5.1.1.1 History

The process of evaluating a product or service with a holistic view of all the life cycle stages evolved from activities in the mid-1900s with respect to purchasing (Novick, 1959). In that report, not only the cost to purchase weapon systems was considered in total cost, but other life stages including use, development, and end of life provide the true cost over the entire lifetime. This kind of life cycle thinking for cost is now very prevalent. Most people have heard about the “cost of ownership” of automobiles, which includes purchase price in addition to other costs like fuel, maintenance, and trade-in value. The principle of life-cycle analysis in the 1970s was a starting point for life-cycle analysis applied to other issues such as environmental impacts of products. Dissatisfaction with some of the early LCA studies included ambiguous methods and results that had questionable motivations. LCAs performed with poor methods and hidden motivations unfortunately discredited LCAs for some time. The issues surrounding methods and motivations needed to be addressed to enable creditable studies to inform product designers and decisions makers. This motivated the standardization of LCA methods using a predefined framework and detailed methodological guidance. The International Organization of Standards created two guidance documents for performing LCAs that are defined in standards ISO 14040 (2006) and ISO 14044 (2006). More recently, an additional ISO method was developed to provide guidance on accounting for water use and consumption in LCAs – ISO 14046 (2014).

Triple Bottom Line

Triple bottom line refers to the evaluation of environmental, economic, and social considerations for a product or service (Figure 5.1). Product or service need to consider and address all three pillars of sustainability to be considered sustainable from the perspective of triple bottom line. Environmental LCA contributes to the triple bottom line reporting by quantifying the ecological and human health performance of competing products and services. Of the three measures, the economic evaluation is the most certain and objective. Environmental and health LCA analyses are often less certain and less objective, with the social analysis the most difficult to produce objective, clear, and certain results.

5.1.2.1 Eco-labels

In general, consumers collectively have great influence on products and services that are prevalent. It is logical to think that different people from different countries, cultures, religions, and economic status would have different opinions and preferences. These groups would have different opinions on what is important with respect to the environment. In the end, we all share the planet and collectively should care for it. To the average consumer, products and service environmental performance is performed mainly through environmental labels. An eco-label is a statement that discloses the environmental performance of products based on an environmental LCA. These are important in that similar to a nutrition label, an eco-label is how a consumer may evaluate alternative products. These eco-labels can be powerful tools in gaining larger shares of a market. With data-based, fair, and objective eco-labels, consumers can make better choices.

Eco-labels are classified into three types, each with different levels of rigorous LCA backing. Type 2 eco-labels are environmental self-declaration claims, usually focusing on a single claim. Examples might be the labeling of a product as natural, biodegradable, or recyclable. These often are not independently verified and so there is a risk of “green-washing,” the deliberate misrepresentation of an environmental performance. These labels should be looked at with a critical eye. Type 1 eco-labels are third-party-certified multicriteria environmental labeling that ensure that a set of predetermined requirements are met. Within a product category, this can allow a consumer to understand which products are preferable. Examples of this are the Energy Star program that promotes energy efficiency and the Forestry Stewardship Council that promotes sustainable management of forests.

The most rigorous eco-label is type 3, in which quantified environmental information on the life cycle of a product is performed using standard LCA procedures so that fair and objective comparison of products can be made. For a type 3 eco label or environmental product declaration, a program operator must define LCA protocols for a product category that is reviewed and commented on by interested stakeholders. With this consistent methodology and reporting, competing entities with products can perform LCAs on their products using the defined methods enabling comparable results that can ultimately provide consumers with creditable information to make purchasing decisions.

5.3.1.1 Goal Definition

The first step of an LCA is defining the goal of the study. In this part of the LCA, the aim of the study and breadth and depth of the study are communicated. There are two types of LCA purposes: descriptive and change-oriented. A descriptive LCA generally looks at broader aspects of an issues, for example, how much of the world's global warming impact can be attributed to transportation. These larger environmental questions are answered by descriptive LCAs. The second purpose of an LCA is a change-oriented LCA where two decision options for fulfilling a function are compared. Some typical examples of change-oriented LCAs are paper versus plastic bags, and flying versus driving your car. These types of studies can guide the audience in ways to reduce environmental impacts through changing behaviors based on the findings of the LCA.

