Research paper

profilelordsaldana
2018RCRLCAportableanaerobicDigestionMixedCardoordAndFOODwaste.pdf

Contents lists available at ScienceDirect

Resources, Conservation & Recycling

journal homepage: www.elsevier.com/locate/resconrec

Full length article

Life cycle assessment of portable two-stage anaerobic digestion of mixed food waste and cardboard

Claudia Isolaa, Heidi L. Sieverdinga, Caitlin M. Asatob, Jorge Gonzalez-Estrellab, David Litzenc, Patrick C. Gilcreaseb, James J. Stonea,⁎

a Department of Civil and Environmental Engineering, South Dakota School of Mines & Technology, Rapid City, SD, 57701, USA b Department of Chemical and Biological Engineering, South Dakota School of Mines & Technology, Rapid City, SD, 57701, USA c Litzen Process Consulting Inc., Rapid City, SD, 57701, USA

A R T I C L E I N F O

Keywords: Biogas production Anaerobic digestion Life cycle assessment Organic waste treatment Waste reduction

A B S T R A C T

Biogas produced from organic waste can reduce waste and produce renewable energy and is a viable waste treatment alternative for remote encampments. Portable, small-scale anaerobic digestion (AD) units can be used to sustainably produce biogas in remote areas and reduce landfilled waste. This project investigated the life cycle impacts of a portable AD unit and the effects of organic loading rate (8–32 g chemical oxygen demand (COD) L−1

d−1) and waste composition (food versus cardboard waste ratios of 35:65 and 65:35) on biogas production efficiency. Optimal biogas production was obtained using a 65% food to 35% cardboard waste mixture and a mid-range organic loading rate (16 g COD L−1 d−1); this scenario also yielded the lowest climate change impact [37.4 ± 0.7 g CO2 eq per kg COD waste] due to greater biogas conversion efficiency. However, the overall life cycle impacts of biogas production were not significantly affected by waste mixtures or feed rates in the AD portable system and experiments evaluated. Life cycle impacts due to portable AD processing were overall agnostic to feedstock variability. Thus, waste type and volume variations generated by encampments with fluctuating populations can likely be accommodated by the portable AD system without substantially affecting short term process sustainability. Portable AD system biogas generation rates were comparable to conventional, full-scale waste to energy facilities, while combustion impacts were more sustainable than those associated with conventional fossil fuels. Portable AD units represent a sustainable energy resource, waste reduction, and landfill alternative for remote areas.

1. Introduction

Solid waste is a critical global issue and is epitomized by United States’ (US) waste production. In 2014, the US generated 258 million tonnes of municipal solid waste (MSW), averaging two kg (4.4 lbs.) per person per day (USEPA, 2016). Most this MSW was organic waste in the form of food waste (FW) and paper and paperboard (PPB), 15% and 27% respectively. Organic wastes composed 36% of US landfilled MSW (136 million tonnes per year) after removal of recyclables (USEPA, 2016). Landfilling has potential consequences such as greenhouse gas emissions, gas and leachate generation, as well as possible health ha- zards, fires and explosions, damage to vegetation, landfill settlement, and groundwater and air pollution (El-Fadel et al., 1997; Emberton and Parker, 1987). Therefore, biodegradable organic waste represents not only a potential environmental hazard, but also wasted energy poten- tial. Producing energy from waste is an alternative management option

which may reduce the environmental impact of waste disposal. Anae- robic digestion (AD) is an organic waste treatment option from which renewable biogas (i.e. methane) can be collected in the absence of oxygen to use in energy and/or heat generation (Tagliaferri et al., 2016). According to El-Fadel et al. (1997), AD has advantages over landfilling with gas collection because it is more efficient (volatiles elimination efficiency of ∼70% for AD versus ∼7% for landfilling with gas collection), occupies a smaller footprint, and presents smaller pol- lution risks. Therefore, a portable AD system may be an alternative option for waste treatment in small and isolated encampments such as military forward operating bases, temporary refugee camps, and dis- aster areas. Installation of advanced solid waste treatments in isolated areas is hampered by community investment, ability, and willingness as well as camp site duration (Medina and Waisner, 2011). Portable AD treatment technologies are an opportunity to reduce solid waste volume and generate energy in minimal or damaged infrastructure areas,

https://doi.org/10.1016/j.resconrec.2018.08.008 Received 7 July 2017; Received in revised form 9 February 2018; Accepted 13 August 2018

⁎ Corresponding author. E-mail address: [email protected] (J.J. Stone).

Resources, Conservation & Recycling 139 (2018) 114–121

Available online 22 August 2018 0921-3449/ © 2018 Elsevier B.V. All rights reserved.

T

reducing also the need to transport fuel across conflict or disaster zones (Asato et al., 2016).

