Assessing anaerobic co-digestion of pig manure with agroindustrial wastes: The link between environmental impacts and operational parameters

Science of the Total Environment 497–498 (2014) 475–483

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Assessing anaerobic co-digestion of pig manure with agroindustrial wastes: The link between environmental impacts and operational parameters Ivan Rodriguez-Verde ⁎, Leticia Regueiro, Marta Carballa, Almudena Hospido, Juan M. Lema Department of Chemical Engineering, Institute of Technology, University of Santiago de Compostela, E-15782 Santiago de Compostela, Spain

H I G H L I G H T S • • • • •

The environmental consequences of using wastes as co-substrates were evaluated. Anaerobic co-digestion enhanced the environmental profile of mono-digestion. LCA results were strongly affected by energy production and ammonia emissions. Readily biodegradable co-substrates are advised under an environmental perspective. LCA results were clearly linked to operational parameters of anaerobic co-digestion.

a r t i c l e

i n f o

Article history: Received 25 February 2014 Received in revised form 31 July 2014 Accepted 31 July 2014 Available online xxxx Editor: E. Capri Keywords: Biogas Co-substrates Digestate Environmental impact LCA Statistical correlations

a b s t r a c t Anaerobic co-digestion (AcoD) is established as a techno-economic profitable process by incrementing biogas yield (increased cost-efficiency) and improving the nutrient balance (better quality digestate) in comparison to mono-digestion of livestock wastes. However, few data are available on the environmental consequences of AcoD and most of them are mainly related to the use of energy crops as co-substrates. This work analysed the environmental impact of the AcoD of pig manure (PM) with several agroindustrial wastes (molasses, fish, biodiesel and vinasses residues) using life cycle assessment (LCA) methodology. For comparative purposes, mono digestion of PM has also been evaluated. Four out of six selected categories (acidification, eutrophication, global warming and photochemical oxidation potentials) showed environmental impacts in all the scenarios assessed, whereas the other two (abiotic depletion and ozone layer depletion potentials) showed environmental credits, remarking the benefit of replacing fossil fuels by biogas. This was also confirmed by the sensitivity analysis applied to the PM quality (i.e. organic matter content) and the avoided energy source demonstrating the importance of the energy recovery step. The influence of the type of co-substrate could not be discerned; however, a link between the environmental performance and the hydraulic retention time, the organic loading rate and the nutrient content in the digestate could be established. Therefore, LCA results were successfully correlated to process variables involved in AcoD, going a step further in the combination of techno-economic and environmental feasibilities. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Anaerobic digestion (AD) is an appropriate technology for the treatment of livestock wastes, such as pig manure (PM), resulting in two added value products: biogas and digestate (Nasir et al., 2012). The former constitutes a renewable energy source that can be used to replace fossil fuels, while the latter is a higher quality fertiliser than raw livestock manure, by enhancing the nutrient availability and pathogen reduction (Xie et al., 2011). However, AD of PM as a sole substrate offers low biogas ⁎ Corresponding author. Tel.: +34 881 816021; fax: +34 881 816702. E-mail address: [email protected] (I. Rodriguez-Verde).

http://dx.doi.org/10.1016/j.scitotenv.2014.07.127 0048-9697/© 2014 Elsevier B.V. All rights reserved.

yields due to its low organic matter content (Angelidaki and Ellegaard, 2003) and high ammonium concentration (2–3 g N-NH+ 4 /L), which can result in ammonia toxicity events (Regueiro et al., 2012). In order to overcome these limitations, the anaerobic co-digestion (AcoD) of PM with other fermentable bio-wastes has been set as a common practice in agro-industrial biogas plants improving their technoeconomic profitability (Álvarez et al., 2010; Pantaleo et al., 2013). Energy crops represent a suitable co-substrate for AcoD because of their high energy density (Holm-Nielsen et al., 2009); however, the use of energy crops shows some drawbacks such as the consumption of water, the fertiliser usage or the controversy derived from the competition with food and feed (Gonzalez-Garcia et al., 2013; Pantaleo et al.,

476

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

2013). Other common co-substrates are wastes coming from agroindustry. They constitute appropriate residues because of their high biogas yield (Mata-Alvarez et al., 2013) and they must be selected depending on their availability in the geographical area of the biogas plants. The co-digestion of PM with molasses (MW) (Fang et al., 2011; Misi and Forster, 2001), fish wastes (FW) (Álvarez et al., 2010; Serrano et al., 2013), biodiesel wastes (BW) (Astals et al., 2012; Regueiro et al., 2012) or vinasses wastes (VW) (García-Gen et al., 2013; Riaño et al., 2011) was proved to increase methane production from 36 to 96% in comparison to mono-digestion of PM. In contrast, the environmental performance of AcoD process is not much reported and the few studies applying the life cycle assessment (LCA) methodology (ISO, 14040, 2006; ISO, 14044, 2006) used mostly energy crops as co-substrate (Bacenetti et al., 2013; De Vries et al., 2012; Lansche and Müller, 2012; Poeschl et al., 2012). There, AcoD of livestock manure with energy crops was proven as an appropriate technology to diminish the greenhouse gas emissions compared to the use of PM as sole substrate (Bacenetti et al., 2013; Lansche and Müller, 2012), but when a life cycle perspective was used and cultivation stage included, the global warming potential was clearly prejudiced (De Vries et al., 2012). Moreover, the cultivation stage reported worse performance in terms of acidification (AP) and eutrophication (EP) potentials when comparing AcoD vs. mono-digestion (De Vries et al., 2012; Lansche and Müller, 2012). Poeschl et al. (2012) and De Vries et al. (2012) also evaluated livestock manure AcoD with organic wastes and they similarly conclude better environmental performances in comparison to the co-digestion of energy crops and mono-digestion of livestock wastes in all the categories under assessment. These authors stated that higher proportions of agricultural wastes and organic wastes should be introduced in the co-digestion system in order to improve environmental sustainability. Following this line, the main objective of this work was to assess the environmental performance of AcoD of PM with several agro-industrial wastes. In addition, the effect of co-substrate characteristics on the environmental profile was addressed. Furthermore, the relationship between LCA results and the process variables involved in AcoD was determined with the aim of establishing some correlations between environmental and operational parameters. 2. Materials and methods 2.1. Functional unit definition The functional unit (FU) is usually defined according to the main function of the assessed process or system (Baumann and Tillman, 2004). When dealing with waste management systems, the FU is commonly defined in terms of the system input, i.e. the waste to be managed (McDougall et al., 2001). In this case, the management of a fixed amount of untreated liquid PM as base substrate by anaerobic codigestion was defined as the function of the system. The corresponding reference flow was 110,000 ton/year of PM, which provided an electric power of 500 kWe, a common value for a full-scale digester (Pantaleo et al., 2013) taking into account the biogas potential of PM (Regueiro et al., 2012). The use of co-substrates in the co-digestion scenarios will result in an increase of this electrical power. 2.2. Scenarios description and system boundaries definition Related to the defined FU, the PM treated in the plant was set as 110,000 ton/year. This PM production belongs to a total herd of 50,000 pigs and, therefore, a centralised system collecting and treating the PM from 17 farms with an average size of 3,000 heads (Eurostat, 2013) was assumed. The location selected was the Mountain Region in Galicia (northwest of Spain) (Fig. 1), where the PM production ranges from 70,000 to 203,000 ton/year, approximately.

