (SPIs) associated with each SPC along with their measurement methods and relative importance weights have been determined using the literature and expert opinions. The concept of performance level (PL) was introduced for normalization of the calculated SPIs with respect to their benchmark values. Then, the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) MCDA method has been used to develop the

1.2. Sustainability of modular buildings A literature review of modular construction revealed that there were only a very few studies that had been conducted on the environmental life cycle assessment of modular buildings [36]. This work also showed that there was no sustainability performance assessment research that addressed all triple bottom line (TBL) sustainability dimensions of modular buildings over their life cycle. Therefore, it is necessary to comparatively evaluate the life cycle sustainability of buildings built by modular construction method and its counterparts (conventional 22


M. Kamali et al.

Fig. 1. Proposed framework for sustainability assessment of modular buildings.

sustainability performance criteria (SPCs) in this paper. For example, ‘material’ can be represented by material consumption, waste management, etc. In general, SPCs are employed to assess the sustainability of a product or process. In the case of buildings, SPCs can be used for different purposes, such as sustainability comparison among similar buildings, performance evaluation, decision making, among others. Each SPC itself can be presented by a number of measurable sub-criteria, namely sustainability performance indicators (SPIs). For example, ‘Waste management’ is a SPC within the environmental category of sustainability that can include different SPIs such as recycled content, reused components, among others. According to Robert et al. [37], indicators are used to measure program, status, and change towards

sustainability indices by aggregation of the calculated SPIs and weights. Finally, these steps were applied to a case study of modular building and the environmental sustainability indices were developed and benchmarked. In this methodology section, the above steps have been explained in details. 2.1. Establishing sustainability performance criteria There are different areas that can significantly contribute to the life cycle sustainability of buildings, such as energy, material, cost, and so forth. To evaluate the sustainability of a building, each area can be broken down into a number of assessment criteria, namely 23


M. Kamali et al.

when assessing the sustainability of modular versus conventional buildings? CII – Data Availability: Regardless of the given SPC whether quantitative or qualitative, is the data available to measure it? CIII – Data Accuracy: How accurate is the available data to measure the given SPC? The relative importance weights of these criteria were determined through a group decision making process using the AHP method [51]. The participant group consisted of a number of researchers at the University of British Columbia, Canada, who were familiar with sustainability assessment of industrial projects. A questionnaire survey was designed to facilitate experts' pairwise comparisons between the evaluation criteria. The relative importance of one criterion over the other was asked to be judged using a rating scheme ranging from 1 (Equally important) to 9 (Extremely more important). In this research, the questionnaires were delivered to and collected from participants separately; therefore, the AHP's aggregating individual priorities (AIP) aggregation method was used to deal with the outcomes of the survey [52–54].

achieving the goals of sustainability. By measuring and aggregating all the SPIs associated with a SPC, it can be quantified. Numerous sustainability assessment criteria and indicators have been reported in the literature for the built environment [32,38–42]; however, many of them may not be suitable for a given construction project. Therefore, to efficiently appraise the sustainability performance of every construction project, first, a set of appropriate SPCs that suit the conditions of the project, should be identified and selected. Similarly, several criteria have primarily been developed for sustainability evaluation of conventional buildings. These criteria should be reviewed in the context of modular buildings. This section identifies and ranks the most appropriate criteria for life cycle sustainability assessment of residential modular buildings by refining the existing TBL SPCs and subsequently evaluating and ranking the refined SPC categories using suitable evaluation criteria. 2.1.1. Primary potential SPCs A comprehensive literature review was conducted to compile and select the most applicable SPCs for conventional buildings [43]. Different rating systems such as LEED, Green Globes, and LBC, as well as journal/conference articles that discussed sustainability criteria were searched and reviewed. Exploration of the differences between the modular and regular construction (i.e., different life cycle phases and also benefits and challenges) enables in-depth understanding of the significant life cycle sustainability criteria to comparatively evaluate these two construction methods. Thus, another study was performed by focusing on such differences to ensure the comprehensiveness of the selected SPCs [36]. Finally, the selected SPCs were reviewed by experts through informal interviews and their opinions applied in finalizing the initial SPC list. In order to reduce the number of criteria, similar, overlapped, or related ones were grouped under one main SPC in the screening process. Eventually, a total of 33 SPCs were categorized into the main sustainability categories: environmental, economic, and social, as presented in Table 1. The table also shows the acronyms of SPCs that are sometimes used in this paper for simplicity purposes. Descriptions of the SPCs are presented in Appendix A.

2.1.3. Ranking SPCs The compiled environmental, economic, and social SPCs were evaluated by the construction industry experts against the evaluation criteria to rank them according to their suitability for assessing sustainability of modular buildings. Depending on the circumstances, there might be some limitations on the selection of sustainability performance criteria and hence the analyzer should keep the number of the criteria to a minimum. Therefore, the costs of data collection and implementation of the proposed framework can be minimized (monetary and time-related) [55]. Thus, importance ranking of SPCs assists in the selection of the most suitable areas that need to be measured to assess the sustainability performance of the subject building. Using the developed SPC categories, two questionnaire surveys, called Applicability and Measurability surveys in this paper, were designed with the help of Adobe LiveCycle Designer which provided the respondents an interactive environment. In both surveys, the objective, advantages, confidentiality, duration, completion guidance, contact information, and consent form were included. The surveys also included questions on the participant's background, such as the profession, the years of experience, and the amount of involvement in modular construction projects. The core section of both surveys was intended to evaluate the developed SPCs with respect to the evaluation criteria. Through the Applicability survey and a number of informal interviews, the applicability (relevance) of each SPC for sustainability assessment of modular buildings was examined. In the Measurability survey, the SPCs were evaluated against data availability and data accuracy criteria. In both surveys, the SPCs and their descriptions along with the descriptions of the evaluation criteria were listed. The participants were asked to outline their preferences by scoring each SPC with respect to the

2.1.2. Criteria for evaluation of SPCs In order to identify the most appropriate SPCs for sustainability assessment of a building project, the potential SPCs should be evaluated against suitable evaluation criteria. In general, performance criteria/ indicators should be applicable, adequate, understandable, measurable, and verifiable [44–49]. In a recent work, the suitability of the compiled SPC categories for sustainability assessment of modular buildings was evaluated against applicability only [50]. The study described in this paper, employs both applicability and measurability (i.e., data availability and data accuracy) to evaluate and rank the SPCs. Descriptions of the evaluation criteria are as follows: CI – Applicability (Relevance): How important and relevant is the SPC Table 1 Primary potential sustainability performance criteria proposed in this paper. Environmental SPCs

Economic SPCs

Social SPCs

Site selection (SS) Alternative transportation (AT) Site disruption and appropriate strategies (SD) Renewable energy use (RE) Energy performance and efficiency strategies (EP) Embodied energy (EE) Water and wastewater efficiency strategies (WE) Regional (local) materials (RM) Renewable and environmentally preferable products (REP) Waste management (WM) Greenhouse gas emissions (GE) Material consumption in construction (MC)

Design and construction time (DCT) Design and construction costs (DCC) Operational costs (OC) Maintenance costs (MC) End of life costs (EC) Durability of building (DB) Investment and related risks (IR) Flexibility of building (FB) Integrated management (IM)

Health, comfort, and well-being of occupants (HO) Influence on the local economy (ILE) Functionality and usability of the physical space (FU) Aesthetic options and beauty of building (AB) Workforce health and safety (WHS) Community disturbance (CD) Influence on local social development (ISD) Cultural and heritage conservation (CHC) Affordability (A) Safety and security of building (SSB) User acceptance and satisfaction (UAS) Neighborhood accessibility and amenities (NAA)

24


M. Kamali et al.

SPIs to be normalized, aggregated and compared to benchmarks. The PLs are established between 10 and 100, where 10 representing the threshold of very low performance and 100 representing the most desirable performance of a SPI. Theoretically, the PL of zero indicates the least desirable performance. However, assigning a value of zero to any of the SPIs of a building does not seem justifiable because it can certainly be designed, constructed, used, and demolished/renovated even though the SPCs and corresponding SPIs are not meeting the least desirable performances. The benchmark values for each and every SPI are established through the literature and expert consultations. Then, each calculated SPI is transformed (normalized) into its corresponding performance level. In cases where the values of some of the SPIs lie outside the applicable range, the performance levels of 10 and 100 can be used for lower and higher values, respectively. The PL values and weights of all SPIs associated with a SPC are then aggregated to develop the aggregated sustainability index for the SPC through performing suitable MCDA methods. However, aggregation of PLs with the simple weighted average (SWA) method without considering to what extent a PL value is close to or far from the desirable performance can be misleading. In other words, a suitable aggregation approach based on the synthesizing criterion is required, by which the relative performance of the SPIs under each SPC is considered. In this regard, the TOPSIS MCDA method which is based on the relative closeness to the best performance and relative remoteness from the worst performance, provides a more realistic benchmarking approach. Thus, the TOPSIS method has been used in this study to develop the sustainability indices. The indices developed by TOPSIS are based on the concept of similarity (i.e., relative closeness) to the positive-ideal solution (PIS) and the remoteness from the negative-ideal solution (NIS) [58]. A step-by-step procedure of the TOPSIS method followed in this paper has been described in Appendix D.

evaluation criterion by comparing the sustainability of modular and conventional construction methods. In this research, ordinal scales were chosen to capture the construction professionals' opinions. Primary construction practitioners, such as engineers, architects, construction managers, and manufacturers, as well as academically affiliated experts (originally engineers/architects) were searched as the potential participants for the first questionnaire and informal interviews. In this connection, an attempt was made to identify those practitioners that had experience in both modular and conventional building projects with the focus on North American construction industry. Because of space limitations, the details of the surveys including the scoring systems, and the demographic characteristics of the respondents have been provided in Appendix B. The data collected through the surveys and interviews for scoring the SPCs against the evaluation criteria was combined with the weights of the evaluation criteria using the ELECTRE 1 analyses. Developed by Benayoun et al. [56], ELECTRE 1 method is one of the most known MCDA outranking methods that have been extensively employed in different decision making problems. In this paper, the ELECTRE 1 was employed to analyze and rank the developed SPCs within each sustainability category. In the solution algorithm of this method, dissimilar to the other compensatory MCDA methods, weights are not viewed as the direct criteria substitution rates, but rather the absolute power of each individual criterion toward reaching the final goal, hence making the method non-compensatory [57]. In addition, when an oral scale is used to evaluate criteria, it might be difficult to establish preferences between different alternatives (i.e., SPCs). In such circumstances (e.g., this research), this method can resolve the problem by accumulating slight differences of scorings between different alternatives (i.e., SPCs) with regard to each evaluation criterion; hence, distinct outranking relations between different alternatives can be established [44]. In ELECTRE method, the concordance and discordance sets are produced to form outranking relationships between alternatives. In fact, the concordance and discordance sets represent the level of satisfaction and dissatisfaction of a decision maker (i.e., a survey participant in this paper), respectively, when he/she gives preference to one alternative over the others [58]. Details about the ELECTRE method and its step-by step procedure can be found in references [57,58] and Appendix C.

