Environ. Sci. Technol. 2010, 44, 2624–2631
Accounting for Ecosystem Services in Life Cycle Assessment, Part II: Toward an Ecologically Based LCA YI ZHANG, ANIL BARAL, AND BHAVIK R. BAKSHI* William G. Lowrie Department of Chemical & Biomolecular Engineering, The Ohio State University, Columbus, Ohio 43210
Received February 19, 2009. Revised manuscript received November 1, 2009. Accepted January 28, 2010.
Despite the essential role of ecosystem goods and services in sustaining all human activities, they are often ignored in engineering decision making, even in methods that are meant to encourage sustainability. For example, conventional Life Cycle Assessment focuses on the impact of emissions and consumption of some resources. While aggregation and interpretation methods are quite advanced for emissions, similar methods for resources have been lagging, and most ignore the role of nature. Such oversight may even result in perverse decisions that encourage reliance on deteriorating ecosystem services. This article presents a step toward including the direct and indirect role of ecosystems in LCA, and a hierarchical scheme to interpret their contribution. The resulting Ecologically Based LCA (Eco-LCA) includes a large number of provisioning, regulating, and supporting ecosystem services as inputs to a life cycle model at the process or economy scale. These resources are represented in diverse physical units and may be compared via their mass, fuel value, industrial cumulative exergy consumption, or ecological cumulative exergy consumption or by normalization with total consumption of each resource or their availability. Such results at a fine scale provide insight about relative resource use and the risk and vulnerability to the loss of specific resources. Aggregate indicators are also defined to obtain indices such as renewability, efficiency, and return on investment. An Eco-LCA model of the 1997 economy is developed and made available via the web (www.resilience.osu.edu/ecolca). An illustrative example comparing paper and plastic cups provides insight into the features of the proposed approach. The need for further work in bridging the gap between knowledge about ecosystem services and their direct and indirect role in supporting human activities is discussed as an important area for future work.
Introduction Accounting for the direct and indirect role of ecosystem goods and services in supporting human activities is of urgent importance due to at least two reasons: these goods and services are the basis of sustainability and are deteriorating rapidly, as identified by the Millennium Ecosystem Assessment (2); as societies become wealthier, the public’s ecological knowledge diminishes as the role of ecosystems becomes indirect and less apparent (3). Life cycle oriented methods aim to account for direct and indirect consumption of * Corresponding author e-mail:
[email protected]; tel: 1-614-2924904; fax: 1-614-292-3769. 2624
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
resources and environmental impact, but as reviewed in Part I of the series (4), none of the existing methods account for the role of ecosystems. This is due to challenges such as difficulty in getting adequate information about ecosystem services and networks, and need for aggregate indicators that can help in interpreting high dimensional data while accounting for substitutability between resources. Among ecosystem services, information about many provisioning services is most readily available, and used in many life cycle oriented methods. Knowledge about many other ecosystem services is available from ecology and natural sciences. For example, information about supporting services such as biogeochemical cycles is available in terms of exergy flow, and that of net primary productivity and its human appropriation from recent studies, as discussed in Part I (4). However, information about many other services, such as regulating services, is not easy to find, particularly in quantitative terms. Some such information is included in distinct life cycle oriented methods, indicating a need for greater integration between these methods. This article represents a step toward meeting this goal to result in an Ecologically Based LCA (Eco-LCA). With respect to interpreting inventory data and aggregation methods, conventional LCA has been successful in standardizing life cycle inventory data and developing methods to quantify the impact of emissions on human health, ecosystem quality, and resources. Steps in impact analysis include classification, characterization, and normalization. Midpoint aggregation methods yield impact categories such as global warming potential, human toxicity, and acidification potential (5). Endpoint impact assessment methods further combine these categories into a smaller number of categories and even a single aggregate indicator (6). These methods address the important challenge of assuming substitutability between items being aggregated or compared. One way of satisfying this assumption is to represent different variables in terms of a common numeraire that accounts for their relative quality. Economics does this via monetary valuation, while life cycle impact assessment methods rely on characterization factors that represent emissions in equivalents of another chemical to capture their relative impact. For example, the impact due to all greenhouse gases is represented in