Evaluation of Ecotoxicity Effect Indicators for Use in LCIA (10+4 pp)

Goal, Scope and Background The paper describes different ecotoxicity effect indicator methods/approaches. The approaches cover three main groups, viz. PNEC approaches, PAF approaches and damage approaches. Ecotoxicity effect indicators used in life cycle impact assessment (LCIA) are typically modelled to the level of impact, indicating the potential impact on 'ecosystem health'. The few existing indicators, which are modelled all the way to damage, are poorly developed, and even though relevant alternatives from risk assessment exist (e.g. recovery time and mean extinction time), these are unfortunately at a very early stage of development, and only few attempts have been made to include them in LCIA. Methods The approaches are described and evaluated against a set of assessment criteria comprising compatibility with the methodological requirements of LCIA, environmental relevance, reproducibility, data demand, data availability, quantification of uncertainty, transparency and spatial differentiation. Results and Discussion The results of the evaluation of the two impact approaches (i.e. PNEC and PAF) show both pros and cons for each of them. The assessment factor-based PNEC approaches have a low data demand and use only the lowest data (e.g. lowest NOEC value). Because it is developed in tiered risk assessment, and hence makes use of conservative assessment factors, it is not optimal, in its present form, to use in the comparative framework of LCIA, where best estimates are sought. The PAF approaches have a higher data demand but use all data and can be based on effect data (PNEC is no-effect-based), thus making these approaches non-conservative and more suitable for LCIA. However, indiscriminate use of ecotoxicity data tends to make the PAF-approaches no more environmentally relevant than the assessment factor-based PNEC approaches. The PAF approaches, however, can at least in theory be linked to damage modelling. All the approaches for damage modelling which are included here have a high environmental relevance but very low data availability, apart from the 'media recovery-approach', which depends directly on the fate model. They are all at a very early stage of development. Conclusion Recommendations and Outlook. An analysis of the different PAF approaches shows that the crucial point is according to which principles and based on which data the hazardous concentration to 50% of the included species (i.e. HC50) is estimated. The ability to calculate many characterisation factors for ecotoxicity is important for this impact category to be included in LCIA in a proper way. However, the access to effect data for the relevant chemicals is typically limited. So, besides the coupling to damage modelling, the main challenge within the further development and improvement of ecotoxicity effect indicators is to find an optimal method to estimate HC50 based on little data.     

Addressing speciation in the effect factor for characterisation of freshwater ecotoxicity—the case of copper

Purpose  

Determination of the ecotoxicity effect factor (EF) in life cycle impact assessment (LCIA) is based on test data reporting the total dissolved concentration of a substance. In spite of the recognised influence of chemical speciation and physico-chemical characteristics of the aquatic systems on toxicity of dissolved metals, these properties are not considered when calculating characterization factors (CFs) for metals. It is hypothesised that the main cause of the variation in reported EC50 values of Cu among published test results lies in different speciation patterns for Cu in the test media, and that the toxicity of Cu is predominantly caused by the free Cu²⁺ ion. Hence, the free Cu²⁺ ion concentration should substitute the total dissolved metal concentration when determining the EF.     

Predictive Value of Species Sensitivity Distributions for Effects of Herbicides in Freshwater Ecosystems

In this article we present a review of the laboratory and field toxicity of herbicides to aquatic ecosystems. Single-species acute toxicity data and (micro)mesocosm data were collated for nine herbicides. These data were used to investigate the importance of test species selection in constructing species sensitivity distributions (SSDs), and in estimating hazardous concentrations (i.e., HC5) protective for freshwater aquatic ecosystems. A lognormal model was fitted to toxicity data (acute EC50s and chronic NOECs) and the resulting distribution used to estimate lower (95% confidence), median (50% confidence), and upper (5% confidence), HC5 values. The taxonomic composition of the species assemblage used to construct the SSD does have a significant influence on the assessment of hazard and only sensitive primary producers should be included for the risk assessment of herbicides. No systematic difference in sensitivity between standard and non-standard test species was observed. Hazardous concentrations estimated using laboratory-derived acute and chronic toxicity data for sensitive freshwater primary producers were compared to the response of herbicide-stressed freshwater ecosystems using a similar exposure regime. The lower limit of the acute HC5 and the median value of the chronic HC5 were protective of adverse effects in aquatic micro/mesocosms even under a long-term exposure regime. The median HC5 estimate based on acute data was protective of adverse ecological effects in freshwater ecosystems when a pulsed or short-term exposure regime was used in the microcosm and mesocosm experiments. There was also concordance between the predictions from the effect model PERPEST and the concentrations at which clear effects started to emerge in laboratory and field studies. However, compared to the SSD concept, the PERPEST model is able to provide more information on ecological risks when a common toxicological mode of action is evaluated as it considers both recovery and indirect effects.     

