Quantifying system uncertainty of life cycle assessment based on Monte Carlo simulation

Background, aim, and scope

Many studies evaluate the results of applying different life cycle impact assessment (LCIA) methods to the same life cycle inventory (LCI) data and demonstrate that the assessment results would be different with different LICA methods used. Although the importance of uncertainty is recognized, most studies focus on individual stages of LCA, such as LCI and normalization and weighting stages of LCIA. However, an important question has not been answered in previous studies: Which part of the LCA processes will lead to the primary uncertainty? The understanding of the uncertainty contributions of each of the LCA components will facilitate the improvement of the credibility of LCA.

Methodology

A methodology is proposed to systematically analyze the uncertainties involved in the entire procedure of LCA. The Monte Carlo simulation is used to analyze the uncertainties associated with LCI, LCIA, and the normalization and weighting processes. Five LCIA methods are considered in this study, i.e., Eco-indicator 99, EDIP, EPS, IMPACT 2002+, and LIME. The uncertainty of the environmental performance for individual impact categories (e.g., global warming, ecotoxicity, acidification, eutrophication, photochemical smog, human health) is also calculated and compared. The LCA of municipal solid waste management strategies in Taiwan is used as a case study to illustrate the proposed methodology.

Results

The primary uncertainty source in the case study is the LCI stage under a given LCIA method. In comparison with various LCIA methods, EDIP has the highest uncertainty and Eco-indicator 99 the lowest uncertainty. Setting aside the uncertainty caused by LCI, the weighting step has higher uncertainty than the normalization step when Eco-indicator 99 is used. Comparing the uncertainty of various impact categories, the lowest is global warming, followed by eutrophication. Ecotoxicity, human health, and photochemical smog have higher uncertainty.

Discussion

In this case study of municipal waste management, it is confirmed that different LCIA methods would generate different assessment results. In other words, selection of LCIA methods is an important source of uncertainty. In this study, the impacts of human health, ecotoxicity, and photochemical smog can vary a lot when the uncertainties of LCI and LCIA procedures are considered. For the purpose of reducing the errors of impact estimation because of geographic differences, it is important to determine whether and which modifications of assessment of impact categories based on local conditions are necessary.

Conclusions

This study develops a methodology of systematically evaluating the uncertainties involved in the entire LCA procedure to identify the contributions of different assessment stages to the overall uncertainty. Which modifications of the assessment of impact categories are needed can be determined based on the comparison of uncertainty of impact categories.

Recommendations and perspectives

Such an assessment of the system uncertainty of LCA will facilitate the improvement of LCA. If the main source of uncertainty is the LCI stage, the researchers should focus on the data quality of the LCI data. If the primary source of uncertainty is the LCIA stage, direct application of LCIA to non-LCIA software developing nations should be avoided.     

Bridge life cycle assessment with data uncertainty

Purpose

Although life cycle assessment (LCA) has been employed to analyze the environmental impacts of bridges, the uncertainties associated to LCA have not been studied, which have a profound effect on the LCA results. This paper is intended to provide a comprehensive environmental impact assessment of bridge with data uncertainty, by assigning probability distributions on the considered parameters, assessing the variability in the acquisition of inventory and identifying the key parameters with significant environmental impacts.

Methods

A life cycle assessment of a bridge in Shanxi Province of China was conducted in a cradle-to-grave manner, by considering the source of the uncertainty of LCA. A statistical method was applied to quantify the uncertainty of measured inventory data and to calculate the probability distribution of the data. The uncertainty propagation was conducted through using a Monte Carlo simulation. Finally, the factor which is of vital importance to the assessment result was identified by sensitivity analysis.

Results and discussion

For the case of bridge, normal distribution can be adopted to fit environmental substances and environmental impact in steel production. The distributions of the weighted value of human health damage, ecological system damage, and resource and energy consumption can be represented by an approximate similar normal distribution function. The coefficient of variance (COV) of each weighted value is about 40, 30 to 40, and about 6 %, respectively. The COV for the total environmental impact is about 10 % in all stages of the bridge’s life cycle, except the operation stage, which is up to 22.67 %. By conducting sensitivity analysis, PM₁₀, NO_x, and oil consumption was found to have a great influence on the result of human health damage, ecological system damage, and resource and energy consumption, respectively.

