Life Cycle Inventory & Assessment Report:Source Separation of Dairy Cattle Manure, Finland

Baltic Forum for Innovative Technologies for Sustainable Manure Management KNOWLEDGE REPORT Life Cycle Inventory & Assessment Report:Source Separation of Dairy Cattle Manure, Finland By Juha Grönroos, Katri Rankinen, José E. Cano-Bernal, Lauri Larvus and Laura Alakukku WP5 Assessing Sustainability of Manure Technology Chains December 2013

Baltic Manure WP5 Assessing Sustainability of Manure Technology Chains Life Cycle Inventory & Assessment Report: Source Separation of Dairy Cattle Manure, Finland By Juha Grönroos, Katri Rankinen and José E. Cano-Bernal Finnish Environment Institute SYKE Lauri Larvus and Laura Alakukku University of Helsinki 2013 1

Preface This report presents the inventory data, results and interpretation of a consequential life cycle assessment carried out for the technique Source Separation, as applied to dairy cattle manure, in the context of Finland. It was produced as part of work package 5 of the project Baltic Forum for Innovative Technologies for Sustainable Manure Management (Baltic Manure). The long-term strategic objective of the Baltic Manure project is to change the general perception of manure from a waste product to a resource, while also identifying its inherent business opportunities with the most suitable manure handling technologies and policy framework, for the Baltic Sea Regions (BSR). Baltic Manure is partly financed by the European Union (), through the Baltic Sea Region Programme 2007-2013. The report was performed and edited by Juha Grönroos and Katri Rankinen (Finnish Environment Institute SYKE), and Lauri Larvus and Laura Alakukku (University of Helsinki). Internal review for the inventory analysis of the LCA was performed by Marianne Wesnæs from the University of the Southern Denmark (SDU). December 2013 The authors 2

Table of Contents 1 Introduction... 5 1.1 Background and objective (overall Baltic Manure project)... 5 1.2 Scope and objectives of this report... 6 1.3 Organization & Participants... 6 2 Scope... 6 2.1 Methodology... 6 2.2 Background and objective... 7 2.3 Basis for the comparison: The functional unit... 7 2.4 System Boundaries... 7 2.4.1 System boundaries for the reference scenario... 8 2.4.2 System boundaries for the source separation scenario... 9 2.5 Temporal, geographical and technological coverage... 10 2.6 Data... 10 2.7 Impact categories... 11 3 Life Cycle Inventory data for Reference scenario... 11 4 Life Cycle Inventory data for Source Separation scenario... 14 4.1 Urine and faeces in the housing units... 14 4.1.1 NH 3 emissions... 14 4.1.2 CH 4 and CO 2 emissions... 14 4.1.3 N 2 O emissions... 14 4.1.4 Energy consumption... 14 4.1.5 Life cycle inventory data for the processes Urine in the housing units and Faeces in the housing units... 14 4.1.6 Mass balances for the processes Urine in the housing units and Faeces in the housing units... 16 4.2 Urine and Faeces in the outdoor storage... 17 4.2.1 NH 3 emissions... 17 4.2.2 CH 4 and CO 2 emissions... 17 4.2.3 N 2 O emissions... 18 4.2.4 Life cycle inventory data for the processes Outdoor storage of urine and Outdoor storage of faeces... 18 4.2.5 Mass balances for the processes Outdoor storage of urine and Outdoor storage of faeces... 19 4.3 Applying urine and faeces to the field... 20 4.3.1 CH 4 and CO 2 emissions... 20 4.3.2 Fertilizer substitution... 22 4.3.3 Nitrate leaching... 22 4.3.4 Phosphorus leaching... 23 4.3.5 Life cycle inventory data for the processes Urine application to field and Faeces application to field... 23 5 Life Cycle Assessment Results and Interpretation... 25 5.1 Life Cycle Assessment Results... 25 3

5.2 Discussion... 27 5.3 Conclusions... 28 6 References... 29 4

1 Introduction 1.1 Background and objective (overall Baltic Manure project) In 2009, the European Union Strategy for the Baltic Sea Region (EUSBSR), along with its Action Plan, was approved by the European Council, making it the first macro regional strategy in Europe. As part of the Action Plan, the Strategy promotes Flagships Projects which fall within the scope of the overall objectives of the Strategy, namely: Save the Sea, Connect the Region and Increase Prosperity. Baltic Manure, which involves 18 partners from eight BSR countries (Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland and Sweden), is one of these Flagship projects. The long-term strategic objective of the Baltic Manure project is to change the general perception of manure from a waste product to a resource, while also identifying its inherent business opportunities with the most suitable manure handling technologies and policy framework. The project is divided into 7 work packages: WP1: Project management and administration WP2: Communication WP3: Innovative technologies for manure handling WP4: Standardisation of manure types with focus on phosphorus WP5: Assessing sustainability of manure technology chains WP6: Energy potentials of manure WP7: Business innovation The results presented in this document are the outcome of WP5. The objectives of WP5 are twofold: To assess the environmental consequences of different manure management technology chains of relevance for the BSR in order to provide a support for prioritization of these technologies in the different BSR countries: To propose a common platform for Life Cycle Assessment (LCA) of manure management in the BSR. One key outcome expected from WP5 consists of the production of Life Cycle Inventory reports for selected manure processing technology chains that can be used as a support for policy instruments. As a result, a myriad of such reports were made for a selection of different combinations of manure processing technologies, manure types and BSR countries. An overview of the combinations assessed is available in WP5 final report. 5

1.2 Scope and objectives of this report This report presents the inventory data, results and interpretation of the life cycle assessment carried out for the technique Source Separation, as applied to dairy cattle manure, for the BSR country Finland. It aims to highlight, in a so-called whole-system perspective, the environmental consequences of using this manure management technology, as compared to the status-quo (or reference) manure management situation, where the dairy manure is simply stored (in-house and outdoor) and then applied to soil as an organic fertilizer as slurry. 1.3 Organization & Participants Baltic Manure is partly financed by the European Union (), through the Baltic Sea Region Programme 2007-2013. The project is led by MTT Agrifood Research (Finland), with a total budget of 3.7 million. This 3-y project started in 2011 and ended in 2013. The participants of WP5 include: Lorie Hamelin, Henrik Wenzel, Marianne Wesnæs & Henrik Saxe; University of Southern Denmark Juha Grönroos, Katri Rankinen & José E. Cano-Bernal; Finnish Environment Institute (SYKE) Andras Baky; Swedish Institute of Agricultural and Environmental Engineering (JTI) Sirli Pehme; Estonian University of Life Sciences Laura Alakukku & Lauri Larvus; University of Helsinki Ksawery Kuligowski, Dorota Skura, Marek Ziółkowski & Andrzej Tonderski; Pomeranian Centre for Environmental Research and Technology (POMCERT) More details about the Baltic Manure project and the overall participants can be found on the project website; www.balticmanure.eu. 2 Scope 2.1 Methodology This report is based on the Life Cycle Assessments method (LCA) described in the Danish EDIP method by Wenzel et al. (1997) and further updates of this method (Hauschild & Potting (2005), Weidema et al. (2004), Weidema (2004), Stranddorf et al. (2005)). The method used is based on the consequential LCA approach (CLCA). The purpose of the CLCA approach is to show the environmental consequences of the decision that is assessed by the LCA. The LCA shall reflect that choosing one alternative over another involve an increasing demand for that alternative and the environmental consequences of this choice; in this case the consequences of choosing source separation in the manure management chain as a replacement for the slurry management method. This is done through system expansion and the use of marginal data, striving to include only what is affected by a change in demand for the alternative technology. 6