A life cycle impact assessment procedure with resource groups as areas of protection

Goal and Background  Current Life Cycle Impact Assessment (LCIA) procedures have demonstrated certain limitations in the South African manufacturing industry context. The aim of this paper is to propose a modified LCIA procedure, which is based on the protection of resource groups. Methods  A LCIA framework is introduced that applies the characterisation procedure of available midpoint categories, with the exception of land use. Characterisation factors for land occupation and transformation is suggested for South Africa. A distanceto-target approach is used for the normalisation of midpoint categories, which focuses on the ambient quality and quantity objectives for four resource groups: Air, Water, Land and Mined Abiotic Resources. The quality and quantity objectives are determined for defined South African Life Cycle Assessment (SALCA) Regions and take into account endpoint or damage targets. Following the precautionary approach, a Resource Impact Indicator (RII) is calculated for the resource groups. Subjective weighting values for the resource groups are also proposed, based on survey results from the manufacturing industry sector and the expenditure trends of the South African national government. The subjective weighting values are used to calculate overall Environmental Performance Resource Impact Indicators (EPRIIs) when comparing life cycle systems with each other. The proposed approaches are evaluated with a known wool case study. Results and Discussion  The calculation of a RJI ensures that all natural resources that are important from a South African perspective are duly considered in a LCIA. The results of a LCIA are consequently not reliant on a detailed Life Cycle Inventory (LCI) and the number of midpoint categories that converge on a single resource group. The case study establishes the importance of region-specificity, for LCIs and LCIAs. Conclusions  The proposed LCIA procedure demonstrates reasonable ease of communication of LCIA results. It further allows for the inclusion of additional midpoint categories and is adaptable for specific regions. Recommendations and Outlook  The acceptance of the LCIA procedure must be evaluated for different industry and government sectors. Also, the adequate incorporation of Environmental Performance Resource Impact Indicators (EPRIIs) into decision-making for Life Cycle Management purposes must be researched further. Specifically, the application of the procedures for supply chain management will be investigated.     

Comparative evaluation of Life Cycle Impact assessment methods with a South African case study

Goal and Background  LCIA procedures that have been used in the South Africa manufacturing industry include the CML, Ecopoints, EPS and Eco-indicators 95 and 99 procedures. The aim of this paper is to evaluate and compare the applicability of these European LCIA procedures within the South African context, using a case study. Methods  The five European methods have been evaluated based on the applicability of the respective classification, characterisation, normalization and weighting approaches for the South African situation. Impact categories have been grouped into air, water, land and mined abiotic resources for evaluation purposes. The evaluation and comparison is further based on a cradle-to-gate Screening Life Cycle Assessment (SLCA) case study of the production of dyed two-fold wool yarn in South Africa. Results and Discussion  Where land is considered as a separate category (CML, Eco-indicator 99 and EPS), the case study highlights this inventory constituent as the most important. Similarly, water usage is shown as the second most important in one LCIA procedure (EPS) where it is taken into account. However, the impact assessment modelling for these categories may not be applicable for the variance in South African ecosystems. If land and water is excluded from the interpretation, air emissions, coal usage, ash disposal, pesticides and chrome emissions to water are the important constituents in the South African wool industry. Conclusions  In most cases impact categories and procedures defined in the LCIA methods for air pollution, human health and mined abiotic resources are applicable in South Africa. However, the relevance of the methods is reduced where categories are used that impact ecosystem quality, as ecosystems differ significantly between South Africa and the European continent. The methods are especially limited with respect to water and land resources. Normalisation and weighting procedures may also be difficult to adapt to South African conditions, due to the lack of background information and social, cultural and political differences. Recommendations and Outlook  Further research is underway to develop a framework for a South African LCIA procedure, which will be adapted from the available European procedures. The wool SLCA must be revisited to evaluate and compare the proposed framework with the existing LCIA procedures.     