3. Results and discussion 3.1. Ranked SPC categories By implementing the AHP group decision making process described in the methodology section, the weights of applicability, data availability, and data accuracy, were determined to be 59.65%, 20.54%, and 19.81%, respectively. This indicates that participants believed that the applicability of a SPC is more important than its measurability. To examine the impact of the weights of the evaluation criteria on the overall rank of the SPCs, a sensitivity analysis has been conducted and presented in this section. As stated earlier, the data collected from the surveys along with the above weights were analyzed to rank the SPCs. The ELECTRE 1 steps (Appendix C) were carefully followed for each sustainability category to separately determine the overall rank of the environmental, economic, and social SPCs. As an example, the ELECTRE calculations for the economic SPC category have been included in Appendix C.

2.2. Aggregated sustainability indices The proposed framework in this paper leads to development of sustainability indices for the SPCs by which the performance of the given modular building can be benchmarked. A SPC's sustainability performance index can be developed through calculating and aggregating the corresponding SPIs. To determine appropriate SPIs under each selected SPC, their measurement methods, and their relative importance weights, another study required to be conducted. Similar to what has been done for compiling SPCs, the literature regarding rating systems and other published studies were reviewed. Through a few screening processes, all SPIs related to a SPC were established. As an example, the SPIs related to the SPC ‘Renewable energy use’ were determined to be ‘Renewable electricity generation systems’ and ‘Solar hot water heating systems’. The approach followed in this research involves choosing the simplest methods of measurement for SPIs by which each SPI can be measured by having the minimum amount of data. Therefore, in the cases where the same SPI existed in different references with (different names and) different methods of measurement, the one that provided easiest method and needed less data was selected. The established SPIs under a SPC do not necessarily have the same unit of measurement. For instance, while all the SPIs under ‘Embodied energy’ are of the same unit, this is not the case for the two SPIs under ‘Renewable energy use’. Therefore, this paper has introduced and employed the concept of performance level (PL) to enable the calculated

3.1.1. Environmental SPCs The overall rank order of each SPC within the environmental category is shown in Table 2. Each SPC was ranked based on its net concordance and net discordance indices (Cp and Dp). In cases where the rankings of Cp and Dp are inconsistent (e.g., ‘Waste management’), a method was proposed and used in this paper to determine the final ranking of the SPC (see Appendix E for details). As Table 2 illustrates, the most significant SPC among the top prioritized environmental criteria is ‘Energy performance and efficiency strategies’. The literature revealed the use phase to be strongly prevailing in terms of energy consumption and environmental impacts [59–61]. For example, between 70% and 98% of a building total energy is consumed in its occupancy phase [62–64]. Two reasons may be offered as to why the respondents chose ‘Energy performance and 25


M. Kamali et al.

Table 2 Net outranking of the environmental sustainability performance criteria. SPC

Cp

Dp

Ranking of Cp

Ranking of Dp

Final ranking of SPC

Energy performance and efficiency strategies (EP) Waste management (WM) Material consumption in construction (MC) Site disruption and appropriate strategies (SD) Embodied Energy (EE) Renewable and environmentally preferable products (REP) Greenhouse gas emissions (GE) Renewable energy use (RE) Regional (local) materials (RM) Site selection (SS) Water and wastewater efficiency strategies (WE) Alternative transportation (AT)

9.807 7.772 7.807 0.983 2.228 −0.174 −0.840 −1.772 −6.176 −6.809 −5.832 −6.994

−10.773 −8.178 −8.008 −3.014 −1.401 −0.127 1.145 1.139 6.837 6.645 7.860 7.876

1 3 2 5 4 6 7 8 10 11 9 12

1 2 3 4 5 6 8 7 10 9 11 12

1 2 3 4 5 6 7 8 9 10 11 12

foundations work [1,71]. This SPC is also capable to be measured precisely as time is a crisp measure with minimum uncertainty. The next priority SPC identified by the respondents was ‘Design and construction costs’. As Table 3 illustrates, the costs associated with the construction activities (e.g., design, coordination, material, labor, and so forth) were given more attention by the construction practitioners than the costs of occupancy and end of life phases. This is because the costs associated with the initial phases of a building life cycle can be perceived as short-term costs; therefore, they are more perceptible. Interestingly, the costs related to the end of life phase are long-term costs and have been rated as the least priority SPC within the economic category. Accordingly, ‘Operational costs’ and ‘Maintenance costs’, which are both mid-term costs, were located somewhere between ‘Design and construction costs’ and ‘End of life costs’. Another reason for such a ranking can be the data availability to measure the criteria. Compared to ‘Design and construction costs’, there is less information available related to criteria ‘Operational costs’, ‘Maintenance costs’, and ‘End of life costs’. Modular construction can effectively influence the other SPCs from the economic point of view, such as ‘Durability of the building’. The modular construction method offers higher quality than the traditional counterpart due to controlled manufacturing environment [69,72,73]. Furthermore, higher finished building quality can be achieved due to ‘less material exposure to harsh weather’ on the final project site. Regardless of using extra materials when transporting modules to ensure the required structural strength, in many cases, high quality, lightweight, and durable materials are utilized in the construction of each module itself [2,74–76].

efficiency strategies’ as the top environmental priority SPC: i) the modular and conventional buildings' dissimilar designs of operational energy efficiency strategies, e.g., insulation systems, etc.; and ii) accessibility and accuracy of data, in which the performance of this criterion can be measured. In recent years, buildings have been becoming more energy efficient over the occupancy phase due to employing efficient technologies. Consequently, other life cycle phases of the buildings should be prone to gain growing importance, owing to the commitment to reduce the energy consumption within these phases [65–67]. This is why both ‘Waste management’ which can save energy in both the design and construction phase and the end of life phase and also ‘Embodied energy’ which represents energy in the design and construction phase (and even before in material acquisition phase) were assigned good ranks of second and fifth by the experts, respectively. The other reason that ‘Waste management’ was highlighted as the second major environmental concern is the potential capability of providing more efficient waste management strategies in modular factory environments, contrary to on-site (conventional) construction [68,69]. For example, materials can be precisely cut in the factory environment which results in less construction waste. As another example, different modules can be used in other projects by disassembling, relocating, or refurbishing them in the end of the building lifetime [70]. Nevertheless, up to 15% additional materials are consumed in modular construction when transporting the fabricated modules to the final site to ensure their ‘required integrity and structural strength’ [4]. This also confirms the difference in the ‘Material consumption in construction’ between modular and conventional construction which was emphasized by the construction practitioners in this research. The SPCs ‘Site selection’, ‘Water and wastewater efficiency strategies’, and ‘Alternative transportation’ were all dominated by all other SPCs within the category. The construction experts believed that there is no significant difference between modular and conventional buildings with respect to these SPCs; therefore, they assigned the lowest scores to the applicability of these SPCs among all the 12 SPCs. This is the reason why even increasing the weights of the other two evaluation criteria (i.e., data availability and data accuracy) did not improve the rank order of these SPCs (see the Sensitivity analysis section below).

3.1.3. Social SPCs Results of the final preferences for the SPCs within the social category are reported in Table 4. According to this table, ‘Workforce health and safety’ was ranked first among all the 12 social criteria. This SPC received highest applicability and data accuracy scores by the respondents among all social SPCs. This was anticipated because: i) according to the literature, the modular and conventional construction methods can provide extremely different degrees of workforce health and safety in construction projects. For example, on-site reportable accidents can be reduced up to 80% by transferring the majority of construction activities to manufacturing centers when using modular construction [77,78]; and ii) if the information associated with the construction accidents is available, it should be accurate (crisp) by nature as it reports the number of injuries and fatalities due to construction activities in a specific period of time. Although ‘Workforce health and safety’ was of paramount importance from the construction experts' perspectives; ‘Health, comfort and well-being of occupants’ was not (10th). The respondents believed that both modular and conventional buildings can provide the end users (i.e., occupants) similar level of health, comfort and well-being; therefore, they did not score this SPC as a significant criterion for

3.1.2. Economic SPCs The outcomes resulting from the economic criteria ranking based on the ELECTRE method are reviewed below. According to Table 3, the top priority economic criteria that outranked all other economic SPCs, was identified as ‘Design and construction time’. Supporting this, the literature reveals that a significant difference among modular construction and its counterpart is the fast turnaround between the breaking of ground and occupancy when using modular construction. This is mainly because in this method, manufacturing a building's modules can be executed parallel to preparation of the building's final site, e.g., 26


M. Kamali et al.

Table 3 Net outranking of the economic sustainability performance criteria. SPC

Cp

Dp

Ranking of Cp

Ranking of Dp


Design and construction time (DCT) Design and construction costs (DCC) Investment and related risks (IR) Durability of building (DB) Integrated management (IM) Operational costs (OC) Flexibility of building (FB) Maintenance costs (MC) End of life costs (EC)

7.604 6.396 1.614 2.203 1.186 −2.587 −3.420 −4.995 −8.000

−7.583 −6.417 −3.478 −1.540 −0.736 3.333 3.564 4.875 8.000

1 2 4 3 5 6 7 8 9

1 2 3 4 5 6 7 8 9

1 2 3 4 5 6 7 8 9

• WS : applicability (50%), data availability (25%), data accuracy

sustainability assessment. In addition, another reason may be the measurability of the latter SPC compared to the former as there is fewer data available and it needs to be communicated with the occupants of the buildings after the occupancy to quantify this criterion. The next high importance criteria within the social category were ‘Safety and security’ and ‘Community disturbance’. Safety and security performance assessment of a building constructed by modular construction method seemed important to the construction experts. However, this does not necessarily mean that one of the two construction methods provides safer and more secure buildings. Regarding the next SPC in Table 4, modular construction has been claimed to provide cheaper buildings; therefore, ‘affordability’ of buyers becomes a major consideration when choosing a building constructed by modular or conventional method. In addition, the required data to measure this SPC (e.g., overall affordability of a household to buy a house, the sale price of similar modular/conventional houses) is available and accurate based on the statistical reports prepared by different governmental or private organizations each year. In the case of ‘Community disturbance’, the respondents believed that there is a significant difference between modular and conventional construction in terms of minimizing the impacts of on-site construction activities on both occupants and surrounding local communities. According to the literature, the majority of the construction activities (approximately 90%) that are needed to construct a modular building is undertaken in factory environments [1]. Consequently, on-site construction activities' negative impacts on surrounding neighborhoods and families, such as construction noise, traffic congestion, dust, among others, can be substantially reduced.