The consequential approach requires that the LCA is comparative, i.e. that alternatives are compared. The consequential and comparative approach ensures that all compared alternatives are equivalent and provide the same services to society, not just regarding the primary service, which is the main function of the system, which is in this study management of manure from dairy cattle, but also on all secondary services. Secondary services are defined as products/services arising e.g. as co-products from processes in the studied systems. In this study, secondary functions are for example the nutrient value of the manure (that can replace mineral fertilizers). See further explanation of comparative and consequential LCA in Hamelin (2013), Wenzel (1998), Ekvall and Weidema (2004) and Weidema (2004). Internal peer review for the inventory analysis part of the LCA was performed by Marianne Wesnæs from the University of the Southern Denmark (SDU). 2.2 Background and objective Urine is rich in nitrogen, and faeces rich in phosphorus. If managed separately, urine can be used as a nitrogen fertilizer on the fields with high phosphorus content situating usually close to the animal housing. Faeces can be transported to the fields further away where phosphorus is needed. In a slurry system urine and faeces are mixed together but the slurry N:P ratio is not optimal with regard to the nutrient requirements of the plants. The objective of this LCA was to compare, in terms of life cycle environmental impacts, dairy cattle manure system with source separation of urine and faeces, to a reference system, dairy cattle slurry system. 2.3 Basis for the comparison: The functional unit In order to make a reasonable comparison it is fundamental to perform the LCA in relation to the same function, i.e. the same service i.e. the Functional Unit. The source separation scenario was compared to the reference scenario based on the functional unit 1000 kg of manure ex-animal, i.e. right after excretion. The composition of the reference manure (slurry) is further specified in the description of the reference system (Hamelin et al., 2013a). 2.4 System Boundaries In principle, an LCA covers all environmental impacts from all processes in the entire chain; however, when comparing alternatives, it is not necessary to include processes that are identical in the compared systems. In this study, focus was put on the differences, and the processes, that are identical for the reference scenarios and the alternative technologies were left out. Common for all the scenarios in this study are all the processes upstream of the slurry excretion, i.e. production of cows, production of feed, medicine, hormones, housing systems etc. In other words, the system starts when the manure leaves the animal and hits the floor in the housing system. Gaseous emissions (e.g. CH 4 through enteric fermentation or CO 2 through respiration) from the animals are not included within the system boundaries, as changed slurry management has no influence on the enteric fermentation and on the respiration. It is not claimed that the processes 7

upstream (i.e. before the manure excretion) have no environmental significance - it is just outside the frame of this study. In the source separation scenario, urine and faeces are kept separate after excretion, i.e. not mixed together as in the slurry system serving as the reference system. Included within the system boundary are all processes related to manure - urine and faeces - handling: e.g. manure storage (in-house, outdoor storage, separately for urine and faeces), possible manure treatment, energy needed for handling urine and faeces (pumping, stirring, loading, transport) and fertilization operations (application of urine and faeces, and their fates in the soil). A reference crop rotation was established in order to estimate the ammonia emissions in the period after application in the field. 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 manure management. In the LCA, biogenic CO 2 was included in the climate impact assessment. The reason for this was the aim to demonstrate how different manure management systems affect the fate of manure carbon (C) and the potential C sequestration capacity of them. If the biogenic C-loss - as CO 2 - in a manure management system is small, the potential C sequestration into soil after manure application is high, and vice versa. The more C is sequestered into soil for a long time period, the higher is the benefit from the climate point of view. This can be seen as a low climate impact value caused by the biogenic CO 2 emissions. 2.4.1 System boundaries for the reference scenario The reference scenario used in this study reflects the conventional manure management practices - the slurry system - for dairy cattle manure in Finland. The reference scenario can be summarized as the following three main stages: In-house storage, outdoor storage and application to field. Once excreted, the cow slurry is not stored in-house for a long time, but is removed once a day or continuously. On a regular basis, the slurry pits are emptied to an outdoor storage tank, made of concrete. It is assumed that the storage tank is covered with mixed roofing according to the estimated roofing ratios in Finland, based on the national ammonia emission model for agriculture (Grönroos et al., 2009): no cover 30% of slurry, tight roof (concrete) 20%, semi-tight roof (floating covers) 10%, natural crust 30%, tent roof 10%. Slurry remains in the storage tank until the suitable period for field fertilization. When suitable, the slurry is pumped from the storage tank, transported to the field and applied to the fields to be fertilized (see Figure 2.1). 8

Figure 2.1: System Boundaries for the reference system (dairy cattle slurry) 2.4.2 System boundaries for the source separation scenario The technique covered by this report is source separation of manure from dairy cows. The manure segregation takes place in the housing unit right after excretion. In the animal house, urine flows into a gutter and through a pipeline system to an outdoor storage, made of concrete. It is assumed that the storage tank is covered with mixed roofing according to the estimated roofing ratios in Finland, based on the national ammonia emission model for agriculture (like most of the information on manure management in Finland used in this study; Grönroos et al., 2009): no cover 25%, tight roof (concrete) 75%. This differs from the roofing ratios of slurry tanks in the reference scenario because covering of urine and slurry tanks differs in Finland. Urine remains in the storage tank until the suitable period for field fertilization. When suitable, urine is pumped from the storage tank, transported to the field and applied to the fields to be fertilized. See Figure 2.2. Faeces is removed by a scraper and stored in an outdoor storage made of concrete. It is assumed that 30% of solid manure storages are covered with a roof or similar cover to prevent rainwater entering the storage/manure (MMM/Tike, 2013). Additionally, part (30%) of solid manure is covered to reduce ammonia emissions with a separate cover or a natural crust which also has some ammonia reduction properties. Urine is applied on arable land (50%), on plant-covered land (25%), and on stubble (25%) using similar application methods as is used when cattle slurry (reference manure) is applied. Solid manure is applied on arable land or stubble (96%; of which no measures (10%), incorporation with ploughing <12 hrs (20%), >12 hrs (20%), incorporation with harrowing <4 hrs (10%), <12 hrs (20%), > 12 hrs (20%)), or on plant covered land (4%; broadcast spreading, no emission reduction measures). The system boundaries for the Source separation scenario are shown in Figure 2.2. In the basic scenario of the source separation, not all urine is managed separately. Ca. 23% of urine excreted is mixed with faeces and straw in animal house, while ca. 77% is collected 9

separately. In a sensitivity analysis, the effect of 100% segregation of urine on the final results was studied. The ratio 23%:77% is based on the amount of litter used for dairy cows in a barn where urine is separated (4 m 3 of peat/animal/year according to Iivonen 2008), and the liquid infiltration capacity of the litter material (for peat: 660 l/m 3 ). Figure 2.2: System Boundaries for the scenario source separation of dairy cattle manure. 2.5 Temporal, geographical and technological coverage The study was based on data from the most recent year for which consistent data are available. It is the intention, that data used for this study should apply for 2011 and 5-7 years ahead. The scenario in this report covers manure management under Finnish conditions (e.g. housing systems, storage facilities, soil types, application methods, energy production and legislation regarding fertilization and nutrient substitution). Furthermore, the manure composition varies significantly within the European countries due to differences in on-farm management, e.g. for feeding. Accordingly, it is not possible to transfer the results of this study directly to other European countries without adjustments. For the Source Separation scenario and the reference scenario, the technological coverage is based on average technology and represents the state of the year 2011. 2.6 Data The Finnish-specific manure management data for the reference system and source separation system were based on the expert opinions of the experts in MTT Agrifood Research Finland and University of Helsinki. Manure composition data ex-animal were based on the calculations of MTT Agrifood Research Finland (Nousiainen, 2013). Data for the source separation technique were mainly based on information acquired from the Viikki research farm of the Faculty of Agriculture and Forestry, University of Helsinki. Data on energy systems, energy and chemicals production and the technology used in the agricultural production systems were based on the Ecoinvent database (v2.2) (Frischknecht & Rebitzer, 2005). 10

The Finnish ammonia emission model for agriculture (Grönroos et al., 2009) was used to calculate ammonia emissions for the studied manure management systems. Emission factors for methane and nitrous oxide are based on the IPCC Guidelines for National Greenhouse Gas Inventories from the Intergovernmental Panel on Climate Change (IPCC, 2006). Emission factors for nitrogen monoxide and nitrogen is based on EMEP-EEA air pollutant emission inventory guidebook 2009 - Technical guidance to prepare national emission inventories (EMEP-EEA, 2009) from the European Environment Agency (and the 2010 update of this). When needed, these factors have been combined with data from various literatures; see the references at the end of the report. The Life Cycle Assessment was facilitated with the LCA software SimaPro 7.3.3. 2.7 Impact categories Four main impact categories are included: Global Warming, Acidification and Nutrient Enrichment (distinguishing between N and P being the limiting nutrient for growth), these being seen as the most relevant for agricultural biomass systems (see further explanation in the Main Report by Hamelin et al. (2013b)). 3 Life Cycle Inventory data for Reference scenario The Life Cycle Inventory data for the reference scenario are described in a report Reference life cycle assessment scenarios for manure management in the Baltic Sea regions - An assessment covering six animal production, five countries, and four manure types (Hamelin et al., 2013a). This section contains a summary only. The main preconditions for the reference system for dairy cattle, Finland are: A cubicle housing system with slatted floor. Manure removed by gravity to slurry storage. No pumping needed. Slurry channels are emptied to an outdoor storage tank, made of concrete. It is assumed that the storage tank is covered with mixed roofing according to the estimated roofing ratios in Finland. The transport distance from storage to application to fields is estimated to 5 km. The slurry is applied on the fields using the estimated ratios of the different application measures of pig slurry in Finland. The manure composition data are given in Table 3.1 below. A detailed description of the algorithms and assumptions behind the manure composition and mass balances are given in Hamelin et al. (2013a). Life Cycle Inventory data for the reference system are shown in Table 3.2. A detailed description of the algorithms and assumptions for these are given in Hamelin et al. (2013a). 11