Ecological Toxicity Methods and Metals. An examination of two case studies (8 pp + 1)

Goal, Scope and Background The Apeldoorn Workshop (April 15th, 2004, Apeldoorn, NL) brought together specialists in LCA and Risk Assessment to discuss current practices and complications of the life cycle impact assessment (LCIA) ecological toxicity (ecotox) methodologies for metals. The consensus was that the LCIA methods currently available do not appropriately characterize impacts of metals due to lack of fundamental metals chemistry in the models. A review of five methods available to perform ecotox impact assessment for metals has been prepared to provide Life Cycle Assessment (LCA) practitioners with a better understanding of the current state of the science and potential biases related to metals. The intent is to provide awareness on issues related to ecotox impact assessment. Methods In this paper two case studies, one a copper based product (copper tube), the other a zinc-based product (gutter systems), were selected and examined by applying freshwater ecological toxicity impact models – USES-LCA, Eco-indicator 99 (EI 99), IMPACT 2002, EDIP 97, and CalTOX-ETP. Both studies are recent, comprehensive, cradle-to-gate, and peer-reviewed. The objective is to review the LCIA results in the context of the practical concerns identified by the Apeldoorn Declaration, in particular illustrating any inconsistencies such as chemical characterization coverage, species specificity, and relative contribution to impact results. Results and Discussion The results obtained from all five of the LCIA methods for the copper tube LCI pointed to the same substance as being the most important – copper. This result was obtained despite major fundamental differences between the LCIA methods applied. However, variations of results were found when examining the freshwater ecological toxicity potential of zinc gutter systems. Procedural difficulties and inconsistencies were observed. In part this was due to basic differences in model nomenclature and differences in coverage (IMPACT 2002+ and EDIP 97 contained characterization factors for aluminium that resulted in 90% and 22% contribution to burden respectively, the other three methods did not). Differences were also observed relative to the emissions source compartment. In the case of zinc, air emissions were found to be substantial for some ecotox models, whereas, water emissions results were found to be of issue for others. Conclusions This investigation illustrates the need to proceed with caution when applying LCIA ecotox methodologies to life cycle studies that include metals. Until further improvements are made, the deficiencies should be clearly communicated as part of LCIA reporting. Business or policy decisions should not without further discussion be based solely on the results of the currently available methods for assessing ecotoxicity in LCIA. Outlook The outlook to remedy deficiencies in the ecological toxicity methods is promising. Recently, the LCIA Toxic Impacts Task Force of the UNEP/SETAC Life Cycle Initiative has formed a subgroup to address specific issues and guide the work towards establishment of sound characterization factors for metals. Although some measure of precision of estimation of potential impact has been observed, such as in the case of copper, accuracy is also a major concern and should be addressed. Further investigation through controlled experimentation is needed, particularly LCIs composed of a variety of inorganics as well as organics constituents. Support for this activity has come from the scientific community and industry as well. Broader aspects of structure and nomenclature are being collectively addressed by the UNEP/SETAC Life Cycle Initiative. These efforts will bring practical solutions to issues of naming conventions and LCI to LCIA flow assignments.     

Using Bioassays and Species Sensitivity Distributions to Assess Herbicide Toxicity towards Benthic Diatoms

Although benthic diatoms are widely used in ecological studies of aquatic systems, there is still a dearth of data concerning species sensitivities towards several contaminants. Within the same community, different species may respond differently depending on their physiological and ecological characteristics. This lack of knowledge makes specific appropriate risk assessment impossible. To find out whether species sensitivity distribution (SSD) could be used to estimate the risk of herbicide toxicity for diatoms, we need to know whether their sensitivity depends on their physiological and ecological characteristics. We carried out single-species bioassays on 11 diatom species exposed to 8 herbicides. Dose-responses relationships were used to extrapolate the Effective Concentration 5 (EC(5)) and the Effective Concentration 50 (EC(50)) for each exposure. These data were used to fit a SSD curve for each herbicide, and to determine the Hazardous concentration 5 (HC(5)) and 50 (HC(50)). Our results revealed a high level of variability of the sensitivity in the set of species tested. For photosystem-II inhibitor (PSII) herbicides, diatoms species displayed a typical grouping of sensitivity levels consistent with their trophic mode and their ecological guild. N-heterotroph and "motile" guild species were more tolerant of PSII inhibitors, while N-autotroph and "low profile" guild species were more sensitive. Comprehensive SSD curves were obtained for 5 herbicides, but not for sulfonylurea herbicides or for dimetachlor, which had toxicity levels that were below the range of concentration tested. The SSD curves provided the following ranking of toxicity: diuron> terbutryn> isoproturon> atrazine> metolachlor. The HC that affected 5% of the species revealed that, even at the usual environmental concentrations of herbicides, diatom assemblages could be affected, especially by isoproturon, terbutryn, and diuron.     