CO2 equivalents, and that of acidifying emissions is represented in terms of their SO2 equivalents. Thus, to aggregate a ton each of CO2 and CH4 emissions, the latter is converted into CO2 equivalents by a multiplier of 23 if a 100-year time horizon is considered (5). Further aggregation to endpoint indicators relies on similar factors such as those that convert the human impact of emissions into disability adjusted life years. Other aggregation approaches have also been suggested in other methods (4). This article proposes a hierarchical approach that complements LCIA as a step toward accounting for ecosystem services. This paper contains at least three contributions directed toward bridging the gap between LCA and natural capital assessment. First, it presents a framework for an Eco-LCA to include the role of natural capital in the life cycle of industrial goods and services. This framework can account for ecosystem goods and services that support the processes or economic sectors in the selected life cycle, and relies on existing methods in LCA at the process and economy scales. The focus of this article is on provisioning services, including water from different sources, sunlight, minerals, fossil fuels, and different types of biomass; supporting services such as soil formation and erosion, earth cycles and photosynthesis; and on some regulating services such as pollination and air 10.1021/es900548a
2010 American Chemical Society
Published on Web 02/24/2010
quality regulation. Ongoing research is using the proposed framework to include the role of other ecosystem services, as more knowledge becomes available. Second, unlike most existing approaches for resource accounting where the emphasis is on representing resource flows in terms of a common unit, we propose a hierarchy of metrics for understanding the role of ecological resources via multiple levels of aggregation. This hierarchy consists of individual resource flows forming the bottom layer, which may be normalized by respective national or regional flows. The resulting dimensionless quantities permit identification of the limiting resources to which the selected product may be most vulnerable. Resources are then characterized and categorized into midpoint groups. These are further combined to yield endpoint indicators such as cumulative consumption of resources, efficiency, renewability, and return on investment. Several popular physical aggregation schemes are considered to enable easy comparison and insight that should permit a rational evaluation of the pros and cons of each approach. Monetary valuation or ecological footprint may also be used for aggregation if relevant data are available. This approach can retain the information at different levels in an open and transparent structure, and can assist decision-makers at different levels to access the information quickly and add more information at any level of the hierarchy (7). Third, the approach described in this article has been used to develop an Eco-LCA model of the 1997 U.S. economy and a user-friendly software package (1). This approach and tool combine methods such as EIOLCA (8) and emergy (9) and includes new data for several ecosystem services that have not been considered before. In the rest of this paper, the next section presents the Eco-LCA approach including its data requirements, calculations, and resource aggregation. The following section demonstrates the method via an illustrative case study comparing paper and plastic cups. Most previous studies of this popular comparison have focused mainly on emissions and fossil fuel use. Our application provides novel and complementary insight about resource intensity and role of ecosystem goods and services. Other more comprehensive hybrid Eco-LCA studies have also been completed (10).
Approach for Ecologically Based LCA Since Eco-LCA is meant to complement and extend conventional LCA, its methodological framework is deliberately kept similar to that of traditional LCA, but the details in most steps are different, as described in this section. Goal and Scope Definition. The focus of Eco-LCA is on understanding the direct and indirect dependence of human activities on ecological resources to guide their wise management. Scope defines the boundary of the selected system, that is, what processes are included in the study. The main difference in Eco-LCA, as compared to conventional LCA, is that the analysis boundary extends to include ecosystems. Inventory Acquisition. The goods and services selected by the Millennium Assessment (MA) (2) are of vital importance for providing benefits that people derive from ecosystems. They are shown in Table 1, along with an indication of the extent to which each item is included currently in Eco-LCA. This table is an extension of Table 1 in Part I (4). MA considers only services whose quality and formation can be altered by human beings. Consequently, resources such as solar energy and minerals, which are both essential in any technological life cycle, are not part of the MA report since their generation does not depend on the presence of living organisms (2). However, such resources are included in EcoLCA because they are of significant interest for life cycle studies. Food, fiber, and biomass fuels were examined in MA, but they are products of managed ecosystems, derived
TABLE 1. Ecological Goods and Services and Their Inclusion in Eco-LCA service
included?