Four different wastes (MW, FW, BW and VW) were selected as cosubstrates according to the national project Probiogas (2010) in terms of biomethane potential and generation rate. Once the wastes were selected, the distance for their transport was calculated and then fixed as the distance between the nearest point of generation and the location of the biogas plant. • Around 170,000 ton of FW (Probiogas, 2010) are produced yearly in Galicia due to the importance of canning industry in this region (Regueiro et al., 2012). A fraction of the FW, such as viscera or fish heads, can be recovered and utilised as co-substrate in AcoD due to their high protein and lipid content (Table 1). • MW are by-products of the sugar-extraction process and are often used as animal feeding and for biofuel production, both bioethanol and biogas (Jiménez et al., 2004). Their use in anaerobic digestion is justified because of their great organic load (723 g COD/kg, Table 1). • BW are residues obtained during biodiesel production and they contain mainly glycerol. Glycerol market is nowadays saturated due to the growing industry of biodiesel in the last years (Quispe et al., 2013), and therefore, this waste is available for other uses. Although this residue is readily biodegradable, instability can occur due to volatile fatty acids (VFAs) accumulation, and hence, no more than 6% of the total feeding mixture is advisable (Regueiro et al., 2012). • VW can be used for different purposes, such as animal feeding or biotechnological processes (Devesa-Rey et al., 2011). Animal feeding could not be an appropriate option for the low nutrient content of VW, however, their treatment by AcoD is strongly recommended due to the high organic load (Riaño et al., 2011). Three scenarios were considered: co-digestion of PM with MW and FW (C1), co-digestion of PM with BW (C2) and co-digestion of PM with VW (C3). Moreover, a reference scenario based on mono-digestion of PM was also included. Table 2 collects the average operational conditions in steady state of each scenario. The operational data were obtained in experimental periods of more than 55 days, corresponding to 2–4 HRT depending on the scenario assessed. In all scenarios, the digester volume and the PM volumetric flow-rate (the FU) were kept constant at 7,000 m3 and 351 m3/d, respectively, but the organic loading rate (OLR) and the hydraulic retention time (HRT) varied due to the use of co-substrates and, consequently, the electric power was different among scenarios. The process subsystems and boundaries for the four scenarios are displayed in Fig. 2. The impacts related to PM and co-substrates (FW, MW, BW and VW) production were excluded since they were considered residues and all the environmental burdens associated to their productive processes have been allocated to the main products (such as pork in pig production). Therefore, the analysis started with the transport of substrates from the generation point to the centralised biogas plant and its performance in steady state was considered for the LCA assessment. 2.3. Inventory analysis In this stage, the raw materials consumed, the energy used, the products and co-products obtained, as well as the emissions to air, water and soil, were identified and quantified for each scenario. The missing data were extracted from literature sources after their evaluation and considering the worst-case approach (i.e. the most unfavourable from an environmental point of view and realistic situation) (Table SI1). The whole process was divided into 6 subsystems (Fig. 2) as described below: 2.3.1. Subsystem 01: manure and co-substrates provision and storage As previously stated, PM was provided by 17 farms: one farm where the digester was installed plus 16 farms located at an average distance of 25 km. The collected PM was characterised (Table 1) in terms of pH, density, total (TS) and volatile (VS) solid content, chemical oxygen demand (COD), total Kjeldahl nitrogen (TKN) and ammonium concentration

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

477

Fig. 1. Density of pig manure production in Spain and location of the biogas plant (red box) in the northwest of the country (Mountain Region, 1,642 km2). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) adapted from Probiogas (2010)

(NH+ 4 ) according to Standard Methods (Clesceri et al., 1998). PM was collected from open chambers and therefore, weather conditions cause high fluctuations in some parameters, such as COD (65.0 ± 25.7 g O2/kg, Table 1) or TS (29.8 ± 17.2 g/kg, Table 1). PM was stored in an open chamber of 25,500 m3 that guarantees the storage for a 3-month PM production according to the Spanish legislation (RD 3483/2000, 2000). Air emissions taking place during the PM storage were estimated according to the emission factors provided by the national government in accordance to the IPCC (Intergovernmental Panel on Climate Change) Directive (PRTR, 2013, Table 3). In the case of methane emissions, the considered factor includes the emissions from both raw manure storage and spreading, with a major contribution of the storage. Regarding the emissions to water, NO− 3 , P and K leaching were considered insignificant due to the legal requirements applicable for manure storage (RD 3483/2000, 2000) forcing the facilities to be built with impermeable materials.

