Abstract
Static life cycle assessment (LCA) methodologies fail to consider the temporal profiles of system inputs and outputs (including emission timing), such that they underestimate the benefits of temporarily stored biogenic carbon in bioproducts, such as cotton. This research focuses on greenhouse gas emission timing and applies dynamic emission accounting to the life cycle of cotton woven pants. The significance of temporary biogenic carbon storage and emission timing is illustrated by converting the 2017 Cotton Incorporated static LCA to a dynamic model using the Dynamic Carbon Footprinter (baseline scenario). A reduction in cumulative radiative forcing for dynamic relative to static modeling of 22%, 5%, and 2% are observed at 10-years, 30-years, and 100-years, respectively. Alternative scenarios analyzed include converting cotton woven pants at end of life to bioenergy, to compost, or to building insulation, an alternative cotton production scenario using regenerative agricultural practices, and two pants extended lifetime scenarios. The regenerative agricultural practice scenario provides reductions in cumulative impacts compared to the baseline scenario of 96%, 69%, and 105% after 10, 30, and 100-years, respectively. A 3x extension in the lifetime of pants provides a benefit in reduced cumulative impacts of 31%, 40%, and 41%, after 10, 30, and 100-years, respectively. This case study with cotton demonstrates that dynamic LCA is a useful tool for assessing the benefits of biobased products, and it allows for more nuanced analysis of reductions in climate impacts in both the short- and long-term time horizons.
Download PDF
Full Article
Evaluating Cotton Apparel with Dynamic Life Cycle Assessment: The Climate Benefits of Temporary Biogenic Carbon Storage
Steven T. Pires,a,b,* Allan Williams,c Jesse Daystar,b William Joe Sagues,d Kai Lan,a and Richard A. Vendittia,*
Static life cycle assessment (LCA) methodologies fail to consider the temporal profiles of system inputs and outputs (including emission timing), such that they underestimate the benefits of temporarily stored biogenic carbon in bioproducts, such as cotton. This research focuses on greenhouse gas emission timing and applies dynamic emission accounting to the life cycle of cotton woven pants. The significance of temporary biogenic carbon storage and emission timing is illustrated by converting the 2017 Cotton Incorporated static LCA to a dynamic model using the Dynamic Carbon Footprinter (baseline scenario). A reduction in cumulative radiative forcing for dynamic relative to static modeling of 22%, 5%, and 2% are observed at 10-years, 30-years, and 100-years, respectively. Alternative scenarios analyzed include converting cotton woven pants at end of life to bioenergy, to compost, or to building insulation, an alternative cotton production scenario using regenerative agricultural practices, and two pants extended lifetime scenarios. The regenerative agricultural practice scenario provides reductions in cumulative impacts compared to the baseline scenario of 96%, 69%, and 105% after 10, 30, and 100-years, respectively. A 3x extension in the lifetime of pants provides a benefit in reduced cumulative impacts of 31%, 40%, and 41%, after 10, 30, and 100-years, respectively. This case study with cotton demonstrates that dynamic LCA is a useful tool for assessing the benefits of biobased products, and it allows for more nuanced analysis of reductions in climate impacts in both the short- and long-term time horizons.
DOI: 10.15376/biores.19.3.5074-5095
Keywords: Biogenic carbon; Dynamic life cycle assessment; Climate change; Cotton sustainability; Circular economy; Regenerative agriculture; Textile sustainability
Contact information: a: North Carolina State University, Department of Forest Biomaterials, 2820 Faucette Dr., Raleigh, NC 27607, USA; b: Cotton Incorporated, 6399 Weston Parkway, Cary, NC 27513, USA; c: Cotton Research and Development Corporation, 2 Lloyd Street, Narrabri, NSW 2390 Australia; d: North Carolina State University, Department of Biological & Agricultural Engineering, 3110 Faucette Dr., Raleigh, NC 27607, USA; *Corresponding authors: richardv@ncsu.edu; spires@cottoninc.com
INTRODUCTION
Life cycle assessment (LCA) is an environmental accounting tool that has been widely used to identify hotspots and evaluate the impacts of products and services across their life cycle. The first ever published LCA was a cradle-to-gate assessment conducted by Coca-Cola in 1969 (Bauman and Tillman 2004). The original use case for the tool was centered on facility assessments. Over five decades later, LCA has been applied to a wide range of product systems including agricultural commodities and other bioproducts with much broader system boundaries and complex physical processes to model. For LCAs applied to these natural systems, the set of assumptions and methodological decisions used can dramatically influence the results. Some key methodological decisions include defining the system boundaries, functional unit selection, co-product allocation, defining the temporal scope, life cycle inventory data collection approach, and accounting for biogenic carbon uptake and emission timing.