provisioning services fossil fuels & minerals renewable energy land crops, livestock and fiber wild fish and aquaculture wild plant and animal food timber nonwood forest products biomass fuel genetic resources natural medicines fresh water
yes yes yes indirectly partially no yes no partially no no yes
regulating services air quality regulation climate regulation water regulation erosion regulation water purification disease & pest regulation waste processing pollination natural hazard regulation
partially no no partially partially no partially partially no
supporting services soil formation photosynthesis primary production nutrient cycling water cycling cultural services
yes yes partially partially partially no
from the collective efforts of other ecosystem services and human labor. Instead of including these ecosystem products directly, the present version of Eco-LCA (1) considers the ecological resources that are necessary for producing these products such as sunlight, land, water, and fertile soil. Among the four types of ecological services, many provisioning, regulating, and supporting services are included in the EcoLCA model. Figure 1 depicts the flows used in the Eco-LCA model of the 1997 U.S. economy. Provisioning services are usually accompanied by concomitant material or energy flow and are included in EcoLCA via physical data about specific resources. For the process model-based Eco-LCA approach, such data would have to be obtained for each process included in the life cycle, and may have some overlap with existing methods such as LCA and energy analysis. For the economic model-based EcoLCA, data about resource use in specific sectors are compiled from public sources that are listed in detail in the Supporting Information. Many provisioning services are shown as red dashed lines in Figure 1. Raw data of provisioning services are usually obtained in mass or energy units. Conversion ratios from mass or energy to exergy values for many resources are also available (11, 12). Accounting for regulating services is more challenging, particularly if they are to be quantified. These services, shown as solid gray lines in Figure 1, include climate regulation, water purification, and flood protection, etc. The role of ecosystem services for pollution dissipation and absorption, depicted at the right bottom corner of Figure 1, is considered partially by estimating the uptake of CO2 by ecosystems through the total amount of primary production on Earth (13). Pollination services by honey bees are also included. The duration and the number of honeybee hives for one acre of various crops are known (14), which multiplied with the crop land area gives the contribution of honeybees. The unit VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2625
FIGURE 1. Flows of ecosystem services to the 1997 U.S. economy. They form the basis of the Eco-LCA software (1). of pollination service is “hive-days”. Accounting for other regulating services is part of ongoing work, and could benefit from other efforts like those in refs 15 and 16. It is common in ecology and natural sciences to represent natural and ecological systems as networks of exergy flow. Such efforts have quantified many supporting services including biogeochemical cycles, photosynthesis, and nutrient cycling (9, 11, 17, 18). Quantifying the contribution of such services to specific ecological resources requires allocation of these ecological flows to ecosystem products and services. This challenge is similar in principle to the allocation issues in LCA, and has been addressed via emergy (9) or ecological cumulative exergy consumption (19). The EcoLCA model uses this approach to quantify the role of supporting services for producing many other services, as shown in Figure 1 by the blue dotted lines. Here, detrital matter, nitrogen, and phosphorus mineralization are contributors to soil formation. Solar energy is absorbed by plants during photosynthesis. Biogeochemical cycles, including mineral and water cycles, are considered via their exergy flows. Following the emergy approach, these flows are represented in terms of solar equivalents. Such conversion to a common numeraire can be useful to account for quality differences as discussed in the introduction, but is not essential if the goal is only to appreciate the role of supporting services. Studies in systems ecology have resulted in ratios of the cumulative exergy consumed in ecosystems to make a natural resource to the exergy or mass of the resource. These are often called the resource’s transformity (9). In the emergy approach, the transformity has been considered to represent the quality of the resource based on its position in the ecological hierarchy. These ratios can assist in converting all the flows, expressed in their raw units to solar equivalent joules (sej) easily. This permits easy accounting for the supporting services for making natural resources and com2626
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
paring them on the same basis of solar equivalent joules. This common basis also helps in validating the assumption of substitutability between resources made in all aggregation methods. Figure 1 shows the currently available data about services that support the 1997 U.S. economy. These data are expressed in their original units. References and details about converting them into other physical units are included in the Supporting Information. Ongoing efforts are directed toward finding such information for other services. Network Algebra. Since the life cycle of any product or process is usually a large and complicated network, various approaches have been developed for obtaining a reasonable approximation of this network. These include process modelbased LCA and economic model-based LCA. Relevant network algebra has also been developed for calculating the direct and indirect resource use. The methods for processmodel-based LCA (20) may be used directly for process model-based Eco-LCA. This network is often relatively straightforward and even linear. For economic model-based LCA, a variety of methods have been developed for allocation of physical flow information via the monetary flow network (21). These include models for EIOLCA, energy and exergy analysis, ecological footprint analysis, land use analysis, etc. (22, 23). The Eco-LCA model at the process or economy scale proposed in this article relies on all these methods. For economic model-based LCA, physical flow information is usually available at the “demand side” in the form of emissions and resource consumption at the point of use. However, information about the use of ecosystem goods and services is usually available in the form of flows entering specific processes or sectors. For example, honeybee pollination services enter the economy via agricultural sectors that grow honeybee pollinated fruits. This “supply side” information needs to be allocated or distributed through the economic network to compute the direct and indirect resource consumption. Network algebra techniques have