Effect Factors for marine eutrophication in LCIA based on species sensitivity to hypoxia

Hypoxia is an important environmental stressor to marine species, especially in benthic coastal waters. Increasing anthropogenic emissions of nutrients and organic matter contribute to the depletion of dissolved oxygen (DO). Biotic sensitivity to low levels of DO is determined by the organisms’ ability to use DO as a respiratory gas, a process depending on oxygen partial pressure. A method is proposed to estimate an indicator of the intensity of the effects caused by hypoxia on exposed marine species. Sensitivity thresholds to hypoxia of an exposed ecological community, modelled as lowest-observed-effect-concentrations (LOEC), were compiled from literature for 91 benthic, demersal and benthopelagic species of fish, crustaceans, molluscs, echinoderms, annelids, and cnidarians, and converted to temperature-specific benthic (100 m depth) LOEC values. Species distribution and LOEC values were combined using a species sensitivity distribution (SSD) methodology to estimate the DO concentration at which the potentially affected fraction (PAF) of the community's species having their LOEC exceeded is 50% (HC50_LOEC). For the purpose of effect modelling in Life Cycle Impact Assessment (LCIA), Effect Factors (EF [(PAF) m³ kgO₂⁻¹]) were derived for five climate zones (CZ) to represent the change in effect due to a variation of the stressor intensity, or EF = ΔPAF/ΔDO = 0.5/HC50_LOEC. Results range from 218 (PAF) m³ kgO₂⁻¹ (polar CZ) to 306 (PAF) m³ kgO₂⁻¹ (tropical CZ). Variation between CZs was modest so a site-generic global EF of 264 (PAF) m³ kgO₂⁻¹ was also estimated and may be used to represent the average impact on a global ecological community of marine species exposed to hypoxia. The EF indicator is not significantly affected by the major sources of uncertainty in the underlying data suggesting valid applicability in characterisation modelling of marine eutrophication in LCIA.     

Challenges of including human exposure to chemicals in food packaging as a new exposure pathway in life cycle impact assessment

Purpose

Limiting exposure to potentially toxic chemicals in food packaging can lead to environmental impact trade-offs. No available tool, however, considers trade-offs between environmental impacts of packaging systems and exposure to potentially toxic chemicals in food packaging. This study therefore explores the research needs for extending life cycle impact assessment (LCIA) to include exposure to chemicals in food packaging.

Methods

The LCIA framework for human toxicity was extended for the first time to include consumer exposure to chemicals in food packaging through the product intake fraction (PiF) metric. The related exposure pathway was added to LCIA without other modifications to the existing toxicity characterization framework used by USEtox®, i.e., effect factor derivation. The developed method was applied to a high impact polystyrene (HIPS) container case study with the functional unit of providing 1 kg of yogurt in single servings. Various exposure scenarios were considered, including an evidence-based scenario using concentration data and a migration model. Human toxicity impact scores in comparative toxic units (CTU_h) for the use stage were evaluated and then compared to human toxicity impact scores from a conventional LCIA methodology.

Results and discussion

Data allowed toxicity characterization of use stage exposure to only seven chemicals in HIPS out of fourty-four identified. Data required were the initial concentration of chemicals in food packaging, chemical mass transfer from packaging into food, and relevant toxicity information. Toxicity characterization demonstrated that the combined CTU_h for HIPS material acquisition, manufacturing, and disposal stages exceeded the toxicity scores related to consumer exposure to previously estimated concentrations of the seven characterizable chemicals in HIPS, by about two orders of magnitude. The CTU_h associated with consumer exposure became relevant when migration was above 0.1% of the European regulatory levels. Results emphasize missing data for chemical concentrations in food contact materials and a need to expand the current USEtox method for effect factor derivation (e.g., to consider endocrine disruption, mixture toxicity, background exposure, and thresholds when relevant).

Conclusions

An LCIA method was developed to include consumer exposure to chemicals in food packaging. Further study is required to assess realistic scenarios to inform decisions and policies, such as circular economy, which can lead to trade-offs between environmental impacts and potentially toxic chemicals in packaging. To apply the developed method, data regarding occurrence, concentration, and toxicity of chemicals in food packaging are needed. Revisiting the derivation of effect factors in future work could improve the interpretation of human toxicity impact scores.