Conclusions

The COV for the total environmental impact is 22.67 % in the bridge’s operation stage; it is important to establish a reasonable maintenance strategy to decrease the uncertainty of the bridge’s environmental impact. The COVs of the weighted value for human health damage and resource and energy consumption have a quite modest difference among the four stages of the bridge’s life cycle. However, The COV of the weighted value for ecological system damage shows large difference among the four stages of the bridge’s life cycle; construction stage has the greatest uncertainty. In addition, different values of PM₁₀, NO_x, and oil consumption have a profound influence on the result of human health damage, ecological system damage, and resource and energy consumption, respectively.

     

Part II: Dealing with parameter uncertainty and uncertainty due to choices in life cycle assessment

Results of product assessments are often criticised as to their handling of uncertainty. Therefore, it is necessary to develop a comprehensive methodology reflecting parameter uncertainty in combination with uncertainty due to choices in the outcome of LCAs. This paper operationalises the effect of combined parameter uncertainties in the inventory and in the characterisation factors for global warming and acidification for the comparison of two exemplary types of roof gutters. For this purpose, Latin Hypercube sampling is used in the matrix (inventory) method. To illustrate the influence of choices, the effect on LCA outcomes is shown of two different allocation procedures in open-loop recycling and three time horizons for global warming potentials. Furthermore, an uncertainty importance analysis is performed to show which parameter uncertainties mainly contribute to uncertainties in the comparison and the separate environmental profiles of the product systems. These results can be used to prioritise further data research.     

Analyzing uncertainty in a comparative life cycle assessment of hand drying systems

Purpose

The goal of this study is to evaluate and compare the environmental impact (with a focus on global warming potential) of five hand drying systems: hands-under (HU) dryers, high-speed hands-under (HSHU) dryers, high-speed hands-in (HSHI) dryers, cotton roll towels, and paper towels. Another objective is to incorporate uncertainty into this comparative life cycle assessment (LCA) as a means of understanding the statistical robustness of the difference between the environmental impacts of the hand drying systems.

Methods

We conducted a life cycle assessment in accordance with the ISO 14040/14044 standards using data primarily from publicly available reports. As part of the study, we performed a parameter uncertainty analysis for multiple scenarios to evaluate the impact of uncertainty in input data on the relative performance of products. In addition, we conducted a probabilistic scenario analysis of key drying system parameters in order to understand the implications of changing assumptions on the outcomes of the analyses.

Results and discussion

The scope of the analyses enabled us to draw robust conclusions about the relative environmental performance of the products. We can say with a high degree of confidence that the high-speed dryers have a lower impact than paper towels and cotton roll towels. Differentiating the performance of the hand dryers requires being more specific about framing assumptions. Under certain conditions, the HSHI dryer is expected to have a lower impact than the HU and HSHU dryers. However, under other conditions, one cannot say that the HSHI dryer is clearly better than the other dryers. We cannot differentiate the performance between the HU dryer, cotton roll towels, and paper towels.

Conclusions

This work demonstrates the importance of going beyond traditional uncertainty analyses for comparative LCAs that are used for assertions of relative product environmental impact. Indeed, we found instances where the conclusions changed as a result of using the probabilistic scenario analysis. We outline important elements that should be included in future guidance on uncertainty analyses in comparative LCAs, including conducting parameter and scenario uncertainty analyses together and then using the outcomes to guide selection of parameters and/or choices to analyze further.     

A model for streamlining and automating path exchange hybrid life cycle assessment

The International Journal of Life Cycle Assessment - Life cycle assessment (LCA) is inherently complex and time consuming. The compilation of life cycle inventories (LCI) using a traditional...     