Table 3.1: Manure composition for the reference system (dairy cattle slurry, Finland). All data per 1000 kg of slurry (at the respective manure stage, i.e. ex-animal, ex-housing or ex-outdoor storage). Manure stage Comments ex-animal exhousing ex-outdoor storage ex-animal ex-housing ex-outdoor storage Total mass (ton) 1 1 1 Mass balance a Mass balance b Dry matter (DM) (kg) 104.4 96.1 80.2 Nousiainen Mass balance c Mass balance c (2013) Ash content (kg) 20.9 19.3 18.0 DM minus VS DM minus VS DM minus VS Volatile solids (VS) (kg) 83.5 76.7 62.2 Same loss, in absolute, as DM Same loss, in absolute, as DM Based on an average of Finnish data, VS is 80% of TS for dairy slurry) Carbon (C ) (kg) 45.8 41.7 35.0 Mass balance d Mass balance d C:DM ratio is 28.1/64 for dairy slurry, based on Knudsen and Birkmose (2005) Total N (kg) 6.25 5.16 4.4 Nousiainen (2013) Mass balance e Mass balance e NH4 + -N (kg) 4.05 2.92 2.52 Based on N:NH4 + ratio in Viljavuuspalvelu, 2013 Phosphorus (P) (kg) 0.96 0.86 0.81 Nousiainen Mass balance f Mass balance f (2013) Potassium (K) (kg) 6.18 5.58 5.21 Nousiainen Mass balance g Mass balance g (2013) a Change in total mass during in-house storage: +3.95 kg added straw + 112.3 kg added water minus change in DM. b Change in total mass during outdoor storage: +80 kg added water minus change in DM. c Change in DM: + DM added by straw minus DM losses. DM in straw: 850 kg DM/ton straw (Møller et al., 2000). DM losses: 3.16 kg VS loss per kg CH 4 loss (3.16 gram VS loss per gram CH 4 is based on the Buswell equation (Symons and Buswell (1933), using the same principles as in chapter 5 of the Supporting Information of Hamelin (2013)). d Change in Total-C: C from added straw minus emissions of CO 2 -C and CH 4 -C. e Change in Total-N: N from straw (same as reference system) minus emissions of NH 3 -N, N 2 O-N, NO-N and N 2 -N (indirect emissions of N 2 O-N not included). N added by straw: 0.00528 kg N/kg dm: Møller et al. (2000). f P from added straw as in reference system. 0.0009 kg P/kg dm: Møller et al. (2000). g K from added straw as in reference system. 0.015 kg K/kg dm: Møller et al. (2000). 12

Table 3.2: Life Cycle Inventory data for the reference system, dairy cattle slurry, Finland. Life cycle stage Emissions In-house Outdoor storage Comments Field In-house Outdoor storage Field per 1000 kg manure ex-animal per 1000 kg manure ex-housing per 1000 kg manure ex-storage NH 3 -N 0.44 0.34 0.61 7% of N ex-animal (Grönroos et al., 2009) 7% of N ex-housing (Grönroos et al., 2009) 14% of N ex-storage (Grönroos et al., 2009) NH 3 -N, at application - - 0.01 0.5% of TAN applied (Hansen et al., 2008) N 2 O-N 0.0084 0.02 0.04 0.005 kg N 2 O-N per kg N ex-animal (IPCC, 2006), distributed 30% to in-housing and 70% to outdoor storage, see explanation in Hamelin et al. (2013a) NO-N (representing NO x ) 1.70E-04 1.27E-04 0.0044 0.0001 kg NO per kg TAN ex-animal (EMEP-EEA (2009), Table 3.9) NO 3 -N 0 0 0.20 Assumption: No leaching, as leakages from animal housing units are prohibited in Finland N 2 -N 0.0109 0.0082 0.003 kg NO per kg TAN ex-animal (EMEP-EEA (2009), Table 3.9) CO 2 -C 0.59 1.73 33.23 2.13 kg CO 2 /kg CH 4 for dairy slurry, based on Hamelin et al., 2011, but adjusted with updated Buswell, which accounts for protein VS 0.0001 kg NO per kg TAN ex-housing (EMEP-EEA (2009), Table 3.9) Assumption: No leaching, as leakages from animal housing units are prohibited in Finland 0.003 kg NO per kg TAN ex-housing (EMEP-EEA (2009), Table 3.9) 2.13 kg CO 2 /kg CH 4 for dairy slurry, based on Hamelin et al., 2011, but adjusted with updated Buswell, which accounts for protein VS CH 4 -C 0.13 2.24 0 IPCC (2006) algorithm, distributed 5% to in-house storage and 95% to outdoor storage, see explanation in Hamelin et al. (2013a) P leaching 0 0 0.0193 Assumption: No leaching, as leakages from animal housing units are prohibited in Finland Indirect N 2 O-N (due to emissions of NH 3 and NO X ) Indirect N 2 O-N (due to NO 3 leaching) 0.0044 0.0034 0.0062 0.01 kg N 2 O N per kg (NH 3 N + NO X N) (exanimal) (IPCC, 2006) Table 11.3) Assumption: No leaching, as leakages from animal housing units are prohibited in Finland 0.01 kg N 2 O N per kg (NH 3 N + NO X N) (exhousing) (IPCC, 2006) Table 11.3) 0 0 0.0015 No leaching, see above No leaching, see above 1% of N applied (IPCC, 2006, chapter 11) 0.1 N 2 O-N, based on (Nemecek and Kägi, 2007) Based on empirical model of Simmelsgaard and Djurhuus (1998) Not included as it is not needed for the mass balanced Based on Danish C- TOOL model (Hamelin, 2013, section 8). Assumed negligible for aerobic conditions. Modelled by use of an empirical model (relating P balance and P loading) (Ekholm et al., 2005) 0.01 kg N 2 O N per kg (NH 3 N + NO X N) (exstorage) (IPCC, 2006) Table 11.3) 0.0075 kg N 2 O-N per kg N leaching (IPCC, 2006). 13

4 Life Cycle Inventory data for Source Separation scenario The life cycle inventory for the Source Separation scenario is to a great extent a modification of the reference scenario dairy cow slurry, Finland. Accordingly, it is recommended to read the description of the reference system first. Similarly as in the reference scenario, the starting point of the mass balance and emission calculations in the inventory analysis was the urine and faeces excretion rates and composition ex-animal, based on the studies of MTT Agrifood Research Finland (Nousiainen, 2013). 4.1 Urine and faeces in the housing units 4.1.1 NH 3 emissions As in the reference manure scenario, ammonia emission estimates from dairy cow manure management system are based on the Finnish ammonia emission model (Grönroos et al., 2009). In the model, the flow of nitrogen in animal manure is followed during the entire manure management chain. Evaporation of ammonia is based on the data on animal manure management in Finland, including information on which ammonia abatement methods are used, on ammonia emissions from the different manure management phases without emission reduction measures, and on emission reduction potentials of the emission reduction measures. 4.1.2 CH 4 and CO 2 emissions Methane and carbon dioxide emissions from urine management chain were based on the same algorithms as were used in the reference (slurry) scenario. Methane emissions from the faeces management chain were based on the same IPCC algorithms as were used for urine and slurry, but using adjusted factors. For carbon dioxide emissions from faeces it was assumed that in animal house no emissions occur because manure is quickly moved to an outdoor storage after excretion. 4.1.3 N 2 O emissions Nitrous oxide emissions from urine and faeces management chains were based on the same IPCC algorithms as were used in the reference (slurry) scenario. 4.1.4 Energy consumption In house, electricity is needed to operate the manure scraper (23 kwh/1000 kg of manure exanimal). The electricity was modelled using the SimaPro process Electricity, hard coal, at power plant/marginal Finland. 4.1.5 Life cycle inventory data for the processes Urine in the housing units and Faeces in the housing units In Table 4.1 below, the life cycle inventory data for the processes Urine in the housing units and Faeces in the housing units are shown. The data in Table 4.1 are the data entered in SimaPro. 14