     

Regional LCA in a global perspective. A basis for spatially differentiated environmental life cycle assessment

Background, aim and scope  

In the context of environmental life cycle assessment (LCA), life cycle impact assessment (LCIA) is one of the central issues with respect to modelling and methodological data collection. The thesis described in this paper focusses on the assessment of toxicity-related impacts, and on the collection of normalisation data. A view on the complementary roles of LCA toxicity assessment on the one hand and human and environmental risk assessment (HERA) on the other is presented, and the global, spatially differentiated LCA toxicity assessment model GLOBOX for the assessment of organics and metals is described. Normalisation factors for the year 2000 are calculated on a global as well as on a European level.     

Freshwater ecotoxicity characterization factors for aluminum

Purpose

Aluminum (Al) is an abundant, non-essential element with complex geochemistry and aquatic toxicity. Considering its complex environmental behavior is critical for providing a reasonable estimate of its potential freshwater aquatic ecotoxicity in the context of Life Cycle Impact Assessment (LCIA).

Methods

Al characterization factors (CFs) are calculated using the following: (1) USEtox^? model version 2.1 for environmental fate, (2) MINEQL+ to estimate the distribution of Al between the solid phase precipitate and total dissolved Al, (3) WHAM 7 for Al speciation within the total dissolved phase, and (4) Biotic Ligand Model (BLM) and Free Ion Activity Model (FIAM) for ecotoxicity estimation for seven freshwater archetypes and default landscape properties for the European continent. The sensitivity of the CFs to aquatic chemistry parameters is calculated. New CFs are compared with Dong et al. (Chemosphere 112:26–33, 2014) and default CF calculated by USEtox 2.1.

Results and discussion

Al CFs vary over 5 orders of magnitude between the seven archetypes, with an arithmetic average CF_ave of 0.04 eq 1,4-DCB (recommended for use), geometric mean CF_geo of 0.0014 eq 1,4-DCB, and weighted average CF_wt of 0.026 eq 1,4-DCB. These values are lower (less toxic) than those for Cu, Ni, Zn, and Pb (with one exception). The effect factor (EF) contributed most to this variability followed by the bioavailability factor (BF), varying over 8 and 4 orders of magnitude, respectively. These revised CFs are 2–6 orders of magnitude lower than those presented by Dong et al. (Chemosphere 112:26–33, 2014) mainly because of consideration of Al precipitation.

Conclusions

Freshwater archetype-specific Al CFs for freshwater ecotoxicity that address the effect of Al speciation on bioavailability (BF) and ecotoxicity (EF) have been calculated, and a CF of 0.04 eq 1,4-DCB is recommended for use in generic LCA. For site-specific LCA, the choice of water chemistry and, in particular, pH, and consideration of metal precipitation could significantly influence results.

Practical implications

Incorporating estimates of metal speciation and its effect on aquatic toxicity is essential when conducting LCIA. Along with metal speciation estimates, the values derived from the definition of water chemistry parameters must also be included into LCIA. For site-generic assessments, we recommend using the arithmetic average of metal CFs. We also recommend using FIAM as a suitable alternative to BLM to estimate EF if the latter is not available. Consideration of metal speciation is essential for providing more realistic estimates of Al freshwater ecotoxicity in the context of LCIA.

     

Comparison between three different LCIA methods for aquatic ecotoxicity and a product environmental risk assessment