Temporal boundary selection (e.g., the timeframe used to evaluate the impacts) is critical, since society is focused on reducing our overall climate impacts in the near term. For greenhouse gas reporting, using the concept of global warming potential (GWP) at the 100-year time horizon (GWP-100) has become the standard within the scientific community. The Intergovernmental Panel on Climate Change (IPCC) developed the concept of GWP, which can be calculated for any gas (i) using Eq. 1 (Myhre et al. 2013a),
where the Absolute Global Warming Potential (AGWP) of greenhouse gas (i) is given by the integral of the Radiative Forcing (RF) of a pulse emission of gas (i) at a specific point in time (t) relative to the integral of the RF of a 1 kg pulse emission of CO2 at time (t) and is provided in Watt years per square meter per kilogram of gas (W yr m-2 kg-1). The RF of a given gas (i) is calculated by Eq. 2 (Lan and Yao 2022),
where Ai is the radiative efficiency and Ci is the amount of gas left in the atmosphere after the initial pulse emission.
The early adoption of the 100-year GWP by the Kyoto Protocol cemented its use in the scientific community and by other industry stakeholders (Breidenich et al. 1998; Shine 2009). However, both the 100-year time horizon and the concept of GWP have sparked debate in the literature. Kleinberg (2020) argued that using GWP to evaluate climate change impacts is “unphysical, unintuitive, arbitrary, ignores the time dependence of emission sources, and is in some cases misleading”. They argue that beyond a 20-year time horizon, the GWP measure becomes meaningless. Lynch et al. (2020) indicate the GWP-100 metric falls short in accurately representing the contrasting impacts of short- and long-lived climate pollutants, leading to ambiguities in its relationship with actual global warming. With many organizations and nations striving to reach their climate commitments by 2050, these short-term differences in individual greenhouse gas species radiative efficiency become even more important (Lynch et al. 2020).
A significant methodological decision in the LCA process is determining how to account for the timing of climate-changing emissions within the study. Traditional static LCAs sum the global warming impacts (GWIs) occurring over a specified timeframe, typically 100-years, and model them as if they all occur in the first year (Daystar et al. 2017). This static LCA approach causes inconsistent temporal boundaries, which can distort the real-world implications of the GHG emissions occurring over the chosen analytical time horizon (Levasseur et al. 2013; Daystar et al. 2017). This inconsistency ignores the actual short-term reduction in cumulative radiative forcing due to temporarily stored biogenic carbon, which is shown to be important in bioproducts products with longer lifespans (Levasseur et al. 2013; Daystar et al. 2017). Durable cotton textiles provide longer-term biogenic carbon storage as well, the benefits of which have gone unrealized under existing LCAs applying static GHG accounting approaches. Further, when applying static LCA methodologies, emissions occurring later in the time horizon (e.g., landfill emissions following garment disposal after an 8-year life span) are incorrectly accounted for within a 100-year (or other) time horizon as all emissions are assumed to occur at year 1.
A more comprehensive approach to evaluating the benefits of biogenic carbon uptake, storage, and emission timing is a system known as dynamic life cycle assessment (Levasseur et al. 2010, 2013). This concept was originally developed by Levasseur et al. (2010) where dynamic characterization factors (DCF) are used to evaluate the radiative forcing impact of a GHG through time, given by Eq. 3,
where the dynamic characterization factor at time t is given in Watts per year per square meter (W/yr/m2). The instantaneous radiative forcing per unit of mass increase in the atmosphere, a, in Watts per square meter per kilogram (W/m2/kg) is then multiplied by the atmospheric load over of the GHG at time t after the emission occurs (kg). The DCFs are then used to calculate the immediate climate impact of the greenhouse gasses considered, known as the Instantaneous Global Warming Impact (GWIinst), as shown by Eq. 4:
In this example, the expression provides the GWIinst at time t for carbon dioxide and methane; however, the expression would expand to match the number of GHGs being evaluated. The authors’ example equation sums the products of emission factors for CO2, gCO2(i), and CH4, gCH4(i), with their respective DCFs: DCFCO2(t-1) and DCFCH4(t-1) from time 0 to time t. The dynamic GHG inventory result (the sum of the positive and negative emissions) for year i (in kg) is represented by the expression gCO2(i) and gCH4(i), for carbon dioxide and methane, respectively, in this example. These results are used to evaluate the cumulative impact the gasses have on the atmosphere (GWIcum) over the specified time horizon by using Eq. 5,
where the aggregate of all the instantaneous global warming impacts (GWIinst) is analyzed for each preceding year up to the current time t.
