An Index of Completeness (IoC) of life cycle assessment: Implementation in the building sector

Highlights

• An Index of Completeness (IoC) is proposed to describe to what extent the processes and data are included.

• An IoC calculation tool is developed for the building sector.

• The case study of a residential building project is conducted.

• The IoC of human toxicity is as low as 85%, caused by the exclusion of plywood formwork in the original model.

Abstract

Over the past three decades, life cycle assessment (LCA) has been increasingly employed to evaluate the environmental impacts of products and buildings. However, the reliability of LCA is frequently weakened by the practice of arbitrary omission of processes and materials and the problem of insufficient data, both leading to inconsistencies in the LCA results. This study proposes an Index of Completeness (IoC) to account for the completeness of an LCA study for a variety of impact categories. An IoC calculation tool is developed to cover the entire life cycle of buildings, including stages of product, construction, use and end-of-life as well as the benefits from recycling and reuse. The IoC helps to identify and remove the inconsistencies in LCA and to enhance its comparability. Residential building projects are tested as case studies. It is found that the IoCs of most impact categories are as high as 98%–99%, whereas of the category of human toxicity the IoC is 85%, mainly due to the fact that plywood formwork was not included in the original model. Sensitivity analysis reveals that dataset changes in an LCA model can not only affect the impact assessment results, but also possibly have influence on the completeness of the LCA model. The LCA practitioners are suggested to combine the newly developed IoCs with the standard four phases of LCA when applying to complex products. Future research should focus on case studies implementing the IoC calculation tool, inclusion of more impact categories in the IoC model, and continuing development to include other complex products.

Introduction

Life cycle assessment (LCA) is a technique to assess the environmental impacts of a product throughout its entire life cycle stages. The practice of LCA can be traced back to 1960s (Hunt et al., 1992), and the first peer reviewed LCA papers and early international standards appeared in 1990s (McManus and Taylor, 2015). With a rapid development in the last three decades, LCA has been adopted to the evaluation of environmental impacts for wide ranges of products and organizations (Hellweg and Llorenç, 2014). LCA is indispensable to carbon labeling (Weidema et al., 2008), environmental product declaration (Schau and Annik, 2008), eco-design (Ahmadi and Ligia, 2015), environmental policy (Earles et al., 2013), etc. Besides the environmental aspect, LCA has recently been extended to the economic and social aspects, thus embracing all the three pillars of the concept of sustainability (Dong and Ng, 2016). Nowadays, LCA becomes a valuable tool of decision making in urban (Mirabella et al., 2019) and national scales (Kasumu et al., 2018).

Although being popular, LCA has long been criticized for its reliability (Igos et al., 2019). LCA itself is a complex collection of methods and data that are related to a huge number of research fields. According to ISO 14040 (ISO, 2006a), LCA consists of four phases: (i) goal and scope definition, (ii) inventory analysis, (iii) impact assessment, and (iv) interpretation. For a relatively complex product, the inconsistent definitions of system boundary can cause significant differences in subsequent LCA results (Säynäjoki et al., 2017). According to Tillman et al. (1994), an ideal system boundary is between the technological system and nature, and the inputs and outputs at the boundary are elementary flows. It is necessary to let all the significant processes be included in the study scope to make LCA results reliable (Weidma, 2000). Questions as which processes should be included within a system boundary or what activities are discarded as being insignificant have long been under debate among the LCA community (Tillman, 2000). To decide which processes are included within a system boundary in a practical model, the so-called cut-off criteria are recommended (ISO, 2006b), allowing the omission of other unit processes or life cycle stages with less significant environmental impacts. In practice, however, decisions on whether to include or exclude a process typically are not made on a scientific basis (Suh et al., 2004). Astrup et al. (2015) reviewed 136 journal articles of waste-to-energy technology and concluded that very few of them provided sufficient descriptions of goal and scope of the assessment. Gargiulo et al. (2017) examined 16 LCA studies on electricity networks and only 3 of them specified cut-off criteria. Finnegan et al. (2018) conducted a review on LCA studies of cheese production, in which it was found that without a sound rationale the downstream of the processing factory gate was often omitted. In general, the common practice to omit processes or data is mainly caused by the lack of information, the difficulty of prediction, and the assumed insignificant impact of these life cycle processes (Oregi et al., 2017).