Table 1 Main characteristics of pig manure (PM), fish wastes (FW), biodiesel wastes (BW), molasses wastes (MW) and vinasses wastes (VW) used as co-substrates. Parameter

PMa

pH Density (kg/L) TSb (g/kg) VSb (g/kg) CODb (g O2/kg) N-TKNb (g N/kg) N-NH+ 4 (g N/kg)

7.6 1.0 29.8 17.8 65.0 4.3 3.5

± ± ± ± ± ± ±

0.5 0.0 17.2 12.7 25.7 0.4 0.8

FW

BW

MW

VW

7.1 1.0 304 282 567 19.0 0.7

7.3 1.0 593 557 1,679 0.0 0.0

8.1 1.1 835 707 723 57.0 14.8

3.3 1.0 0 0 174 0.12 0.02

a Standard deviations are only shown for PM since several batches of this substrate were used along the experiments. b TS: total solids, VS: volatile solids, COD: chemical oxygen demand, and TKN: total Kjeldahl nitrogen.

The four co-substrates (FW, BW, MW and VW) were transported from several generation points in Galicia in 3.5–7.5 ton trucks and stored in the plant in closed chambers with low storing times, so no air emissions were considered. The transport distances were 90, 110, 150 and 90 km for FW, MW, BW and VW, respectively. 2.3.2. Subsystem 02: pre-treatment Prior to anaerobic digestion, pasteurisation (1 h, 70 °C) was conducted in order to comply with the Spanish legislation (RD 1528/ 2012, 2012). The thermal energy required for this step was provided from the heat produced in the combined heat and power (CHP) unit installed in the biogas plant and only electricity use for pumping from storage tanks to pasteurisation unit was considered, ranging from 0.03 to 0.05 kWh/ton (Enconsult, 2004). 2.3.3. Subsystem 03: anaerobic (co-)digestion A continuous stirred tank reactor (CSTR) of 7,000 m3 operating in mesophilic range (37 °C) with a HRT of 15–20 days was considered. This implied a daily flow ranging from 351 to 447 m3/d, depending on the scenario considered (Table 2). The energy required for pumping the feeding into the digester varied from 0.12 to 0.15 kWh/ton depending on the inlet flow (Table 2) and 220 MWh were needed for stirring the digester (Enconsult, 2004). The reactor temperature was maintained at 37 °C and the losses produced were compensated by the influent coming from the prior pasteurisation (70 °C). The experimental data used in this work were obtained in lab (M, C1, C2) and pilot scale (C3) operations and average values are shown in Table 2. Biogas yielded from 1.74 × 106 (M) to 6.01 × 106 m3/FU (C2), with an average methane content ranging from 55 to 67% (Table 2). Likewise, the content of CO2 varied from 32 to 44%. For all the scenarios, the content of N2, H2S and NH3 in biogas was set at 1%, 0.05% and 0.01%, respectively. A small fraction of the biogas produced (1.5%) was assumed to be lost and released to the atmosphere

478

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

Table 2 Average operational parameters for the steady state of each scenario (M: mono-digestion of pig manure (PM); C1, C2, C3: co-digestion of PM with different co-substrates; FW: fish wastes; MW: molasses wastes; BW: biodiesel wastes; VW: vinasses wastes; OLR: organic loading rate; HRT: hydraulic retention time; Q: daily flow; COD: chemical oxygen demand). Biogas and digestate characteristics of scenario C2 were estimated by data reported by Regueiro et al. (2012); the characteristics for scenarios C1 and C3 were unpublished data. Scenario

M (n = 24) C1 (n = 29) C2 (n = 35) C3 (n = 65) a

Co-substrates

– FW, MW BW VW

Digester performance

Biogas characteristics

OLRPM (kg COD/m3 d)

OLRco-substrates (kg COD/m3 d)

HRT (d)

Q (m3/d)

CODremoval (%)

Biogas (×106 m3/FU)

CH4 (%)

2.4 2.4 2.4 2.4

– 1.6 3.6 2.4

20.0 19.0 19.0 15.4

351 369 369 447

52 75 79 79

1.74 3.65 6.01 4.34

55 63 60 67

± ± ± ±

Digestate characteristics

6 2 3 1

CO2 (%)

Energy (MWhe)

COD (g O2/kg)a

N-NH+ 4 (g N/kg)a

P-PO3− 4 (mg P/kg)a

44 36 39 32

7,196 9,036 14,122 11,298

23.0 12.0 14.6 18.0

2.70 2.50 2.56 3.05

50.8 124 48.3 186

Parameters expressed in wet basis.

(IPCC, 2006). COD and ammonium content in the digestate did not present large differences among the scenarios (12–23 g O2/kg and 2.50–3.05 g N-NH+ 4 /kg, respectively); however, phosphate levels were quite different, varying from 48 mg/kg (C2) to 186 mg/kg (C3). 2.3.4. Subsystem 04: energy recovery unit The biogas obtained was energetically valorised in a CHP unit, with electric and heat efficiencies of 40 and 50%, respectively (Pertl et al., 2010). Air emissions of CO, CO2, CH4 and non-methane volatile organic compounds (NMVOC) from biogas combustion were included in the analysis by using emission factors reported in literature (Doka, 2007). All the electricity generated was exported to the grid, displacing the production of energy according to the Spanish profile (Red Eléctrica, 2009). The produced heat was used on-site in the pasteurisation unit, which requires about 7.86 kWh/ton (Banks et al., 2011), and the remaining amount was either employed for the heating of the animal facilities (around 1,456 MWh/year) or exported to the factories located next to the farm (wood and metal industries). 2.3.5. Subsystem 05: digestate storage Digestate was stored in an open pool of 25,500 m3 (guaranteeing a storage time of 3 months) before its application on agricultural land. Electrical energy was required for pumping the digestate from the anaerobic digester to the storage tank, ca. 0.02 kWh/ton (Enconsult, 2004). Residual biogas production is likely to occur during digestate storage due to the presence of significant amounts of undigested organic matter and to the dissolved methane in the digestate (Gioelli et al., 2011). A value of 10% of the biogas yield achieved in the digester has been assumed for the total methane emissions during digestate storage in accordance with Poeschl et al. (2012). In addition, N-emissions

also occurred during digestate storage and they have been considered equal to those associated to PM storage (Table 3).