Environmental assessment of brownfield rehabilitation using two different life cycle inventory models

Goal, Scope and Background  

The principal aim of this paper is to evaluate the environmental attributes and consequences of a ‘rehabilitation for residential redevelopment’ scenario. It is contrasted to a non-intensive and low-cost ‘exposure minimization’ scenario, assumed to be the default intervention option to obtain compliance. This paper also aims to (1) quantitatively evaluate the relative environmental significance of primary, secondary and tertiary impacts, and (2) to compare conclusions obtained from attributional and from consequential LCA of the same decision.     

Environmental assessment of brownfield rehabilitation using two different life cycle inventory models

Goal, Scope and Background The primary goal of this paper is to present a LCI modelling approach that allows the inclusion of all three types of impacts. The approach is based on consequential LCA (CLCA) rather than more common attributional LCA (ALCA). In CLCA, system boundaries are expanded in order to include all significantly affected activities. In addition we show how changing from an attributional to a consequential approach alters how the impacts are evaluated, and discuss the applicability of these two distinct approaches to brownfield rehabilitation decision support. The paper is restricted to urban and contaminated brownfields that are the result of industrial use and whose rehabilitation is aimed at allowing residential redevelopment. Main Features The approach is based on an analogy between the open-loop recycling of material resources and brownfield rehabilitation. Brownfield rehabilitation is associated with two functions: (1) managing the legacy of past occupations on the site, analogous to a waste management function, and (2) providing redevelopable land, analogous to a commodity production function. The consequential system is expanded to cover the subsequent occupation life cycle of the brownfield and the effects on the occupation life cycles of other sites. The proposed model quantifies effects on sites competing to supply the same occupation function. Two approaches are proposed to determine the nature of the sites that are affected and to what extent they are affected: the first resembling a closed-loop approximation, and the second based on economic partial-equilibrium models. Results and Conclusions The scope of the CLCA is far more complex than that of the ALCA. It requires additional data that are associated with important sources of uncertainty. It does allow, however, for the inclusion of tertiary impacts, making it suitable for the evaluation of the often cited environmental benefits of reintegrating the site in the economy. In addition, the ALCA methodology seems to be inappropriate to compare brownfield management options that result in different subsequent uses of the site. Since the effects of this fate are included within the scope of CLCA, however, virtually any brownfield management option available to a decision-maker can aptly be compared. The evaluation of primary and secondary impacts also differs when the consequential approach is used rather than the attributional approach. It is impossible to anticipate the effects of these methodological differences on the results based on the qualitative discussion presented in this paper. Perspectives The complexity and uncertainty introduced by switching to a consequential approach is very high: it is therefore recommendable to evaluate the significance in the gain of environmental information in an actual case study to determine if system expansion is recommendable. Such a case study is presented in Part II to this paper. [39] Lesage P, Ekvall T, Deschênes L, Samson R (20006): Environmental Assessment of Brownfield Rehabilitation Using Two Different Life Cycle Inventory Models. Part 2: Case Study. Int J LCA, OnlineFirst (DOI: )     

A hybrid life cycle assessment of water treatment chemicals: an Australian experience

Purpose

Life cycle assessment (LCA) of chemicals is usually developed using a process-based approach. In this paper, we develop a tiered hybrid LCA of water treatment chemicals combining the specificity of process data with the holistic nature of input–output analysis (IOA). We compare these results with process and input–output models for the most commonly used chemicals in the Australian water industry to identify the direct and indirect environmental impacts associated with the manufacturing of these materials.

Methods

We have improved a previous Australian hybrid LCA model by updating the environmental indicators and expanding the number of included industry sectors of the economy. We also present an alternative way to estimate the expenditure vectors to the service sectors of the economy when financial data are not available. Process-based, input–output and hybrid results were calculated for caustic soda, sodium hypochlorite, ferric chloride, aluminium sulphate, fluorosilicic acid, calcium oxide and chlorine gas. The functional unit is the same for each chemical: the production of 1 tonne in the year 2008.

Results and discussion

We have provided results for seven impact categories: global warming potential; primary energy; water use; marine, freshwater and terrestrial ecotoxicity potentials and human toxicity potential. Results are compared with previous IOA and hybrid studies. A sensitivity analysis of the results to assumed wholesale prices is included. We also present insights regarding how hybrid modelling helps to overcome the limitations of using IO- or process-based modelling individually.