2

(25%)

Compared to WSGDM (i.e., applicability: 59.65%, data availability: 20.54%, and data accuracy: 19.81%), in WS1, more weight was assigned to the applicability criterion, contrary to the case of WS2 where the weights of data availability and data accuracy criteria were increased. Results of repeating the ELECTRE analyses for each sustainability category using the three different weight sets are presented in Fig. 2. It can be seen from the net outranking of the SPCs that changing the established weights of the evaluation criteria (WSGDM) within at least the defined range (WS1↔WS2), does not affect the rankings of the economic and social SPCs (Fig. 2b and Fig. 2c). In other words, all the economic and social SPCs were assigned identical ranks using the three weight sets except CD and A with rank orders of third and fourth for WS1 (they are fourth and third, respectively, for both WSGDM and WS2). In contrast, as Fig. 2a demonstrates, there are some minor discrepancies between the environmental SPC rankings when using the different weight sets. However, these inconsistencies cannot be significant because (1) all the SPC rankings follow the same trend, which means when switching from a WS to another, the rank order of a SPC changed locally; and (2) there is no change in the rank order of the top priority SPCs (i.e., EP, WM, and MC) for all three WSs. 3.2. Case study In every building project, decision makers or any users of the proposed sustainability assessment framework have limitations and priorities (e.g., data collection cost and duration, priorities on sustainability categories and/or life cycle phases). In other words, it is difficult to choose all the SPCs, quantify the associated SPIs and develop the aggregated sustainability indices. Therefore, according to the SPC rank orders, starting from the top to down, maximum possible number of the SPCs within the intended sustainability category is selected and used in the sustainability performance assessment process. In this section, the proposed framework was applied to a case study of a residential modular building to benchmark its environmental performance.

3.1.4. Sensitivity analysis In addition to the weight set assigned through the AHP-based group decision making process (WSGDM), two additional hypothetical weight sets were applied to examine the sensitivity of the SPC ranking results to the weights of the evaluation criteria. These additional weight sets included:

• WS : applicability (70%), data availability (15%), data accuracy 1

(15%)

Table 4 Net outranking of the social sustainability performance criteria. SPC

Cp

Dp

Ranking of Cp

Ranking of Dp


Workforce health and safety (WHS) Safety and security of building (SSB) Affordability (A) Community disturbance (CD) Functionality and usability of the physical space (FU) User acceptance and satisfaction (UAS) Aesthetic options and beauty of the building (AB) Influence on the local economy (ILE) Neighborhood accessibility and amenities (NAA) Health, comfort and well-being of occupants (HO) Influence on local social development (ISD) Cultural and heritage conservation (CHC)

9.386 8.418 5.619 5.361 2.409 0.965 −1.980 −1.600 −3.953 −6.599 −8.012 −10.002

−10.360 −9.238 −6.038 −5.292 −3.119 0.760 0.935 2.330 4.491 7.078 8.080 10.372

1 2 3 4 5 6 8 7 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

27


M. Kamali et al.

Fig. 2. Net outranking of (a) Environmental category; (b) Economic category; (c) Social category, for different weight sets of the evaluation criteria.

of the building in which the modules are highlighted by dashed lines. According to the builder, the finished modules of the building are then transported to the final site which can be located in different places throughout BC. As stated before, to develop a sustainability index for a SPC, corresponding SPIs should be determined, calculated (i.e., performance levels), and aggregated into their weights. Each SPI can be calculated

3.2.1. Description of the case study building The case study modular building is a common single family home built in the Okanagan, BC, Canada, by one of the known modular home builders in Canada. This wooden one-story building consists of three modules with the total floor area of 138 square meters (m2 or 1480 square foot [ft2]). It includes three bedrooms, two bathrooms, one dining room, living room, kitchen, and den. Fig. 3. shows the floor plan

Fig. 3. Floor plan of the case study modular building. 28


M. Kamali et al.

Table 5 The environmental SPCs and corresponding SPIs used for sustainability assessment of the case study modular building. SPI

SPI weight

Units of data variable

Source

1- Energy performance and efficiency strategies (EP) EP1- Envelope insulation EP2- Air filtration EP3- Windows and glass doors EP4- Space heating and cooling equipment EP5- Heating & cooling distribution system EP6- Efficient domestic hot water equip. EP7- Efficient lighting EP8- Efficient appliances EP9- Residential refrigerant management

0.077 0.107 0.107 0.107 0.143 0.214 0.107 0.107 0.036

improved insulation % ACH@50PA ER or Ufactor (W/m2.K) SEER, COP, AFUE, etc. heating/cooling features hot water equipment Y/N to conditions Y/N to requirements Y/N to conditions

LEED, literature LEED, literature ENERGY STAR, literature LEED, EO LEED, EO LEED ENERGY STAR, LEED ENERGY STAR, LEED LEED

2- Construction waste management (CWM) CWM1- Material efficient consumption CWM2- Waste diversion from landfill CWM3- Recycled aggregates

0.46 0.27 0.27

Y/N to strategies % of waste diversion % recycled aggregate

Green Globes, LEED, EO Green Globes, LEED, BREEAM BREEAM, EO

3- End of life waste management (EWM) EWM1- Reuse of façades EWM2- Reuse of structural systems, etc. EWM3- Reuse of non-structural elements

0.374 0.313 0.313

% of reuse % of reuse % of reuse

Green Globes, EO Green Globes, EO Green Globes, EO

4- Site disruption and appropriate strategies (SD) SD1- Construction activity pollution prevention SD2- Landscaping SD3- Heat island effects SD4- Rainwater management SD5- Non-toxic pest control

0.056 0.389 0.056 0.389 0.111

Y/N Y/N Y/N Y/N Y/N

to to to to to

LEED, LEED, LEED, LEED, LEED,

5- Renewable and environmentally preferable products (REP) REP1- Types of exterior wall materials REP2- Types of floor materials REP3- Types of foundation materials REP4- Types of interior wall & ceilings materials REP5- Types of landscape materials REP6- Types of roof materials REP7- Types of roof, floor and wall materials REP8- Types of miscellaneous materials

0.133 0.133 0.067 0.133 0.067 0.133 0.133 0200

% % % % % % % %

REP REP REP REP REP REP REP REP

6- Renewable energy use (RE) RE1- Renewable electricity generation system RE2- Solar hot water heating systems

0.769 0.231

renewable energy % Y/N, equipment type

Energuide, LEED CAN/CSA, LEED

7- Regional materials (RM) RM1- Exterior wall materials RM2- Floor materials RM3- Foundation materials RM4- Interior wall & ceilings materials RM5- Landscape materials RM6- Roof materials RM7- Roof, floor and wall materials RM8- Miscellaneous materials

0.133 0.133 0.067 0.133 0.067 0.133 0.133 0200

% % % % % % % %

BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM,

of of of of of of of of

of of of of of of of of

conditions features, % conditions, % features, % conditions

local local local local local local local local

content content content material content content content content content

material material material material material material material material

EO EO EO EO EO

BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM, BREEAM,

LEED, LEED, LEED, LEED, LEED, LEED, LEED, LEED,

LEED, LEED, LEED, LEED, LEED, LEED, LEED, LEED,

EO EO EO EO EO EO EO EO

EO EO EO EO EO EO EO EO

Note: LEED = leadership in energy and environmental design, BREEAM = building research establishment environmental assessment method, LBC = living building challenge, EO = expert opinions.

1. The weights of all SPIs were determined through a literature review. The main focus was on the references the SPIs had been selected. The results are presented in Table 5. 2. In this paper, the performance levels of all SPIs have already been established as the benefit criteria ranging from 10 to 100; hence, there was no need for normalization. All the SPIs were first calculated using the collected data, and then transformed into equivalent normalized performance levels. For example, Table 6 shows the performance levels for all 9 SPIs under the EP SPC. 3. The weighted values for the SPIs (vij) were calculated by multiplying the performance level of each SPI and its weight (Table 6). 4. The positive-ideal solution (PIS) and negative-ideal solution (NIS) for all the SPIs were calculated in terms of weighted performance levels by substituting the most and least desirable performance values (100 and 0, respectively). For instance, the PIS and NIS for EP3 were calculated as:

by collecting the required data. To evaluate the environmental performance of the case study building, a questionnaire was designed to collect the data required to calculate the SPIs. The participating modular home builder was initially contacted by email and phone explaining the overall research objectives and was invited to participate in this study. Following the initial contacts, the questionnaire was reviewed in detail through a number of meetings in the modular factory to eliminate any ambiguity. Table 5 presents the SPCs selected by the modular builder. Accordingly, the corresponding SPIs have been listed in this table. 3.2.2. Case study results and discussion The aggregated sustainability indices for the case study building were developed through step-by-step implementation of the TOPSIS method. Due to space limitations, only the details of the sustainability index calculation for ‘Energy performance and efficiency strategies’ (EP) are presented; however, final results for the other sustainability indices have been included (further descriptions of the steps and equations can be found in Appendix D).

