|
Life Cycle Assessment of Biogas from Separated slurry 2 Scope
2.1 MethodologyThe methodology used in this study is Life Cycle Assessment, and is standardised by international standards ISO 14044 (ISO, 2006a) and ISO 14040 (ISO, 2006b). This methodology assesses, for the whole life-cycle of a given product or service, the environmental impacts generated as a consequence to this product or service. According to the ISO standards 14044 and 14040 (ISO, 2006a; ISO, 2006b), a LCA is an iterative methodology including four main phases:
The present study is a comparative study; all four biogas scenarios will be compared to the reference scenario established and described by Wesnæs et al. (2009). The primary data for the technologies in this study are delivered by the producers of these technologies. Background data are from the Ecoinvent database (v2.0) as it is acknowledged that this is the most reliable and high-quality database for Life Cycle Inventory data, providing transparent, independent and consistent data for a large variety of processes. The Ecoinvent database structure and supporting material is described in more details by Frischknecht et al. (2007). Access to the Ecoinvent database requires a licence (Ecoinvent Centre, 2009). The modelling has been carried out by the use of the Simapro 7.1 LCA software. Details about the software can be found in PRé Consultants (2009). The use of this software also requires a license. 2.2 System overview: reference scenarios and alternatives2.2.1 Reference Scenario (Scenario A)The reference scenario used in this study reflects the conventional slurry management practices for both fattening pig slurry and dairy cow slurry, i.e. the slurry is used as an organic fertiliser and is spread to land without any prior treatment. For both fattening pig and dairy cow slurry, the reference scenario can be summarised as the following three main stages:
The in-housing slurry composition, which is the very basis for the various changes slurry undergoes according to the alternative studied, is described in the reference scenario (Annex A of Wesnæs et al., 2009 for an comprehensive description; Annex A of this study for a summarised description). The reference scenario used in the present study is the same as extensively described in Wesnæs et al. (2009), unless when otherwise specified. Some of the key processes described in Wesnæs et al. (2009) for the reference scenario are also described in Annex A of the present report. All the alternative biogas treatments investigated in this study will be compared to this reference scenario. 2.2.2 Biogas from raw pig slurry and fibre fraction from chemical-mechanical separation (Scenario F)This scenario considers the production of biogas with the two following inputs:
These fractions do not necessarily come from the same farm (and most probably they do not), but they both end up at the biogas plant. Once at the biogas plant, these fractions are mixed according to their composition and to their degradability in order to achieve realistic production conditions. This scenario shall be seen as a scenario including “best available technologies” for biogas production. This scenario can be summarised with the 4 following processes:
This biogas scenario is extensively described in Annex F of this report, including all mass balances, assumptions and detailed calculations. 2.2.3 Biogas from raw cow slurry and fibre fraction from chemical-mechanical separation (Scenario G)This scenario considers the production of biogas with the two following inputs:
This scenario shall be seen as a scenario including “best available technologies” for biogas production. This scenario can be summarised by the exact same processes as described in section 2.2.2 (however, the slurry origins from dairy cows instead of fattening pigs). An extensive description of all mass balances, assumptions and calculations involved in this scenario is presented in Annex G of the present report. 2.2.4 Biogas from raw pig slurry and fibre fraction from mechanical separation (Scenario H)This scenario considers the production of biogas from the two following inputs:
The mechanical separation considered is the screw press separation technology extensively described in Annex C of Wesnæs et al. (2009). This scenario can be summarised with the 4 following processes:
This biogas scenario is extensively described in Annex H of this report, including all mass balances, assumptions and detailed calculations. 2.2.5 Biogas from raw slurry and processed fibre pellets (Scenario I)This scenario considers the production of biogas from the two following inputs:
The fibre pellets process is the same as extensively described in Annex D of Wesnæs et al. (2009). This scenario can be summarised with the 4 following processes:
This biogas scenario is extensively described in Annex I of this report, including all mass balances, assumptions and detailed calculations. 2.2.6 Overview of the 4 alternative biogas scenarios and of the reference scenarioSome similarities can be noted between the biogas scenarios described in section 2.2.2 to 2.2.5. Scenario F and G are in fact mostly identical, the only difference being that scenario F assesses fattening pig slurry and scenario G assesses dairy cow slurry. The particularity of these two scenarios is that they use a chemical-mechanical separation process prior to the biogas production, which is anticipated to produce a fibre fraction rich in VS and thereby suitable for biogas production. These scenarios also involved a second separation after the biogas production, which is expected to produce a fibre fraction rich in phosphorus. Scenario H differs from scenario F as this second separation is not performed and as the first separation does not involve the use of polymer. Scenario I, as opposed to scenarios H and F, does not use fibre fraction as an input for biogas production but fibre pellets (produced as a result of further processing of the mechanical fraction). Figure 2.1 schematised the 4 alternative biogas scenarios and the reference scenario. Figure 2.1. Simplified illustration of the reference scenario and the 4 alternative biogas scenarios considered 