Conclusions and recommendations

The advantages of using hybrid modelling have been demonstrated for water treatment chemicals by expanding the boundaries of process-based modelling and also by reducing the sensitivity of IOA to fluctuations in prices of raw materials used for the production of these industrial commodities. The development of robust hybrid life cycle inventory databases is paramount if hybrid modelling is to become a standard practice in attributional LCA.     

A hybrid life cycle assessment model for comparison with conventional methodologies in Australia

Background, aim, and scope  

One barrier to the further implementation of LCA as a quantitative decision-support tool is the uncertainty created by the diversity of available analytical approaches. This paper compares conventional (‘process analysis’) and alternative (‘input–output analysis’) approaches to LCA, and presents a hybrid LCA model for Australia that overcomes the methodological limitations of process and input–output analysis and enables a comparison between the results achieved using each method. A case study from the water industry illustrates this comparison.     

Using GaBi 3 to perform life cycle assessment and life cycle engineering

The growing availability of software tools has increased the speed of generating LCA studies. Databases and visual tools for constructing material balance modules greatly facilitate the process of analyzing the environmental aspects of product systems over their life cycle. A robust software tool, containing a large LCI dataset and functions for performing LCIA and sensitivity analysis will allow companies and LCA practitioners to conduct systems analyses efficiently and reliably. This paper discusses how the GaBi 3 software tool can be used to perform LCA and Life Cycle Engineering (LCE), a methodology that combines life cycle economic, environmental, and technology assessment. The paper highlights important attributes of LCA software tools, including high quality, well-documented data, transparency in modeling, and data analysis functionality. An example of a regional power grid mix model is used to illustrate the versatility of GaBi 3.     

Electric car life cycle assessment based on real-world mileage and the electric conversion scenario

Purpose

While almost all life cycle assessment (LCA) studies published so far are based on generic vehicles, type approval energy consumption as well as emission data, and application scenarios related to standardized laboratory-based driving cycles, this projects aims at quantifying the LCA based on a real-world vehicle composition and energy consumption data measured before and after the electric conversion of a mini class car. Furthermore, consequences of a second life of a vehicle’s glider on the environmental impact were investigated.

Methods

After having driven 100,000 km, a Smart was converted from combustion to electric in a laboratory project. The inventory was developed grounded upon materials data from laboratory measurements during the conversion process as well as on real-world energy consumption data prior and after the conversion. Three base models are compared in this life cycle impact assessment: a conventional new Smart (combustion engine), a new electric Smart, and a Smart converted from combustion engine to electric. Together with two sensitivity analyses (four different electricity mixes as well as urban vs. mixed driving conditions) and two EOL treatments, 36 scenarios have been quantified. The inventory is based on Ecoinvent database v 2.2 as a background system and includes raw material extraction.

Results and discussion

In urban use, the modeled battery electric vehicle has a favorable environmental impact compared to the ICEV even when charged with the German electricity mix of the year 2013. The advantage in summed up endpoints of the converted Smart is 23 % vs. the new electric Smart on average for the mixed driving conditions and 26 % for the urban driving conditions, respectively. Over a variety of impact categories, electricity consumption during battery cell production in China as well as impacts due to microelectronic components dominated the life cycle. Results for 18 midpoint categories, endpoints for damages to human health, to resource quality and to ecosystem quality as well as the Single score endpoints are reported.

Conclusions

This investigation points out that real-world treatments in inventory development can more specifically outline the environmental advantages of the electric car. The electric conversion of a used combustion engine vehicle can save an additional 16 % (CO₂-eq) and 19 % (single score endpoints) of the environmental impact over a lifetime, respectively, when compared with the new BEV.