PISEP3 = 100 × 0.107 = 10.70

29


M. Kamali et al.

Table 6 Performance levels, weighted values, positive-ideal soltions, and negative-ideal solutions for the SPIs of EP. SPI

EP1

EP2

EP3

EP4

EP5

EP6

EP7

EP8

EP9

PL vij PIS NIS

(0.071) 10 0.71 7.10 0.00

(0.107) 10 1.07 10.70 0.00

(0.107) 70 7.49 10.70 0.00

(0.107) 55 5.89 10.70 0.00

(0.143) 100 14.30 14.30 0.00

(0.214) 25 5.35 21.40 0.00

(0.107) 10 1.07 10.70 0.00

(0.107) 70 7.49 10.70 0.00

(0.036) 10 0.36 3.60 0.00

Note: Values in parentheses represent the corresponding weights of SPIs.

manufactured within 800 km of the building's final location. ‘Construction waste management’ is the next SPC showed High performance. This was expected because, according to the literature, modular construction provides better waste management results in terms of control (reduce), reuse, recycle, and waste disposal in manufacturing centers [1,68,71]. However, the ‘End of life waste management’ did not show a satisfactory performance (EWMi = 22.83). By investigating the performance levels of the SPIs under this SPC, it was realized that only a small portion of the façades and none of the structural and non-structural systems of the existing buildings (that are at the end of their life), are reused in the construction of new buildings in the modular factory. The worst performing SPC among all the environmental assessment criteria is ‘Renewable energy use’. According to the participating modular builder, no strategies are being adopted in the design and production of the building for using renewable energy sources, such as installation of solar hot water heating systems. Therefore, the environment cannot benefit from less consumption of non-renewable energy sources, especially when taking into account the total energy consumption during over 50 years of building's occupancy phase. The other SPC that needs improvement is ‘Renewable and environmentally preferable products’ with Medium performance (REPi = 34.58). Investigation of the performance levels of the SPIs under this SPC (REP1-REP8 in Table 5), indicated that materials such as FSC-certified or reclaimed wood and materials with at least 25% recycled content, were used in the components of only two assemblies of the building (flooring, as well as insulation of roof, floor, and wall). If the builder pursues the strategy of consuming renewable energy and environmentally friendly materials, it can result in less consumption of raw materials and also less requirement for energy associated with extracting, processing, and manufacturing of these materials.

NISEP3 = 0 × 0.107 = 0.00 Likewise, the PIS and NIS were calculated for all the SPIs as presented in Table 6. 5. In this step, the distance of the building performance, with regard to each SPC, from the positive and negative solutions (separation measures) was calculated using the n-dimensional Euclidean distance. For example, the separation measures for the EP SPC was measures as:

S+EP =

(0.71 − 7.10)2 + (1.07 − 10.70)2 +…+(0.36 − 3.60)2 = 23.19

S−EP =

(0.71 − 0.00)2 + (1.07 − 0.00)2 +…+(0.36 − 0.00)2 = 19.57

The results of separation measures for all the selected SPCs are provided in Table 7. 6. As the last step, the aggregated sustainability index for each SPC was developed by calculating similarities to PIS. The sustainability index for the ‘Energy performance and efficiency strategies’ for the subject modular building was obtained as:

S− 19.57 ⎞ × 100 = 45. 76 EPi = ⎜⎛ − EP + ⎟⎞ × 100 = ⎛ ⎝ 19.57 + 23.19 ⎠ ⎝ SEP + SEP ⎠ To develop each sustainability index between 10 and 100, the result was multiplied by 100 (as seen in the above calculation). Similarly, other aggregated sustainability indices (i.e., CWMi, EWMi, SDi, REPi, REi, and RMi) were developed and presented in Table 7. To facilitate in taking timely actions for improvement of the building performance in accordance with the developed sustainability indices, a SPC performance scale ranging from ‘Very Low’ to ‘Excellent’ has been proposed in this study. In this regard, the performance scales provided by known environmental sustainability rating systems including LEED, BREEAM, Green Globes, along with expert opinions were used to establish the SPC performance scale as presented in Table 8. By plotting the aggregated sustainability indices against the proposed SPC performance scale, the performance of the case study building with respect to different environmental criteria was illustrated (Fig. 4). From Fig. 4 it is evident that the overall life cycle environmental performance of the building is satisfactory, with sustainability indices lying in the Medium, Good, High, and even Excellent ranges (sustainability indices > 29), except for two SPCs (i.e., ‘Renewable energy use’ and ‘End of life waste management’). The top SPC with Excellent performance is ‘Regional materials’ (RMi = 81.05). This is due to the fact that the materials required for almost all components of this building's assemblies (e.g., exterior and interior walls, floor, foundation) were extracted, processed, and

3.3. Research limitations Similar to any other research, there are a number of challenges related to the analyses and factors that could affect the results. The most important limitations have been discussed in this section. The sample sizes of the surveys in this paper were beyond the minimum requirement (according to Cochran [79]). Since the result of a survey depends on its respondents whose answers are subjective and not repeatable, better conclusions could be drawn if there were more participants. However, it is difficult to expect experts to spend their time on longer surveys in today's busy industry. It should be mentioned that to reduce uncertainty, the questionnaires were sent to people very experienced in both modular and conventional construction industries. The other difficulty faced by this research was the data collection challenges for case study analyses. Among four modular builders in the Okanagan, BC, only one provided data for one of its typical residential modular buildings. However, having more case studies provides useful information to establish benchmarks among modular buildings. The proposed framework in this study has been developed after completing a comprehensive review of the previous studies, conducting interviews with experts in both modular and conventional construction industries, and employing suitable MCDA methods. Therefore, the

Table 7 Separation measures and sustainability indices for all the selected environmental SPCs. SPC

EP

CWM

EWM

SD

REP

RE

RM

S+ S− Sustainability Index (SPCi)

23.19 19.57 45.76

25.88 44.66 63.31

46.67 13.82 22.83

28.43 33.59 54.16

29.36 15.52 34.58

72.26 8.03 10.00

6.69 28.62 81.05

30


M. Kamali et al.

Table 8 Proposed SPC performance scale and improvement actions. Sustainability Index (SPCi)

Performance

Suggested action

SPCi < 11 11 ≤ SPCi < 29 29 ≤ SPCi < 43 43 ≤ SPCi < 55 55 ≤ SPCi < 69 69 ≤ SPCi < 100

Very Low Low Medium Good High Excellent

Unsatisfactory performance. Urgent and detailed improvements required Detailed investigations required for major and/or minor improvements of SPIs Investigations required to improve many of the SPIs Satisfactory performance but improvements required for underperforming SPIs Very good performance. Minor improvement of some SPIs recommended if doesn't cost much No action required. Consistency required to obtain similar results

Fig. 4. Aggregated sustainability indices and corresponding performance for the selected SPCs.

assessment framework for residential modular buildings that adequately addresses the existing research gaps. Metric benchmarking with the use of system of sustainability criteria, which is the most widely used and recognized method of benchmarking, is employed as the benchmarking approach in this research. In the proposed framework, first, potential sustainability performance criteria (SPCs) to be used in the sustainability assessment process are compiled and grouped into three categories of sustainability dimensions. The construction industry experts' perceptions of these SPCs with respect to three evaluation criteria; applicability, data availability, and data accuracy are captured and ELECTRE 1 and AHP MCDA methods are used in analyses. Consequently, the SPCs are ranked within their associated sustainability categories that determines the priorities in selection of SPCs for sustainability assessment of modular buildings. Depending on the circumstances and limitations of each building project, the decision maker/assessor chooses maximum possible number of SPCs according to their rank orders and uses them in the assessment process. Second, the aggregated sustainability criteria for benchmarking modular buildings are developed. This is performed by determining suitable sustainability performance indicators (SPIs) associated with each SPC, their measurement methods, their relative importance weights, and their performance benchmark values. By collecting the required data, each SPI is calculated and then transformed (normalized) into a performance level (PL) between 10 and 100 depending on the benchmark value of the SPI. Finally, through an aggregation process, a sustainability index is developed for each SPC using the TOPSIS MCDA method. The TOPSIS method which is based on the relative closeness to the best performance and relative remoteness from the worst performance, provides a more realistic benchmarking approach. The proposed framework has been validated through a case study of a modular building in the Okanagan, BC, Canada. Aggregated sustainability indices that are developed through

framework can be effectively used for performance benchmarking of typical residential modular buildings for any region in the world by collecting data for measuring as many sustainability criteria as possible and developing the aggregated sustainability indices. Nevertheless, the benchmark values (i.e., performance level relationships) and the SPC performance scale established in this research can never replace the actual benchmarking processes based on the performance of many existing modular buildings in a region. Therefore, users can import the methodology, and if required, review, revise, and recalibrate the SPIs' benchmark values and tweak the SPC performance scale considering the site specific socio-economic and geographical conditions. This should be performed over time through a continuous benchmarking process and might take a few years to compile a comprehensive data base for benchmarking of modular buildings as the modular construction industry is relatively new. 4. Conclusions Modular construction offers many advantages, such as speed of work, less cost, less waste, and so forth, that can effectively contribute to sustainable construction. However, this method of construction also faces some challenges, such as transportation constraints, more complicated engineering and planning processes, the public's negative perceptions of new construction methods, among others. Therefore, it is imperative to comparatively appraise the sustainability of modular versus conventional buildings over the entire life cycle through an assessment framework. This assessment process should be capable of addressing all triple bottom line (TBL) sustainability dimensions, i.e., economic, environmental, and social as well as all life cycle phases of buildings. Currently, there is no major sustainability assessment studies in the literature to benchmark modular buildings by addressing all sustainability dimensions over the life cycle. This paper proposes and validates a benchmarking based life cycle sustainability performance 31


M. Kamali et al.

associated with data variables and expert opinions. The proposed framework in this paper can also be used to develop sustainability indices for each life cycle phase, for each sustainability dimension, and for the overall performance of the subject building [80]. It can also be adopted for sustainability performance assessment in other construction practices or by researchers in other fields to evaluate a process or product's performance. This can be accomplished by selecting suitable criteria, measuring them, and establishing the associated benchmarks.

implementation of the proposed sustainability assessment framework in this research, describe the performance of the subject modular building with regard to each SPC within the sustainability category it belongs to. Furthermore, the contribution of each life cycle phase to sustainability of the building can be explored. Therefore, the results of sustainability assessment can be used to improve the life cycle sustainability performance of the building by identification of low performing SPCs and further investigation of the corresponding SPIs. In addition, the results can assist the construction practitioners, such as building designers, architects, decision makers, and developers to make informed decisions on the selection of the best construction methods. The guidelines established in this work provide a baseline to initiate the benchmarking process for residential modular buildings and involve uncertainties. Therefore, for future research, the proposed framework is recommended to be improved upon to handle the uncertainties

Acknowledgement The authors gratefully acknowledge Mr. Lloyd Dehart of Moduline Industries for providing the case study data.