While there are many life cycle assessment studies focused on textiles, the majority of these apply traditional static LCA methodologies and do not consider emission timing. Liu et al. 2024 evaluates the greenhouse gas emissions across the life cycle of t-shirts made from cotton, polyester, and viscose. Their research finds that yarn manufacturing and product use phases are the primary contributors to the carbon footprint due to the associated energy consumption. The study highlights the significant role of carbon sequestration in plant-derived fibers such as cotton and viscose, which become more beneficial as the product’s service life increases (e.g., longer carbon storage). The study recommends technological innovation in production, greater reliance on renewable energy, and encouraging consumers to prolong the life cycle of products as ways to reduce climate impacts of textile products. However, while the research alludes to the potential benefits of storing carbon in clothing such as cotton, it does not directly quantify this metric.
Biogenic carbon uptake and emission timing has been an ongoing focus of the building and forest products sectors due to their relatively long-term carbon sequestration potential compared to many other products. Hoxha et al. (2020) explore this concept in detail in their review of various LCA accounting methods for evaluating biogenic carbon uptake and emissions in an energy-efficient timber building. The authors compared traditional static LCA approaches against three alternative biogenic carbon accounting methods, including a carbon neutrality approach (e.g., the 0/0 approach), the -1/+1 approach, and a dynamic modelling approach with two scenarios: carbon uptake by the forest before extraction and carbon uptake by the forest after extraction. Results showed substantial differences in the global warming values from each method, ranging from 16% when evaluating the building as a whole and up to a 200% difference when evaluating individual building components. Overall, the study concluded that the dynamic approach provided the most reliable and understandable results of the methods evaluated.
It is important to acknowledge there are many other methods for addressing biogenic carbon uptake and emission timing in LCA such as the Lashof method, GWPbio, Müller-Wenk and Brandão, and the carbon footprint method (Brandão et al. 2018; Liu et al. 2023), which each offer unique ways to account for the temporal aspects of greenhouse gas emissions. These methods vary in their application and rely on the use of carbon dioxide equivalents, which discount the unique physical impacts of each individual GHG species as previously discussed. Using a dynamic approach better evaluates the environmental impacts of products by accounting for the timing of emissions within a consistent temporal scope (Daystar et al. 2017). Applying dynamic life cycle assessment methodologies to cotton textiles can allow for a better understanding of the real-world climate benefits associated with a reduction in radiative forcing over a defined period of time. Through converting the Cotton Incorporated LCA from traditional nondynamic accounting to dynamic accounting, it is shown in this work how important this reduction in radiative forcing can be, even at the garment level. This work evaluates the benefits of stored biogenic carbon in textiles at different time horizons by highlighting different levers for change across the cotton supply chain.
EXPERIMENTAL
This paper leverages existing data from Cotton Incorporated’s LCA (Cotton Incorporated 2017) and applies a dynamic GHG emission timing profile to the life cycle of cotton woven pants to contrast the static emission model with various scenarios across 10, 30, and 100-year time horizons. The global cotton LCA provides full cradle to grave impacts of cotton garments (t-shirt, knit polo shirt, and casual woven pants) at each production stage including production (from seed to finished ginned bale), manufacturing (including cut and sew and transportation), consumer use (including washing, drying, ironing), and product end of life (incineration and landfilling) (Cotton Incorporated 2017; Daystar et al. 2019). To better understand the discrete impacts and timing of emissions across a cotton textile products’ life cycle, the Dynamic Carbon Footprinter (Levasseur 2010) was utilized with a focus on woven pants. The functional unit for the cotton production phase in the Cotton Incorporated LCA is 1,000 kg of ginned cotton lint manufactured into 2,049 pairs of casual woven pants. The use phase assumes use impacts (washing, drying, duration of service, etc.) and end of life (EOL) impacts following approximately 4 years of use (Cotton Incorporated 2017; Daystar et al. 2019).
Six scenarios were established to illustrate the importance of emission timing, biogenic carbon uptake, and enhanced sustainability practices across the cotton supply chain: 1) dynamic life cycle assessment approach, 2) regenerative cotton production, 3) long term carbon storage at product end-of-life, 4) garment composting at product end-of-life, 5) combustion with energy recovery, and 6) extended product lifetime. The majority of the climate impacts in Cotton Incorporated’s global cotton LCA are driven by carbon dioxide, methane, and nitrous oxide and thus are the focus of this work and the dynamic modeling approach.