Table 4.1: Life cycle inventory data for the processes Urine in the housing units and Faeces in the housing units. Note: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. All emissions Comments per 1 000 kg of manure (urine + faeces) ex-animal Input Urine "ex-animal" 315 kg The input to this process is 315 kg of urine ex-animal, which together with the amount of faeces forms the study s functional unit (1 000 kg of manure ex-animal). The emissions are calculated relative to the FU, but separately for urine and faeces. Faeces "ex-animal" 685 kg The input to this process is 685 kg of faeces ex-animal, which together with the amount of urine forms the study s functional unit (1 000 kg of manure ex-animal). The emissions are calculated relative to the FU, but separately for urine and faeces. Straw 21.21 kg Per 1 000 kg of manure (all to faeces -> 21.21 kg per 685 kg of faeces). Water 0 kg No water added because cleaning waters are managed separately. Output Manure "ex-housing" 1 020.8 kg Increased due to addition of straw, reduced caused by DM loss, see Table 4.2. Energy consumption 23 kwh All to scraping faeces. No energy for urine pumping as urine is led to the outdoor storage by gravity. Emission to air (altogether, and urine+faeces in brackets) Carbon dioxide (CO 2 ) 0.21 kg CO 2 -C (0.21+0.00) For urine same algorithms as in reference system: 2.13 kg CO 2 per 1 kg CH 4. For faeces, no C-loss assumed. Methane (CH 4 ) 0.103 kg CH 4 -C Based on the IPCC guidelines and algorithms. (0.001+0.102) Ammonia (NH 3 -N) 0.54 kg NH 3 -N (0.38+0.16) According to the Finnish ammonia emission model for agriculture (Grönroos et al., 2009); 14% (urine) and 4% (faeces) of tot N Nitrous oxide (N 2 O-N), 0.0094 kg N 2 O-N Based on the IPCC guidelines and algorithms. direct emissions Nitrous oxide (N 2 O-N), indirect emissions (0.004+0.005) 0.0054 kg N 2 O-N (0,0038+0.0016) Nitrogen monoxide 0.00015 kg NO-N (NO-N) (representing (0.0009+0.00006) total NO X ) Nitrogen dioxide Included in the above (NO 2 -N) Nitrogen (N 2 -N) 0.0095 kg N 2 -N (0.0055+0.004) Discharge to water None Discharge to soil None Indirect emissions from volatilization, same algorithms as in reference system: 0.010 kg N 2 O-N per kg NH 3 -N + 0.010 kg N 2 O-N per kg NO X -N volatilized. Algorithm as in reference system: 0.0001 kg NO per kg TAN (EMEP- EEA (2010), Table 3.9). As TAN same as reference system, NO-N is the same, i.e. 0.000196 kg NO-N. Included in the above. Algorithms as in reference system: 0.0030 kg N 2 per kg TAN. Assumed to be zero, as leakages from housing systems are prohibited in Finland. Assumed to be zero, as leakages from housing systems are prohibited in Finland. 15

4.1.6 Mass balances for the processes Urine in the housing units and Faeces in the housing units Mass balances for the process Urine in the housing units and Faeces in the housing units can be followed in Tables 4.2 and 4.3. It should be kept in mind that in the basic scenario of the source separation system, not all urine is managed separately. Ca. 23 % of urine excreted is mixed with faeces and straw in animal house, while ca. 77% is collected separately. In the sensitivity analysis, the effect of 100% segregation of urine on the final results was studied. Table 4.2. Mass balances for the process Urine in the housing units. Urine composition kg per ton manure ex-animal Mass balance: Change during indoor storage kg Mass balance: Amount after indoor storage kg Urine composition kg per ton urine ex-housing h Total mass 315.45 0.00 a 315.45 1000.00 DM 12.21 0.00 b 12.20 38.69 VS 9.77 0.00 c 9.76 30.95 Total N 2.82-0.39 d 2.43 7.71 Phosphorus (P) 0.03 0.00 e 0.03 0.08 Potassium (K) 3.69 0.00 f 3.69 11.70 Carbon (C) 5.36-0.21 g 5.15 16.33 Important: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. a Change in total mass: 0 kg added straw Change in DM. No water added in-house. b Change in DM: + DM added by straw (0) DM losses. DM losses: 3.16 kg VS loss per kg CH 4 loss (3.16 gram VS loss per gram CH 4 is based on the Buswell equation (Symons and Buswell (1933), using the same principles as in chapter 5 of the Supporting Information of Hamelin (2013)). c Same absolute reduction as DM. d Change in Total-N: N from straw (same as reference system) minus emissions of NH 3 -N, N 2 O-N, NO-N and N 2 -N (indirect emissions of N 2 O-N not included). N added by straw: 0.00528 kg N/kg dm: Møller et al. (2000). e No P from added straw. f No K from added straw. g Change in Total-C: emissions of CO 2 -C and CH 4 -C. h The manure composition "ex-housing" per 1000 kg of urine. 16

Table 4.3. Mass balances for the process Faeces in the housing units. Faeces composition kg per ton manure ex-animal Mass balance: Change during indoor storage kg Mass balance: Amount after indoor storage kg Faeces composition kg per ton faeces ex-housing h Total mass 684.55 20.79 a 705.34 1000.00 DM 92.19 17.60 b 109.79 155.66 VS 73.76 13.99 c 87.75 124.40 Total N 3.43-0.07 d 3.35 4.76 Phosphorus (P) 0.94 0.02 e 0.95 1.35 Potassium (K) 2.49 0.27 f 2.76 3.92 Carbon (C) 40.48 8.12 g 48.60 68.91 Important: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. a Change in total mass: +21.21 kg added straw (as in reference system) Change in DM. No water added in-house. b Change in DM: + DM added by straw DM losses. DM in straw: 850 kg DM/ton straw (Møller et al., 2000). DM losses: 3.16 kg VS loss per kg CH 4 loss (3.16 gram VS loss per gram CH 4 is based on the Buswell equation (Symons and Buswell (1933), using the same principles as in chapter 5 of the Supporting Information of Hamelin (2013)). c Same absolute reduction as DM. d Change in Total-N: N from straw (same as reference system) minus emissions of NH 3 -N, N 2 O-N, NO-N and N 2 -N (indirect emissions of N 2 O-N not included). N added by straw: 0.00528 kg N/kg dm: Møller et al. (2000). e P from added straw as in reference system. 0.0009 kg P/kg dm: Møller et al. (2000). f K from added straw as in reference system. 0.015 kg K/kg dm: Møller et al. (2000). g Change in Total-C: C from added straw (same amount as in reference system minus emissions of CO 2 -C and CH 4 -C. 0.4563 kg C/kg DM (Mean value from Biolex database) (www.biolexbase.dk). h Faeces composition "ex-housing" per 1000 kg of faeces. 4.2 Urine and Faeces in the outdoor storage The emissions for the outdoor storage of urine and faeces were calculated using generally the same principles as were used for the reference (slurry) system. Some different principles were used for solid manure. 4.2.1 NH 3 emissions Ammonia emission estimates from dairy cow manure management system were based on the Finnish ammonia emission model (Grönroos et al., 2009), as was described in section 4.1.1. 4.2.2 CH 4 and CO 2 emissions Methane and carbon dioxide emissions from urine management chain were based on the same IPCC algorithms as were used in the reference (slurry) scenario. Methane emissions from the faeces management chain were based on the same algorithms as were used for urine and slurry. Petersen et al. (1998) found that CO 2 -C loss from pig solid manure during storing was 50% of the initial C-content, but from cow manure C loss was negligible. However, we supposed that some spontaneous composting takes place also in cow manure and 10% C-loss was used. 17