Appendix A. Descriptions of the TBL sustainability performance criteria Table A1 Environmental sustainability performance criteria (adopted from [50]). SPC

Description

Site selection (SS)

Avoiding development of inappropriate sites (e.g., farmlands, greenfields, etc.), constructing in urban areas with existing infrastructure, and so forth. Alternative transportation (AT) Availability of public transportation access, cycling facilities, among others. Site disruption and appropriate strategies Promoting natural biodiversity (e.g., providing adequate open space), planning for stormwater (SD) management, avoiding blocking fresh air or sunlight or natural waterways for adjacent developments, and so forth. Renewable energy use (RE) Reducing the negative impacts of non-renewable energy consumption by using renewable energy sources (e.g., solar). Energy performance and efficiency Energy monitoring, minimizing space for heating/cooling, integrating daylight to reduce the need for strategies (EP) electrical lighting, among others. Embodied energy (EE) Energy consumed by all of the processes associated with the construction of the building (e.g., processing of natural resources, manufacturing, transportation, etc.). Water and wastewater efficiency Reducing the potable water consumption by reuse and recycling systems, water monitoring, among strategies (WE) others. Regional (local) materials (RM) Using building materials and products which are extracted, processed, and manufactured within the region. Renewable and environmentally Increasing demand for building components that maximize renewable and environmentally preferable preferable products (REP) material consumption Waste management (WM) Strategies for diverting waste from disposal in landfills and incineration facilities (Reuse, Recycling, etc.). Greenhouse gas emissions (GE) The amount of greenhouse gas emissions leading to global warming, ozone depletion, etc. Material consumption in construction The amount of any product or natural resource used during the design and construction phase of the (MC) building.

Table A2 Economic sustainability performance criteria (adopted from [50]). SPC

Description

Design and construction time (DCT) Design and construction costs (DCC) Operational costs (OC) Maintenance costs (MC) End of life costs (EC) Durability of building (DB)

Total duration of the building design and construction (e.g., planning, manufacturing, installation, finishing, etc.).

Pre-construction and construction costs (e.g., coordination, drawings, materials, workforce, transportation, defect/ damage to the product before final completion, etc.) Costs of the building operation during the occupancy phase. Costs of repair and maintenance of the building during the occupancy phase. Costs of dismantling and waste treatment (recycling, disposal, etc.). Specifying durable and low-maintenance building materials and assemblies in order to have a building with a long usable life leading to economic benefits. Investment and related risks The speed of return on loans or other investments and the associated risks. (IR) (continued on next page) 32


M. Kamali et al.

Table A2 (continued) SPC

Description

Flexibility of building (FB)

Compatibility of the product and adaptability to accommodate substantial changes in the future at a lower cost (e.g., using fastening systems that allow for easy disassembly). Ways of handling the management functions and procedures that are conducted during the design and construction of the building.

Integrated management (IM)

Table A3 Social sustainability performance criteria (adopted from [50]). SPC

Description

Health, comfort, and well-being of occupants (HO) Influence on the local economy (ILE) Functionality and usability of the physical space (FU) Aesthetic options and beauty of building (AB) Workforce health and safety (WHS) Community disturbance (CD)

Health, comfort and well-being of the end users in the occupancy phase of the building life cycle (e.g., indoor air pollutants). Influence on the region of construction (e.g., job market, employment, workforce stability, etc.). Usability of the physical space for engineering systems as well as future occupants (e.g., physical spans, openings, etc.). Containing design features intended for human delight, spirit and place appropriate to its function, internal and external beauty, and visual appearance. Risks of any health and safety issues in the workplace (e.g., injury, damage, chronic health risks, etc.) Impacts of the construction activities on occupants and surrounding local communities (e.g., construction noise and dust, traffic congestion, etc.). Influence on local social development Influence on culture, interaction of people, and development of new local communities. (ISD) Cultural and heritage conservation Considering local cultural heritage sites. (CHC) Affordability (A) Ability to purchase the building. Safety and security of building (SSB) Providing adequate measures and equipment that promote low risk, safe and secure use of the building (e.g., fire/seismic resistance). User acceptance and satisfaction The amount of occupant satisfaction when interacting with the building. (UAS) Neighborhood accessibility and Proximity to local facilities (e.g., recreational centre, parks, etc.). amenities (NAA)

Appendix B. Scoring systems of the Applicability and Measurability surveys and demographic characteristics of the respondents The scoring systems applied to evaluate the SPCs against the evaluation criteria should be clearly defined to the respondents. In this research, ordinal scales were chosen to capture the construction professionals' opinions as shown in Table B1. The applicability of the SPCs were ranked on a five-point scale with an interval of one ranging from 1 as ‘very low’ to 5 as ‘very high’; whereas, the data availability and data accuracy for the SPCs were ranked using a three-point scale (1 as ‘low’, 2 as ‘medium’, and 3 as ‘high’). Table B1 Scoring system used for evaluating the SPCs with respect to the evaluation criteria. Evaluation criterion

Score

CI – Applicability (Relevance)

SPC is the least important and seems to be irrelevant for sustainability assessment of modular vs. conventional buildings. SPC is fairly important and has low relevance to sustainability assessment of modular vs. conventional buildings. SPC is important and has adequate relevance to sustainability assessment of modular vs. conventional buildings. SPC is very important and highly applicable for sustainability assessment of modular vs. conventional buildings. SPC is extremely important and has to be chosen for sustainability assessment of modular vs. conventional buildings. 1: Low No or limited data is available for measuring the variables of the SPC. 2: Medium Some data is available for measuring the variables of the SPC. 3: High Enough data is available for measuring the variables of the SPC.

CII – Data Availability

Description

1: Very Low 2: Low 3: Medium 4: High 5: Very High

CIII – Data Accuracy (continued on next page)

33


M. Kamali et al.

Table B1 (continued) Evaluation criterion

Score

Description

Data variables of the SPC cannot be accurately measured. 1: Low 2: Medium Data variables of the SPC can be measured with medium accuracy (with some uncertainties). Data variables of the SPC can be accurately measured (all the variables have absolute values). 3: High Primary construction practitioners, such as engineers, architects, construction managers, and manufacturers, as well as academically affiliated experts (originally engineers/architects) were searched as the potential participants for the first questionnaire and informal interviews. In this connection, an attempt was made to identify those practitioners that had experience in both modular and conventional building projects with the focus on North American construction industry. Once the questionnaires were designed and the potential participants were identified, the next step was to implement the surveys. The Applicability survey was disseminated to the identified potential participants through delivering online (emailing the interactive version) or delivering offline (distributing the paper version). In total, 210 potential respondents were sent the questionnaire and consent forms. Consequently, two follow up reminders were emailed to those who delayed their response. Among the 53 forms returned, 48 participants fully completed the questionnaire forms. The participants were affiliated with diverse types of organizations, such as engineering companies, modular manufacturers, general contractors, and academic institutions. Moreover, the number of employees in each organization was different. The professional experience of experts (regardless of modular or conventional projects) can play a significant role in evaluating a SPC against its measurability (i.e., data availability and data accuracy). Approximately 60% of the participants in the measurability survey had over 20 years of professional experience compared to 45% in the Applicability survey. In addition, the participants of the Measurability and Applicability surveys who had been involved in at least 6 modular projects were 65% and 54%, respectively. Table B2 shows demographic characteristics of the surveys' respondents including the profession and experience in either the construction industry or academia. It should be mentioned that, different professions were grouped into a single category in the table, depending on their similarities in job characteristics. Table B2 Profession and years of professional experience of respondents. Parameter

Parameter

Applicability Survey

Measurability Survey

Profession

Client, Developer Engineer, Architect Contractor, Construction Manager, Off-site manufacturer Academic researcher (originally engineer/ architect) Over 30 20–30 10–20 Less than 5

9% 16% 40%

5% 19% 52%

35%

24%

25% 20% 39% 16%

40% 20% 25% 15%

Experience in the construction industry/academia (years)

Appendix C. Electre 1 MCDA method Electre 1 steps In this appendix, the step-by step procedure of the ELECTRE 1 MCDA method used in this paper is described (adopted from [58]). Step 1. Normalized rating matrix First, the normalized rating matrix is developed by using the values of alternatives with regard to attributes (i.e., the score assigned by the survey participants) as:

r ⋯ r1n ⎤ ⎡ 11 Rij = ⎢ ⋮ ⋱ ⋮ ⎥ r r ⋯ mn ⎦ ⎣ m1 rij =

x ij m

∑i = 1 x ij2

(C.1)

, i = 1,2, …, m andj = 1,2, …, n (C.2)

Where, xij is the value of alternative i with respect to attribute (criterion) j. Since attributes can have different measurement scales, the rating matrix is normalized to enable their values to be comparable. In addition, it should be mentioned that, if an attribute is not a benefit (the more, the better) criterion, e.g., cost criterion, the value of should be reversed in the above equation.

34


M. Kamali et al.

Step 2. Weighted normalized rating matrix The second step is to multiply the entries of the normalized matrix by the weights of corresponding attributes. Thus, the weighted normalized rating matrix is obtained as:

r w ⋯ r1n wn ⎤ ⎡ 11 1 Vij = ⎢ ⋮ ⋱ ⋮ ⎥ ⎣ rm1 w1 ⋯ rmn wn ⎦

(C.3)

Step 3. Concordance and discordance sets As stated above, the concordance and discordance sets are formulated for each pair of alternatives Ap and Aq (p, q = 1, 2, …, m, and p ≠ q). The concordance set includes all attributes for which Ap is preferred to Aq, can be expressed as:

C (p , q) = {j|vpj ≥ vqj}

(C.4)

Where vpj is the weighted normalized rating of alternative Ap with respect to the jth attribute (Eq. (C.3)). In other words, C (p, q) is the collection of attributes where Ap is better than or equal Aq. The discordance set, D (p, q), which is the compliment of concordance set, comprises all attributes for which Ap is worse than Aq and can be stated as:

D (p , q) = {j|vpj < vqj}

(C.5)

Step 4. Concordance and discordance indices The concordance index, Cpq, represents the relative power of each concordance set. In other words, Cpq indicates the degree of confidence in the pairwise judgment of two alternatives (Ap → Aq) and can be computed as:

Cpq =

∑ w j∗ j∗

(C.6)

Where j are attributes included in the concordance set, i.e., j ε C (p, q). In fact, Cpq, is the sum of the weights of all attributes contained in Eq. (C.4). Conversely, the discordance index (Dpq), represents the relative power of each discordance set and measures the degree of disagreement in pairwise judgment, Ap → Aq. Two main equations have been proposed for Dpq in the literature. According to Yoon and Hwang [58], Dpq can be calculated as: *