Background and Objective  In the OMNIITOX project 11 partners have the common objective to improve environmental management tools for the assessment of (eco)toxicological impacts. The detergent case study aims at: i) comparing three Procter &c Gamble laundry detergent forms (Regular Powder-RP, Compact Powder-CP and Compact Liquid-CL) regarding their potential impacts on aquatic ecotoxicity, ii) providing insights into the differences between various Life Cycle Impact Assessment (LCIA) methods with respect to data needs and results and iii) comparing the results from Life Cycle Assessment (LCA) with results from an Environmental Risk Assessment (ERA). Material and Methods  The LCIA has been conducted with EDIP97 (chronic aquatic ecotoxicity) [1], USES-LCA (freshwater and marine water aquatic ecotoxicity, sometimes referred to as CML2001) [2, 3] and IMPACT 2002 (covering freshwater aquatic ecotoxicity) [4]. The comparative product ERA is based on the EU Ecolabel approach for detergents [5] and EUSES [6], which is based on the Technical Guidance Document (TGD) of the EU on Environmental Risk Assessment (ERA) of chemicals [7]. Apart from the Eco-label approach, all calculations are based on the same set of physico-chemical and toxicological effect data to enable a better comparison of the methodological differences. For the same reason, the system boundaries were kept the same in all cases, focusing on emissions into water at the disposal stage. Results and Discussion  Significant differences between the LCIA methods with respect to data needs and results were identified. Most LCIA methods for freshwater ecotoxicity and the ERA see the compact and regular powders as similar, followed by compact liquid. IMPACT 2002 (for freshwater) suggests the liquid is equally as good as the compact powder, while the regular powder comes out worse by a factor of 2. USES-LCA for marine water shows a very different picture seeing the compact liquid as the clear winner over the powders, with the regular powder the least favourable option. Even the LCIA methods which result in die same product ranking, e.g. EDIP97 chronic aquatic ecotoxicity and USES-LCA freshwater ecotoxicity, significantly differ in terms of most contributing substances. Whereas, according to IMPACT 2002 and USES-LCA marine water, results are entirely dominated by inorganic substances, the other LCIA methods and the ERA assign a key role to surfactants. Deviating results are mainly due to differences in the fate and exposure modelling and, to a lesser extent, to differences in the toxicological effect calculations. Only IMPACT 2002 calculates the effects based on a mean value approach, whereas all other LCIA methods and the ERA tend to prefer a PNEC-based approach. In a comparative context like LCA the OMNIITOX project has taken the decision for a combined mean and PNEC-based approach, as it better represents the ‘average’ toxicity while still taking into account more sensitive species. However, the main reason for deviating results remains in the calculation of the residence time of emissions in the water compartments. Conclusion and Outlook  The situation that different LCIA methods result in different answers to the question concerning which detergent type is to be preferred regarding the impact category aquatic ecotoxicity is not satisfactory, unless explicit reasons for the differences are identifiable. This can hamper practical decision support, as LCA practitioners usually will not be in a position to choose the ’right’ LCIA method for their specific case. This puts a challenge to the entire OMNIITOX project to develop a method, which finds common ground regarding fate, exposure and effect modelling to overcome the current situa-tion of diverging results and to reflect most realistic conditions.     

User needs analysis and development of priorities for life cycle impact assessment

The development of the LCIA programme of the UNEP/SETAC Life Cycle Initiative started with a global survey of LCA practitioners. There were 91 LCIA-specific responses from all global regions. Respondents gave an indication of how they use LCA with respect to both the stage of LCA that they base decisions on (LCI, LCIA or a combination of both) as well as the types of decisions which they support with LCA information. The issues requiring immediate attention within the UNEP SETAC Life Cycle Initiative identified from this User Needs analysis are the need for transparency in the methodology, for scientific confidence and for scientific co-operation as well as the development of a recommended set of factors and methodologies. Of interest is the fact that results from the different regions highlighted the need for different impact categories. Based on this information proposals were made for new impact categories to be included in LCA (and thus LCIA). The LCIA programme aims to enhance the availability of sound LCA data and methods and to deliver guidance on their use. More specifically, it aims to 1) make results and recommendations widely available for users through the creation of a worldwide accessible information system and 2) establish recommended characterisation factors and related methodologies for the different impact categories, possibly consisting of sets at both midpoint and damage level. The work of the LCIA programme of the UNEP/SETAC Life Cycle Initiative has been started within four task forces on 1) LCIA information system and framework, 2) natural resources and land use, 3) toxic impacts, and 4) transboundary impacts. All participants willing to contribute to these efforts are invited to contact the LCIA programme manager or to join the next LCIA workgroup meeting that will take place in at the world SETAC congress in Portland on Thursday 18 November 2004.     

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.     

Comparison of two equivalency factor approaches with simplified risk assessment for lcia of toxicity impact potential

Three approaches recommended for characterization of toxicity impact potential in a life cycle impact assessment (LCIA) are tested on a case study and compared. The two equivalency factor methods are the Persistence, Bioaccumulation, and Toxicity (PBT) method and the Multimedia Fate Modeling (MFM) method using a Mackay Level III model with state-specific environmental data. The simplified risk assessment (SRA) method involved dispersion modeling using site-specific environmental data. The life cycle inventory information evaluated by all three methods was limited to manufacturing of the RDX-based explosive in Kingsport, Tennessee. The effort to collect site-specific environmental data and conduct air dispersion modeling for the SRA method required about 24 times more effort than the PBT method and about 4 times more effort than the MFM method. Direct comparison of impact potential scores for the three approaches were limited to inhalation toxicity scores for nine air pollutants modeled by SRA. Correlations were made on the rank order of the impact potential scores for the nine air emissions evaluated for all three LCIA methods. Although the number of chemicals compared is very limited, the best correlation coefficient (0.96) was between the rank orders for the MFM and the SRA methods. The minimal effort and reduced accuracy of the PBT approach make it best suited for screening large numbers of chemicals for further evaluation of the highest ranked chemicals. The intermediate effort and reasonable accuracy (includes transfers to other media) of the MFM approach make it well suited for LCIAs involving comparative assertions or governmental policy decisions. The maximum effort and assumption of highest accuracy make the SRA approach suitable only after limiting the locations of interest to a few sites by screening with the other two approaches.     