been developed for both situations. Since the demand side model has been studied widely (8), this section focuses on how to model the supply side information. Additional details and theoretical insight into the relation between various approaches is in ref 24. The resources from nature are essential inputs for economic production, and may be considered as physical “value added” (25). The Input-Output (IO) technique, in essence, distributes resources through a matrix of interindustry transactions against a net output sold to consumers. This approach requires information about the services entering the economy and their entry points or peripheral sectors that rely directly on the services. These inputs are distributed to other sectors in proportion to the monetary flow between sectors. Then, the following equation may be derived for calculating cumulative resource consumption (23, 26) Xph ) (I - GT)-1Vph where (I - GT)-1 is the Ghosh inverse, the matrix G is a normalized form of transaction table (Z) over total output vector (X) calculated by gij ) zij/xi. Each element of Vph, expressed as Vph.ij, represents the amount of the j-th resource entering the i-th sector. The contribution of the j-th resource for one dollar of the i-th sector’s final demand (Rio.ij) can be calculated via normalization of the sectoral economic throughput as ˆ -1 Rio ) (I - GT)-1VphX
(1)
ˆ is the diagonalized matrix of vector X. Depending Here, X on the unit used for Vph, the result can be readily expressed in terms of cumulative consumption of mass, energy, or exergy. Equation 1 is based on the Ghosh inverse, and can be shown to be identical to other approaches based on the Leontief inverse (24). Given the producer price of a product, multiplying it with the resource intensity calculated in eq 1 gives the cumulative resource consumption for the product. An illustrative example about how to carry out this calculation is in the Supporting Information. Although the Input-Output Model can represent the whole economy with all products sold on the market, it also has limitations due to the highly aggregated sectors, age of available models, and assumption of price homogeneity. Some of these shortcomings may be overcome by performing hybrid studies that combine the completeness of the EIO model with the accuracy of process models as shown in ref 10. Quantifying Resource Use. Interpreting the inventory about use of ecosystem goods and services requires analysis of multivariate data. The approach proposed here for EcoLCA is analogous to the impact assessment step in LCA, but focuses on inputs from nature as opposed to emissions to nature. We propose a novel hierarchical structure where the finest scale retains all the detailed resource consumption but permits comparison between resources and helps in identifying resource scarcity or vulnerability. Such insight is not available via existing life cycle oriented methods. These data may be aggregated via different schemes to form coarser scales, and further aggregated to single indicators or metrics. Raw and Normalized Data. At the finest scale, representing raw data in different units can shed light on different aspects of the system being studied. For example, mass and energy values represent relatively local and short-term effects of materials and fuel use, respectively. Industrial Cumulative Exergy Consumption (ICEC) accounts for exergy consumption only in industrial systems in a life cycle. It permits comparison across more resources than mass or energy because of its ability to capture the useful work of material
and energy flows, while Ecological CEC (ECEC) quantifies broader implications including the role of supporting ecosystem services (19). The conversion between units is facilitated by the exergy and emergy literature. In addition to interpreting these raw data, normalizing them to obtain dimensionless numbers can also provide insight about limiting resources and potential resource vulnerabilities. Depending on the goal and scope of the study, normalization can be designed from the perspective of global, national, or policy considerations (27). In this work, normalized metrics are calculated by comparing cumulative consumption of a resource with its annual flow at the national level. This denominator is chosen because of ready availability of data and because it matches the economy scale Eco-LCA model. Other denominators such as total reserve of each resource or carrying capacity of ecosystem services may also be used. The insights from various units and normalized results are explained in the case study in the next section and in the Supporting Information. Classification. Various classification schemes may be defined for ecosystem goods and services. The most common distinction between natural resources is in terms of their renewability versus nonrenewability. Such categorization is useful for calculating aggregate metrics such as a renewability index. Resources can also be classified on the basis of their origin as biotic versus abiotic, or as materials versus energy, or in terms of their ecological origin from one of four ecospheres: lithosphere (land), hydrosphere (water), atmosphere (air), and biosphere (living flora and fauna), and other services (solar energy, pollination). The most appropriate classification scheme depends on user preferences, since unlike the classification scheme for emissions, classification of resources is not standardized and could benefit from more research. Aggregation. An important challenge faced by Eco-LCA is how to interpret the multiple attributes of various services. Aggregation is meant to assist by consolidating multivariate data but also leads to loss of information, and if not done carefully, can be misleading. Aggregation often involves