The intended audience is another aspect of the goal and scope, which is important to communicate. The audience can include interest groups such as policy makers, company marketing groups, or product development teams. Additionally, the involved interest groups and parties should be identified. These include companies, funding sources, target audiences, and expert reviewers. It is noted that the intended use of the LCA can be different from the end use as the information may be relevant to other decisions and analysis beyond the original intent.

One specific type of LCA used to compare two different products is a “comparative assertion disclosed to the public.” In this type of study, “environmental claim regarding the superiority or equivalence of one product vs a competing product which performs the same function” are communicated to the public. These types of studies must follow the ISO 14044 standards including the nine steps required for a “comparative assertion.”

5.3.1.2 Scope Definition

The scope definition serves the purpose of communicating to the audience what is included and what is excluded from the study. Depending on the goal of the study, there are several types of scopes including cradle-to-gate, cradle-to-grave, and gate-to-gate. There are other words that are commonly used to describe these scopes but the ideas are similar.

• Cradle-to-grave includes all flows and impacts from raw material extraction to disposal and reuse.
• Cradle-to-gate includes all flows and impacts from raw material to production and excludes product use and end of life.
• Gate-to-gate only includes flows from production or material processing steps of a product life cycle.

The scope should be carefully selected considering the potential implications of not including product stages or phases in the scope of the work. For example, a product may have lower production emissions but has a shorter lifetime than an alternative product that would not be considered if a cradle-to-gate boundary was selected. The scope of the study is often best communicated in a process stage diagram as seen in Figures 5.2 and 5.3. These types of diagrams list the major unit steps that are considered within a study and clearly show what is not included in the study.

Temporal boundaries are also set in the scope definition. Temporal assumptions, or assumptions relating to time, can have large influences on the results of a study. It is important to pick a study timeframe that will best capture the impacts of the product or processes being studied. Impacts occurring in 100 years is a common temporal boundary for global warming potential (GWP). In a 100-year temporal boundary, impacts occurring after 100 years would not be calculated and included in the results. There is an emerging field of dynamic LCA, which can more accurately model emissions through time for product systems lasting over many years (Daystar et al., 2016; Levasseur et al., 2010). This new method improves LCA temporal consistency that enables a better determination of the global impacts over a given time period.

Other aspects to be included in the scope are technology types and geographical regions. Many studies are spatially dependent, and the overall results and conclusions may not be broadly applied to other regions. Product manufacturing technology can also be important to the study results. Products or services from older technologies often have different impacts than the most current technology. For this reason, it is important to clearly communicate the type and stage of the technology under analysis. In addition to the aspect listed, allocation procedures impact assessment methods used should be reported in the scope of an LCA.

5.3.1.3 Functional Unit

A functional unit is the primary measure of the product, service, or good you are studying. ISO states “the functional unit defines the quantification of the identified functions (performance characteristics) of the product. The primary purpose of a functional unit is to provide a reference to which the inputs and outputs are related. This reference is necessary to ensure comparability of LCA results” (ISO 14044, 2006). This can be a service, mass of material, or an amount of energy. Selecting appropriate functional units is critical to creating an unbiased analysis. For example, comparing paper milk cartons to glass milk bottles may not be the best option due to the different possible sizes of each container. The real purpose of the container would be to deliver a quantity of fresh milk to a consumer. For this example, a better functional unit may be impacts of a container delivering 8 ounces of milk to a consumer. The results will then be normalized to a quantity of milk, which is what the consumer really wants not the container it comes from.

5.3.1.4 Cutoff Criteria

Data collection for an LCA is the most time-intensive and laborious step. To expedite this process, cutoff criteria are often used. A cutoff criteria defines a level of product content or other parameter which the study will not consider. For example, material contents less than 1% of the total product mass are often not considered in the LCA. This allows the LCA practitioner to focus on data from the main flows while systematically eliminating flows that may not influence the results. An LCA practitioner should perform cutoff decisions carefully as some materials can produce emissions and environmental impacts disproportionate to their component weight.