Simplified fate modelling in respect to ecotoxicological and human toxicological characterisation of emissions of chemical compounds

Purpose  

The impact assessment of chemical compounds in Life Cycle Impact Assessment (LCIA) and Environmental Risk Assessment (ERA) requires a vast amount of data on the properties of the chemical compounds being assessed. The purpose of the present study is to explore statistical options for reduction of the data demand associated with characterisation of chemical emissions in LCIA and ERA.     

A simple methodology for elaborating the life cycle inventory of agricultural products

Goal, Scope and Background  A methodological approach for representing agricultural products in terms of life cycle inventory is suggested in this paper. This approach was developed during the conduction of an LCA study for two perennial crops of important Brazilian exportation products: green coffee and orange juice, which included tillage cultivation by commercial farms, harvest, as well as product processing when pertinent. The published papers on agricultural products LCA usually discuss the final results in terms of LCIA, being not very clear what methodology or principles were applied on the LCI phase. The aim of this paper is to present a simple methodology that would be employed by different stakeholders as farmers, environment managers and decision makers for evaluating the environmental performance of their products. In recent years, many researchers have tried to make a worldwide effort in order to reach comparable results of LCA studies developed in different countries. So, the proposed methodology has also the aim of isolating the site-dependency of the results that are not strictly related to the agricultural production. The time coverage suggested is the period can be considered as an average for the specific tillage under evaluation, usually two crops, since there is a large variation on the inputs in every other crop, including the higher and subsequent lower productive periods. Method  The functional unit recommended is 1,000 kg of the specific product, being recommended to distinguish the energy used for the cultivation from that used by the processing stage. There are several specific considerations to transform the data collected through the questionnaires in an inventory data set of fertilizers (macro and micro nutrients), correctives, fillers and pesticides further detailed. Water used for chemicals preparation, in the cleaning and processing stages of the harvested crop is also considered. Land use refers to the area used land for cultivation divided by the medium life period of the tillage. The stoichiometric balance is performed based on the elementary composition of the products. An average carbohydrate formula is established for the products considering the relationship among the carbon, hydrogen and oxygen contents of them. The carbohydrate formula (output) is balanced with carbon dioxide and water (inputs) according to the basic principles of the photosynthesis reaction. The differences among the mineral composition of the products and the total content of these elements (N, P, K, Ca, Mg and micronutrients elements) for all the crop inputs (fertilizers, pesticides, correctives) are allocated as outputs of the system. The pesticides is counted in two forms: grouped in classes (herbicide, fungicide, acaricide, bactericide and inseticide) and specified by the chemical name of the active ingredient. Results and Discussion  A simplified inventory useful for different purposes is generated with the principles described in this paper. The exact fate of each pesticide, fertilizer or corrective or assumptions can be further associated to impact categories as nutriphication, human health, natural resources depletion, ecological toxicity, etc. In this approach the mass balance was focused in the grain or fruit growth and not in the plant or tree as a whole, considering basically the elementary composition of the product and the photosynthesis principle. Despite agricultural LCAs performed in different countries have been published, neither of them considers the carbon capture by the agricultural products during their growth. Conclusions  This method is based on well accepted universal principles of stoichiometry applied to the grain or fruit growth. Minimum estimations were introduced in this approach, which produces ‘clean inventories’, with comparable results between different studies. The generated inventory can be gradually improved as the understanding about each emission fate is known, producing a valid methodology for actual and future knowledge about the fate of tillage emissions. The inventory results of this method can be employed by different stakeholders as farmers, environment managers, decision makers and traders, with valuable environmental parameters for evaluating the environmental performance of their products and also for introducing improvements on their systems, without however to exhibit any particular data.     