     

Screening life cycle impact assessment with weighting methodology based on simplified damage functions

The procedure of screening LCIA with weighting methodology and the result of a case study have been described. The weighting methodology incorporates the impacts related with input and output by the simplified damage functions. Through the dominant analysis by this methodology, we can detect the significant substances and environmental problems in life cycle of the product. With this result, LCA practitioners can concentrate on the analysis for these items to improve the reliability of investigation effectively in the following step. According to the result of case study, an imaginary copy machine, the primary consideration in the foreseeable study should put emphasis on the analysis of the consumption of natural gas and wood, and the emission of carbon dioxide.     

Analytical uncertainty propagation in life cycle inventory and impact assessment: application to an automobile front panel

Background, aim, and scope

Uncertainty information is essential for the proper use of life cycle assessment (LCA) and environmental assessments in decision making. So far, parameter uncertainty propagation has mainly been studied using Monte Carlo techniques that are relatively computationally heavy to conduct, especially for the comparison of multiple scenarios, often limiting its use to research or to inventory only. Furthermore, Monte Carlo simulations do not automatically assess the sensitivity and contribution to overall uncertainty of individual parameters. The present paper aims to develop and apply to both inventory and impact assessment an explicit and transparent analytical approach to uncertainty. This approach applies Taylor series expansions to the uncertainty propagation of lognormally distributed parameters.

Materials and methods

We first apply the Taylor series expansion method to analyze the uncertainty propagation of a single scenario, in which case the squared geometric standard deviation of the final output is determined as a function of the model sensitivity to each input parameter and the squared geometric standard deviation of each parameter. We then extend this approach to the comparison of two or more LCA scenarios. Since in LCA it is crucial to account for both common inventory processes and common impact assessment characterization factors among the different scenarios, we further develop the approach to address this dependency. We provide a method to easily determine a range and a best estimate of (a) the squared geometric standard deviation on the ratio of the two scenario scores, “A/B”, and (b) the degree of confidence in the prediction that the impact of scenario A is lower than B (i.e., the probability that A/B<1). The approach is tested on an automobile case study and resulting probability distributions of climate change impacts are compared to classical Monte Carlo distributions.

Results

The probability distributions obtained with the Taylor series expansion lead to results similar to the classical Monte Carlo distributions, while being substantially simpler; the Taylor series method tends to underestimate the 2.5% confidence limit by 1-11% and the 97.5% limit by less than 5%. The analytical Taylor series expansion easily provides the explicit contributions of each parameter to the overall uncertainty. For the steel front end panel, the factor contributing most to the climate change score uncertainty is the gasoline consumption (>75%). For the aluminum panel, the electricity and aluminum primary production, as well as the light oil consumption, are the dominant contributors to the uncertainty. The developed approach for scenario comparisons, differentiating between common and independent parameters, leads to results similar to those of a Monte Carlo analysis; for all tested cases, we obtained a good concordance between the Monte Carlo and the Taylor series expansion methods regarding the probability that one scenario is better than the other.

Discussion

The Taylor series expansion method addresses the crucial need of accounting for dependencies in LCA, both for common LCI processes and common LCIA characterization factors. The developed approach in Eq. 8, which differentiates between common and independent parameters, estimates the degree of confidence in the prediction that scenario A is better than B, yielding results similar to those found with Monte Carlo simulations.

Conclusions

The probability distributions obtained with the Taylor series expansion are virtually equivalent to those from a classical Monte Carlo simulation, while being significantly easier to obtain. An automobile case study on an aluminum front end panel demonstrated the feasibility of this method and illustrated its simultaneous and consistent application to both inventory and impact assessment. The explicit and innovative analytical approach, based on Taylor series expansions of lognormal distributions, provides the contribution to the uncertainty from each parameter and strongly reduces calculation time.     

Knowledge diffusion and collaboration networks on life cycle assessment

Purpose  

The interest in life cycle assessment (LCA) studies has increased over the years, and one of the main ways of disseminating these studies is through the publication of articles in scientific journals. Coauthorship relations form a social network where it is possible to identify how research is organized and structured in a specific field of knowledge. This paper aims to show the spread of these studies and the configuration of a collaboration network based on coauthorship relations between researchers of LCA considering some properties of social networks.     