assumptions and should be used with care to avoid misinterpretation of the results. This challenge is particularly severe for ecological resources due to wide differences in their qualities. Although mass, energy, and exergy have been popular for representing resources, they cannot capture all the resources. For example, mass cannot capture solar energy, minerals do not have energy content, and pollination services are represented in units of “hive-days”, with no clear conversion into mass, energy, or exergy. Similar challenges exist for other aggregation methods including monetary valuation and ecological footprint. Therefore, existing aggregation methods tend to ignore goods and services with inconvenient units or difficult-to-find conversion factors. In practice, there is no ideal aggregation approach that can fully encompass all the desired qualities. This motivates the hierarchical approach proposed in this work. It can retain detailed information at the raw and normalized data level, and build multiple aggregation schemes and indices to assist decision making. As discussed in Part I (4), the challenges in aggregating resources are quite significant. Many different aggregation methods have been proposed, but few studies compare their performance. In the absence of any consensus, in this article and the present implementation of Eco-LCA (1), we consider aggregation by mass, energy, ICEC, and ECEC, since each approach has its benefits. For example, for materials or fossil energy intensive processes, aggregation in mass or energy units, respectively, makes sense. ICEC may be appropriate for aggregating mass and energy resources, when their quality differences are small. ECEC seems most appropriate for situations with a wide variety of resources ranging from VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2627
renewable materials and fuels to nonrenewables. The usefulness of these different aggregation methods may also be appreciated in terms of the scale of their analysis or their use (28). Representing resources in terms of mass, energy, or ICEC provides information about short-term harm and a local scale, while ECEC considers longer-term issues and the contribution from ecosystems. Aggregation methods for each physical representation are discussed in more detail in refs 19 and 23. As mentioned in ref 4, aggregation implies a strong sense of substitutability (29). Replacing crude oil by natural gas or steel by aluminum can be achieved but only to a certain degree. However, natural resources exhibit too many attributes such as electric conductivity, heat resistance, hardness, corrosion inhibition tailored to specific industrial use, visual appearance, and appeal to human preferences. In practice, it is not possible to aggregate based on all these properties. Ideally, a case by case examination on a single or a small group of resources is more appropriate for such a purpose (29, 30), but this is quite challenging. Metrics and Indices. Challenges in defining metrics have been widely discussed, and hierarchical (7) or multiple parallel metrics have been suggested (31). Many types of metrics may be defined with the Eco-LCA model, and some of them are discussed here. • Resource Intensity (R) may be defined for each of the four units as material intensity (RCMC), energy intensity (RCEnC), ICEC intensity (RICEC), and ECEC intensity (RECEC). These represent cumulative resource consumption per unit of product or per dollar of economic output and may also be calculated for various classification schemes discussed elsewhere. • Efficiency (η) is the ratio of the physical “value” of the products (Y) to the total resource use in the same unit. Y R
η)
Y is the quantity of the raw materials converted to the desired product. The remaining resources in R are used for processing the raw materials into products (Rproc). The two parts of R can be expressed as R ) Y + Rproc. • Renewability Index (RI) is the ratio of the consumption of renewable resources to total consumption. It lies between 0 and 1, and can be calculated for various units. RI )
Rren R
This index is related to some existing metrics such as the emergy loading ratio (32) and breeding factor (31). • Physical Return on Investment (pROI) represents the ratio of the value of the product (Y) to the quantity of resources needed to convert the raw materials into the product (Rproc). pROI )
Y Rproc
Such a metric has been popular in Net Energy Analysis and is related, but not identical, to the emergy yield ratio (32). Equipped with the vast amount of information, many other indicators can build beyond these conventional engineering indicators. For example, the carbon footprint of a product may be evaluated by comparing its greenhouse gas emission with the amount absorbed by the relevant sequestration services. Other examples are land use, water footprint, and nutrient balance. Other thermodynamic metrics based on abatement exergy and resource reserves may also be calculated (33–35). As more data about ecosystem 2628
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
services are included, the Eco-LCA framework can enable the calculation of many other metrics. These metrics are important for assessing and comparing the life cycle resource consumption of economic products, and their use is illustrated in the next section and in ref 10. Interpretation of aggregate metrics is not necessarily straightforward, and significant care must be exercised to avoid misleading conclusions. Practical case studies can play an important role in understanding and interpreting the hierarchy of metrics.