5.3.1.5 Problems Set – Goal and Scope Definition

5.3.2.1 Preparation of Data Collection Based on Goal and Scope

The goal and defined system boundary for an LCI will define which life-cycle stages and unit processes data are collected for. For life-cycle stages within the defined system boundary, unit processes should be identified. A unit process is a transformation of material or service performed. Many unit processes can make up a larger process or life-cycle stage. Identifying the material flows into and out of the defined system is the first step of an LCI performed by the LCA practitioner.

Elementary flows originate in the environment and are mined or retrieved to be used in a process or flows that are released from processes that are released to the environment and are not used by other processes. One can think of these elementary flows as the actual material used and materials released to the environment as a result of the studied product system. In Figure 5.4, the two types of LCI data can be seen. On the top half of the figure, technosphere flows such as products, services, and other goods are listed. The lower half of figure lists the elementary flows such as chemicals released to soil or air.

After the materials flows have been determined through interviews, literature searches, and measurement, LCA software can be used to track the material process flows back elementary flows to and from the environment.

5.3.2.2 Data Collection

Biogenic and Anthropogenic Carbon

In performing LCIs for forest biomaterials and other materials that include some form of plant materials, it is important to understand the differences in accounting methods of carbon derived from biological resources and petroleum resources. Carbon contained in a piece of wood is typically termed biogenic carbon as it was taken in from the environment during tree growth through photosynthesis. The natural uptake of carbon during tree growth and eventual decay of wood carbon during natural decomposition in forests or in a landfill is often termed the carbon cycle. This uptake and release of carbon within a relatively short time period, often less than 100 years, is in contrast to anthropogenic carbon emissions that are produced from oil and other carbon sources that have been affixed in some type of material for thousands of years. When oil or coal is burned, additional CO₂ is released and added to the environment, which over times increases the net concentration of CO₂ as the coal or other source did not capture the carbon in recent time periods.

To determine the CO₂ absorbed during a biomaterials growth, the following equation can be used.

In this equation, the percent carbon in the material, for instance, piece of loblolly pine with a carbon content of ∼50%, can be multiplied by both the mass of the material and the molecular weight of CO₂ then divided by the molecular weight of carbon. This calculation can determine the CO₂ taken in from the atmosphere during growth and the potential for the material to release CO₂ to the environment during decomposition.

There are different methods that can be used to track both biogenic and anthropogenic carbon. Some studies do not track biogenic carbon as they assume that biogenic CO₂ is “carbon neutral” meaning has no effect on global climate change. This standpoint is controversial and is not as transparent as tracking the uptake and the later potential releases. Tracking biogenic carbon can add additional work, however, is more a more robust method and provides additional transparency.

Basic LCA Software Structure

LCA software can be split into several components: (i) the software package, (ii) data sets, and (iii) LCIA methods (which is explained in detail in the next section). The software package such as SimaPro, openLCA, and Gabi can be thought of as a framework or a calculator that keeps track of data and performs intensive numerical calculations. With the many flows and detailed data, much effort has been invested in creating efficient calculation methods to speed up analysis time. This framework, however, is not useful without inventory data. There are many premade secondary data sets provided from sources such as Ecoinvent, Gabi, and United States Department of Agriculture (USDA) that contain previous LCI results for various chemicals, materials, energy, services, and waste treatment processes. LCA software can access this previously developed data and allow an LCA practitioner to include a chemical or other process from a data set in their LCA without the need to perform an entire LCA on that particular material or process. This fundamental aspect of LCA, the leveraging of previous study results for new studies, is a key benefit of LCA software and can save countless hours on the LCI step. LCIA methods are procedures and conversions that are used in performing an LCIA such as GWP characterizations and weighting methods. There are many accepted LCIA methods that calculate LCA results using different impact categories, types of impacts, and weighting methods. Further discussion surrounding LCIA is provided later.

Figure 5.5 visually depicts how the different components of LCA software and data interact. The LCI step requires data from data sets (e.g., Ecoinvent) and primary data gathered by the LCA practitioner surrounding the process or product under analysis. An LCI is calculated with the combination of these two types of data and the use of LCA software calculations. The LCI data can then be used to perform an impact assessment using the LCIA methods (e.g., TRACI) (Table 5.1).