Dpq =

(∑

*

j∗

vpj∗ − vqj∗

)

⎛⎜ ∑ v − v ⎟⎞ qj j pj ⎝ ⎠

(C.7)

Where j ε (1, 2, …, n) and j are attributes that contained in the discordance set, i.e., j ε D (p, q). Dpq can also be calculated by the following equation [57,81]: *

Dpq =

*

max j∗ vpj∗ − vqj∗ max j vpj − vqj

(C.8)

In this paper, Eq. (C.7) was used to calculate Dpq values. Step 5. Outranking relationships The power of the dominance relationship of alternative Ap over alternative Aq depends on how high is the concordance index (Cpq) and how low is the discordance index (Dpq). The outranking relationships are built by comparing the concordance and discordance indices with specified limits (thresholds) for concordance and discordance. In fact, Ap outranks Aq if:

Cpq ≥ c

(C.9)

and

Dpq < d

(C.10)

Where c and d are the concordance and discordance thresholds, respectively. The more severe the threshold values, the more difficult it is to pass the tests. The values of c and d can be calculated based on the results of Cpq and Dpq, or constant values previously chosen by the decision maker/analyst. For example, Yoon and Hwang [58] defined c and d as the averages of Cpq and Dpq values, respectively, while Collette and Siarry [81] suggested c = 0.7 and d = 0.3. By defining the threshold values c and d , the outranking relationships between alternatives can be established. However, the impact of the selected threshold values upon the ultimate ranking can be significant and this is one of the weaknesses of the ELECTRE 1 method [58]. Therefore, to avoid defining the threshold values, a complementary version of ELECTRE 1 was introduced by van Delft and Nijkamp [82] by defining net concordance and net discordance indices for each alternative. The net concordance and discordance indices provide an effective numerical measure to sort all the alternatives from the best to the worst [57]. The complementary analysis of the ELECTRE 1 method is described below as the last step of the calculations.

35


M. Kamali et al.

Step 6. Net outranking relationships for ranking the alternatives As stated above, through this last step, the overall ranks of alternatives are established using the net outranking relationships; therefore, the selection process of suitable alternatives becomes easier. The net outranking relationships are obtained by calculating the net concordance and the net discordance indices for each alternative. The net concordance index (Cp) estimates the degree to which the dominance of an alternative (e.g., Ap) over all other alternatives exceeds the dominance of other alternatives over the given alternative. Cp can be computed as: thereof m

Cp =

m

∑ Cpk − ∑ Ckp; k=1

k≠p (C.11)

k=1

Similarly, the net discordance (Dp) indicates the relative feebleness of an alternative with regard to the others and can be calculated as: m

Dp =

m

∑ Dpk − ∑ Dkp ; k=1

k≠p (C.12)

k=1

By calculating and checking the values of Cp and Dp for all the alternatives, the net outranking relationships can be developed and thereof final ranking of the alternatives is established. Namely, the higher Cp and lower Dp values control the final ranking order of the alternatives. In other words, the alternative with the maximum Cp and minimum Dp values is the most preferred alternative, and the other alternatives are ranked accordingly. Calculation example for final ranking of the economic SPCs Based on the literature review, interviews, and screening process, 9 SPCs were selected as potential representatives of the economic sustainability. Then, the construction experts evaluated these SPCs against applicability, data availability, and data accuracy. Table C1 shows the resulting rating matrix. As stated in Section 2.1.2 of this main manuscript, the weights of the evaluation criteria were allocated using a group decision making process and AHP method. Accordingly, the normalized weighted rating matrix was developed using Eq. (C.1), Eq. (C.2), and Eq. (C.3) as presented in Table C2. Table C1 The rating matrix for the economic category. SPC

Applicability (CI)

Data availability (CII)

Data accuracy (CIII)

Design and construction time (DCT) Design and construction costs (DCC) Operational costs (OC) Maintenance costs (MC) End of life costs (EC) Durability of building (DB) Investment and related risks (IR) Flexibility of building (FB) Integrated management (IM)

4.37 4.32 3.80 3.73 3.32 3.93 3.85 3.78 3.88

2.41 2.35 1.88 1.82 1.65 1.94 2.18 1.76 1.88

2.38 2.41 1.88 1.94 1.59 1.94 2.19 2.00 2.00

Table C2 The normalized weighted rating matrix for the economic category. SPC

Applicability (CI)

Data availability (CII)

Data accuracy (CIII)

Weights of the evaluation criteria (%) Design and construction time (DCT) Design and construction costs (DCC) Operational costs (OC) Maintenance costs (MC) End of life costs (EC) Durability of building (DB) Investment and related risks (IR) Flexibility of building (FB) Integrated management (IM)

59.65 0.223 0.220 0.194 0.190 0.169 0.200 0.197 0.193 0.198

20.54 0.082 0.080 0.064 0.062 0.056 0.066 0.074 0.060 0.064

19.81 0.076 0.078 0.061 0.063 0.051 0.063 0.070 0.064 0.064

Using the normalized weighted values of the SPCs (Table C2), Eq. (C.4), and Eq. (C.5), the concordance and discordance sets are obtained as shown in Tables C3 and C4.

36

C C C C C C C C

37

(2,3) = 0 (2,4) = 0 (2,5) = 0 (2,6) = 0 (2,7) = 0 (2,8) = 0 (2,9) = 0 (2,1) = {1,2}

D D D D D D D D

D D D D D D D D

(1,3) = 0 (1,4) = 0 (1,5) = 0 (1,6) = 0 (1,7) = 0 (1,8) = 0 (1,9) = 0 (1,2) = {3}

D (2,1) = {1,2}

D (1,2) = {3}

D D D D D D D D

(3,2) = {1,2,3} (3,4) = {3} (3,5) = 0 (3,6) = {1,2,3} (3,7) = {1,2,3} (3,8) = {3} (3,9) = {1,3} (3,1) = {1,2,3}

D (3,1) = {1,2,3}

C C C C C C C C

D D D D D D D D

(4,2) = {1,2,3} (4,3) = {1,2} (4,5) = 0 (4,6) = {1,2} (4,7) = {1,2,3} (4,8) = {1,3} (4,9) = {1,2,3} (4,1) = {1,2,3}

C C C C C C C C

(5,2) = 0 (5,3) = 0 (5,4) = 0 (5,6) = 0 (5,7) = 0 (5,8) = 0 (5,9) = 0 (5,1) = 0

C (5,1) = 0

D D D D D D D D

(5,2) = {1,2,3} (5,3) = {1,2,3} (5,4) = {1,2,3} (5,6) = {1,2,3} (5,7) = {1,2,3} (5,8) = {1,2,3} (5,9) = {1,2,3} (5,1) = {1,2,3}

D (5,1) = {1,2,3}

(4,2) = 0 (4,3) = {3} (4,5) = {1,2,3} (4,6) = {3} (4,7) = 0 (4,8) = {2} (4,9) = 0 (4,1) = 0

C (4,1) = 0

D (4,1) = {1,2,3}

(3,2) = 0 (3,4) = {1,2} (3,5) = {1,2,3} (3,6) = 0 (3,7) = 0 (3,8) = {1,2} (3,9) = {2} (3,1) = 0

C (3,1) = 0

Table C4 The discordance sets for SPCs in the economic category.

(2,3) = {1,2,3} (2,4) = {1,2,3} (2,5) = {1,2,3} (2,6) = {1,2,3} (2,7) = {1,2,3} (2,8) = {1,2,3} (2,9) = {1,2,3} (2,1) = 3

C C C C C C C C

C C C C C C C C

(1,3) = {1,2,3} (1,4) = {1,2,3} (1,5) = {1,2,3} (1,6) = {1,2,3} (1,7) = {1,2,3} (1,8) = {1,2,3} (1,9) = {1,2,3} (1,2) = {1,2}

C (2,1) = 3

C (1,2) = {1,2}

Table C3 The concordance sets for SPCs in the economic category.

(6,2) = 0 (6,3) = {1,2,3} (6,4) = {1,2,3} (6,5) = {1,2,3} (6,7) = {1} (6,8) = {1,2} (6,9) = {1,2} (6,1) = 0

D D D D D D D D

(6,2) = {1,2,3} (6,3) = 0 (6,4) = 0 (6,5) = 0 (6,7) = {2,3} (6,8) = {3} (6,9) = {3} (6,1) = {1,2,3}

D (6,1) = {1,2,3}

C C C C C C C C

C (6,1) = 0 (7,2) = 0 (7,3) = {1,2,3} (7,4) = {1,2,3} (7,5) = {1,2,3} (7,6) = {2,3} (7,8) = {1,2,3} (7,9) = {2,3} (7,1) = 0

D D D D D D D D

(7,2) = {1,2,3} (7,3) = 0 (7,4) = 0 (7,5) = 0 (7,6) = {1} (7,8) = 0 (7,9) = {1} (7,1) = {1,2,3}

D (7,1) = {1,2,3}

C C C C C C C C

C (7,1) = 0

(8,2) = 0 (8,3) = {3} (8,4) = {1,3} (8,5) = {1,2,3} (8,6) = {3} (8,7) = 0 (8,9) = {3} (8,1) = 0

D D D D D D D D

(8,2) = {1,2,3} (8,3) = {1,2} (8,4) = {2} (8,5) = 0 (8,6) = {1,2} (8,7) = {1,2,3} (8,9) = {1,2} (8,1) = {1,2,3}

D (8,1) = {1,2,3}

C C C C C C C C

C (8,1) = 0

(9,2) = 0 (9,3) = {1,2,3} (9,4) = {1,2,3} (9,5) = {1,2,3} (9,6) = {3} (9,7) = {1} (9,8) = {1,2,3} (9,1) = 0

D D D D D D D D

(9,2) = {1,2,3} (9,3) = 0 (9,4) = 0 (9,5) = 0 (9,6) = {1,2} (9,7) = {2,3} (9,8) = 0 (9,1) = {1,2,3}

D (9,1) = {1,2,3}

C C C C C C C C

C (9,1) = 0

M. Kamali et al.



M. Kamali et al.

In the next step, using Eq. (C.6), Eq. (C.7), and the results of concordance and discordance sets, the concordance and discordance indices were developed for all the SPCs and presented in Tables C5 and C6 below. Table C5 The concordance index for SPCs in the economic category. Cpq

