Description of the system under study
The case study used for testing the different methods was packaging made from polyethylene (PE) based on biomass. Specifically, polyethylene produced from wood (from Swedish boreal forest) and from sugarcane (from Brazil) was considered (see Fig. 1). In total, three PE production routes were included: (1) two different hypothetical routes from wood to PE, via fermentation to ethanol followed by dehydration to ethylene and via gasification yielding syngas that is converted to methanol and subsequently to ethylene and (2) one existing route from sugarcane to PE, via fermentation to ethanol followed by dehydration to ethylene. The end-of-life scenario applied was incineration, i.e. complete oxidation of the PE packaging. All three routes were investigated in a series of preceding papers (Liptow and Tillman 2012, Liptow et al. (2015), Liptow et al. (2013)). These three studies included inventory data for biogenic CO2 emissions, that is, CO2 emissions originating from biomass, and were used as a database to quantify emissions of fossil and biogenic CO2 in the life cycles of each case. In addition, data for the case of the fossil based PE production route were used for comparison. The selected assessment methods were then applied to these inventory data.
Selection and description of assessment methods
A literature review was done on methods for the assessment of global warming due to land use and biogenic CO2 emissions in LCA. There are different indicators to express the impact on global warming of biogenic carbon (see, e.g. Ericsson et al. (2013)); the review however considered methods using global warming potential (GWP) only because of its ubiquitous use compared to other indicators such as global temperature potential (GTP). Furthermore, methods that are theoretically related but not directly applicable in an LCA context (e.g. Bernier and Paré (2013), Holtsmark (2012), Zanchi et al. (2012)) were outside the scope of this study. Such methods do not provide characterization factors for the impact due to biogenic carbon flows, for instance, but rather focus on the variation over time of these flows. This can be considered as a part of the inventory analysis in an LCA but not as part of an impact assessment method. Also, this study only considered the attributional assessment of global warming, and therefore, methods focusing on effects of indirect land use change (iLUC) (e.g. Kløverpris and Mueller (2013), Schmidt et al. (2015)) were outside the scope of this study.
Three carbon cycle methods (see Introduction), the GWPbio method (Cherubini et al. 2011a), the GWPnetbio method (Pingoud et al. 2012) and the WF method (Väisänen et al. 2012), were selected. Of these three methods, two (GWPbio and GWPnetbio methods) calculate characterization factors (CFs) for assessing the climate impact of biogenic CO2 flows. Like for any other CF, these CFs are applied to inventory flows, in this particular case to the biogenic CO2 flows per tonne of PE produced (which is the functional unit used in the LCA case study (see Sect. 2.1)). In contrast, the WF method calculates a weighting factor (WF) which is applied to the inventoried amount of biogenic CO2 before applying the CFs commonly used for fossil CO2 in the life cycle impact assessment. The three selected methods are now described in more detail.
This method (Cherubini et al. 2011a) is currently one of the most discussed impact assessment methods (see, e.g. Holtsmark (2012) and Cintas et al. (2017)). Similar to the already used GWP characterization factors (IPCC 2013), the GWPbio CF represents the cumulative radiative forcing of a greenhouse gas relative to the forcing of a pulse of CO2 and thus also is a relative measure of the potential effects of that greenhouse gas on the climate. However, in contrast to the commonly used GWP CFs, a CF calculated with the GWPbio method explicitly considers the timing of biogenic CO2 emissions and uptake during a rotation period and thus is time-dependent. A GWPbio CF is derived by assuming that (1) the CO2 that is taken up due to biomass growth is forward looking. This means that CO2 uptake is considered to occur during the re-growth of the biomass after harvest, rather than during the growth before harvest and (2) the CO2 uptake by re-growing biomass is considered at a single stand level. This means that the CO2 emitted from an oxidizing product (in the case study used (see Sect. 2.1), incineration of PE produced from wood or sugarcane) is assumed to be taken up by re-growth occurring on the same stand from which the biomass was originally harvested. A GWPbio CF is calculated with the following three steps:
The change in atmospheric CO2 concentration over time, due to the emission and uptake of CO2, is calculated. This is done by combining the difference of emission and uptake of CO2 with a function that represents the decay of CO2 in the atmosphere via the removal of CO2 by the ocean and terrestrial biosphere sinks (Cherubini et al. 2011b).