Software	Licensing	Data sets	Software features	Website
openLCA	Open source and free	Ecoinvent, Gabi, USLCI, CML, and others	Fast calculation engine, easily shares models, no yearly subscription, process based on transparent data, used for USDA digital commons LCA data development	www.openlca.org
SimaPro	Paid licensing	Ecoinvent, USLCI, CML, and others	Process based on transparent data, good customer support, robust uncertainty analysis	www.pre-sustainability.com/simapro
Gabi	Paid licensing	Gabi Dataset, Ecoinvent, USLCI	Robust data set, visual process flow-based modeling, ease of use, data frequently updated	www.gabi-software.com

5.3.2.3 Data Quality

Mass and Component Balance Data Check

Much of the LCI data comes from the product manufacturing process. These data can be collected directly from a company, from literature, or from process conversion models. When examining a process, a mass balance should be performed to ensure that all process streams are accounted captured. A mass balance should equal zero when adding all of the material inputs to a system subtracted by all the materials exiting in the system. In reality and especially in complicated production processes, the difference of the in–out flows may not be zero. The percent mass closure can be calculated by

Providing mass balances and listing percent closure is a good practice that leads to data transparency. Ideally, the percent closure system should be 100%; however, in reality this is often not the case due to measurement errors, fugitive emissions, and other modeling errors. In practice, mass balances above 95% can often provide meaningful data suitable for use.

Similar to a total mass balance a component balance can be performed to ensure proper tracking of an element within a system. For instance, carbon balances track the mass of carbon flowing into and out of a process. The in and out should be equal or the percent closure should be near 100%. If this is not the case, the data should be reexamined to find the error or missing data.

When using secondary data from LCI databases or literature, it is also important to perform data quality checks. Mass balance is also a valid approach to checking secondary data surrounding unit processes. Further analysis should examine the spatial data surrounding the secondary data. For instance, a material produced in China using an average electrical grid that relies heavily on coal power will have different environmental impacts than a product produced in the Northwest United States where hydroelectric power is more dominant in the average electrical grid. One way to overcome these regional differences is to change the electricity type used for the secondary data and recalculate the overall impacts. Doing this can provide a better representation of the impacts of the product under study. Technology is another important factor. Often, secondary data are out of date and are based on older technology than what is currently employed in the industry. The record date for the secondary data can give some indication of whether the technology is current or not, but further analysis of the documentation should be performed to determine if the technology used is representative of the technology used to produce the product under study.

5.3.2.4 Coproduct Treatment – Allocation

Some production processes produce more than one product, and the emissions of the process cannot be easily attributed to a single product from the process. Coproduct treatment methods as defined by ISO 14044 are used to properly account for emissions from the production of multiple products. The ISO standard states that wherever possible, allocation should be avoided by

1. 1. dividing the unit process to be allocated into two or more subprocesses and collecting the input and output data related to these subprocesses;
2. 2. expanding the product system to include the additional functions related to the coproducts.

If allocation cannot be avoided by these two methods, the allocation method should

1. 1. partition inputs and outputs of the system between its different products or functions in a way that reflects the underlying physical relationships between them;
2. 2. partition input and output data between coproducts in proportion to the economic value of the products.

The first route to avoiding allocation by process subdivision can remove the need to account for multiple products from the overall system by splitting it into additional subprocesses. With a more detailed process flow diagram and data, some products may be produced through separate subprocess and thus can be accounted for individually.

Process subdivision is not always possible and expanding the product system may be necessary to account for the coproducts. Using this method, all impacts associated with production are assigned to the primary product of interest and a credit or negative emission is used to account for the displacement of the coproduct production in other manufacturing processes. The system expansion method works only when the manufacturing process is not the primary route to the coproduct. The life-cycle handbook (Curran, 2012) gives the example of a hydrocracking unit that produces ethylene, propylene, other hydrocarbons, fuel gas, and heat. In this example, energy and gas can be accounted for using system expansion by giving a displacement credit. However, system expansion does not work on the other products, as hydrocracking is the primary commercial route to these products. The other products of the hydrocracking process are accounted for using allocation.