2.3.6. Subsystem 06: digestate application on land Due to the significant content of ammonium as well as the presence of phosphate, digestate is considered a suitable fertiliser for agricultural purposes as long as its application is done under controlled conditions (Calderón et al., 2011). The corresponding avoided fertilisers were ammonium sulphate ((NH4)2SO4) and diammonium phosphate ((NH4)2HPO4) as generic N and P2O5 sources, respectively, selected after a sensitivity analysis of 10 N-based and 6 P-based fertilisers from the Ecoinvent database (Rodriguez-Garcia et al., 2011). A plant uptake of 50% and 70% for the N and P contained in the digestate was assumed (Bengtsson et al., 1997). Those amounts of N and P not uptaken by the plant (50% and 30%, respectively) were susceptible of being the source of nutrient-related emissions. The latter were estimated by means of emission factors from literature: N to air (as NH3, N2O and N2) and to water as NO− 3 (Brentrup et al., 2000) (Doka, 2009). Moreover, the legal requirements and P to water as PO3− 4 related to organic fertilisers application (less than 170 kg N/ha·year, Directive 91/676/EC) were taken into account in all scenarios, resulting in a digestate dosage of 50 ton/ha.

2.3.7. Energy and materials provision Background data related to the production of chemicals, fertilisers and energy were obtained from the Ecoinvent database v2 (Ecoinvent, 2004). Regarding energy requirements, electricity used along the system came from the grid and national profile was consequently updated (Red Eléctrica, 2009).

Fig. 2. Process subsystems and boundaries. Dashed lines represent avoided products (energy or fertiliser).

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483 Table 3 National factors used for direct emissions from manure and digestate storage (fattening pigs between 20 and 100 kg). from PRTR (2013). Pollutant

479

3. Results and discussion 3.1. Environmental performance of AcoD of PM with wastes vs. PM monodigestion

Emission factor

a

CH4 (%COD) NH3 (%N-NH+ 4 ) N2O (%N-NH+ 4 )

5.7 23.9 0.03

a The provided factor includes both outdoors storage and agricultural application; however, the great share of the emission would take place during storage and therefore all the emissions have been allocated there.

2.4. Life cycle impact assessment The methodology developed by the Centre of Environmental Science of Leiden University (CML 2001) was used (Guinée et al., 2002), considering six impact categories widely consolidated in the LCA of AD: AP, abiotic depletion potential (ADP), EP, global warming potential (GWP), ozone layer depletion potential (ODP) and photochemical oxidation potential (PCOP). The software used was the SimaPro, version 7.3.2 (http://www.pre.nl/simapro).

2.5. Sensitivity analysis In order to check the robustness of the environmental profile of the assessed scenarios, a sensitivity analysis was carried out on the type of avoided energy sources and on the COD concentration of pig manure.

2.6. Dependence between LCA results and operational parameters The square Pearson's correlation coefficient (R2) between a LCA selected category and each process variable was calculated for every subsystem (Fig. 2). To do so, for the subsystems, the results of each impact category in the four scenarios were plotted against a selected 3− set of process variables (HRT, OLR, NH+ 4 and PO4 ) and a linear relation was assumed when Pearson's correlation coefficient was higher than 0.95 (Gitelson et al., 1996). Fig. 3 shows a practical example of the procedure for the ADP obtained for subsystem 01 and HRT. In the case of linear correlation, the slope provides the qualitative and quantitative influence of a change in the process variable on the LCA result. A positive slope denotes an increase in the environmental impact (or decrease in the environmental benefit) by increasing a process variable, whereas a negative slope means a decrease in the environmental impact (or increase in the environmental benefit) with the increment of the process variable. Moreover, the value of the slope gives information about the magnitude of the change in the environmental result, but in order to establish comparisons among the different impact categories and subsystems, a normalisation step is required. To do so, each obtained slope was divided by the absolute result of the reference scenario (M) (Fig. 3).

In general, the AcoD of PM with different co-substrates (scenarios C1, C2, C3) showed better net environmental performance than mono-digestion of PM (scenario M), except when AP, EP and GWP were assessed in scenario C3 (Fig. 4). The mitigation of the impacts obtained by AcoD were in accordance with the studies dealing with anaerobic co-digestion of livestock manure with residues (De Vries et al., 2012; Poeschl et al., 2012). Detailed LCA results for all the scenarios are given in the Supplementary information (Table SI2). In all LCA categories, the best results corresponded to scenario C2, probably due to the highest and readily biodegradable COD content of BW (1,679 g O 2 /kg, Table 1), and the highest OLR applied (6 kg COD/m 3 d), enabling a higher biogas production. Below, a more detailed analysis of each category is presented: a. AP and EP: Both categories displayed a similar profile (Fig. 5a and b), showing higher impacts (subsystems 01, 05 and 06) than environmental credits (subsystem 04: CHP), with the exception of the digestate application in C3, whose contribution is much higher in EP (58%) than in AP (43%). No differences were observed among scenarios concerning subsystem 01 (Fig. 5a and b) as PM load was the same in all the scenarios and no emissions during the storage of co-substrates were considered. Moreover, these categories were not so much influenced by the substrate transport, but mostly by the direct emissions, mainly ammonia, produced during PM storage. Regarding both digestate storage and application, scenarios M, C1 and C2 had similar contributions, explained by analogous nitrogen concentration in the digestate and HRT applied during AcoD (2.50– 3 2.70 g N-NH+ 4 /L and 351–369 m /d, respectively, Table 2). Nevertheless, because of the contribution by digestate application, scenario C3 exhibited greater net impacts than even scenario M. The higher + NH+ 4 content in the digestate (3.05 g N-NH4 /L, Table 2) and the 3 higher volume of co-substrate treated (96 m /d, Table 2) in scenario C3 worsened the environmental profiles of subsystems 05 and 06 by increasing the emissions of ammonia (AP and EP) and nitrate leaching (EP). Therefore, AP and EP were clearly affected by changes in HRT (related to flow-rate) and ammonium concentration. The relevance of digestate application in EP has been already highlighted by some authors (Börjesson and Berglund, 2007; Dressler et al., 2012) and, in particular, nitrate leaching was identified as the main contributor to EP by Rehl et al. (2012). Finally, although phosphate discharge can play an important role on EP, in this case it was not significant due to the low phosphate concentrations in the digestate (Table 2). b. ADP and ODP. Both categories showed similar profiles (Fig. 5c and d), presenting net environmental credits, which ranged from − 43,434 to − 89,323 kg Sb eq./FU for ADP and − 0.386 to − 0.816 kg CFC-11 eq./FU for ODP. The subsystem 04 (energy

Fig. 3. Practical example of how the linear dependence between LCA results and certain process variable has been calculated.