4.2.3 N 2 O emissions Nitrous oxide emissions from urine and faeces management chains were based on the same IPCC algorithms as were used in the reference (slurry) scenario. 4.2.4 Life cycle inventory data for the processes Outdoor storage of urine and Outdoor storage of faeces The life cycle inventory data for the processes Outdoor storage of urine and Outdoor storage of faeces are shown in Table 4.4. These are the data entered in SimaPro. Table 4.4: Life cycle inventory data for the processes Outdoor storage of urine and Outdoor storage of faeces. Note: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. All emissions Comments per 1 000 kg manure (urine + faeces) ex-housing Input Urine "ex-housing" 309 kg The input of urine to this process is 309 kg of manure ex-housing. The emissions are calculated relative to the sum of urine+manure. Faeces "ex-housing" 691 kg The input of faeces to this process is 691 kg of manure ex-housing. The emissions are calculated relative to the sum of urine+manure. Straw 0 kg No straw added during storing. Water 130 kg Water from precipitation during storing: 23 litres to urine, 107 kg to faeces. Output Manure "ex-storage" 1119.6 kg Increase due to precipitation, reduction due to DM loss. Energy consumption 8.2 MJ Fuel consumption: 8,2 MJ (0.2 kg) light fuel oil/ton manure and considering a LHV of 41.2 MJ/kg for light fuel oil. Emission to air Carbon dioxide (CO 2 ) 5.14 kg CO 2 -C (0.02+5.12) For urine, the same algorithm than for slurry in the reference case. For faeces, 10% of C is lost as CO 2 -C. Methane (CH 4 ) 1.93 kg CH 4 -C Based on the IPCC guidelines and algorithms. (0.02+1.91) Ammonia (NH 3 -N) 0.86 kg NH 3 -N (0.27+0.58) According to the Finnish ammonia emission model for agriculture (Grönroos et al., 2009): 11% (urine) and 16% (faeces) of tot N. Nitrous oxide (N 2 O-N), 0.0214 kg N 2 O-N See text in the section N 2 O emissions above this table. direct emissions Nitrous oxide (N 2 O-N), indirect emissions Nitrogen monoxide (NO-N) (representing total NO x ) Nitrogen dioxide (NO 2 - N) (0.0097+0.0117) 0.0086 kg N 2 O-N (0.0028+0.0058) 0.00015 kg NO-N (0.00006+0.0009) Included in the above Table continued to the next page. Indirect emission from volatilization, same algorithms as in reference system: 0.010 kg N 2 O-N per kg NH 3 -N + 0.010 kg N 2 O-N per kg NO X -N volatilized. Algorithm as in reference system: 0.0001 kg NO per kg TAN (EMEP- EEA (2010), Table 3.9). 0.75 kg TAN per kg N ex-hosing (Poulsen, 2008, Table 9.7, p.17). Total-N ex-housing: see Table 4.3 above. Included in the above. 18

Table 4.4 continued from the previous page. All emissions per 1 000 kg manure (urine + faeces) ex-housing Emission to air Nitrogen (N 2 -N) 0.0096 kg N 2 -N (0.004+0.0056) Discharge to water Discharge to soil None None Comments Algorithm as in reference: 0.0030 kg N 2 per kg TAN and 0.75 kg TAN per kg N ex-hosing (Poulsen, 2008, Table 9.7, p.17). Total-N exhousing: see Table 4.3 above. Assumed to be zero, as leakages from slurry tanks are prohibited in Finland. Assumed to be zero, as leakages from slurry tanks are prohibited in Finland. 4.2.5 Mass balances for the processes Outdoor storage of urine and Outdoor storage of faeces Mass balances for the processes Outdoor storage of urine and Outdoor storage of faeces can be followed in Tables 4.5 and 4.6. Table 4.5: Mass balances for the process Outdoor storage of urine. Urine composition kg per ton urine ex-housing Mass balance: Change during outdoor storage kg Mass balance: Amount after storage kg Urine composition kg per ton urine ex-storage h Total mass 1000 74.74 a 1074.17 1000 DM 38.69-0.28 b 38.41 35.75 VS 30.95-0.28 c 30.67 28.55 Total N 7.71-0.935 d 6.771 6.303 Phosphorus (P) 0.08 0.000 e 0.08 0.07 Potassium (K) 11.70 0.000 f 11.70 10.89 Carbon (C) 16.33-0.117 g 16.22 15.10 Important: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. a Change in total mass: +75 kg added water Change in DM. No straw added during storing. b Change in DM: + DM added by straw (0) DM losses. DM losses: 3.16 kg VS loss per kg CH 4 loss (3.16 gram VS loss per gram CH 4 is based on the Buswell equation (Symons and Buswell (1933), using the same principles as in chapter 5 of the Supporting Information of Hamelin (2013)). c Same absolute reduction as DM. d Change in Total-N: N from straw (same as reference system) minus emissions of NH 3 -N, N 2 O-N, NO-N and N 2 -N (indirect emissions of N 2 O-N not included). N added by straw: 0.00528 kg N/kg dm: Møller et al. (2000). e No P from added straw. f No K from added straw. g Change in Total-C: emissions of CO 2 -C and CH 4 -C. h Urine composition "ex-outdoor storing" per 1000 kg of urine. 19

Table 4.6: Mass balances for the process Outdoor storage of faeces. Faeces composition kg per ton faeces ex-housing Mass balance: Change during outdoor storage kg Mass balance: Amount after storage kg Faeces composition kg per ton faeces ex-storage h Total mass 1000 140.11 a 1139.88 1000 DM 155.66-11.62 b 144.04 126.37 VS 124.40-11.62 c 112.79 98.95 Total N 4.76-0.870 d 3.886 3.410 Phosphorus (P) 1.35 0.000 e 1.35 1.18 Potassium (K) 3.92 0.000 f 3.92 3.44 Carbon (C) 68.91-10.166 g 58.74 51.53 Important: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. a Change in total mass: +150 kg added water Change in DM. No straw added during storing. b Change in DM: + DM added by straw (0) DM losses. DM losses: 3.16 kg VS loss per kg CH 4 loss (3.16 gram VS loss per gram CH 4 is based on the Buswell equation (Symons and Buswell (1933), using the same principles as in chapter 5 of the Supporting Information of Hamelin (2013)). c Same absolute reduction as DM. d Change in Total-N: N from straw (same as reference system) minus emissions of NH 3 -N, N 2 O-N, NO-N and N 2 -N (indirect emissions of N 2 O-N not included). N added by straw: 0.00528 kg N/kg dm: Møller et al. (2000). e No P from added straw. f No K from added straw. g Change in Total-C: emissions of CO 2 -C and CH 4 -C. h Faeces composition "ex-outdoor storing" per 1000 kg of faeces. 4.3 Applying urine and faeces to the field The emissions from applying the urine and faeces to the field are based on the same algorithms as for the reference scenario, see the description of this for further details. However, transport distances of urine (1.5 tkm) and faeces (5 tkm) differed from that of slurry in the reference scenario (5 tkm). Additionally, the application of faeces differed from that of slurry and urine. For faeces, a SimaPro process Solid manure loading and spreading, by hydraulic loader and spreader, adopted for FI was used for application, including manure loading. In a sensitivity analysis, a ten times longer transport distance (50 tkm instead of 5 tkm) for faeces (solid manure) was applied. 4.3.1 CH 4 and CO 2 emissions The CH 4 emissions on the field are assumed to be negligible, as the formation of CH 4 requires an anaerobic environment, which is, under normal conditions, not the case in the top soil. Soils have an equilibrium C content which is the result of a balance between inflows (e.g. plant matter from above- and below-ground residues, manure, etc.) and outflows (e.g. decomposition, erosion, leaching of soluble C, etc.) to the soil pool. If outflows are greater than inflows, soil C decreases, while soil C increases, if inflows are greater than outflows. Output flows are to a great extent determined by climate-specific parameters like temperature and precipitation, where 20