When allocation is to be performed, physical parameters and relationships should be used to attribute the total impacts to individual products. Such parameters may include mass of final product, raw material ratio required to produce the final products, energy content of products, or other physical relationships. If a physical relationship cannot be determined, economic allocation can be performed by attributing the impacts in accordance with the revenue associated with each product. For instance, in a process producing products A, B, and C, if product A generates 98% of the revenue from a process, 98% of the total impacts would be assigned to product A. Economic allocation should be avoided as product prices can change and thus changes the overall LCA results even when the production process and overall emissions remain constant.

5.3.2.5 Relating Data to the Unit Process

Much of the data collected for an LCI will not be in correct or the most meaningful units. Often, there may be inventories based on a month's production and could have units such as energy use per number of units produced in a month. On most occasions, data surrounding one unit are needed, and thus some data manipulation is required.

As an example, Table 5.2 lists the inputs and outputs of dry rough lumber, at kiln, US PNW as developed by the US LCI database. The outputs from this process are listed in the top portion of the table and the outputs on the bottom. Note that there are product flows and elementary flows as discussed earlier in this chapter. Under the amount column, the total quantity of product is 23 units. This would indicate that all the emissions listed in the amount column are per 23 units of product, which in this case is 1 kg of dry rough lumber at kiln in the US Pacific Northwest. To make these secondary data more useful, it is helpful to relate all the flows to one unit, 1 kg, of output so that in later processes this can be scaled to meet the needs of one product. In this example, all the numbers in the “Amount” column are divided by 23 to get the inputs and outputs per one unit of product.

5.3.2.6 Relating Data to the Functional Unit

The next step of LCI is similar to relating to unit process step, but instead this time the data are related to the functional unit defined in the goal and scope. For instance, if the functional unit of the LCA was a rustic chair, that chair might require several kilograms of wood as well as other materials. In relating the data to the functional units, all the inputs and outputs are scaled to the quantity of material/product that is required to fulfill the requirements of the functional unit; this flow is called the reference flow. This step can be performed in Microsoft Excel worksheet or in an LCA software package. The results after this step may include numbers such as energy use per functional unit or CO₂ emission per functional unit. Elementary, waste, and product flows may be listed at this point; however, they would all be listed in relation to the required amount per functional unit.

5.3.2.7 Data Aggregation

When performing an LCI, many calculations are required for the different life-cycle stages (remember: product production, product use, end of life) that may be useful to analyze separately before combining. Often, the final results of both LCIs and LCAs are reported by life-cycle stages as well as the total impacts. Since the final total number is required, the LCI data are summed across all the life-cycle stages. For instance, if electricity was used by five different processes, the total electric usage may be summed for all these processes and reported.

5.3.2.8 LCI Data Interpretation

Inventory data interpretation is an important step within the larger interpretation of the whole study. Throughout the LCI development process, some level of interpretation must be performed. For example, when collecting data, the practitioner must interpret the available data and make a judgment call on the quality and relevance to the goal and scope. Often when performing an LCI, it will become clear that the goal and scope are at times not appropriate given the availability of data, time, and resources available to complete the study. Developing a high-quality LCI is the most time-consuming part of an LCA, which often experiences hang-ups and delays that are in many cases beyond the control of the practitioner. In these cases, where data are just not available, the goal and scope can be adjusted so that the available data can support the goal and scope and eventually the overall study conclusions.

Another aspect of interpretation is uncertainty in data and modeling assumptions. Though the use of a sensitivity analysis, a variety of study assumptions, and data can be tested to determine the influence on the overall LCI results. For instance, an assumption on a process yield where incoming material is converted to a product material can be varied depending on incoming material composition that varies with time. To determine how this yield that can often change influences the energy or other LCI parameters, the yield could be adjusted up or down a set percentage, for example, 25%, or adjusted according to some statistical measure associated with the value such as standard deviation. Providing a range of values and understanding how different values influence the results and eventually conclusions are far more valuable than providing a static one-number answer without deeper insights into what is driving the overall results.