480

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

These impacts were addressed by the flow treated (Table 2), i.e., the lower the HRT, the higher the impact. Digestate storage and application completed this impact, mainly due to CH4 and N2O emissions. Methane emissions during the digestate storage are related to the biogas production (1.74 × 106 (M)–6.01 × 106 (C2) m3 biogas/FU, Table 2), since the former was considered a fraction of the latter. This explains the higher contribution of digestate storage to GWP and PCOP for C2. Digestate application was only relevant in GWP, mainly due to the indirect emissions of N2O produced during digestate spreading.

3.2. Sensitivity analysis

Fig. 4. Comparative results of the four scenarios under study, using scenario M as baseline (i.e. 100% for AP, EP, GWP and PCOP and −100% for ADP and ODP).

recovery unit) provided the most of benefits by replacing other energy sources since these categories are directly related to energy production. The environmental benefits obtained during the biogas recovery (subsystem 04) followed the same trend (C2 N C3 N C1 N M) as the OLR applied in each scenario (Table 2). Considering appropriate COD removal efficiencies (50–75%), the higher the OLR applied, the greater the biogas production (Table 2). The environmental consequences of producing electricity were calculated, demonstrating that ADP reached a global benefit ranging from − 625,952 (C3) to − 695,761 kg Sb eq./MWhe_produced (C2) and for ODP varied from −4.79 to −6.36 kg CFC-11 eq./MWhe_produced. The critical subsystem for these categories was subsystem 01, because of the transport of the required (co-)substrates to the plant. C3 presented the highest impacts, due to the larger volumes of VW treated and, although C1 and C2 had the same HRT (Table 2), the transport distance was higher in C2, resulting in greater impacts. c. GWP and PCOP. Both categories exhibited similar profiles and net environmental impacts (Fig. 5e and f), especially provoked by the feed transport and storage (up to 45% in GWP and 57% in PCOP).

Since the source of energy export assumed and the high variability of the organic matter concentration in PM may affect the environmental profile, both aspects were subjected to a sensitivity analysis with the aim of assessing the robustness of the results.

3.2.1. Energy export The replaced electricity was assumed to be produced according to the current energy production mix in Spain (Red Eléctrica, 2009), but results are likely to be affected if the replacement of a particular energy source is considered. Following a marginal approach (Weidema et al., 1999), the replacement of hard coal was assumed and recalculations on subsystem 04 (biogas recovery unit) were done and included in the whole process. As expected, the environmental profile improved with the substitution of hard coal (Table 4a), except for ODP, where impact values increased by 50% (Table 4a) due to related emissions of halon generation (Halon 1301) derived from the electricity production within the energy mix. The highest benefit of this change (negative values in Table 4a) was observed in those categories where the subsystem 04 had a major contribution (ADP, GWP and PCOP, Fig. 5c, e and f, respectively) and hardly changed in those where the recovery biogas step had small or not influence (AP and EP, Fig. 5a and b, respectively). Overall, the marginal approach enhances the environmental profile of the process and this analysis shows that it is important to study the current energy market in order to establish accurately the environmental impacts of anaerobic co-digestion.

Fig. 5. Comparison of the four scenarios studied (M, C1, C2, and C3) among subsystems and impact categories. Subsystem 03 was excluded due to the negligible values obtained.

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

3.3. Identification of key operational parameters of AD under LCA perspective Taking into account that the operational parameters (HRT, OLR…) were different among the scenarios, the effect of the cosubstrate on the environmental performance of AcoD could not be clearly identified. Furthermore, considering that the co-substrates had similar biodegradability rates and extensions (70–80%)(Regueiro et al., 2014) and the digestate characteristic were in the same range (excluding phosphate concentration, but it displayed a negligible contribution on the environmental performance), the effect of the cosubstrate nature was uncoupled to the LCA results. Thus, the relationship between the operational parameters and the environmental performance was established. Taking into account the aforementioned results, four operational parameters were identified as crucial: HRT, OLR and both NH+ 4 and content in the digestate. Table 5 shows the R2 values for the four PO3− 4 process variables established in each subsystem and impact categories, being linear dependence highlighted in colour. Subsystems 01 and 02 were exclusively dependent on HRT, while subsystem 04 was complete3− conly related to OLR. Subsystem 06 was influenced by NH+ 4 and PO4 tent in the digestate and subsystem 05 was linked to HRT and OLR. Only three cases were not correlated: subsystem 03, categories AP, GWP and PCOP, but a relatively high dependency with OLR was found for GWP and PCOP (values close to 0.9 in Table 5). The values of the slopes associated to the identified lineal dependencies (Table 5) are presented in the Supplementary information (Table SI3), while Table 6 shows the normalised slopes. 3.3.1. HRT All correlations with HRT showed a negative slope, i.e. the higher the HRT applied, the lower the impact attained. This could be explained since a higher HRT implies small feeding flow-rates, and consequently, lower transport and energy requirements, therefore, lower emissions. ADP and ODP were the most sensitive categories to HRT for subsystem 01 (−0.613 d−1 and −0.595 d−1, respectively), because both categories were dominated by the transport of substrates. On the contrary, all categories were affected equally by HRT in the subsystems 02, 03

Table 4 Effect of a marginal energy perspective (a) and the organic concentration of pig manure (CODPM, b) on LCA results (expressed as the percentage change respect to the results obtained for scenario M).