Illustrative Example: Eco-LCA of Cups LCA of drinking cups has been popular for many years and is presented here to demonstrate the features of Eco-LCA. These results are only at the economy scale making them coarse, and should not be interpreted for decision making. They are meant to be purely illustrative and can be easily reproduced via the Eco-LCA software available online (1). A more comprehensive hybrid Eco-LCA of cups based on combining process information with the EIO model is in ref 24. The focus here is on paper versus plastic single-use cups, which have also been studied earlier (22, 36). The services identified in Figure 1 have been linked to the sectors of the 1997 U.S. Input-Output model. Cumulative resource consumption is calculated via eq 1 for each of the 500 sectors in the model. For this study, the industrial sectors corresponding to these cups are “Paperboard container manufacturing” (NAICS number 322210) and “Foam product manufacturing” (3261A0). Producer prices are found to be 1.2 U.S. cents and 1 cent for one paper and one polystyrene (PS) cup, respectively. The functional unit is one cup used only once. Figure 2 displays one typical hierarchy of results from Eco-LCA. At the bottom is the normalized cumulative consumption of selected resources. Several bars for specific minerals and some services are not shown here for maintaining clarity, and because they show similar normalized values due to their entering the economy via the same sector. This figure shows the greater reliance of PS cups on crude oil and natural gas; while paper cups are more dependent on coal and water. Paper cups also dominate in the use of common nonmetallic minerals, such as stone and sand. The greater reliance on resources related to silvicultural practices such as wood, timber land, and solar energy irradiation on forests is as expected. Land use, mainly in the form of timberland, is another key factor for paper cups and is more than 10 times that of the land used for PS cups. Information about other ecological resources such as agricultural inputs and minerals are also available from Eco-LCA, as shown. Such insight about resource use is not available in conventional LCA or other life cycle oriented methods, and can be useful for ecologically conscious decision making. Based on accounting for carbon sequestration in ecosystems, EcoLCA indicates that a paper cup is nearly carbon neutral while a PS cup emits about 7 g of CO2 per cup. This calculation does not include the potential emissions from paper decomposition and from disturbance of soil. Loss of natural capital such as soil erosion due to construction and farming are also quantified. This figure also shows the reliance of the two cups on nonrenewable resources. Additional plots and data are in the SI. The upper levels of Figure 2 show the aggregated results based on the physical units of mass, energy, ICEC, and ECEC, in Figure 2b-e, respectively. At intermediate levels of aggregation the results are categorized in two ways: as four ecospheres (lithosphere, hydrosphere, atmosphere, biosphere) and other services including soil erosion and nonmaterial inputs; and as renewable versus nonrenewable
FIGURE 2. Comparison of paper and polystyrene cups based on IO model. (a) Normalized cumulative resource consumption (b) Cumulative Mass Consumption; (c) Cumulative Energy Consumption; (d) Industrial Cumulative Exergy Consumption; (e) Ecological Cumulative Exergy Consumption. RI: Renewability Index; η: efficiency. resources. Three indicators are calculated at the coarsest level: cumulative resource use, renewability index, and efficiency. • Figure 2b only captures the mass of resources and is dominated by the use of water in the hydrosphere. The high reliance on fossil fuels and mineral use in the lithosphere can also be seen but as being less important than water use. • Fossil fuels from the lithosphere, wood in the biosphere, and solar energy in ecological services are the major contributors when aggregation is done in energy units, as shown in Figure 2c. Although fossil energy is usually the emphasis of conventional LCA, the biggest energy contributor is sunlight. This is also true for PS cups despite solar energy being an indirect input. This is because of the large quantity of solar energy entering the economy via forestry and agricultural sectors. • ICEC accounts for material and energy resources, and the result is shown in Figure 2d. These results are still dominated by solar energy followed by lithospheric resources. • Accounting for supporting ecosystem services via ECEC indicates the largest “value” of lithospheric resources, and the significantly smaller role of sunlight, as shown in Figure 2e. This is because the requirement of solar energy to make resources like fossil fuels available is several orders of magnitude bigger than plain sunlight. Aggregation in terms of renewable and nonrenewable resources shows the larger consumption of the former when
aggregated in terms of mass, energy, and ICEC. However, for ECEC the trend is reversed due to extending the boundary to include ecosystems, and because ecosystems need to do significant work to produce nonrenewable resources and make them available. In terms of coarser aggregation metrics, cumulative resource consumption in all four units is much larger for paper cups than PS cups. This indicates the larger support required to make paper cups regardless of the analysis boundary. This need not be a negative since the resources may be renewable and plentiful. Note that ECEC is about 3 orders of magnitude larger than ICEC, indicating the large contribution from supporting services. The renewability index is nearly 100% in terms of mass, energy, and ICEC, but much smaller in terms of ECEC. Both cups being nearly completely renewable, as indicated by metrics other than ECEC, is clearly misleading. This result is due to the dominance of low quality resources, namely water and sunlight, and the assumption that they are substitutable with other higher quality resources such as fossil fuels. This index in terms of ECEC shrinks remarkably, and paper cups are found to be more renewable, but still have a 90% reliance on nonrenewable resources. Physical return on investment is less applicable for cups, and is neglected from the case study. Life Cycle efficiency is calculated as the ratio of mass to CMC or fuel value to CEnC, ICEC, and ECEC. This indicates the benefits if the material or fuel content of the cups is recovered at the end of life. Paper cups show higher efficiency in terms of mass, but lower VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2629
efficiency in terms of energy and ICEC. Without including ecological work, PS cups exhibit high energy efficiency because their energy input is small due to it being derived from concentrated high-quality fossil fuels. However, in terms of ECEC, efficiency of PS cups becomes smaller than that of paper cups because of accounting for quality differences between fossil fuels and wood. This analysis indicates that PS cups do seem to have benefits in terms of some aggregate indicators over paper cups, but both have their vulnerabilities to ecological resources, which should be considered in making any decision.