In some studies, the goal may be to compare one product to another. This type of study is referred to as a comparative LCA, and when a company wants to publically communicate such results they must be first certified by through a peer-reviewed process as defined by the ISO 14044 standards. In these types of studies, it is important in the LCI phase to determine if the available data and models can reasonably calculate the differences in environmental flows and impacts. To reasonably claim a difference between flows for competing products, a general rule of thumb is that the values are not significantly different unless they are at least 25% different. This 25% different rule is often used as there is inherently uncertainty in the data, modeling, and other factors that are not possible to completely model. This 25% rule at times could be too high and a robust uncertainty analysis could be performed to further determine the certainty and the probability that one product will produce lower flows and environmental impacts than another.

Another important part of LCI interpretation is determining and communicating the “hotspots” or process areas and process flows that influence the overall LCI the most. For instance, in dried rough lumber production, the electricity used during drying would be an energy use hotspot as well as a large contributor to GHG emissions. Insights surrounding the hotspots are often some of the most important and actionable information that is attainted by performing an LCA, and time should be dedicated to understanding the driving factors behind the values of the most important environmental flows.

For more information surrounding LCI methods, please refer to www.lcatextbook.com and navigate to Chapter 5. This textbook provides additional details and resources that could not be included in this brief introduction to LCA and is a free book available to all.

5.3.2.9 Problems Set – Life-Cycle Inventory

Life-Cycle Impact Assessment

LCIA is the third sequential step of an LCA. The purpose of an LCIA “is to provide additional information to assess LCI result and help users better understand the environmental significance of natural resource use and environmental releases”. The LCIA helps provide significance and simplify results for easier decision making; however, it is important to understand that it does not directly measure the impacts of chemical releases to the environment as an environmental risk assessment does. The third step of LCIA follows sequentially after the LCI using the many flows to and from the environment developed in the LCI. These LCI flows without an impact assessment step are not easily interpreted, and understanding the significance of emissions can be impossible (Figure 5.6).

The LCIA as previously mentioned is different from a risk assessment measuring absolute values of environmental impacts; rather, the LCIA helps determine the significance of emissions and impacts in relation to the study scope. The absolute value of the impacts cannot be determined by the LCIA due to (Margni and Curran, 2012)

• the relative expression of potential environmental impacts to a reference unit;
• the integration of environmental data over space and time;
• the inherent uncertainty in modeling environmental impact;
• the fact that some possible environmental impact occur in the future.

Even though the LCIA has limitations, it is useful in determining what impacts matter the most, what unit processes are contributing most through hot spot analysis, and help identify best-scenario options when environmental trade-offs occur.

According to ISO, there are three mandatory processes of an LCIA including selection of impact categories, classification, and characterization (Figure 5.7).

5.3.2.10 Mandatory Elements

Selection of Impact Methods

The selection of impact methods should reflect the intent and methods outlined in the goal and scope of the study. The impact indicators of the LCIA method must reflect the purpose of the study and examine the resources or impacts to the environment that the study is addressing. For instance, if quantifying fossil fuel usage of a product is a stated goal, the impact assessment method must calculate this to enable the interpretation of results.

There are primarily two main types of impact assessment methods: midpoint and end point impact assessment methods (Goedkoop and Spriensma, 2001; Bare et al., 2006). The midpoint indicator methods are closely tied to science and are based on more exact models. As there are fewer assumptions associated with midpoint indicators, they generally have less uncertainty than end point indicators. End point indicators are useful in that they are easier to understand and are more appealing to a general audience (Figure 5.8).

5.3.2.11 Classification

Classification is the second of the ISO mandatory LCIA steps where emissions are sorted into groups that have an impact on a midpoint indicator. Figure 5.9 lists LCI data of different elemental flows and then shows arrows grouping the emissions to the corresponding impact categories. Often, an emission can impact more than one category (Table 5.3).