1

2

3

4

5

6

7

8

9

1 2 3 4 5 6 7 8 9

– 0.198 0.000 0.000 0.000 0.000 0.000 0.000 0.000

0.802 – 0.000 0.000 0.000 0.000 0.000 0.000 0.000

1.000 1.000 – 0.198 0.000 1.000 1.000 0.198 1.000

1.000 1.000 0.802 – 0.000 1.000 1.000 0.795 1.000

1.000 1.000 1.000 1.000 – 1.000 1.000 1.000 1.000

1.000 1.000 0.000 0.198 0.000 – 0.404 0.198 0.198

1.000 1.000 0.000 0.000 0.000 0.597 – 0.000 0.597

1.000 1.000 0.802 0.205 0.000 0.802 1.000 – 1.000

1.000 1.000 0.205 0.000 0.000 0.802 0.404 0.198 –

Table C6 The discordance index for SPCs in the economic category. Dpq

1

2

3

4

5

6

7

8

9

1 2 3 4 5 6 7 8 9

– 0.792 1.000 1.000 1.000 1.000 1.000 1.000 1.000

0.208 – 1.000 1.000 1.000 1.000 1.000 1.000 1.000

0.000 0.000 – 0.752 1.000 0.000 0.000 0.582 0.000

0.000 0.000 0.248 – 1.000 0.000 0.000 0.315 0.000

0.000 0.000 0.000 0.000 – 0.000 0.000 0.000 0.000

0.000 0.000 1.000 1.000 1.000 – 0.189 0.877 0.704

0.000 0.000 1.000 1.000 1.000 0.811 – 1.000 0.928

0.000 0.000 0.418 0.685 1.000 0.123 0.000 – 0.000

0.000 0.000 1.000 1.000 1.000 0.296 0.072 1.000 –

Using the concordance and discordance indices, Eq. (C.11), and Eq. (C.12), the net concordance index (Cp) and net discordance index (Dp) for all the SPCs were calculated; consequently, the net outranking relationships were developed. The net concordance and discordance indices and the net outranking of the economic sustainability performance criteria have presented in the main text (Table 3) and not repeated here. Appendix D. TOPSIS MCDA method In this appendix, the step-by step procedure of the TOPSIS MCDA method used in this paper is described (adopted from [58]). Step 1. Weights of SPIs The relative importance weights of the sustainability performance indicators (SPIs) under each sustainability performance criterion (SPC) should be determined. Step 2. Normalized SPIs Once the SPIs are calculated (performance score or xij), if the xij values are not already normalized, they need to be normalized (rij). The vector normalization can be used for normalization:

rij =

x ij m ∑i = 1

x ij2

i = 1, …, mj = 1, …, n (D.1)

Step 3. Weighted normalized matrix The weighted normalized matrix should be developed. The weighted normalized performance score of each SPI is calculated as:

vij = wj rij

(D.2)

where wj = corresponding weight of that SPI. Step 4. Positive-Ideal and Negative-Ideal Solutions In this step, the positive-ideal solution (PIS) and negative-ideal solution (NIS) are identified. X+ and X− are defined as the PIS and NIS, 38


M. Kamali et al.

respectively, in terms of weighted performance scores as follows:

X+ = {v1+, v2+, …, v+j , …, vn+} = {(max i vij jεJ1), (mini vij jεJ2), i = 1,2, …, m} X−

=

{v1−,

v2−, …,

v−j , …, vn−}

(D.3)

= {(mini vij jεJ1), (max i vij jεJ2), i = 1,2, …, m}

(D.4)

where J1= set of benefit attributes; and J2= set of cost attributes. Step 5. Separation measures In this step, the distance of the subject building from PIS and NIS values (i.e., separation measures) is calculated. The distances of all the performance levels of SPIs associated with each SPC are measured using the n-dimensional Euclidean distance. The separation measure of each SPC from the PIS can be calculated as: n

Si+ =

∑ (vij − v+j )2,

i = 1,2, …, m (D.5)

j=1

and separation measure of each SPC from the NIS can be calculated as n

Si− =

∑ (vij − v−j )2,

i = 1,2, …, m (D.6)

j=1

Step 6. Aggregated indices The aggregated sustainability index for each SPC (e.g., PEi, CWMi) is developed by calculating similarities to PIS using the following equation:

SPCi =

Si− Si− + Si+

(D.7)

Appendix E. The proposed method to determine the final ranking of SPCs In the cases of some SPCs such as EP, the ranking of net concordance and net discordance are consistent (identical); therefore, finding the overall rank order of these SPCs is not difficult. However, some discrepancies were recognized in the cases of other SPCs as their net concordance ranks were different from their net discordance ranks (e.g., WM and MC); consequently, this is a challenge to determine the final ranking of these criteria. Following step 6 of ELECTRE method (Appendix C), this issue can be addressed by plotting the SPCs using their net concordance vs. net discordance values, as shown in Fig. E1(a). By projecting the SPCs on the -45° line, and eventually, calculating the distance of the projected points from the origin, final ranking of alternatives can be obtained. For those SPCs located in the fourth quadrant, higher distances means better ranks. Contrary, in the cases of SPCs located in the second quadrant, lower distances indicate better ranks. Hence, by calculating and comparing these distances, all the SPCs alternatives were ranked as reflected in the last column of Table 2 in the main text.

Fig. E1. Net Concordance (Cp) and net discordance (Dp) indices for the sustainability performance criteria; (a) Environmental category; (b) Economic category; (c) Social category.

Similarly, based on the values of Cp and Dp indices, the economic SPCs were ranked (Table 3 in the main text and Fig. E1(b)). Despite the environmental SPCs, almost all the economic SPCs were consistent in terms of their Cp and Dp rankings (except DB and IM). In the case of the social category, for majority of the social SPCs, the net concordance and net discordance rankings were identical. The only inconsistencies were noted in AB and ILE. The final ranking of these two SPCs were determined by using Fig. E1(c) and comparing the distance of each SPC from the origin.

offsite modern methods of construction, Constr. Manage. Econ 25 (2007) 183–194. [4] P.J. Cameron, N.G. Di Carlo, Piecing Together Modular: Understanding the Benefits and Limitations of Modular Construction Methods for Multifamily Development, Master Thesis Massachusetts Institute of Technology, Cambridge, 2007. [5] Y.H. Ahn, K. Kim, Sustainability in modular design and construction: a case study of 'the stack, Int. J. Sustainable Build. Technol. Urban Dev. 5 (4) (2014) 250–259. [6] MBI, Permanent Modular Construction, Modular Building Institute, Charlottesville, VA, 2012. [7] W. Nadim, J.S. Goulding, Offsite production in the UK: the way forward? A UK construction industry perspective, Constr. Innovation 10 (2) (2010) 181–202.

References [1] L.R. Kawecki, Environmental Performance of Modular Fabrication: Calculating the Carbon Footprint of Energy Used in the Construction of a Modular Home, Ph.D. Thesis Arizona State University, Phoenix, 2010. [2] M. O'Brien, R. Wakefield, Y. Belivean, Industrializing the Residential Construction Site, Department of Housing and Urban Development, Office of Policy Development and Research, Washington, DC, 2000. [3] W. Pan, A.G.F. Gibb, A.R.J. Dainty, Perspectives of UK housebuilders on the use of