Discussion The Ecologically Based Life Cycle Assessment approach and model developed in this article are able to account for many more ecosystem services than existing life cycle oriented methods. It includes more provisioning services than existing methods, some regulating services, and several supporting services. This is indicated by comparing Table 1 in this article with Table 1 in Part I (4). This approach combines data, knowledge, and insight from multiple existing methods including conventional LCA, emergy or ECEC analysis, exergy analysis, and network algebra, along with the Millennium Ecosystem Assessment. Such integrative research is essential for bridging the gaps that currently exist between environmentally conscious decision making in engineering, business, and government; ecological knowledge; and life cycle methods. Only with such integration will it become possible to consider the true life cycle impact and sustainability of human activities. The data and approach in this article overcome some of the challenges for incorporating the role of ecosystem services in LCA that were identified in Part I. However, this work only represents an early step in the path toward accounting for ecosystem services in LCA, since many challenges remain. Although Eco-LCA includes new data about ecosystem services in LCA such as pollination services and carbon sequestration, as can be seen from Table 1, many resources are ignored or considered only partially. Accounting for these services requires getting relevant data about the contribution of ecosystem services to economic activities at the scale of individual processes as well as economic sectors. Such data, when available, can be easily incorporated in the proposed framework. However, data about some services may not be easy to find and may only be available in a qualitative manner (15). New methods will be required to combine quantitative and qualitative information. Representing the results in a hierarchical manner provides insight for different types of decision makers and combines the raw data in a transparent manner. Other variations to this scheme are certainly possible and will be developed as more data become available. A variety of other aggregation schemes may also be included. If information about the carrying capacity and state of various services is available, it may also be included in the hierarchy to understand the risk to specific products due to ecological deterioration. Practical case studies are also needed to generate useful insight. Eco-LCA relies mainly on physical and thermodynamic methods for quantifying and aggregating various flows. These methods can provide complementary insight, as illustrated in Part I. It also shows how the substitutability assumption can provide misleading insight, and the ability of ECEC to provide intuitive results when a wide range of resources are being aggregated. However, ECEC (or emergy) also faces many challenges due to lack of ecological knowledge and challenges in allocation. The framework in this article can be used to explore other aggregation schemes. For example, the emergy based ecological footprint can be easily included since it would be a direct extension of the current approach. 2630
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
Ecosystem services contributed to the 1997 U.S. economy are compiled and synthesized via economic model-based Eco-LCA algorithm. The case study demonstrates the use of this model. It is hoped that this work and public domain software (1) encourage further research and applications that consider the pivotal role of ecosystems and evaluate the strengths and weaknesses of various accounting methods to ultimately result in a standardized approach that augments the current practice of LCA by building a bridge between industrial ecology and ecosystem ecology. Such work also needs to be combined with economic and social considerations for a more useful and holistic evaluation.
Acknowledgments Partial financial support for this work was provided by the National Science Foundation (ECS-0524924, CBET-0829026) and the Environmental Protection Agency (R832532).
Supporting Information Available Details about the Eco-LCA methodology and the cups example. This information is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) PSE Group and Center for Resilience, The Ohio State University. Ecologically Based Life Cycle Assessment, 1997 U.S. benchmark model, beta version; www.resilience.osu.edu/ecolca 2009; accessed July 13, 2009. (2) Millennium Ecosystem Assessment; www.millenniumassessment. org; accessed April 10, 2008. (3) Pilgrim, S. E.; Cullen, L. C.; Smith, D. J.; Pretty, J. Ecological knowledge is lost in wealthier communities and countries. Environ. Sci. Technol. 2008, 42, 1004–1009. (4) Zhang, Y.; Singh, S.; Bakshi, B. R. Accounting for ecosystem services in life cycle assessment, part I: a critical review. Environ. Sci. Technol. 2009; doi: 1021/es9021156. (5) Handbook on Life Cycle Assessment: Operational Guide to the ISO Standards; Guine´e, J., Ed.; Kluwer Academic Publishers, 2002; Database for impact assessment method is available at http://www.leidenuniv.nl/interfac/cml/pmo/index.html; accessed April 10, 2008. (6) Goedkoop, M.; Spriensma, R. The Eco-indicator 99, A damage oriented method for life cycle impact assessment; Technical Report, 2001; www.pre.nl; accessed April 10, 2008. (7) Yi, H.-S.; Hau, J. L.; Ukidwe, N. U.; Bakshi, B. R. Hierarchical thermodynamic metrics for evaluating the environmental sustainability of industrial processes. Environ. Prog. 2004, 23, 302–314. (8) Hendrickson, C.; Lave, L.; Matthews, H. Environmental Life Cycle Assessment of Goods and Services: An Input-Output Approach; Resources for the Future: Washington, DC, 2006. (9) Odum, H. T. Environmental Accounting: Emergy and Environmental Decision Making; John Wiley and Sons: New York, 1996. (10) Urban, R. A.; Bakshi, B. R. 1,3-Propanediol from fossils versus biomass: a life cycle evaluation of emissions and resource use. Ind. Eng. Chem. Res. 2009, 48, 8068–8082. (11) Szargut, J. Anthropogenic and natural exergy losses (exergy balance of the Earth’s surface and atmosphere). Energy 2003, 28, 1047–1054. (12) De Meester, B.; Dewulf, J.; Janssens, A.; Van Langenhove, H. An improved calculation of the exergy of natural resources for exergetic life cycle assessment (ELCA). Environ. Sci. Technol. 2006, 40, 6844–6851. (13) Ayres, R.; Ayres, L. Accounting for Resources, 1: Economy-Wide Application of Mass-Balance Principles to Materials and Waste; Edward Elgar Publishing: Northampton, MA, 1998. (14) Delaplane, K.; Mayer, D. Crop Pollination by Bees; CABI Publishing: New York, 2000. (15) Hanson, C.; Ranganathan, J.; Iceland, C.; Finisdore, J. The Corporate Ecosystem Services Review: Guidelines for Identifying Business Risks and Opportunities Arising from Ecosystem Change; Technical Report, 2008; http://www.wri.org/publication/ corporate-ecosystem-services-review; accessed March 21, 2008. (16) Daily, G. C.; Polasky, S.; Goldstein, J.; Kareiva, P. M.; Mooney, H. A.; Pejchar, L.; Ricketts, T. H.; Salzman, J.; Shallenberger, R. Ecosystem services in decision making: time to deliver. Front. Ecol. Environ. 2009, 7, 21–28.