Impact category	Scale	Examples of LCI data (i.e., classification)
Global warming	Global	Carbon dioxide , nitrous oxide , methane , chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), methyl bromide
Stratospheric ozone depletion	Global	Chlorofluorocarbons (CFCs), hydrochlorofluorocarbons (HCFCs), halons, methyl bromide
Acidification	Regional, local	Sulfur oxides , nitrogen oxides , hydrochloric acid (HCl), hydrofluoric acid (HF), ammonia
Eutrophication	Local	Phosphate , nitrogen oxide (NO), nitrogen dioxide , nitrates, ammonia
Photochemical smog	Local	Non-methane hydrocarbon (NMHC)
Terrestrial toxicity	Local	Toxic chemicals with a reported lethal concentration to rodents
Aquatic toxicity	Local	Toxic chemicals with a reported lethal concentration to fish
Human health	Global, regional, local	Total releases to air, water, and soil
Resource depletion	Global, regional, local	Quantity of minerals used, quantity of fossil fuels used
Land use	Global, regional, local	Quantity disposed off in a landfill or other land modifications
Water use	Regional, local	Water used or consumed

5.3.2.12 Characterization

The characterization step relates the emission flow to the potential impact in the impact category in a common unit relating multiple flows to a reference flow. Environmental models are used to determine the potential impact each flow has to the corresponding impact category. One common example of this is the relationship between global warming gases of carbon dioxide and methane. The impact of 1 kg of methane, according to the Intergovernmental Panel on Climate change 2013, is 34 times greater than 1 kg of carbon dioxide over a 100-year analytical time horizon. This value is calculated using models to predict the warming effect of greenhouse gasses based on insulating capacity of each gas as well as gas degradation patterns. Table 5.4 provides a list of contributors to GWP and the characterization factors for both a 20-year time horizon and a 100-year time horizon. Note the characterization factor is listed in kg CO₂ equivalents per kg of substance. The kg of the emission times the characterization factor provides the equivalent emission in CO₂ to the emission of the other compound. These and many other emissions have a different potential environmental impact and characterization factor, depending on the time in which they are considered, or analytical time horizon. The characterization factors are often standardized for a set impact assessment method such as Ecoinvent or the TRACI Method. Both of these methods are further discussed in the following sections.

5.3.2.13 Optional Elements

Normalization

Normalization is an optimal step of LCIA according to the ISO standards; however, it can provide some additional insights into the relevancy of certain emissions. Since midpoint indicators represent different measure and use different units, relating the quantity of a midpoint indicator to another midpoint indicator can be difficult. Furthermore, it is not always possible to know how significant a 50 kg of CO₂ eq. emitted is compared to 50 kg H⁺ equivalents. You might ask is 50 kg of CO₂ and H⁺ huge emission in comparison to what is currently being emitted in the world?

Relating midpoint indicators to emissions of an exterior reference system is referred to as external normalization. For instance, dividing the emission of 50 kg CO₂ by the CO₂ emissions of a human in the United States over the course of a year allows a comparison of the current system to the total impact that a person may have in a year. This external normalization, defined as impact of study scenario divided by an external reference value in the same units, provides a unitless value that can be further examined in the next optional steps of impact assessment. External normalization is commonly used; however, it has a major weakness. As an example, when the CO₂ emissions of a product are normalized by the CO₂ emissions of a US citizen over a year, the impacts of the product are often trivial in comparison to the yearly total. This yearly total emission that is used for normalization does not have any relation to what the earth or natural environment can sustainably accommodate; rather it is the current and often unstainable level of emissions. By using often large and unsustainable emissions number to normalize midpoint indicators, the influence of certain impact categories is often diminished in the final single score. In this example, 50 kg CO₂ divided by 24,000 kg (emissions of a US citizen for CO₂ per year,) yields 0.021, somewhat trivial and small number (Table 5.5).

	US normalization factors reference year 2008
Impact category	Annual (impact per year)	Per capita (impact per person year)
Ecotoxicity – nonmetals (CTUe)	2.2E+10	7.6E+1
Ecotoxicity – metals (CTUe)	3.3E+12	1.1E+04
Carcinogens – nonmetals (CTUcanc.)	1.7E+03	5.5E−06
Carcinogens – metals (CTUcanc.)	1.4E+04	4.5E−05
Noncarcinogens – nonmetals (CTUnoncanc.)	1.1E+04	3.7E−05
Noncarcinogens – metals (CTUcanc.)	3.1E+05	1.0E−03
Global warming (kg eq)	7.4E+12	2.4E+04
Ozone depletion (kg CFC-11 eq)	4.9E+07	1.6E−01
Acidification (kg eq)	2.8E+10	9.1E+01
Eutrophication (kg N eq)	6.6E+09	2.2E+01
Photochemical ozone formation (kg eq)	4.2E+11	1.4E+03
Respiratory effects (kg eq)	7.4E+09	2.4E+01
Fossil fuel depletion (MJ surplus)	5.3E+12	1.7E+04