AP

EP

ADP

−

−

−

−13%

−8%

−66%

12%

7%

59%

−

−

−12%

−7%

ODP

GWP

PCOP

a Energy mix (base case) Marginal perspective

−

− −58%

−78%

88%

72%

96%

−

−

−

−59%

−88%

−72%

50%

b COD (g O2/kg)

3.2.2. Organic matter concentration in PM The PM considered in this work had a COD concentration of 65.0 ± 25.7 g O2/kg (Table 1). This significant variability is due to the strong dependence on weather conditions, obtaining more diluted streams during winter months (rainy time) and more concentrated ones in summer period (drier time). As expected, all impacts were lower when using the most concentrated PM (90.7 g O2/kg), while diluted streams (39.3 g O2/kg) led to worse environmental results (Table 4b). Four subsystems (1, 3, 4 and 5) were affected by this assumption, which are those with COD-related emissions. The four subsystems were equally influenced in the GWP and PCOP categories due to the methane emissions. On the contrary, AP, EP, ADP and ODP categories had only significance in subsystem 04. Regarding the net impacts for the entire process (Table 4b), the COD content caused a slight effect on AP and EP. Therefore, considering that subsystem 04 (energy recovery unit) was the most sensitive to the aforementioned categories, and both AP and EP were not clear related to energy production, variations of COD concentration were not significant (7–12%) due to these categories. In addition, the resting four categories (ADP, ODP, GWP and PCOP) were clearly affected, ranging from 59 to 96% of change. Results are strongly related to the COD losses, causing methane emissions, during pig manure storage. Therefore, the recommendation of close storage pools must be followed to decrease these losses. Concomitantly this would provoke a reduction on the methane emissions improving the environmental profile of the entire process (Gioelli et al., 2011; Mezzullo et al., 2013).

481

39.3 65.0 (base) 90.7

−96%

and 05 with a slope of − 0.128, − .004 and − 0.061 d−1, respectively (Table 6). 3.3.2. OLR The biogas recovery step (subsystem 04) showed a similar and inverse (negative slope) linear dependence with the OLR applied to the digester in all categories (−0.271 to −0.302 (kg COD/m3 d)−1, Table 6). This is because incrementing OLR provokes the increase of biogas yield (maintaining the same COD removal efficiency) and thus the environmental profile trends to decrease. OLR also affected negatively (positive slope) the GWP and PCOP results associated to digestate storage (subsystem 05): 0.265 (kg COD/m3 d)−1 and 0.270 (kg COD/m3 d)−1, respectively, as a result of greater emissions of methane from the undigested organic matter occurring at high OLR. 3.3.3. NH+ 4 The NH+ 4 content in the digestate had only negative effect (positive slope) on digestate application (subsystem 06) in AP, EP and GWP, but the magnitude of the effect was the greatest obtained: 1.008– 1.360 (g N/L)− 1 (Table 6). Although the higher ammonium content allows an increased production of avoided fertilisers (characterised as environmental benefit), the ammonia emissions and the nitrate leaching dominated this category. 3.3.4. PO3− 4 Digestate application (subsystem 06) was also influenced by phosphate content in the digestate on ADP, ODP and PCOP categories, but in this case, the slopes were negative (Table 6), i.e. the higher the content of the digestate, the lower the impact. The reason lies in PO3− 4 the avoided P-based fertilisers, which had a detectable contribution only in ADP, ODP and PCOP categories (Fig. 5). This approach permits the evaluation of the effect of process variables on the environmental performance of anaerobic co-digestion processes, thus adding the environmental perspective to the technical feasibility analysis. For example, running an anaerobic digester at low HRT allows increasing the volume of waste treated, thus minimising the investments costs. However, increasing the HRT in one or two days would favour the environmental credits of the process, while not affecting the technical performance. 4. Conclusions Anaerobic co-digestion of pig manure with different wastes not only enhances the techno-economic performance of the process, but also the

482

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483

Table 5 Square Pearson's correlation coefficient (R2) established between LCA results and hydrau3− lic retention time (HRT) (a), organic loading rate (OLR) (b) and N-NH+ 4 (c) and P-PO4 (d) concentration in the digestate for each subsystem (see Fig. 2 for subsystem description). The linear relation (Pearson’s correlation coefficient higher than 0.95) was represented by shaded areas.

01

02

03

04

05

06

AP EP ADP ODP GWP PCOP

HRT 0.989 0.989 0.989 0.989 0.989 0.989

0.997 0.997 0.997 0.997 0.997 0.997

0.423 1.000 0.955 0.955 0.008 0.009

0.125 0.125 0.125 0.125 0.125 0.124

0.999 0.999 0.999 0.999 0.125 0.125

0.902 0.881 0.893 0.888 0.866 0.865

AP EP ADP ODP GWP PCOP

OLR 0.230 0.230 0.230 0.230 0.230 0.231

0.131 0.131 0.131 0.131 0.131 0.131

0.891 0.161 0.042 0.042 0.878 0.879

0.961 0.961 0.960 0.959 0.961 0.963

0.149 0.143 0.143 0.143 0.962 0.962

0.037 0.024 0.039 0.037 0.031 0.030

AP EP ADP ODP GWP PCOP

N-NH4+ 0.683 0.683 0.683 0.683 0.683 0.683

a

b

c

d AP EP ADP ODP GWP PCOP

P-PO43− 0.691 0.691 0.691 0.691 0.691 0.690

Acknowledgments This research was supported by the Ministry of Economy and Competitiveness through COMDIGEST (CTM2010-17196) project and the Ramón y Cajal contract (RYC-2012-10397) to Dr. Marta Carballa. The authors belong to the Galician Competitive Research Group, programme co-funded by FEDER (GRC 2013-032). Authors also want to thank Carolina Alfonsín and Santiago García-Gen for their support in the LCA results management and data provision, respectively. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2014.07.127. References