2.3 Consequential approachAs in the first publication of this LCA foundation for slurry management technologies, the modelling approach adopted in this LCA study is the consequential LCA approach. Comprehensive details about this methodology can be found in Wenzel (1998), Ekvall and Weidema (2004) and Weidema (2004). The consequential modelling approach is set up in order to ensure that the results reflect the environmental consequences of implementing a given technology, product or service as compared to the implementation of a given reference scenario. Since the present study aims to highlight the environmental consequences of implementing different biogas technologies for slurry management instead of the conventional slurry management practices (described in section 2.2.1), the consequential approach was the most appropriate. In the consequential approach, system delimitation requires to include marginal data only (instead of averaged data). In fact, the approach considers that the interactions, i.e the changes in demand, created on the global market as a result of the implementation of a given scenario is the very starting point for the resulting environmental consequences. This means that the marginal processes and activities involved in the system assessed must be identified (i.e. those affected by a change of demand). This can be done in accordance with the methodology and principles described by Weidema (2003) and by Ekvall and Weidema (2004), which can be summarised by the following general principles:
In the case of the present study, this means that the marginal processes need to be identified for: electricity production, heat production and fertiliser-type (for N, P and K fertilisers), among others. This was performed in Wesnæs et al. (2009). Table 2.1 present the marginal electricity, heat and inorganic fertilisers (N, P and K) used in this study. For heat and electricity, sensitivity analysis was performed with other processes, as elaborated in section 8 of this report. Table 2.1. Marginal electricity, heat and inorganic fertilisers (N, P and K) used in this study
Moreover, the consequential approach ensures system equivalency through system expansion (thereby avoiding any allocation). This is in conformity with the ISO standards (ISO, 2006a; ISO 2006b) which state that “whenever possible, allocation should be avoided”. Interactions from secondary services (i.e. those arising together with the studied service) on the global market are included in the model so the full consequences of the system assessed can be reflected by the model results. As stated in Wesnæs et al. (2009), equivalence on all primary and secondary services is ensured in the consequential approach by identifying and including the displacements of alternative products that will occur when choosing one alternative over the other. 2.4 Basis for the comparison: The functional unitIn order to make a reliable comparison between the different alternatives, it has to be ensured that all alternatives are comparable in terms of the main services provided to society. In order to do so, a functional unit has to be defined (ISO 2006a; ISO 2006b), which shall reflects all services provided. According to ISO (2006a), a functional unit provides “a reference to which the input and output data are normalized”. In the present study, the functional unit was defined as in Wesnæs et al. (2009), i.e. “Managing 1000 kg slurry”. All inputs and outputs must then be linked to this functional unit through a reference flow. In the present study, the same reference flow as in Wesnæs et al. (2009) is used, i.e.: “1000 kg slurry ex-animal”. The functional unit and reference flow are the same as in Wesnæs et al. (2009) as the reference scenario (against which all the studied biogas alternatives will be compared) was defined in Wesnæs et al. (2009). This also ensures consistency and comparability of the various studies forming this life cycle foundation of slurry management technologies. 2.5 System boundariesAs stated in the ISO standards (ISO, 2006a), the system boundary determines which unit process shall be included in the LCA. The system boundary fixed on this study is consistent with the one fixed in Wesnæs et al. (2009). As explained in Wesnæs et al. (2009), the purpose of this study is comparative, i.e. different alternative biogas technologies are compared to a reference scenario. Because the interest of this study lies in the differences between the different alternatives and the reference scenario, the processes common to all the alternative technologies and the reference scenario will not be included within the system boundary. Similarly, all the processes irrelevant for answering the research question of this study (i.e. What are the environmental benefits and disadvantages of introducing slurry management technology X?) were also excluded from the system boundary. Excluded processes are as described in Wesnæs et al. (2009):
Included within the system boundary are all processes related to slurry handling: e.g. slurry storage (in-house, pre-tank, outdoor storage), slurry treatment (separation, biogas production), electricity needed for slurry handling (pumping, stirring, separation, biogas production), transport needed (for slurry transport from farm to biogas plant, or from outdoor storage to fields) and fertilisation operations (slurry application and slurry fate in the soil). One innovative aspect of this study is the inclusion of carbon sequestration in the modelling and the consideration of biogenic CO2 emissions. This was also included in Wesnæs et al. (2009). Biogenic carbon is generally neglected in LCA and considered as “neutral” (e.g. Hansen et al., 2006). Yet, Denmark is committed to include carbon changes in cultivated areas in connection to the Kyoto Protocol (Fødevareministeriet, 2008). Moreover, acknowledging that the carbon in the manure comes from feed (i.e. the portion of C ingested through the feed that was not absorbed by the animal and thereby excreted) and acknowledging the environmental importance of the feed production, it appears important to distinguish between the amount of that C that is returned to