higher temperature and moisture favour the soil biota activity (i.e. decomposition). However, any change affecting the activity of soil biota (e.g. change in oxygen availability due to soil compaction, change in soil ph) will result in greater or smaller decomposition. In this sense, any form of agriculture will disturb the soil equilibrium until a new equilibrium is eventually reached after many years of constant agricultural practices. When manure is applied to soils, part of the C it contains ends up in the soil C pool, while the rest of the C essentially ends up emitted as CO 2 to the atmosphere. A given manure handling technology involving that more C ends up in the soil C pool (in comparison to the reference situation) would thus imply an overall decrease of C ending up in the atmosphere. On the other hand, some manure handling technologies could involve that native soil C is lost (if, for example, they involve a drastic decrease of C applied to soils in comparison to the reference situation), in which case an overall increase of C to the atmosphere would be observed. In order to reflect such a balance, an attempt was made in order to model the soil C changes induced as a consequence of the different manure management technologies studied within Baltic Manure. Table 4.7 presents the breakdown considered for the fate of C in the different types of manure (with and without treatments) fractions involved in the LCAs performed within Baltic Manure (those applied to soil). These values are based on the work of Hamelin et al. (2010; 2013c), where the dynamic soil C model C-TOOL, developed to calculate the soil carbon dynamics in relation to the Danish commitments to UNFCCC, was used. This model is parameterized and validated against long-term field experiments conducted in Denmark, UK and Sweden. Further description of the C- TOOL model is given in Petersen et al. (2002) and Petersen (2010). As opposed to many different soil C models, C-TOOL does not only consider the topsoil, but the whole 0-100 cm profile. The values presented in Table 4.7 should be seen as rough estimates, these could of course be improved by a country-specific breakdown based on each country soil s properties. However, these estimates allow reflecting the complete C balance. Table 4.7: Breakdown of the applied C from the different manure types between the atmosphere and soil pool. Description of the applied material CO 2 -C, as a % of the C applied (from manure ex-storage) C ending up in the soil C pool, as a % of the C applied (from manure exstorage) Raw slurry (pig and dairy) 95% 5% Digestate (mono-digestion) 100% 0% Digestate (co-digestion with solid fraction or solid manure) 90% 10% Digestate (co-digestion with grass) 85% 15% Solid manure (raw; pig, horse and broiler) 75% 25% Solid fraction, from separation (of raw manure and/or 80% 20% digestate) Liquid fraction (from source-segregation and from separation of both raw manure and digestate) 100% 0% 21

4.3.2 Fertilizer substitution The amounts of mineral fertilizer N, P and K substituted by the NH 4 -N, P and K in the urine and faeces are (N in urine + N in faeces): Nitrogen (N): 1.06 + 1.36 = 2.42 (kg/t cow manure ex-storage) Phosphorus (P): (63% x 0.02) + (63% x 0.83) = 0.014 + 0.525 = 0.539 (kg/t cow manure ex-storage) Potassium (K): 3.23 + 2.42 = 5.65 (kg/t cow manure ex-storage) The values are the same as the amounts of the nutrients in one ton of slurry ex-storage except for phosphorus for which a substitution rate of 63% was used. This was based on the differences in P fertilization limits in the Finnish agri-environmental support system between mineral and manure P. In an average situation mineral P application level for cereals is 11 kg/ha but in case of manure P fertilization level of 15 kg/ha can be applied. Based on this it can be argued that the difference (4 kg/ha) is P surplus and does not substitute mineral fertilizer P, resulting in 73% of manure P utilization. For grasses, an average mineral P application level is 16 kg/ha but in case of manure P fertilization level of 30 kg/ha can be used. In that case 53% of manure P is utilized. It was assumed that 50% of dairy cow manure is applied on cereals and 50% on grass resulting in an average manure P utilization rate of 63%. Reduced machine work due to reduced use of mineral fertilizers was not considered because it was assumed that a) manure nutrients do not fully satisfy the nutrient needs of the plants, meaning that mineral fertilizers must be applied on the whole acreage where manure was applied, and b) the all three main nutrients are applied at the same time using compound fertilizers. The fertilizer substitution is modelled in SimaPro as negative values, as the mineral fertilizers are subtracted from the system. Avoided Processes (subtracted from the system) are: Nitrogen fertilizer: Adjusted Ecoinvent process: Calcium ammonium nitrate, as N, at regional storehouse, nitric acid from plant with catalytic tech. with adjusted nitric acid process as described in Hamelin (2013), section 2.5 and with applied emissions from spreading (NH 3, N 2 O, NO X and N leaching). Phosphorus fertilizer: Ecoinvent process: Diammonium phosphate, as P2O5, at regional storehouse (see Hamelin (2013), section 2.5) Potassium fertilizer: Ecoinvent process Potassium chloride, as K2O, at regional storehouse/rer (see Hamelin (2013), section 2.5) 4.3.3 Nitrate leaching Nitrogen loads were calculated by an empirical model which was developed to predict annual average nitrate leaching as affected by the long-term rate of N fertilization and crop type by Simmelsgaard and Djurhuus (1998). The model was calibrated to each country by using countryspecific average N loss from fields and recommended N fertilization levels. In this manure management scenario an N loss of 7.15% for urine and 7.22% for faeces NH4 + -N was used. 22

4.3.4 Phosphorus leaching The P loss from the fields was estimated by the model of Ekholm et al. (2005). The model relates the P surplus (or deficit) in a farm to the edge-of-field losses of algal-available P. Based on longterm fertilizer trials, the model first estimates the change in soil-test P of top soil with the aid of the soil-surface balance of P. Soil-test P is then used to approximate the concentration of dissolved reactive P in surface runoff and drainage flow, as adjusted for different P application types. Particulate P is estimated from specific erosion rates for each soil type and a bioavailability coefficient of 0.16 was used. In this manure management scenario a P loss of 2.4% of total phosphorus was used for urine and faeces. 4.3.5 Life cycle inventory data for the processes Urine application to field and Faeces application to field The life cycle inventory data for the processes Urine application to field and Faeces application to field for the source separation scenario are shown in Table 4.8. These are the data entered in SimaPro. Table 4.8. Life cycle inventory data for the processes Urine application to field and Faeces application to field for the source separation scenario. Note: The number of digits does not reflect the precision, but is only included as the numbers are used for further calculations. All emissions Comments per 1 000 kg manure ex-storage Input Urine "ex-storage" 296 kg The input of urine to this process is 296 kg of manure ex-storage. The emissions are calculated relative to the sum of urine+manure. Faeces "ex-storage" 704 kg The input of faeces to this process is 704 kg of manure ex-storage. The emissions are calculated relative to the sum of urine+manure. Transport of urine to field Transport of faeces to field Spreading of urine on field Spreading of faeces on field 1.5 km*ton Shorter as for slurry in the reference scenario. Based on the Ecoinvent process: Transport, tractor and trailer/ch U. Includes diesel for the spreading and production of tractor, trailer and shed. 5 km*ton Longer as for slurry in the reference scenario. Based on the Ecoinvent process: Transport, tractor and trailer/ch U. Includes diesel for the spreading and production of tractor, trailer and shed. 296 kg As for the reference system. Based on the Ecoinvent process: Slurry spreading, by vacuum tanker / CH U. 704 kg As for the reference system. Based on the Ecoinvent process: Slurry spreading, by vacuum tanker / CH U. Table continued to the next page. 23

Table 4.8 continued from the previous page. All emissions Comments per 1 000 kg manure ex-storage Output Manure on field. NH 4 -N in urine 1.06 kg N N fertilizer replaced. NH 4 -N in faeces 1.36 kg N N fertilizer replaced. P in urine 0.014 kg P P fertilizer replaced, see chapter 4.3.2. P in faeces 0.525 kg P P fertilizer replaced, see chapter 4.3.2. K in urine 3.23 kg K K fertilizer replaced. K in faeces 2.42 kg K K fertilizer replaced. Energy consumption Diesel for transport and spreading of urine and faeces is included in input above. Emission to air Carbon dioxide (CO 2 ) kg 33.5 CO 2 -C Modelled by C-TOOL, see description in for reference system. (4.5+29.0) Methane (CH 4 ) 0 kg CH 4 -C The CH 4 emissions on the field are assumed to be negligible. Ammonia (NH 3 -N), at 0.0121 kg NH 3 -N Ammonia emissions at very moment of application. Same algorithms very moment of (0.0053+0.0068) as for reference system: 0.5% of TAN ex-storage for trail hose application application (Hansen, 2008). Ammonia (NH 3 -N) 0.206 kg NH 3 -N (0.152+0.054) Nitrous oxide (N 2 O-N), 0.0427 kg N 2 O-N direct emissions (0.0187+0.0240) Nitrous oxide (N 2 O-N), 0.0021 kg N 2 O-N indirect emissions (0.0015+0.0006) from volatilization Nitrous oxide (N 2 O-N), 0.0013 kg N 2 O-N indirect emissions (0.0006+0.0007) from N leaching Nitrogen monoxide 0.0043 kg NO-N (NO-N) (representing (0.0019+0.0024) total NO x ) Nitrogen dioxide (NO 2 - N) Nitrogen (N 2 -N) 0.469 kg N 2 -N (0.205+0.264) Discharge to water Nitrate leaching Phosphorus leaching 0.174 kg N (0.076+0.098) 0.021 kg P (0.0001+0.020) According to the Finnish ammonia emission model for agriculture (Grönroos et al., 2009): 8% (urine) and 2% (faeces) of tot N. Same algorithms as for reference system: 0.01 kg N 2 O-N per kg N IPCC (2006) emission factor, for any organic amendment. Indirect emission from volatilization, same algorithms as in reference system: 0.010 kg N 2 O-N per kg NH 3 -N + 0.010 kg N 2 O-N per kg NO X -N volatilized (IPCC, 2006). Indirect emission from leaching. Same algorithms as in reference system. From N leaching: 0.0075 kg N 2 O-N per kg N leaching (IPCC, 2006). Algorithm as in reference system: NO X N = 0.1 * direct N 2 O-N (Nemecek and Kägi, 2007). Included in the above. N2 is only calculated in order to implement it in mass balances. According to Vinther and Hansen (2004), the N2/N 2 O ratio for animal manure applied to field is 3.5 for JB3 soils (Vinther and Hansen (2004), page 30, text above Table 4), and accordingly, the N2-N/N 2 O-N factor is 11. See section 4.3.3. See section 4.3.4. 24