0.754 0.754 0.754 0.754 0.754 0.754

0.107 0.695 0.773 0.773 0.052 0.052

0.001 0.001 0.001 0.001 0.001 0.001

0.733 0.737 0.737 0.737 0.001 0.001

0.955 0.950 0.585 0.579 0.957 0.556

0.767 0.767 0.767 0.767 0.767 0.767

0.010 0.790 0.863 0.863 0.048 0.047

0.000 0.000 0.000 0.000 0.003 0.025

0.631 0.664 0.770 0.770 0.056 0.065

0.654 0.651 0.973 0.975 0.605 0.985

environmental profile, remarking that the use of wastes instead of energy crops as co-substrates should be pursued. Furthermore, the storage of pig manure in close chambers is much recommended to decrease the environmental impacts. Although the operational conditions of the scenarios evaluated in this study did not enable to address properly the influence of co-substrate characteristics on the environmental performance, the best environmental profile obtained by treating biodiesel residues as co-substrate anticipates that readily biodegradable substrates with complementary characteristics to pig manure should be used. To the best of our knowledge, this is the first study attempting to correlate the environmental profile of a process with the operational conditions. Although this requires further optimization, LCA results are linked to HRT, OLR and nutrient content in the digestate. This entails a Table 6 Normalised slope calculated dividing the net slope (Table SI3) by the absolute value of reference scenario (Table SI2) for the corresponding category and subsystem ( : HRT, d−1; : OLR, (kg COD/m3 d)−1; : N-NH4+(g N/kg)−1, : P-PO43−, (mg P/kg)−1).

01

02

03

04

05

06

AP

−0.006

−0.128

n.d.

−0.272

-0.061

1.008

EP

−0.007

−0.128

-0.004

−0.271

-0.061

1.360

ADP

−0.613

−0.128

-0.004

−0.291

-0.061

−0.010

ODP

−0.595

−0.128

-0.004

−0.302

-0.061

−0.012

GWP

−0.033

−0.128

n.d.

−0.292

0.265

1.030

PCOP

−0.014

−0.128

n.d.

−0.278

0.270

−0.026

n.d.: no dependency.

step towards the addition of the environmental perspective to the techno-economic aspects in the design, monitoring and control of anaerobic (co-)digestion facilities.

Álvarez JA, Otero L, Lema JM. A methodology for optimising feed composition for anaerobic co-digestion of agro-industrial wastes. Bioresour Technol 2010;101:1153–8. Angelidaki I, Ellegaard L. Codigestion of manure and organic wastes in centralized biogas plants. Appl Biochem Biotechnol 2003;109:95–105. Astals S, Nolla-Ardèvol V, Mata-Alvarez J. Anaerobic co-digestion of pig manure and crude glycerol at mesophilic conditions: biogas and digestate. Bioresour Technol 2012;110: 63–70. Bacenetti J, Negri M, Fiala M, González-García S. Anaerobic digestion of different feedstocks: impact on energetic and environmental balances of biogas process. Sci Total Environ 2013;463–464:541–51. Banks CJ, Chesshire M, Heaven S, Arnold R. Anaerobic digestion of source-segregated domestic food waste: performance assessment by mass and energy balance. Bioresour Technol 2011;102:612–20. Baumann H, Tillman AM. The hitch hiker's guide to LCA — an orientation in life cycle assessment methodology and application. Lund: Studentlitteratur; 2004. Bengtsson M, Lundin M, Molander S. Life cycle assessment of wastewater systems: case studies of conventional treatment, urine sorting and liquid composting in three Swedish municipalities. Chalmers University of Technology; 1997. Börjesson P, Berglund M. Environmental systems analysis of biogas systems—part II: the environmental impact of replacing various reference systems. Biomass Bioenergy 2007;31:326–44. Brentrup F, Küsters J, Lammel J, Kuhlmann H. Methods to estimate on-field nitrogen emissions from crop production as an input to LCA studies in the agricultural sector. Int J Life Cycle Assess 2000;5:349–57. Calderón MPB, Méndez JAA, Muñoz MÁB, Carrillo RC. Guía de utilización agrícola de los materiales digeridos por biometanización: CSIC; 2011 [In Spanish]. Clesceri LS, Greenberg AE, Eaton AD. Standard methods for the examination of water and wastewater; 1998. De Vries JW, Vinken TMWJ, Hamelin L, De Boer IJM. Comparing environmental consequences of anaerobic mono- and co-digestion of pig manure to produce bio-energy — a life cycle perspective. Bioresour Technol 2012;125:239–48. Devesa-Rey R, Vecino X, Varela-Alende JL, Barral MT, Cruz JM, Moldes AB. Valorization of winery waste vs. the costs of not recycling. Waste Manag 2011;31:2327–35. Directive 91/676/EC. Concerning the animal health conditions governing the placing on the market of aquaculture animals and products. European Parlament and Council; 1991. Doka G. Life cycle inventories of waste treatment services. Ecoinvent report no. 7. Düberdorf, Switzerland: Swiss Centre for Life Cycle Inventories; 2007. Doka G. Wastewater treatment. Life cycle inventories on waste treatment. Dübendorf, Switzerland: Swiss Centre for Life Cycle Inventories; 2009. Dressler D, Loewen A, Nelles M. Life cycle assessment of the supply and use of bioenergy: impact of regional factors on biogas production. Int J Life Cycle Assess 2012;17: 1104–15. Ecoinvent. Swiss centre for life cycle inventories (Ecoinvent centre). Ecoinvent database. Dübendorf: Ecoinvent Centre; 2004. Enconsult. A technical feasibility study for Leicestershire. Anaerobic digestion — a renewable energy resource. Leicestershire county. Downloaded from: http://www. leics.gov.uk/ad_feasibility_study_main_report.pdf, 2004. [Last access: November 2013]. Eurostat. http://epp.eurostat.ec.europa.eu/tgm/graph.do?tab=graph&plugin=1&pcode= tag00042&language=en&toolbox=data, 2013. [Last access: November]. Fang C, Boe K, Angelidaki I. Anaerobic co-digestion of desugared molasses with cow manure; focusing on sodium and potassium inhibition. Bioresour Technol 2011; 102:1005–11. García-Gen S, Lema JM, Rodríguez J. Generalised modelling approach for anaerobic co-digestion of fermentable substrates. Bioresour Technol 2013;147:525–33. Gioelli F, Dinuccio E, Balsari P. Residual biogas potential from the storage tanks of non-separated digestate and digested liquid fraction. Bioresour Technol 2011;102: 10248–51.