Cotton Life Cycle Assessment Data Disaggregation
The first scenario takes the data from the static Cotton Incorporated LCA model (Cotton Incorporated 2017), disaggregates the data, and converts the results into a dynamic emission model. Results of the Cotton Incorporated LCA are provided in carbon dioxide equivalents, which is a common industry practice. However, this is not ideal for dynamic modeling. In dynamic LCA modeling, both the timing of emissions and the individual GHG species is important as each individual GHG species (e.g., carbon dioxide, methane, nitrous oxide, etc.) provides varying radiative efficiencies and global warming potentials (Myhre et al. 2013a). In dynamic LCA modeling, understanding the timing of emissions associated with each GHG species is critical. Working with LCA experts from Cotton Incorporated, disaggregated GHG emissions from the original LCA model were exported and utilized to determine the quantity of individual emissions at each production stage. The disaggregated emission data and relative impacts at each life cycle phase is provided in Table 1. The table also provides the timing of emissions associated with each life cycle phase.
Table 1. Woven Pants Disaggregated GHG Emission Values Based on (Cotton Incorporated 2017) with Emission Timing Assumptions
Assumptions were made from each production stage listed in Table 1 to provide the duration of storage and emission timing associated with each stage. Those data provided the foundation for the dynamic LCA model used as the baseline to compare subsequent modeling scenarios.
Static and Dynamic LCA Modeling Approach
Static and dynamic GHG inventories with manufacturing and use phase emissions
The dynamic greenhouse gas inventory for both the dynamic and static LCA is provided in Table 2. In the dynamic LCA inventory, Year 1 emissions are associated with cotton production. Because cotton production naturally sequesters carbon from the atmosphere via the process of photosynthesis, a negative emission is reported. Year 2 includes the impacts associated with woven pants manufacturing, including cutting, sewing, and transportation from the gin to manufacturing facility and ultimate point of sale. Years 3 through 6 include use phase impacts including washing, drying, ironing, etc. for the assumed use duration of four years (Cotton Incorporated 2017; Daystar et al. 2019). Year 7 includes the first year of end-of-life impacts associated with landfill, incineration, and litter impacts. Year 8 includes the second and final year of impacts associated with typical end of life pathways. Garment end-of-life impacts were based on the 2017 CI LCA, which estimated 45.7% of garments were incinerated, 46.8% went sent the landfill, and 7.5% were littered or entered a “wild landfill” situation (Cotton Incorporated 2017).
Table 2. Dynamic and Static LCA GHG Emission Inventory
Static and dynamic GHG inventories without manufacturing and use phase emissions
To better illustrate the benefits of temporarily storing biogenic carbon in cotton products, an alternative baseline scenario was created that removes the manufacturing and use phase emissions from the inventory. Because all textile products have manufacturing and use phase impacts, which are assumed to be similar – these emissions have been removed from selected scenario analysis. This allows for a clearer demonstration of how the temporarily stored biogenic carbon affects the overall reduction of radiative forcing. Manufacturing and use phase impacts are important but fall outside the scope of focus for the current work. Additionally, manufacturing and use phase impacts span across all fiber types, whereas biogenic carbon storage is a unique element of natural fibers, such as cotton. Table 2 also provides the greenhouse gas emission inventories for the baseline scenarios without manufacturing and use phase impacts included. The numbers and associated letters in the following summary tables indicate the year and life cycle stage (e.g., 1P, 2M, 3U, etc.) and are defined in more detail in the notes section of Tables 2, 3, and 4.
Scenario Modelling
This section outlines the scenarios considered in the current work. These scenarios provide optimistic pathways that illustrate the climate change benefits associated with aspects of regenerative cotton production, alternative end of life scenarios for woven pants, textile recycling, improved product lifespan, and textile to bioenergy pathways. As previously mentioned, manufacturing and use phase emissions have been removed from the following scenario analysis to allow for a clearer understanding of the benefits of biogenic carbon uptake in each scenario. A summary of GHG emissions associated with each scenario is provided in Table 3.