5 Life Cycle Assessment Results and Interpretation 5.1 Life Cycle Assessment Results The results of the LCA of the Source Separation scenario are shown in Figure 5.1 together with the sensitivity analysis. The Reference Scenario is also represented in Figure 5.1. The following sensitivity analyses were carried out: Sensitivity analysis 1: Longer transport distance for faeces: 50 km instead of 5 km Sensitivity analysis 2: All urine separated from faeces (in the basic scenario 77% of urine is managed separately, and 23% is mixed with faeces and straw) Sensitivity analysis 3: S2 + effective ammonia emission reduction measures. All urine storages are tightly covered. Urine is injected in soil or incorporated within 12 hours after application, faeces are incorporated in soil within 12 hours after application. 25

(a) (b) (c) (d) Figure 5.1: Global Warming (a; CO 2 -eq), Acidification (b; m 2 "unprotected ecosystems equivalent) and Aquatic Eutrophication N (c; N-eq) and P (d; P-eq) impact per 1000 kg of dairy cow manure (urine + faeces) ex-animal. 26

5.2 Discussion Source separation of dairy cow manure was compared to the reference scenario Dairy cow slurry system. From the Figure 5.1 it can be seen that: Regarding Global Warming, there are no clear differences between the reference system and the basic scenario of the source separation system. Most of the impact is caused by the biogenic CO 2 - emissions from manure, mainly from the field after manure application. Because for sourceseparated manure biogenic CO 2 -emissions from outdoor storage are higher than for slurry manure, less source-separated manure C is ending up to the fields compared to the slurry. Therefore, the total biogenic CO 2 emission from source-separated manure is slightly higher than from slurry, even though proportionally more manure carbon is sequestered into soil after application in the source separation system. Methane is also important but dinitrogen monoxide and especially fossil carbon dioxide have a relatively small effect on Global Warming impact of the manure management systems studied. Methane emissions are higher in the slurry system, dinitrogen monoxide emissions approximately equal between the systems. Longer transport distance of manure solid fraction has only a slight impact on the Global Warming impact of the source separation system (Sensitivity analysis S1). Increasing the transport distance from 5 to 50 km increases the global warming impact of the total system by 4%. If all urine could be separated (instead of 73% of urine in the basic scenario), the Global Warming impact would decrease due to decreased amount of the solid fraction (less methane emissions; Sensitivity analysis S2). If effective ammonia emission measures were introduced in the manure management chain of the source separation system (covering the storages tightly and incorporating manure rapidly after application), the Global Warming impact could be further, but only very slightly, decreased because of the decreased secondary dinitrogen monoxide emissions (Sensitivity analysis S3). Source separation system gives a slightly higher contribution to the environmental impact category Acidification, which is primarily caused by NH 3 emissions. In the source separation system, the reductions in NH 3 emissions from the field process are overruled due to the increased emissions from outdoor storage caused by spontaneous composting of the solid fraction during storing (Sensitivity analysis S2; all urine separately). Effective ammonia emission reduction measures significantly decrease the emissions from manure storing, and during and after manure spreading (Sensitivity analysis S3). Again, longer solid fraction transport distance has only a relatively small contribution to the total Acidification impact value (Sensitivity analysis S1). Regarding the impact category N Aquatic Eutrophication (nitrate leaching, airborne nitrogen emissions), there are some minor differences between the different manure management scenarios, caused by the differences in nitrate leaching and ammonia emissions. The Sensitivity analysis S3 (reduced NH 3 emissions) has the lowest impact score due to the lowest ammonia emissions. Regarding the impact category P eutrophication (phosphorus leaching from the fields, phosphorus emissions to the waters from the industrial processes), source separation does not affect the content of phosphorus in cattle manure. However, fertilizer industry and electricity 27

generation do contribute to this impact category. Accordingly, the electricity consumption of the manure scraper leads to an increased contribution to this category. When analysing the results in SimaPro, the contributions from the electricity consumption mainly arise from the process Disposal, spoil from coal mining, in surface landfilling. If the marginal electricity was not based on hard coal, as described in the main report, there would be no significant difference between the contributions to P eutrophication for source separation system and the reference scenario, except in the case when solid fraction transport distance is increased. In that case, the increase results from the increased manufacturing of machinery. 5.3 Conclusions As regards Global Warming, differences between the two manure management systems studied were relatively small. Because biogenic CO 2 emissions were responsible for the most - approximately for 60% - of the Global Warming impact, the role of the fossil CO 2 emissions appeared to be small in both systems. However, even if the Global Warming results of the source separation system are considered without biogenic CO 2 emissions, the share of fossil CO 2 is only 6% of the total Global Warming impact. Source or mechanical separation of manure is argued with the possibility of enhanced manure P use in the fields. Compared to slurry, P-rich solid fraction of manure can be transported over longer distances and applied on the low P-value fields. Thus unnecessary P fertilizing of the P-rich fields which usually locate near animal farms and intensive livestock farming regions is avoided. In this study it was not possible to study the long-term effects of this kind of manure P re-allocation on P leaching from the fields. Therefore the impact assessment results presented here do not fully describe the potential environmental benefits of the source separation system. In a source separation system, ammonia losses may become higher during the storage of the two fractions as compared to the slurry system. Therefore it is important to prevent ammonia evaporation effectively by covering urine storages tightly and by preventing of spontaneous composting of solid fraction and/or by covering the solid manure heaps with a tight cover. However, environmental impacts of the reference, i.e. slurry system can also be reduced efficiently by reducing ammonia volatilization e.g. by removing slurry to an outdoor storage quickly after excretion and covering the outdoor storages with tight covers. The impact of source separation on odours was not assessed. Odours are also mainly related to ammonia emissions. Ammonia emission reduction decreases also other harmful environmental impacts than those considered in this study: direct damages to vegetation in high concentrations, secondary particle formation in atmosphere causing e.g. health effects to humans, and terrestrial eutrophication. Inside the animal houses ammonia affects the health of animals and humans. Because of the high importance of ammonia in most impact categories, ammonia can be considered as the most important substance causing environmental impacts from manure management. Due to this, it is urgent to decrease ammonia emissions effectively in all manure management phases in all manure management systems. 28