Comparative life cycle assessment and life cycle costing of lodging in the Himalaya

Purpose

The main aim of the study is to assess the environmental and economic impacts of the lodging sector located in the Himalayan region of Nepal, from a life cycle perspective. The assessment should support decision making in technology and material selection for minimal environmental and economic burden in future construction projects.

Methods

The study consists of the life cycle assessment and life cycle costing of lodging in three building types: traditional, semi-modern and modern. The life cycle stages under analysis include raw material acquisition, manufacturing, construction, use, maintenance and material replacement. The study includes a sensitivity analysis focusing on the lifespan of buildings, occupancy rate and discount and inflation rates. The functional unit was formulated as the ‘Lodging of one additional guest per night’, and the time horizon is 50 years of building lifespan. Both primary and secondary data were used in the life cycle inventory.

Results and discussion

The modern building has the highest global warming potential (kg CO_2-eq) as well as higher costs over 50 years of building lifespan. The results show that the use stage is responsible for the largest share of environmental impacts and costs, which are related to energy use for different household activities. The use of commercial materials in the modern building, which have to be transported mostly from the capital in the buildings, makes the higher GWP in the construction and replacement stages. Furthermore, a breakdown of the building components shows that the roof and wall of the building are the largest contributors to the production-related environmental impact.

Conclusions

The findings suggest that the main improvement opportunities in the lodging sector lie in the reduction of impacts on the use stage and in the choice of materials for wall and roof.

     

Functional unit influence on building life cycle assessment

Purpose

The building sector is one of the most relevant sectors in terms of environmental impact. Different functional units (FUs) can be used in life cycle assessment (LCA) studies for a variety of purposes. This paper aimed to present different FUs used in the LCA of buildings and evaluate the influence of FU choice and setting in comparative studies.

Methods

As an example, we compared the “cradle to grave” environmental performance of four typical Brazilian residential buildings with different construction typologies, i.e., multi-dwelling and single dwelling, each with high and basic standards. We chose three types of FU for comparison: a dwelling with defined lifetime and occupancy parameters, an area of 1 m² of dwelling over a year period, and the accommodation of an occupant person of the dwelling over a day.

Results and discussion

The FU choice was found to bias the results considerably. As expected, the largest global warming indicator (GWi) values per dwelling unit and occupant were identified for the high standard dwellings. However, when measured per square meter, lower standard dwellings presented the largest GWi values. This was caused by the greater concentration of people per square meter in smaller area dwellings, resulting in larger water and energy consumption per square meter. The sensitivity analysis of FU variables such as lifetime and occupancy showed the GWi contribution of the infrastructure more relevant compared with the operation in high and basic standard dwellings. The definition of lifetime and occupancy parameters is key to avoid bias and to reduce uncertainty of the results when performing a comparison of dwelling environmental performances.

Conclusions

This paper highlights the need for adequate choice and setting of FU to support intended decision-making in LCA studies of the building sector. The use of at least two FUs presented a broader picture of building performance, helping to guide effective environmental optimization efforts from different approaches and levels of analysis. Information regarding space, time, and service dimensions should be either included in the FU setting or provided in the building LCA study to allow adjustment of the results for subsequent comparison.

     

Framework for modelling data uncertainty in life cycle inventories

Modelling data uncertainty is not common practice in life cycle inventories (LCI), although different techniques are available for estimating and expressing uncertainties, and for propagating the uncertainties to the final model results. To clarify and stimulate the use of data uncertainty assessments in common LCI practice, the SETAC working group ‘Data Availability and Quality’ presents a framework for data uncertainty assessment in LCI. Data uncertainty is divided in two categories: (1) lack of data, further specified as complete lack of data (data gaps) and a lack of representative data, and (2) data inaccuracy. Filling data gaps can be done by input-output modelling, using information for similar products or the main ingredients of a product, and applying the law of mass conservation. Lack of temporal, geographical and further technological correlation between the data used and needed may be accounted for by applying uncertainty factors to the non-representative data. Stochastic modelling, which can be performed by Monte Carlo simulation, is a promising technique to deal with data inaccuracy in LCIs.