Decision Analysis Utilizing Data from Multiple Life-Cycle Impact Assessment Methods Part I: A Theoretical Basis

Numerous methodologies for the life-cycle impact assessment (LCIA) step of life-cycle assessment (LCA) are currently in popular use. These methods, which are based on a single method or level of analysis, are limited to the environmental fates, impact categories, damage functions, and stressors included in the method or model. Because of this, it has been suggested within the LCA community that LCIA data from multiple methods and/or levels of analysis, that is, end-point and midpoint indicators, be used in LCA-based decision analysis to facilitate better or, at least more informed, decision making. In this (two-part) series of articles, we develop and present a series of LCA-based decision analysis models, based on multiattribute value theory (MAVT), which utilize data from multiple LCIA methods and/or levels of analysis. The key to accomplishing this is the recognition of what LCIA damage indicators represent with respect to decision analysis, namely, decision attributes and, in most cases, proxy attributes. The use of proxy attributes in a decision model, however, poses certain challenges, such as the assessment of decision-maker preferences for actual consequences that are only known imprecisely because of inherent limits of both LCA and scientific knowledge. In this article (part I), we provide a brief overview of MAVT and examine some of the decision-theoretic issues and implications of current LCIA methods. We illustrate the application of MAVT to develop a decision model utilizing damage indicators from a single LCIA methodology; and, we identify the decision-theoretic issues that arise when attempting to combine LCIA indicators from multiple methods and/or levels of analysis in a single decision model. Finally, we introduce the use in our methodology of constructed attributes to combine related end-point damage indicators into single decision attributes and the concept and evaluation of proxy attributes.     

Logarithmic Transformation is Essential for Statistical Analysis of Fungicide EC50 Values

Half maximal (50%) effective concentration (EC50) values are widely used to express fungicide potency and sensitivity of plant pathogens. This study explored the necessity of logarithmic transformation for statistical analysis of EC50 values. The results demonstrated that without logarithmic transformation, none of the five sets of epoxiconazole EC50 data (n = 26–33) against Sclerotinia sclerotiorum fitted a normal distribution. But after logarithmic transformation, four of the five datasets became normally distributed. Of the five sets of pyraclostrobin EC50 data (n = 29–32), only one dataset fitted a normal distribution. After logarithmic transformation, four datasets became normally distributed. Logarithmic transformation transformed the heterogeneity of variance across the five sets of epoxiconazole EC50 data to homogeneity but failed to improve the heterogeneity of variance across the five sets of pyraclostrobin EC50 data. For 150 isolates' EC50 values to epoxiconazole and 153 isolates' EC50 values to pyraclostrobin, the intervals of arithmetic means ± standard deviations (SD) covered 85.3% and 90.2% of data points, respectively, whereas the intervals of geometric means (*) multiplied/divided by the multiplicative SD (S*) covered 69.3% and 70.9% of data points, respectively, which approximated the theoretical value of 68.3%. Distribution normality and homogeneity of variance are prerequisites for analysis of variance (anova ) and the two parameters could be improved by logarithmic transformation, therefore, power and efficiency of statistical tests on EC50 data will be greatly enhanced by this kind of transformation.     

Tolerability and toxicity of prophylactic cranial irradiation in patients with non-small cell lung cancer – Results of a phase II study (with estimation of hematological toxicity,pituitary function and magnetic resonance spectra changes)

AimTo evaluate the tolerability and toxicity of PCI in patients with NSCLC.BackgroundProphylactic cranial irradiation (PCI) is a standard treatment for patients with small cell lung cancer. There are data showing a decreasing ratio of brain metastases after PCI for non-small cell lung cancer (NSCLC-non small cell lung cancer) patients but, so far, there is no evidence for increasing overall survival. The main concern in this setting is the tolerance and toxicity of the treatment.Materials and methodsFrom 1999 to 2007, 50 patients with NSCLC treated with radical intent underwent PCI (30 Gy in 15 fractions). Mean follow-up was 2.8 years. The tolerability and hematological toxicity were evaluated in all patients, a part of participants had done neuropsychological tests, magnetic resonance imaging with ¹H nuclear magnetic resonance spectra, and estimation of pituitary function.ResultsDuring follow-up, 20 patients developed distant metastases, 4-brain metastases. Fourteen (30%) patients had acute side effects: (headache, nausea, erythema of the skin). The symptoms did not require treatment breaks. Six patients complained of late side effects (vertigo, nausea, anxiety, lower extremity weakness, deterioration of hearing and olfactory hyperesthesia). Hematological complications were not observed. Testosterone levels tended to decrease (p = 0.062). Visual-motor function deteriorated after treatment (p < 0.059). Performance IQ decreased (p < 0.025) and the difference between performance IQ and verbal IQ increased (p < 0.011). Degenerative periventricular vascular changes were observed in two patients. Analysis of the spectroscopic data showed metabolic but reversible alterations after PCI.ConclusionPCI in the current series was well tolerated and associated with a relatively low toxicity.