The change in this concentration is then used to calculate the absolute global warming potential (AGWPbio) for the time horizon under assessment. AGWPbio is defined as the cumulative radiative forcing of the biogenic CO2 and measures the absolute potential effect the biogenic CO2 has on the climate. It is calculated in two steps: (a) the change in atmospheric concentration of CO2 is multiplied with the radiative efficiency of CO2, and (b) this product is then integrated over the time horizon that is applied in the method (e.g. 100 years).
The GWPbio CF is then calculated by dividing the AGWPbio with the AGWP of a CO2 pulse, and thus becomes a relative measure.
This method (Pingoud et al. 2012) is similar to the GWPbio method and is based on the same core assumptions regarding CO2 uptake. In addition, it considers the potential impact from lost uptake, i.e. the uptake if the trees had been left standing, and had continued to grow and take up CO2 from the atmosphere. This continued growth may be interpreted as relaxation to a ‘natural’ state, and the absence thereof could thus be interpreted as a kind of land occupation impact, as done by Helin et al. (2014). Furthermore, the method considers avoided burdens, which are displaced fossil emissions if the same product had been produced from a fossil resource.
CF is the sum of two factors, namely, (a) the GWPbio CF, which considers the change in carbon stocks in the forest stand due to biomass harvest and re-growth, as well as lost uptake (note that this is not the same GWPbio as defined by the GWPbio method by Cherubini et al. (2011a)), and (b) the GWPbiouse CF, which considers the abovementioned avoided burdens and the CO2 emissions from the biomass-based product and their timing (in the case study used (see Sect. 2.1), incineration of PE produced from wood or sugarcane). Both factors are determined via the same procedure already described for Cherubini’s GWPbio CF. As pointed out by Helin et al. (2013), the modelling of avoided burdens is usually done in the inventory phase of LCA, but in this method, it is included in the impact assessment.
The underlying core assumptions of the WF (weighting factor) method (Väisänen et al. 2012) with respect to CO2 uptake are similar to those of the GWPbio method, but it models the uptake with a simple linear function. Furthermore, Väisänen et al. (2012) do not explicitly take into account the timing of emissions but instead assume that all CO2 emissions occur as a pulse at the time of harvest. This is considered as a carbon debt, which is paid back by the re-growing biomass. The WF is determined by calculating the proportion of the total carbon uptake due to biomass re-growth at every time step, summing up these proportions over the assessment period considered, and finally dividing the sum by the assessment period (e.g. 100 years). Based on the determined WF, the fraction of carbon released into the atmosphere is calculated. Finally, the average CO2 emission due to biomass harvest is derived by multiplying the amount of C (as CO2) harvested with the determined WF. While the GWPbio and GWPnetbio methods determine a characterization factor, the WF method thus provides an inventory factor that is multiplied with the inventoried amount of biogenic CO2. The resulting amounts of CO2, together with the CO2 emissions during other life cycle stages such as transport and harvesting, are then multiplied by the established CO2 GWP characterization factor (IPCC 2013).
Two land use methods (see Introduction), the GWPsoil method (Brandão et al. 2011) and the CRP (Climate Regulation Potential) method (Müller-Wenk and Brandão 2010), were selected. The CRP method aims at assessing climate impacts due to land use by considering carbon transfer between soil and vegetation and the air. The GWPsoil method only includes the changes in soil organic carbon due to land use in the calculation of the impact thereof.