atmosphere (as CO2 and CH4) and sequestrated in the soils and to account for it. This is the only way to account for the benefits of a slurry management allowing greater C sequestration. This cannot be performed if all biogenic C is ignored and considered as “neutral”. According to Thomson et al. (2009), the assumption of valuing biogenic carbon or not has substantial consequences in environmental assessments, as it can be a crucial factor in determining if the greenhouse gas balance is positive or negative. Moreover, the biogeochemical C cycle is closely related to the N cycle (Nieder and Benbi, 2008), and this interdependence involves interactions that are ignored if the biogenic C is not taken into account. The optimal range of C and N in soil is rather narrow, and so are crops yields below these optimal ranges (Nieder and Benbi, 2008). However, above these ranges, emissions of reactive C and N compounds occur through both the atmosphere (CO2, N2O and NOX) and as discharge to waters (dissolved N and C) (Nieder and Benbi, 2008). In this study, the C/N ratios of the organic fertilisers that are applied to the land differ as a consequence of the different technologies. Therefore, the resulting environmental consequences of this must be reflected through the inclusion of all biogenic C flows in the model. As explained in Wesnæs et al. (2009), a reference crop rotation for both the fattening pigs farm and the dairy cows farm has been established in order to estimate the ammonia emissions in the period after application in the field (for liquid fractions). However, the life cycle of these crops is not included within the system boundary (e.g. sowing and harvesting operations, tillage, management of the crop residues, etc.), as this is not a consequence of the slurry management. In the case of the scenarios affecting the crop yield, the system was expanded in order to reflect the consequences of an increased yield. The consequences regarding extended pig or dairy cow production has not been included. According to Danish law, the introduction of separation with high efficiency will allow the farmer to increase the production, i.e. to have more pigs for the same area of land. With the consequential approach that mean, that the extra pig production should be included, and that pig production somewhere else should be subtracted. Introduction of new technology will not make the consumers eat more pig meat, and therefore, the total production of pigs in the world will not be affected, instead, the least competitive pig producer (“the marginal pig producer”) somewhere will have to close down the production and the marked share is lost. It has not been possible to include this aspect in the study. Conformingly with the approach used by Wesnæs et al. (2009), all emissions and flows with significant environmental impacts have to be included in a life cycle assessment. In case of lack of data, estimates have been made rather than leaving gaps. These estimates were then thoroughly justified in the life cycle inventory annexes. Furthermore, as stated in Wesnæs et al. (2009), all processes “behind” the processes directly assessed are included, e.g. production of diesel for the tractor, extraction of oil and refinery for production of the diesel, production of the tractor itself, production of mineral fertilisers and production of chemicals for these, extraction of minerals for production of these chemicals, electricity needed for this production, etc. The system “behind” the product chain for slurry management is in fact tremendous and comprises hundreds of processes. The inclusion of these processes “behind” is notably eased by the use of the Ecoinvent database, in combination with a LCA software. The slurry management alternatives investigated in this project involve complex processes exhibiting a high degree of spatial and temporal variability, and this particularly applies for field processes. Yet, life cycle assessment, as defined in the ISO standards (ISO 2006a; ISO 2006b), is not a methodology capable neither suitable for the modelling of dynamic processes. In fact, when performing LCA, dynamic data must be translated into a set of discrete values that are carefully chosen in order to represent the system assessed as accurately as possible. Such “translations” were performed as transparently as possible and all assumptions taken in this context were justified thoroughly. 2.6 Temporal, geographical and technological coverageThe temporal, geographical and technological coverage considered in this study is as in Wesnæs et al. (2009). Therefore, data from the most recent years (for which consistent data were available) were used. It is the intention that data used for this study applies for 2008 and 5-7 years ahead. As some of the alternative technologies represented in this study are fairly new, it is likely that ongoing product development will improve these technologies during the next decade. This study covers slurry management under Danish conditions (e.g. housing systems, storage facilities, soil types, application methods, energy production and legislation regarding fertilisation and nutrient substitution). Furthermore, the slurry composition varies significantly within the European countries due to differences in on-farm management, e.g. for feeding (Weidema et al., 2008). Accordingly, it is not possible to transfer the results of this study directly to other European countries without adjustments. For the reference scenario, the technological coverage is based on “average technology” and represents the “state of the year 2008”. The intended technology level for the alternative technologies is “Best available technology” (BAT), as these technologies are representing the future technologies. 