(17) Chen, G. Exergy consumption of the earth. Ecol. Modell. 2005, 184, 363–380. (18) Hermann, W. Quantifying global exergy resources. Energy 2006, 31, 1349–1366. (19) Hau, J. L.; Bakshi, B. R. Expanding exergy analysis to account for ecosystem products and services. Environ. Sci. Technol. 2004, 38, 3768–3777. (20) Rebitzer, G.; Ekvall, T.; Frischknecht, R.; Hunkeler, D.; Norris, G.; Rydberg, T.; Schmidt, W.; Suh, S.; Weidema, B.; Pennington, D. Life cycle assessment part 1: framework, goal and scope definition, inventory analysis, and applications. Environ. Int. 2004, 30, 701–20. (21) Leontief, W. W. Quantitative input-output relations in the economic system of the United States. Rev. Econ. Statistics 1936, 18, 105–125. (22) Lave, L. B.; Cobas-Flores, E.; Hendrickson, C. T.; McMichael, F. C. Using input-output analysis to estimate economy-wide discharges. Environ. Sci. Technol. 1995, 29, 420–426. (23) Ukidwe, N. U.; Bakshi, B. R. Industrial and ecological cumulative exergy consumption of the United States via the 1997 inputoutput benchmark model. Energy 2007, 32, 1560–1592. (24) Zhang, Y. Ecologically-Based LCA—An Approach for Quantifying the Role of Natural Capital in Product Life Cycles. Ph.D. thesis, The Ohio State University, Columbus, OH, 2008. (25) Duchin, F. Input-Output Economics and Material Flows; Rensselaer Working Papers in Economics, 2004; ideas.repec.org/p/ rpi/rpiwpe/0424.html, accessed Aug. 21, 2007. (26) Ghosh, A. Input-output approach in an allocation system. Economica 1958, 25, 58–64.
(27) Bare, J.; Gloria, T. Critical analysis of the mathematical relationships and comprehensiveness of life cycle impact assessment approaches. Environ. Sci. Technol. 2006, 40, 1104–1113. (28) Ulgiati, S.; Raugei, M.; Bargigli, S. Overcoming the inadequacy of single-criterion approaches to Life Cycle Assessment. Ecol. Modell. 2006, 190, 432–442. (29) Cleveland, C.; Kaufmann, R.; Stern, D. Aggregation and the role of energy in the economy. Ecol. Econ. 2000, 32, 301–317. (30) Van Oers, L.; de Koning, A.; Guinée, J.; Huppes, G. Abiotic resource depletion in LCA; Technical Report, 2002; oasis.leidenuniv.nl/ interfac/cml/ssp/projects/lca2/, accessed Dec. 24, 2007 (31) Dewulf, J.; Van Langenhove, H.; Van De Velde, B. Exergy-based efficiency and renewability assessment of biofuel production. Environ. Sci. Technol. 2005, 39, 3878–3882. (32) Ulgiati, S.; Brown, M. Monitoring patterns of sustainability in natural and man-made ecosystems. Ecol. Modell. 1998, 108, 23–36. (33) Szargut, J.; Stanek, W. Influence of the pro-ecological tax on the market prices of fuels and electricity. Energy 2008, 33, 137–143. (34) Dewulf, J.; Van Langenhove, H.; Mulder, J.; van den Berg, M.; van der Kooi, H.; de Swaan Arons, J. Illustrations towards quantifying the sustainability of technology. Green Chem. 2000, 2, 108–114. (35) de Swaan Arons, J.; van der Kooi, H.; Sankaranarayanan, K. Efficiency and Sustainability in the Energy and Chemical Industries; CRC Press: Boca Raton, FL, 2004. (36) Hocking, M. B. Relative merits of polystyrene foam and paper in hot drink cups: Implications for packaging. Environ. Manage. 1991, 15, 731–747.
ES900548A
VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2631