The two selected methods are in accordance with the framework for land use impact assessment within LCA, described by Milà (Milà et al. 2007). The framework focuses on impacts due to changes in land quality on the natural environment (biodiversity, ecological soil quality) and on natural resources (biotic production, climate regulation, substance cycling and buffer capacity). It divides land use into land transformation (sometimes referred to as land use change), i.e. the change of land area according to its dedicated purpose, and land occupation, i.e. the use of this land area without any further transformation. In order to do an assessment according to the framework, the area (how much land is used), the time (duration of occupation and relaxation processes) and the land quality must be determined (Fig. 2). Furthermore, the reference state needs to be decided on. Milà et al. (2007) suggest that if the LCA is attributional, the naturally relaxed state is an adequate reference state (i.e. the state the land will return to if not being used any longer) (indicated with Qrel in Fig. 2), while if the LCA is consequential, then changes in land use with respect to an alternative system are considered and this alternative system may be used as the reference state. This means that in the case of an attributional LCA, the method defines the reference state, whereas in the case of a consequential LCA, the analyst needs to determine this state. The recently published UNEP-SETAC guideline on land use impact assessment on biodiversity and ecosystem services also uses the key elements of this framework (Koellner et al. 2013). The two selected land use methods are now described in more detail.
This method includes changes in soil organic carbon due to land use in the assessment of the global warming potential and clearly differentiates between the impact from land occupation and transformation. The method was presented as part of a case study that compared the cultivation of four energy crops (Brandão et al. 2011). In this study, all land transformations were allocated to the 100 years of subsequent occupation, while assuming that the transformation took place more than 100 years ago. Thus, the land transformation impacts no longer played a role in the assessment. As a consequence, a description on how to calculate the impact due to land transformation with the GWPsoil method is missing. Regarding the impact of land occupation, the carbon stock change during occupation is directly calculated from the carbon sequestration rate (whether positive or negative) in soil (via the carbon uptake by plants) during 1 year of a particular land occupation and is expressed in tonnes of C. In order to assess the related climate impact, this amount of C is converted to tonnes of CO2,eq using the characterization factor 3.67 t CO2,eq/t C. The latter is the stoichiometric conversion factor from carbon to CO2, which is also used for fossil based carbon. This implies that the method equals biogenic carbon from soil with carbon of fossil origin.
Climate regulation potential (CRP) can be defined as the foregone sequestration of carbon due to land use, i.e. carbon that is not stored, compared to a reference land use (Milà et al. 2013). The CRP method takes into account changes in below and above ground carbon stocks (i.e. organic carbon in soil and in vegetation) due to both land occupation and transformation (Müller-Wenk and Brandão 2010). These changes are determined with respect to a reference state, which might either be the historical natural (Qhis in Fig. 2) or a future relaxed state (Qrel in Fig. 2). The method provides no guidance in this respect. The method calculates characterization factors (CFs) for land occupation and transformation, expressed as t CO2,eq/ha year and t CO2,eq/ha, respectively. These CFs are calculated by multiplying the emissions of biogenic carbon per surface area due to changes in the carbon stocks with a so-called duration factor. The main difference between determining the CFs for occupation and transformation is how this duration factor is calculated. It is defined as the ratio of the average stay of biogenic carbon in air and the average stay of fossil carbon in air, which is assumed to be 157 years (calculated with an arbitrary time horizon of 500 years). In the case of land transformation, the average stay of biogenic carbon in air is determined as 50% of the relaxation time (i.e. the time it takes for the land to arrive at the reference state), because “the mean carbon stay in air is approximately the average between zero years and the number of years required for complete relaxation” (Müller-Wenk and Brandão 2010).
In the case of land occupation, the average stay of carbon in air reflects the delay of relaxation (due to the occupation of the land) by 1 year and is thus set to be 1 year. The resulting duration factor thus is 1/157 for all types of occupation. Next, the CFs are multiplied with the inventory flows of land occupation, expressed in ha year, and transformation, expressed in ha. Finally, The CRP due to land use is calculated as the sum of the CRPs due to land occupation and land transformation. It should be noted that the CRP method does not explicitly address the issue of amortizing land transformation impacts to a period of land occupation but that amortization is needed when applying the method.