6 References Ekholm P, Turtola E, Grönroos J, Seuri P, Ylivainio K (2005): Phosphorus loss from different farming systems estimated from soil surface phosphorus balance. Agriculture, ecosystems and environment 110: 266-278. Ekvall T, Weidema BP (2004): System Boundaries and input data in consequential life cycle inventory analysis. Int. J. Life Cycle Assess. 2004, 9(3), 161171, DOI:10.1007/BF02994190. EMEP-EEA (2009): EMEP/EEA air pollutant emission inventory guidebook 2009. Technical guidance to prepare national emission inventories. EEA Technical report No 9/2009. EEA (European Environment Agency). Frischknecht R, Rebitzer G (2005): The Ecoinvent database system: a comprehensive web-based LCA database. Journal of Cleaner Production 13:1337-1343. Grönroos J, Mattila P, Regina K, Nousiainen J, Perälä P, Saarinen K, Mikkola-Pusa J (2009): Development of the ammonia emission inventory in Finland. Revised model for agriculture. The Finnish Environment 8/2009. 60 p. Hamelin L (2013): Carbon management and environmental consequences of agricultural biomass in a Danish Renewable Energy strategy. PhD thesis. University of Southern Denmark. Department of Chemical Engineering, Biotechnology and Environmental Technology. Faculty of Engineering. Hamelin L, Wesnæs M, Wenzel H, Petersen BM (2010): Life cycle assessment of biogas from separated slurry. Copenhagen, DK, Danish Ministry of the Environment, Environmental Protection Agency. Hamelin L, Wesnæs M, Wenzel H, Petersen BM (2011): Environmental Consequences of Future Biogas Technologies Based on Separated Slurry. Environmental Science & Technology 2011;45:5869 77. Hamelin L, Baky A, Cano-Bernal J, Grönroos J, Kuligowski K, Pehme S, Rankinen K, Skura D, Wenzel H, Wesnæs M, Ziolkowski M (2013a): Reference life cycle assessment scenarios for manure management in the Baltic Sea regions - An assessment covering six animal production, five countries, and four manure types. Baltic Manure WP5 Assessing Sustainability of Manure Technology Chains. Available at www.balticmanure.eu Hamelin L, Baky A, Cano-Bernal J, Grönroos J, Kuligowski K, Pehme S, Rankinen K, Skura D, Wenzel H, Wesnæs M, Ziolkowski M (2013b): Life Cycle Assessments of manure management techniques for the Baltic Sea Regions. Baltic Manure WP5 Assessing Sustainability of Manure Technology Chains. Available at www.balticmanure.eu 29

Hamelin L, Naroznova I, Wenzel H (2013c): Environmental Consequences of Different carbon Alternatives for Increased Manure-Based Biogas and Supporting Information for this. (Paper V in PhD Thesis). Accepted in Applied Energy. In print. Hansen MN, Sommer S, Hutchings NJ, Sørensen P (2008): Emissionsfaktorer til beregning af ammoniak-fordampning ved lagring og udbringning af husdyrgødning. Det Jordbrugsvidenskabelige Fakultet, Aarhus. Universitet. DJF Husdyrbrug nr. 84. December 2008. Hauschild M, Potting J (2005): Spatial Differentiation in Life Cycle Impact Assessment The EDIP 2003 Methodology. Copenhagen, Denmark, Danish Ministry of the Environment, Environmental Protection Agency. Iivonen S (2008): Ympäristöturpeet ja niiden käyttö. Helsingin yliopisto, Ruralia-instituutti. Raportteja 32. Available at: http://www.helsinki.fi/ruralia/julkaisut/pdf/raportteja32.pdf IPCC (2006): 2006 IPCC Guidelines for National Greenhouse Gas Inventories Volume 4 Agriculture, Forestry and Other Land Use Chapter 10: Emissions from Livestock and Manure Management. Knudsen L, Birkmose T (2005): Indhold af næringsstoffer i gylle bestemt ved analyser. [Content of nutrients in slurry as determined by analysis]. Planteavlsorientering Nr. 07-540. Landscentret Planteavl, Dansk Landbrugsrådgivning, Landcentret MMM/Tike (2013): Maatalouslaskenta 2010. Kotieläinten elinolot ja lannan varastointi (In Finnish). 15.12.2011. Available at: http://www.maataloustilastot.fi/e-lehti-kotielaimet/ Møller HB, Lund I, Sommer SG (2000): Solid liquid separation of livestock slurry separation efficiency and costs. Biores. Techn. 74, 223-229. Nemecek T, Kägi T (2007): Life Cycle Inventory of Agricultural Production Systems. Data v 2.0. Dübendorf, Switzerland: Swiss Centre for Life Cycle Inventories; 2007. Nousiainen J (2013): Nutrient excretion values for pigs, cattle and poultry - values for Finnish animals in 2013. Unpublished data. E-mail September 6, 2013. Petersen BM (2010): A model for the carbon dynamics in agricultural, mineral soils. Aarhus, Denmark: Faculty of Agricultural Sciences, Aarhus University; 2010. Petersen BM, Olesen JE, Heidmann T (2002): A flexible tool for simulation of soil carbon turnover. Ecological Modelling 2002; 151:1 14. Petersen SO, Lind A-M, Sommer SG (1998): Nitrogen and organic matter losses during storage of cattle and pig manure. J. Agric. Sci. Cambr. 130, 69 79. Poulsen HD (2008): Baggrundstal 2008. 30

http://agrsci.au.dk/institutter/institut_for_husdyrbiologi_og_sundhed/husdyrernaering_og_miljoe /normtal/ Retrieved December 15th, 2011. Simmelsgaard SE, Djurhuus J (1998): An empirical model for estimating nitrate leaching as affected by crop type and the long-term N fertilizer rate. Soil use and management, vol 14, 37 43. Stranddorf HK, Hoffmann L, Schmidt A (2005): Update on Impact Categories, Normalisation and Weighting in LCA Selected EDIP97 Data. Environmental project no. 995. Danish Ministry of the Environment, Environmental Protection Agency, Copenhagen, Denmark. Available at: http://www2.mst.dk/common/udgivramme/frame.asp?http://www2.mst.dk/udgiv/publications/ 2005/87-7614-570-0/html/default_eng.htm (accessed 11 July 2011). Symons GE, Buswell AM (1933): The methane fermentation of carbohydrates. J Am Chem Soc 55:2028 36. Viljavuuspalvelu (2013): Manure composition statistics based on manure samples analysed by the Eurofins Viljavuuspalvelu Oy between 2006-2009. In Finnish. http://www.viljavuuspalvelu.fi/sites/default/files/sites/default/files/tilastot/lantatilasto%202005 %20-%202009.pdf Vinther P, Hansen S (2004): SimDen. en simpel model til kvantificering af N 2 O-emission of denitrifikation, DJF rapport Markbrug nr. 104; Danish Ministery of Food, Agriculture and Fisheries, Danmarks JordbrugsForksning: Copenhagen, Denmark, 2004. http://agro.au.dk/fileadmin/djf/agro/fpv/djf_rapport_nr_104_2004.pdf Weidema BP (2004): The product, functional unit and reference flows in LCA, Environmental News No. 70, The Danish Ministry of the Environment, Environmental Protection Agency, Copenhagen. Weidema BP, Wenzel H, Petersen C, Hansen K (2004): The product, functional unit and reference flows in LCA. Environmental News No. 70. Danish Environmental Protection Agency. Wenzel H (1998): Application dependency of lca methodology: Key variables and their mode of influencing the method. The International Journal of Life Cycle Assessment, 3, 281 288. Wenzel H, Hauschild M, Alting L (1997): Environmental Assessment of Products -1: Methodology, Tools and Case Studies in Product Development. Kluwer Academic Publishers, Massachusetts, USA. 31

This report in brief In the Source Separation system, urine and faeces are kept separately after excretion, and not mixed together as in the slurry system. Separation of manure is argued with the possibility of enhanced use of manure P in the fields. Compared to slurry, P-rich solid fraction of manure can be transported longer distances and applied on the low P-value fields, while the N-rich liquid fraction can be utilized as N fertilizer in high P-value fields. The study showed that using Source Separation system ammonia losses may become higher during the storage of the two fractions than in the slurry system. Therefore it is important to prevent ammonia evaporation effectively by covering urine storages tightly and by preventing spontaneous composting of solid fraction or covering the solid manure heaps tightly. However, environmental impacts of the slurry system can also be reduced efficiently using the same principles. About the project The Baltic Sea Region is an area of intensive agricultural production. Animal manure is often considered to be a waste product and an environmental problem. The long-term strategic objective of the project Baltic Manure is to change the general perception of manure from a waste product to a resource. This is done through research and by identifying inherent business opportunities with the proper manure handling technologies and policy framework. To achieve this objective, three interconnected manure forums has been established with the focus areas of Knowledge, Policy and Business. Read more at www.balticmanure.eu. This report on Source Separation of dairy cattle manure was prepared as part of Work Package 5 on Assessing Sustainability of Manure Technology Chains in the project Baltic Manure. www.balticmanure.eu Part-financed by the European Union ()