39


M. Kamali et al.

habitability, J. Asian Archit. Build. Eng 15 (1) (2016) 49–56. [43] M. Kamali, K. Hewage, Performance indicators for sustainability assessment of buildings, 5th International/11th Construction Specialty Conference (ICSC), Vancouver, BC, 2015. [44] H. Haider, R. Sadiq, S. Tesfamariam, Performance indicators for small and medium sized water supply systems: a review, Environ. Rev. 22 (1) (2014) 1–40. [45] ADB, Handbook for Selecting Performance Indicators for ADB-funded Projects in the PRC, Asian Development Bank, Mandaluyong, Philippines, 2012. [46] G.A. Giff, J. Crompvoets, Performance indicators a tool to support spatial data infrastructure assessment, computers, Environ. Urban Syst 32 (2008) 365–376. [47] M. Lundin, Indicators for Measuring the Sustainability of Urban Water: a Life Cycle Approach, Chalmers University of Technology, Gothenburg, 2002. [48] M.C. Lee, The analytic hierarchy and the network process in multicriteria decision making: performance evaluation and selecting key performance indicators based on ANP model, in: M. Crisan (Ed.), Convergence and Hybrid Information Technologies, ISBN 978-953-307-068-1, InTech, Croatia, 2010, p. 426 http://www.sciyo.com. [49] J. Heilala, M. Myllysilta, S. Vatanen, E. Järvenpää, K. Syrjälä, The Concept for Sustainability Performance Indicators, Reporting and Improvement, (2015) Final report from FIMECC MANU. [50] M. Kamali, K. Hewage, Development of performance criteria for sustainability evaluation of modular versus conventional construction methods, J. Clean. Prod. 142 (4) (2017) 3592–3606. [51] T.L. Saaty, The Analytic Hierarchy Process, McGraw-Hill, New York, 1980. [52] E. Forman, K. Peniwati, Aggregating individual judgements and priorities with the analytic hierarchy process, Eur. J. Oper. Res. 108 (1998) 165–169. [53] R. Ramanathan, L.S. Ganesh, Group preference aggregation methods employed in AHP: an evaluation and intrinsic process for deriving members' weightages, Eur. J. Oper. Res. 79 (1994) 249–265. [54] M.T. Escobar, J.M. Moreno-jiménez, Aggregation of individual preference structures in AHP-group decision making, Group Decis. Negot. 16 (4) (2007) 287–301. [55] H. Alegre, E. Cabrera, W. Merkel, Performance assessment of urban utilities: the case of water supply, wastewater and solid waste, J. Water Supply Res. T. (AQUA) 58 (5) (2009) 305–315. [56] R. Benayoun, B. Roy, N. Sussman, Manual de reference du programme ELECTRE, Note de Synthese et. Formation n.25. Paris, Direction scientifique SEMA, 1966. [57] A.S. Milani, A. Shanian, C. El-Lahham, Using different ELECTRE methods in strategic planning in the presence of human behavioral resistance, J. Appl. Math. Decis. Sci. 2006 (2006) 1–19, http://dx.doi.org/10.1155/JAMDS/2006/10936. [58] K.P. Yoon, C.L. Hwang, Multiple Attribute Decision Making: an Introduction, Sage Publications, Thousand Oaks, 1995. [59] G. Verbeeck, H. Hens, Life cycle inventory of buildings: a contribution analysis, Build. Environ. 45 (2010) 964–967. [60] R.J. Cole, P.C. Kernan, Life-cycle energy use in office buildings, Build. Environ. 31 (4) (1996) 307–317. [61] I. Sartori, A.G. Hestnes, Energy use in the life cycle of conventional and low-energy buildings: a review article, Energy Build. 39 (3) (2007) 249–257. [62] C. Scheuer, G.A. Keoleian, P. Reppe, Life cycle energy and environmental performance of a new university building: modeling challenges and design implications, Energy Build. 35 (10) (2003) 1049–1064. [63] O. Ortiz, F. Castells, G. Sonnemann, Sustainability in the construction industry: a review of recent developments based on LCA, Constr. Build. Mater 23 (1) (2009) 28–39. [64] G.A. Keoleian, S. Blanchard, P. Reppe, Life-cycle energy, costs, and strategies for improving a single-family house, J. Ind. Ecol. 4 (2) (2000) 135–156. [65] S. Copiello, Building energy efficiency: a research branch made of paradoxes, Renewable Sustainable Energy Rev. 69 (2017) 1064–1076. [66] L.F. Cabeza, C. Barreneche, L. Miró, J.M. Morera, E. Bartolí, A.I. Fernández, Low carbon and low embodied energy materials in buildings: a review, Renewable Sustainable Energy Rev. 23 (2013) 536–542. [67] M. Optis, P. Wild, Inadequate documentation in published life cycle energy reports on buildings, Int. J. Life Cycle Assess. 15 (7) (2010) 644–651. [68] M. Zenga, A. Javor, Modular Homes: The Future has Arrived, Fideli publishing, Paragon, 2008. [69] J.P. Cartz, M. Crosby, Building high-rise modular homes, Struct. Eng. 85 (19) (2007) 20–21. [70] X.X. Li, G.L. Li, Exploration of modular build of architectural space, Appl. Mech. Mater. 357 (2013) 338–344. [71] C.T. Haas, J.T. O'Connor, R.L. Tucker, J.A. Eickmann, W.R. Fagerlund, Prefabrication and Preassembly Trends and Effects on the Construction Workforce. Center for Construction Industry Studies, University of Texas, Austin, TX, 2000. [72] MBI, Improving Construction Efficiency and Productivity with Modular Construction, Modular Building Institute, Charlottesville, VA, 2012. [73] A.L. Rogan, R.M. Lawson, N. Bates-Brkljac, Value and Benefits Assessment of Modular Construction, Steel Construction Institute, Berkshire, UK, 2000. [74] J.L. Celine, The Evolution of the Use of Prefabrication Techniques in Hong Kong Construction Industry, Ph.D. Thesis Hong Kong Polytechnic University, Hung Hom, 2009. [75] J.T. Cartwright, Zoning and Designing for Affordability Using Modular Housing, Master Thesis Iowa State University, Ames, 2011. [76] C.T. Haas, W.R. Fagerlund, Preliminary Research on Prefabrication, Pre-assembly, Modularization and Off-site Fabrication in Construction, The Construction Industry Institute, University of Texas at Austin, Austin, TX, 2002. [77] R.M. Lawson, R.G. Ogden, R. Bergin, Application of modular construction in high-

[8] C. Loss, M. Piazza, R. Zandonini, Connections for steel–timber hybrid prefabricated buildings. Part II: innovative modular structures, Constr. Build. Mater 122 (2016) 796–808. [9] E.M. Generalova, V.P. Generalov, A.A. Kuznetsova, Modular buildings in modern construction, Procedia Eng 153 (2016) 167–172. [10] J. Molavi, D.L. Barral, A Construction procurement method to achieve sustainability in modular construction, Procedia Eng 145 (2016) 1362–1369. [11] N. Boyd, M.M. Khalfan, T. Maqsood, Off-site construction of apartment buildings, J. Archit. Eng 19 (1) (2013) 51–57. [12] S. Martinez, A. Jardon, Building industrialization: robotized assembly of modular products, Assem. Autom 28 (2) (2008) 134–142. [13] S. Velamati, Feasibility, Benefits and Challenges of Modular Construction in High Rise Development in the United States: a Developers' Perspective, Master thesis Massachusetts Institute of Technology, Cambridge, 2012. [14] M.M. Rahman, Barriers of implementing modern methods of construction, J. Manage. Eng 30 (1) (2013) 69–77. [15] J. Quale, M.J. Eckelman, K.W. Williams, G. Sloditskie, J.B. Zimmerman, Construction matters: comparing environmental impacts of building modular and conventional homes in the United States, J. Ind. Ecol. 16 (2) (2012) 243–253. [16] USCB, Characteristics of New Housing, Report US Census Bureau (USCB), Washington, DC, USA, 2016 www.census.gov/construction/chars/completed.html , Accessed date: 3 December 2016. [17] D.A. Steinhardt, K. Manley, Adoption of prefabricated housing–the role of country context, Sustain. Cities Society 22 (2016) 126–135. [18] D. O'Neill, S. Organ, A literature review of the evolution of British prefabricated low-rise housing, Struct. Surv. 34 (2) (2016) 191–214. [19] C. Mao, Q. Shen, W. Pan, K. Ye, Major barriers to off-site construction: the developer's perspective in China, J. Manage. Eng 31 (3) (2015) 4014043. [20] L. Na, Investigation of the Designers' and General Contractors' Perceptions of Offsite Construction Techniques in the United States Construction Industry, Ph.D. Thesis Clemson University, Clemson, 2007. [21] C.L. Pasquire, A.G. Gibb, Considerations for assessing the benefits of standardisation and pre-assembly in construction, J. Financ. Manage. Prop. Constr 7 (3) (2002) 151–161. [22] L.D.D. Harvey, A Handbook on Low-energy Buildings and District-energy Systems: Fundamentals, Routledge Publishing, New York, NY, USA, 2006 Techniques and Examples. [23] C.J. Kibert, Sustainable Construction: Green Building Design and Delivery, John Wiley & Sons, Inc., Hoboken, NJ, USA, 2016. [24] U. Berardi, Beyond sustainability assessment systems: upgrading topics by upscaling the assessment, Int. J Sustain. Build. Technol. Urban Dev. 2 (4) (2011) 276–282. [25] R. Hueting, L. Reijnders, Broad sustainability contra sustainability: the proper construction of sustainability indicators, Ecolo. Econ 50 (3–4) (2004) 249–260. [26] WCED, Our Common Future, World Commission on Environment and Development, Oxford University Press, New York, NY, USA, 1987. [27] U. Berardi, Moving to sustainable buildings: paths to adopt Green innovations in developed countries, Walter de Gruyter, 2013. [28] H. Son, C. Kim, W.K. Chong, J.S. Chou, Implementing sustainable development in the construction industry: constructors' perspectives in the US and Korea, Sustain. Dev. 19 (5) (2011) 337–347. [29] R. Chiu, Social equity in housing in the Hong Kong Special administrative region: a social sustainability perspective, Sustain. Dev. 10 (3) (2002) 155–162. [30] A. Parr, M. Zaretsky, New Directions in Sustainable Design, Taylor & Francis, London, UK, 2010. [31] N. Dempsey, G. Bramley, S. Power, C. Brown, The social dimension of sustainable development: defining urban social sustainability, Sustain. Dev. 19 (5) (2011) 289–300. [32] L. Braganca, R. Mateus, H. Koukkari, Building sustainability assessment, Sustain. Times 2 (7) (2010) 2010–2023. [33] K.M. Fowler, E.M. Rauch, Sustainable Building Rating Systems Summary, Pacific Northwest National Laboratory (PNNL), Richland, WA, 2006 Report No.: PNNL15858. [34] R. Reed, A. Bilos, S. Wilkinson, K. Schulte, International comparison of sustainable rating tools, J. Sustain. Real Estate 1 (1) (2009) 1–22. [35] M. Erlandsson, M. Borg, Generic LCA-methodology for buildings, constructions and operation services – today practice and development needs, Build. Environ. 38 (2003) 919–938. [36] M. Kamali, K. Hewage, Life cycle performance of modular buildings: a critical review, Renewable Sustainable Energy Rev. 62 (2016) 1171–1183. [37] K.W. Robert, T.M. Parris, A.A. Leiserowitz, What is sustainable development? Goals, indicators, values and practice. Environ 47 (3) (2005) 8–21. [38] Y. Chen, G.E. Okudan, D.R. Riley, Sustainable performance criteria for construction method selection in concrete buildings, Autom. Constr 19 (2) (2010) 235–244. [39] H. Alwaer, D.J. Clements-Croome, Key performance indicators (KPIs) and priority setting in using the multi-attribute approach for assessing sustainable intelligent buildings, Build. Environ. 45 (4) (2010) 799–807. [40] A. Mwasha, R.G. Williams, J. Iwaro, Modeling the performance of residential building envelope: the role of sustainable energy performance indicators, Energy Build. 43 (9) (2011) 2108–2117. [41] W. Pan, A.R.J. Dainty, A.G.F. Gibb, Establishing and weighting decision criteria for building system selection in housing construction, J. Constr. Eng. Manage 138 (11) (2012) 1239–1250. [42] M.K. Kim, M.J. Kim, Affordable modular housing for college students emphasizing

40


M. Kamali et al.

framework for modular buildings, 6th International Construction Specialty Conference (CSCE/CRC 2017), Vancouver, Canada, 2017. [81] Y. Collette, P. Siarry, Multiobjective Optimization: Principles and Cases Studies, Decision Engineering, Springer, New York, 2003. [82] A. Van Delft, P. Nijkamp, A multi-objective decision model for regional development, environmental quality control and industrial land use, Pap. Reg. Sci. 36 (1) (1976) 35–57.

rise buildings, J. Archit. Eng 18 (2) (2012) 148–154. [78] McGraw-Hill Construction, Prefabrication and Modularization: Increasing Productivity in the Construction Industry, McGraw-Hill Construction, New Orleans, LA, 2011. [79] W.G. Cochran, Sampling Techniques, third ed., Wiley & Sons, New York, NY, USA, 1977. [80] M. Kamali, K. Hewage, Sustainability performance assessment: a life cycle based

41