Table 3. GHG Emission Inventories for Alternative Production and End of Life Scenarios
Regenerative cotton production (Climate Smart Cotton)
This scenario examines the garment level implications of the U.S. Climate Smart Cotton Program, which encourages the adoption of regenerative agricultural practices, such as cover cropping, nutrient management, and no-till farming, in cotton production. This scenario assumes that all cotton being produced for the associated 2,049 casual woven pants from Cotton Incorporated (2017) fully adopted these regenerative agricultural practices. Emission reduction assumptions for the practices were based off the project narrative from the U.S. Climate Smart Cotton Program proposal (U.S. Cotton Trust Protocol 2023). The project narrative estimates a reduction of 0.32 metric tons (MT) of CO2 equivalents (CO2e) for the adoption of cover cropping, 0.19 MT CO2e reduction for no-till adoption, and 0.08 MT CO2e for nutrient management. To translate these benefits to the same functional unit evaluated in this study, a weighted average yield of 824 lb cotton lint per acre was calculated from the Cotton Incorporated LCA. Based on the weighted average yield from that study, 2.7 acres were required to produce 1000 kg of fiber. When these practice changes are applied to 2.7 acres, the result is a 0.8 kg reduction of N2O emissions from nutrient management, 507 kg of CO2 emissions from no-till, and an 854 kg reduction of CO2 emissions from cover cropping. In the dynamic LCA inventory for this scenario it was assumed that emission reductions for nutrient management were from reduced N2O emissions. Cover cropping and no-till adoption were assumed to reduce CO2 emissions.
Long term sequestration via recycled cotton-based building insulation
This scenario assumes that all the cotton woven pants are recycled into a cotton-based insulation material at their end of life. It is assumed that this insulation is used to insulate new construction residential homes, providing long term biogenic carbon sequestration. To calculate the emissions associated with the cotton-based insulation manufacturing, a screening level life cycle inventory was completed and modeled off a commercial recycling facility in Arizona. From this data, 1000 kg of woven pants are converted to 778 kg of cotton-based insulation while producing emissions of 884 kg CO2 and less than 0.005 kg of N2O and CH4. Conversely, to generate the same level of insulation provided by the cotton-based insulation, 453 kg of fiberglass insulation would be required, generating 933 kg CO2. A credit was taken under this scenario where the cotton-based insulation displaces the production of fiberglass, thus providing a negative 49 kg CO2 benefit. Petroleum-based fibers (i.e., polyethylene) used as binders in traditional fiberglass do not contain biogenic carbon and thus do not provide any net removals or storage of carbon from the atmosphere. Any emissions taking place with building end of life are assumed to occur after the 100-year timeframe considered in the study. Thus, the end-of-life emissions typically encountered in the baseline scenario are avoided under this alternative scenario.
Garment composting at product end-of-life
In this scenario, end-of-life emission impacts are reduced by composting the waste cotton textiles. Emission reductions for composting cotton textiles in this scenario are based on emission factors provided by Nordahl et al. (2023) for wet yard waste. For simplicity, the authors assume cotton woven pants will have similar compost emission profiles to wet yard waste and would offset that feedstock to provide a similar carbon to nitrogen (C:N) ratio in the resulting compost recipe. Emission factors from Nordahl et al. (2023) used are as follows: methane (CH4) at 2.06E-03 kg/kg, nitrous oxide (N2O) at 4.54E-05 kg/kg, and carbon dioxide (CO2) at 1.71E-01 kg/kg. The 2,049 woven pants in the Cotton Incorporated LCA are all assumed to be diverted from typical end of life scenarios to the composting environment, resulting in a reduction in end-of-life emissions.
Combustion with energy recovery at product end-of-life
This scenario uses model results from Aspen, in which waste cotton (assumed to be pure cellulose) is burned for bioenergy production. The values generated for bioenergy in this scenario are utilized to offset the average grid emission values in the United States provided by the US Environmental Protection Agency (US EPA) in their “Emission Factors for Greenhouse Gas Inventories” summary document (US EPA 2014). US EPA eGRID data indicates the US average grid energy is 79% non-renewables and 21% renewables (including 1.2% bioenergy). The results from the Aspen model show that the combustion of one ton of cotton biomass for bioenergy production results in the emission of approximately 1,620 kg CO2. Concurrently, this process generates about 0.97 Megawatt-hours (MWh) of energy. In this scenario the conventional end-of-life emissions are replaced by the positive emissions of 1,620 kg CO2 minus the grid emission factor credit of 542 kg CO2 associated with the production of the 0.97 MWh created by cotton bioenergy. Negative emission values of 0.01 kg of methane and nitrous oxide are taken as an avoided emission from typical US average grid energy production. These credits are used in the end-of-life emission reductions associated with the benefits of this scenario.
Table 4. GHG Emission Inventories for Extended Lifetime Scenarios (Including Manufacturing and Use Phase Emissions)