Analysis of selected assessment methods
The methods were analysed based on a framework developed by Helin et al. (2013) which can be summarized with the following five questions:
Does the method use a reference situation?
Does the method account for potential timing differences between emission release and uptake?
Does the method consider all carbon pools (above and below ground) related to the biomass system?
Does the method account for temporary carbon storage in biomass-based products?
Does the method consider product substitution effects, that is, does the method consider avoided environmental burdens due to the replacement of fossil based products?
Helin et al. (2013) stated an additional sixth question, which investigates the type of indicator used to express the climate impact of GHG emissions. However, this question was outside the scope of this study (see also Sect. 2.2). Table 1 gives a comparison of the carbon cycle and land use methods that are discussed above, using these five questions.
In addition to addressing these five method-oriented questions in the analysis, practical aspects of applying the methods were also tested. These aspects included data availability, acceptance of the method and ease of application.
Life cycle inventory
Table 2 presents the key inventory data used for testing the impact assessment methods, based on data from Liptow and Tillman (2012), Liptow et al. (2013) and Liptow et al. (2015). These inventory flows are also depicted in Fig. 1 (and their naming is given in Table 2). The life cycle inventory data were derived using several major modelling choices and assumptions. First of all, an attributional LCA approach was used. Next, all biogenic carbon emissions released during biomass harvest and production, and all carbon emissions related to land use, were accounted for and were assessed as CO2. In the case of sugarcane, which is cultivated in Brazil, pre-harvest burning was included, and in the case of wood, which originates from managed boreal forest in Sweden, the inventory for its acquisition was modified to stem wood using data from Berg and Lindholm (2005). For comparison, the CO2 emissions from a conventional fossil production route are also shown in Table 2. Furthermore, emissions from incineration of the PE (the method of disposal) were included and, for reasons of simplification, completely allocated to the PE.
As required by the tested methods, additional data were collected, e.g. on carbon stocks in soil and vegetation. These data are summarized in Table 3.
Data and modelling choices and assumptions
Impact assessment of sugarcane PE
For the impact assessment of the sugarcane route, several additional assumptions and data were used, with regard to the system under analysis (see Fig. 1) and the assessment methods, respectively, and are described here.
It usually takes 12 to 18 months before a sugarcane plant can be harvested for the first time after planting (de Carvalho et al. 2004). Subsequently, the sugarcane plant is cut (harvested) every year for the next 4 years before a new one is planted (de Carvalho et al. 2004). However, for reasons of simplification, we here assumed a harvest every year, and we assumed that the carbon bound in standing sugarcane biomass before harvest is 20 t C/ha (estimate based on de Carvalho et al. (2004); similar numbers are also presented by de Figueiredo et al. (2010)) (indicated as Cfeedstock in Fig. 1). The harvest is followed by the re-growth of sugarcane, which starts immediately after the harvest and which captures an amount of carbon equal to the amount of carbon harvested plus the amount of carbon burned during pre-harvest operations (CO2,regrowth in Fig. 1). For reasons of simplification, harvest and incineration of PE (oxidation of the biomass-based product, leading to biogenic CO2 emissions (indicated as CO2,burn,bio in Fig. 1)) were assumed occur at the same time since PE packaging has a very short life span.
In the GWPbio and GWPnetbio methods, the carbon content in the standing biomass is considered. The re-growth of the biomass was modelled using a probability density function in these methods, following Cherubini et al. (2011b). In calculating the GWPnetbio CF, also the replaced alternative production of the PE from a fossil feedstock is considered. For this, data for fossil CO2 emissions were used as presented in Table 2 (the related carbon flow is not depicted in Fig. 1). All three carbon cycle methods investigated include the re-growth of the biomass after harvest (indicated as CO2,regrowth in Fig. 1). It was assumed that the full relaxation takes 50 years in total. However, only the re-evolution (relaxation) from sugarcane cultivation to Cerrado in its first year was considered. Obeying these two assumptions and using a linear function to model the relaxation, the lost uptake during the first year of occupation is negligibly small (the lost uptake flow is not indicated in Fig. 1). For the WF method, a complete uptake occurs within 1 year after harvest and represents the re-growth of sugar cane during one rotation period.