Another method of normalization for comparative LCA studies is internal normalization. In such a comparative study, the midpoint indicators between an option A and option B are divided by scenario with the highest value for an individual impact. For instance, if product A produced 20 kg eq. of CO₂ and option B produced 50 kg CO₂ eq., scenario A values would be divided by option B, 20 kg/50 kg. In a comparative study, this method has some advantages as this goal is often to provide information that leads to making a decision that will produce the lowest impacts. Since this goal is based on deciding between one of multiple options, the scenarios can be normalized between each other.

Weighting

Weighting is a subjective methodology where the relative importance of impact categories is determined. This can be useful in simplifying the results of an LCA especially when there are trade-offs between scenarios. For instance, in the production of a paper product two bleaching processes are used and compared in an LCA; however, option A has lower impacts in three categories while option B has lower impacts in four categories. It is often unclear how to compare GWP to acidification potential midpoint indicators. Through weighting normalized midpoint indicators can provide further insights that can be useful in decision making and can be used in a single environmental score.

Weighting values are often determined by stakeholder groups involved in the project, such as a company commissioning the LCA study. Alternatively, standardized weighting values established in other studies can be used. One example of standardized weighting values is the Eco-indicator 99 impact assessment method. For the Eco-indicator 99 method (Ministry of Housing, 2000), a panel of 365 people was asked to rank the importance of ecosystems health, resource use, and human health. The results from this Swiss LCA interest group panel indicated that the human health and ecosystems quality were twice as important as recourse use (Table 5.6). Figure 5.10 shows the wide range of responses of this panel, which further illustrates the subjective nature of weighting. United States-based weighting factors were also determined through survey panel by Gloria et al. (2007).

Impact category	Mean (%)	Rounded (%)	St. deviation (%)	Median (%)
Human health	36	40	19	33
Ecosystem quality	43	40	20	33
Resources	21	20	14	23

Since the act of weighting is subjective, various scientists have sought to create methods that reduce the subject nature of a single score. One method employed by Daystar et al. (2016)) tested the scenario outcomes using 16 different weighting methods established by LCA experts, product producers, and product users. In some results, the weighting factor can play a major role in influencing the results; however, in Daystar et al. (2016), the results were generally the same when different weighting methods were applied. A more robust approach to weighting and single score is the stochastic multiattribute analysis (SMAA) (Prado-Lopez et al., 2014) where all possible weighting factors are used in combination with internal normalization and uncertainty data. This helps provide insight into whether the differences are significant between options as well as the probability of one option resulting in a more favorable outcome than another based on the range of weighting factors tested.

Single Score

A single environmental score result is calculated using both the normalized midpoint results and weighting methods as described by the following equation. Communicating the single score results has the advantage of one data point that is easy to compare across multiple scenario option; however, this one value is inherently subjective and has more uncertainty than midpoint indicators. When communicating results in a single score, it is necessary to also provide midpoint indicator score and properly document the normalization and weighting factors used to determine the single score. Study transparency is a critical to the integrity of any LCA study and requires much documentation of these and other methods and data sources.

5.3.2.14 Life Cycle Impact Assessment Interpretation

The impact assessment methods and impact categories selected in the goal and scope can shape the study conclusions. Omitting or including different impact categories is one way in which the study can be drastically altered. Additionally, there are times in which the study set out to examine a wide list of impacts, but available data are not sufficient to calculate one or several of the impact categories. It is important to again interpret the results from each of the LCA steps and to realign new insights with the goal and scope.

With regard to impact assessment, it is important to evaluate if the LCIA results provide enough data to make decisions or to take action. At times, there can be two options that present environmental trade-offs or where one options will have higher values for some impacts and lower values for others. In these scenarios, LCIA left at a midpoint indicator can do little to help a decision maker. Additional analysis such as single scores or an SMAA can be performed to provide further guidance into the most advisable action or decision (Prado-Lopez et al., 2014).

5.3.2.15 Problems Set –Life-Cycle Impact Assessment