2.7 Environmental Impacts and ResourcesThe impact assessment phase of the life cycle assessment methodology described in section 2.1 consists to relate the substances flow inventory to environmental impact potentials. The ISO standards for life cycle assessment (ISO 2006a; ISO 2006b) requires to clearly specify the impacts categories used for the assessment as well as the characterisation model upon which they are modelled. The environmental impact categories chosen in this study are the same as used and described in the first publication of this LCA foundation for slurry management technologies (Wesnæs et al., 2009). These are primarily based on the Danish EDIP method. Not all impact categories from the EDIP method were included, as shown in table 2.2. All the impacts categories included in this study are indicators, i.e. indicators for impacts on human beings and nature. For example, global warming (climate change) is an environmental concern in itself (mid-point); however, the larger concern is usually the subsequent damages to humans, animals and plants (end-point). Global warming has many impacts, for example drought in some areas, extreme weather conditions, flooding and rising sea levels in other areas, all having potential impact on crop yields and availability of food for humans. The Life Cycle methodology is a general approach focussing on the potential contributions of substances and emissions from the systems assessed to the environmental impacts, and not the actual environmental impacts. This is explained in more details by Wenzel et al. (1997). Accordingly, it is not within the frame of the LCA methodology to include site specific considerations of e.g. nature being particularly sensitive to specific emissions like e.g. ammonia. This is in accordance with both the ISO standards for Life Cycle Assessment (ISO, 2006a; ISO, 2006b) and international consensus, acknowledging that it is in practice impossible to know all reception sites of the various emissions to the environment and all actual exposure pathways of the emitted substances. From the EDIP method, the following categories have been included:
A few categories have been added to the EDIP method:
An attempt to include odour as a separate impact indicator category has been made. Odour emissions, the largest public concern in animal production areas (Blanes-Vidal et al., 2009), is the result of a large number of volatiles compounds. As most of these compounds are by-products of the decomposition of animal slurry (Blanes-Vidal et al., 2009), the slurry management system can have an important influence on the odour emissions. In the present study, technologies acknowledged to have a positive effect on odours (e.g. biogas production and separation) are involved. However, the inclusion of odour in LCA is not simple, and no methodology to include odour in LCA exists. The definition of where the odour measurements should be taken can be discussed. It is probably more the neighbours of the farm that are affected (or bothered) by the odour than the farmer, but the outdoor emissions from housing units to a great degree depend on the distance to the neighbours, the number of animals in the housing units, wind, temperature etc. Furthermore, the odour problem is not “mathematically linear” – an odour of 100*106 OUE for 5 days might be worse than an odour of 500*106 OUE for 1 day. The area where the odour is distributed is very significant, too. Moreover, it has been extremely difficult to find data for odour that can be related to “1000 kg slurry” especially for cattle slurry. The high uncertainty and variability of available data related to odour presents a challenge too. It has thus been decided not to include quantitative data on odour for this study, and odour is not included as an impact category in this study. However, as for the first part of this LCA foundation for slurry management, the database has been prepared for including odour at a later stage. As mentioned in section 2.5, the production of medicine was not included in the system boundary, as it is a process occurring prior to slurry excretion. Yet, the fate of these medicine residues that will end up in the slurry is likely to be influenced by the slurry management technology considered. However, it has not been possible to find adequate quantitative data on these aspects; thus, they were not included. The database for the LCA foundation was however prepared in order to facilitate their inclusions when more information will be available for this. This also applies for possible biological contamination of the slurry as a result of animal diseases, for example. Some of the separation technologies considered in this study involve the use of cationic polyacrylamide (PAM), a polymer highly resistant to biodegradation. As not enough information was available to assess the exact fate of this polymer in the environment, the “accumulation of polymer in the environment” was considered as a discussion point only rather than as an impact category. Table 2.2 presents the different impacts categories considered in this study as well as the methodology used to model each impact. In order to ensure transparency, the impacts categories not considered in this study are also mentioned and their omission is justified. Table 2.2. Included and excluded impact categories.
Table 2.3 shows the main emissions that contribute to the impact assessment categories mentioned in table 2.1. According to Sleeswijk et al. (2008), for LCA environmental impact categories not related with toxicity, 10 main contributors can be highlighted: CO2, CH4, SO2, NOX, NH3, PM10, NMVOC, (H)CFCs emissions to air as well as emissions of N- and P-compounds to fresh water. Nine of these were in fact inventoried in the present study. In the case of slurry management, one additional major contributor may be added to the list of Sleeswijk et al. (2008), namely N2O to air, which is a particularly important contributor to the impact category “global warming”. The emissions in table 2.3 have been included for all the “foreground processes” as far as possible (i.e. for all the processes regarding slurry management for which data have been collected in this study). The “background processes” from the Ecoinvent database contains far more emissions than these. Table 2.3. emissions for the “foreground processes” in this study.
[1] Methane is a volatile organic compound, but due to its exceptionally long lifetime, a distinction is often made between methane and others VOC (called NMVOC, which stands for non methane volatile organic compounds).
|