In the GWPsoil method, only belowground carbon flows (soil organic carbon, indicated as Csoil in Fig. 1) are considered, whereas in the CRP method, also aboveground carbon flows (standing biomass, indicated as Cfeedstock in Fig. 1) are considered (see Sect. 2.2). Both methods used land transformation as an inventory parameter (see Table 2). This flow was calculated based on the approach given by Milà et al. (2013). Data about land use in Brazil from 1992 until 2011 were taken from the FAO statistics website (Food and Agricultural Organization of the United Nations 2014) to assess whether land transformation should be considered, and to calculate the land transformation flow for sugarcane. These flows (expressed in ha/t PE) are different for the two methods because they do not consider the same carbon flows (see Table 2). Both methods also used land occupation as an inventory parameter (see Table 2). In the case of the CRP method, this flow was calculated based on the above- and belowground carbon flows per surface area due to harvesting and the amount of carbon needed to produce 1 t of PE (the functional unit in this study, see Fig. 1). For the calculation of the GWPsoil, the carbon flow during land occupation was based on the carbon sequestration rate during sugarcane cultivation. This rate was assumed to be 0.29 t C/ha year (Anderson-Teixeira et al. 2009). The occupation and transformation flows were then multiplied with their respective characterization factors (see Sect. 2.2). Finally, we calculated the total impact due to the carbon flows as the sum of the impact during land occupation and the impact of land transformation amortized over 20 years.
Impact assessment of wood PE
For the wood based routes, several additional assumptions and data were used as well during the modelling and calculations with regard to forest growth and the assessment methods, respectively, and are described here.
The forest is modelled as an even-aged boreal forest stand that is clear-cut harvested, followed by immediate re-vegetation with the same species. The rotation period of the stand is 100 years, during which the forest captures an amount of carbon equal to the amount of carbon harvested (see Fig. 1). For reasons of simplicity, harvest and incineration of PE (oxidation of the biomass-based product, leading to biogenic CO2 emissions (indicated as CO2,burn,bio in Fig. 1)) occur at the same time because PE packaging has a very short life time. For the assessment of the wood PE with the GWPbio and GWPnetbio methods, a forestry growth model based on the Schnute growth function and its derivative was used. This equation represents an S-curve describing forest growth (see Cherubini et al. (2011a) for more details).
Cherubini et al.’s (2011a) forestry model was also used to derive data for the lost uptake to calculate the GWPnetbio CF. For this purpose, the model was applied for a stand age of 100 to 200 years. This describes the further growth of a mature forest that is not cut down and thus represents the lost uptake (this process and the associated carbon flow is not depicted in Fig. 1). For the GWPnetbio CF calculation, data on fossil CO2 release were used, as presented in Table 2, in order to calculate the avoided burdens of displaced fossil PE production (this flow is not shown in Fig. 1) (see Sect. 2.2). For the assessment with the WF method, growth was modelled by assuming a linear uptake of carbon by the re-growing biomass over 100 years which is a simplification when compared to the models used in the GWPbio and GWPnetbio methods for this uptake.
For the CRP and the GWPsoil methods, it was assumed that land transformation from natural to managed forest took place long ago and that its impact is no longer of relevance (and therefore transformation of the natural to the managed forest is not depicted in Fig. 1). The land occupation flows were calculated based on the aboveground forest growth rate (1.1 t C/ha year, see Table 3 and indicated as Cregrowth in Fig. 1) and the amount of carbon needed to produce 1 t of PE. Furthermore, it was assumed that there is no change in the soil organic carbon content when transforming the natural forest into a managed forest and that there is no further change in this carbon stock if the forest is managed sustainably (Chen et al. 2010, De Simon et al. 2012).