I. Rodriguez-Verde et al. / Science of the Total Environment 497–498 (2014) 475–483 Gitelson AA, Merzlyak MN, Lichtenthaler HK. Detection of red edge position and chlorophyll content by reflectance measurements near 700 nm. J Plant Physiol 1996;148: 501–8. González-García S, Bacenetti J, Negri M, Fiala M, Arroja L. Comparative environmental performance of three different annual energy crops for biogas production in Northern Italy. J Cleaner Production 2013;43:71–83. Guinée JB, Gorrée M, Heijungs R, Huppes G, Kleijn R, de Koning A, et al, editors. Handbook on life cycle assessment: operational guide to the ISO standards. Series: eco-efficiency in industry and scienceDordrecht. Hardbound: Kluwer Academic Publishers 1-40200228-9; 2002. [Paperback, ISBN 1-4020-0557-1]. Holm-Nielsen JB, Al Seadi T, Oleskowicz-Popiel P. The future of anaerobic digestion and biogas utilization. Bioresour Technol 2009;100:5478–84. IPCC (Intergovernmental Panel on Climate Change). N2O emissions from managed soils, and CO2 emissions from lime and urea application. In: Eggleston HS, Buendia L, Miwa K, Ngara T, Tanabe K, editors. IPCC guidelines for national greenhouse gas inventories. National Greenhouse Gas Inventories Programme, Hayama, Japan: IGES; 2006. [chapter 11]. ISO 14040. International organization for standardization, ISO 14040. Environmental management — life cycle assessment — principles and framework, Geneva; 2006. ISO 14044. International Organization for Standardization, ISO 14044. Environmental management — life cycle assessment — requirements and guidelines, Geneva; 2006. Jiménez AM, Borja R, Martı́n A. A comparative kinetic evaluation of the anaerobic digestion of untreated molasses and molasses previously fermented with Penicillium decumbens in batch reactors. Biochem Eng J 2004;18:121–32. Lansche J, Müller J. Life cycle assessment of energy generation of biogas fed combined heat and power plants: environmental impact of different agricultural substrates. Eng Life Sci 2012;12:313–20. Mata-Alvarez J, Dosta J, Fonoll X, Romero M, Peces M, Astals S. Anaerobic co-digestion: a review of achievements and perspectives. Proceedings of the 13th IWA World Congress on anaerobic digestion, June 25–28 2013, Santiago de Compostela, Spain; 2013. McDougall F, White P, Franke M, Hindle P. Integrated solid waste management: a life cycle inventory. Oxford, U. K.: Blackwell Science Ltd; 2001 Mezzullo WG, McManus MC, Hammond GP. Life cycle assessment of a small-scale anaerobic digestion plant from cattle waste. Appl Energy 2013;102:657–64. Misi SN, Forster CF. Batch co-digestion of multi-component agro-wastes. Bioresour Technol 2001;80:19–28. Nasir IM, Mohd Ghazi TI, Omar R. Anaerobic digestion technology in livestock manure treatment for biogas production: a review. Eng Life Sci 2012;12:258–69. Pantaleo A, Gennaro BD, Shah N. Assessment of optimal size of anaerobic co-digestion plants: an application to cattle farms in the province of Bari (Italy). Renew Sustain Energy Rev 2013;20:57–70.

483

Pertl A, Mostbauer P, Obersteiner G. Climate balance of biogas upgrading systems. Waste Manag 2010;30:92–9. Poeschl M, Ward S, Owende P. Environmental impacts of biogas deployment — part II: life cycle assessment of multiple production and utilization pathways. J Clean Prod 2012; 24:184–201. Probiogas. Manual Sobre Co-digestión Anaerobia Agroindustrial. www.probiogas.es, 2010. [In Spanish]. PRTR. Registro estatal de emisiones y fuentes contaminantes. http://www.prtr-es.es/, 2013. [last access: November]. Quispe CAG, Coronado CJR, Carvalho Jr JA. Glycerol: production, consumption, prices, characterization and new trends in combustion. Renew Sustain Energy Rev 2013; 27:475–93. RD 1528/2012. Real Decreto por el que se establecen las normas aplicables a los subproductos animales y los productos derivados no destinados al consumo humano; 2012 [In Spanish]. RD 3483/2000. Real Decreto por el que se establecen normas básicas de ordenación de las explotaciones porcinas; 2000 [In Spanish]. Red Eléctrica. Red Eléctrica de España. La Energía en España 2008; 2009 [Alcobendas, In Spanish]. Regueiro L, Carballa M, Álvarez JA, Lema JM. Enhanced methane production from pig manure anaerobic digestion using fish and biodiesel wastes as co-substrates. Bioresour Technol 2012;123:507–13. Regueiro L, Veiga P, Figueroa M, Lema J, Carballa M. Influence of transitional states on the microbial ecology of anaerobic digesters treating solid wastes. Appl Microbiol Biotechnol 2014;98:2015–27. Rehl T, Lansche J, Müller J. Life cycle assessment of energy generation from biogas— attributional vs. consequential approach. Renew Sustain Energy Rev 2012;16: 3766–75. Riaño B, Molinuevo B, García-González MC. Potential for methane production from anaerobic co-digestion of swine manure with winery wastewater. Bioresour Technol 2011;102:4131–6. Rodriguez-Garcia G, Molinos-Senante M, Hospido A, Hernández-Sancho F, Moreira MT, Feijoo G. Environmental and economic profile of six typologies of wastewater treatment plants. Water Res 2011;45(18):5997–6010. Serrano A, Siles JA, Chica AF, Martín MÁ. Agri-food waste valorization through anaerobic co-digestion: fish and strawberry residues. J Clean Prod 2013;54:125–32. Weidema B, Frees N, Nielsen A-M. Marginal production technologies for life cycle inventories. Int J Life Cycle Assess 1999;4:48–56. Xie S, Lawlor PG, Frost JP, Hu Z, Zhan X. Effect of pig manure to grass silage ratio on methane production in batch anaerobic co-digestion of concentrated pig manure and grass silage. Bioresour Technol 2011;102:5728–33.