The lack of consistency and comprehensiveness observed among the LCA studies makes it necessary to estimate the contributions of the omitted processes to the overall results. It is known to the LCA community that the truncation error due to the omission of processes can introduce significant biases in the results (Lenzen, 2000). However, there were limited studies that made effort in solving the reliability problem. Raynolds et al. (2000) noticed the weakness in terms of the inconsistent system boundaries for comparative LCAs and attempted to develop a method to select processes. Due to the lack of case studies and insufficient knowledge of emissions at the time, the developed method was based on an analysis of only about ten processes - hence could not provide a convincing evaluation for LCA practice. Suh et al. (2004) discussed the system boundary selection by comparing input-output (IO) and process-based LCAs and concluded that IO method can result in larger impact results than process-based method (since process-based method usually arbitrarily omits processes without sufficient data).

Buildings, essential to human lives, are more complex than ordinary products and have multiple environmental impacts of significance. The building sector is one of the major contributors to carbon emissions and energy consumption (Chau et al., 2012). It has been reported that the building sector is responsible for 40% of energy consumption and 60% of material consumption worldwide (Horvath, 2004) and accounts for up to 50% of the global carbon emissions (Ramesh et al., 2010). The concrete production rate is about one tonne per capita per year (Khasreen et al., 2009), while concrete is a carbon-intensive material that contributes to 5–7% of the world carbon emissions (Flower and Sanjayan, 2007). The building sector is also credited with a major share of several environmental impacts during the long life-span (Meyer, 2009).

Previous studies of LCA on buildings focused more on its practical usage in evaluating the environmental impacts, while the consistency and reliability of the LCA analysis are overlooked. The environmental impacts of different life cycle stages of buildings were extensively studies, such as construction materials (Ibbotson and Kara, 2013; Dong et al., 2015), building construction (Hong et al., 2014; Dong and Ng, 2015), operation (Wallhagen et al., 2011), refurbishment (Ghose et al., 2017) and demolition (Butera et al., 2015), as well as the “cradle-to-grave” impacts to cover the entire life cycle of buildings (Marique and Barbara, 2018). Regarding the complexity of buildings, the inaccurate results have already been pointed out in the building LCA research. Pan et al. (2018) proposed an inconsistency ratio of system boundaries to demonstrate the inconsistency in case studies for buildings, whereas the detailed building life cycle processes were not investigated in their study, leaving inconsistency problem (caused by incomplete LCA scopes) untainted. Zhang et al. (2019) reported the simplified system boundary that includes only major materials and energy use can cause potential underestimation of emissions up to 10%. It is noteworthy that, in the life cycle of buildings, although a process or a material is insignificant to one environmental impact category, it can be influential to another (Lasvaux et al., 2015). The likely existence of such pitfalls in the LCA study of a very complex system is understandable and consequently extra attention must be paid to the completeness of LCA system coverage in buildings.

For complex products, buildings in particular, a systematic evaluation of the comprehensiveness and consistency of the LCA studies is of great importance to guarantee reliable results. The environmental impact categories should be selected to let the system scope comprehensive for all the life cycle stages of product. The system boundary needs to be delimited with care to include significant processes and materials, while the omitted processes and materials should be treated with scrutiny and explicit justification. Standardizing the requirements on boundary scoping enhances the usefulness of LCA (Chau et al., 2015), thus making LCA results more comparable. Accordingly, a mathematical description is to be developed to represent the extent to which the life cycle stages are covered in an LCA study. In this regard, the method to be devised in this paper provides an evaluation of the completeness of LCA studies. It is designed with the primary aim to standardize the determination of system boundary for the complex products of building sector. The method is intended to be applicable to other complex products in a general manner.

The aim of this research is to propose an Index of Completeness (IoC) and develop a calculation tool for IoC to be applicable in the building sector. Using the IoC calculation tool, the LCA practitioners may know whether their LCA studies are complete for certain impact categories. The development of IoC consists of four steps. In Step 1, the concept of IoC is proposed and the formula of IoCs are presented. Step 2 is to develop an IoC calculation tool in the Microsoft Excel format. In this step, the impact categories are selected and life cycle impact assessment (LCIA) method is determined. The IoC calculation tool is developed based on the IoC formula and the data collected from literature and databases. Step 3 is to conduct a case study of a residential building project. The emissions and IoCs of this study case are calculated and results from different scenarios are analyzed. Comparative study on different building cases is performed. In Step 4 suggestions on the implementation of IoC calculation tool are offered. The procedures to calculate IoCs based on the standard four phases of LCA are provided in Fig. 1.