Many factors affect AD process design and operational efficiencies, such as feedstock characteristics, reactor construction, and operation conditions (Hawkes, 1980; Zhang et al., 2007). To equate differing feedstock qualities for methane conversion under anaerobic conditions, the chemical oxygen demand (COD) is used because it can be used to compare production potential. Food waste contains high amounts of water-soluble organics that can rapidly convert to volatile fatty acids (VFAs) at early stages of digestion (Cho et al., 1995) and cause a pH drop detrimental to methane generation. Meanwhile, PPB products such as cardboard (CB) are lignocellulosic feedstocks that do not readily biodegrade under anaerobic conditions (Verma, 2002). As a result, FW mixed with CB has been identified as a promising substrate for methane (CH4) production; however, differences in degradability were not con- ducive to synergistic co-digestion in a single reactor (Asato et al., 2016). The faster digestion of mixtures with a high FW composition inhibited methanogenesis due to VFA accumulation (Asato et al., 2016).

This tendency towards VFA accumulation makes multi-stage AD more appropriate to maximize CH4 production. Multi-stage AD is a well-established technology wherein VFA production occurs primarily in a separate reactor from CH4 production, allowing VFAs to accumu- late in the first stage without inhibiting the methanogenic microbes (Pohland and Ghosh, 1971). This allows for higher loading rates of easily-degraded feedstocks such as FW. Numerous benefits and chal- lenges are associated with designing AD systems and stages. Factors such as temperature, pH, C/N ratios, feedstock variability, organic loading rates (OLR), retention time, accelerant use, cost, reactor design, regulations, and treatment targets and goals affect AD design selection and treatment efficiency (Hagos et al., 2017; Mao et al., 2015; Xiao et al., 2018). This portable AD unit prototype was designed with the goals of optimizing CH4 production, simplifying operation, and im- proving process stability.

Methodologies such as LCA are used to assess environmental im- pacts of a product or process from raw material to end of life (“cradle- to-grave”), evaluating all stages of materials processing, manufacturing, distribution, use, and disposal (Finkbeiner et al., 2006; Guinee, 2002). LCA has been used to evaluate various waste treatment scenarios and past studies have shown that AD provides a sustainable waste-to-energy option; however, no study to date has evaluated a portable AD. Re- search has demonstrated that AD waste energy recovery processes from food waste can be an economical (Ahamed et al., 2016) and/or sus- tainable (Bernstad Saraiva Schott et al., 2016) waste avoidance alter- native when compared to landfilling, incinerating, or composting (Arafat et al., 2015; Evangelisti et al., 2014; Opatokun et al., 2017; Xu et al., 2015). However, waste composition, such as high levels (> 5%) of oil or lipids in FW, has been shown to reduce the benefits of AD (Ahamed et al., 2016). Xu et al. (2015) performed an LCA of FW treatment options in China that identified diesel use during transpor- tation as a factor which increased FW disposal impacts, supporting the hypothesis that on-site AD waste processing systems could reduce transportation-related impacts. A review of food waste disposal LCAs showed that energy substitution and system boundary assumptions significantly affected findings, systems which included fossil fuel, en- ergy crop, or manure substitutions showed greater AD greenhouse gas (GHG) avoidances (−2084 to 28 kg CO2 eq per tonne wet FW) than those without (45–71 kg CO2 eq per tonne wet FW) (Bernstad Saraiva Schott et al., 2016). In addition, life cycle impacts are significantly af- fected by AD design (Xiao et al., 2018), waste feedstock blends or co- digestion (Edwards et al., 2017), OLR (Di Maria et al., 2016), and waste feedstock and biogas quality (Chiu and Lo, 2018). Waste is spatially and temporally variable in both quantity and quality, complicating eva- luation of waste management pathways (Pierie et al., 2016). The in- consistency of these factors makes AD comparison between locations and systems difficult, necessitating full life cycle understanding and caution when evaluating potential benefits.

The purpose of this study was to determine and evaluate the life cycle impacts of FW and CB co-digestion using a portable, multi-stage AD to process organic solid waste and generate biogas for local heating with respect to major environmental impacts, cumulative energy de- mand (CED), and waste prevention. The goal was to understand the sensitivity of the portable AD to feedstock and OLR changes, effect of scale, and heat source as well as identifying operating conditions which provided the best biogas yield and lowest impact. The environmental impact of portable AD generated biogas used as a heat source was compared to traditional heat sources such as natural gas, coal, and diesel to assess renewable fuel use. The efficiency of the portable AD was further evaluated by comparing its energy efficiency to traditional, full-scale waste disposal systems with energy recovery operations to assess scale impacts.

2. Methods

An LCA evaluation was conducted using an attributional approach to estimate environmental life cycle impacts of using AD treatment of FW and CB with feedstock quality expressed as COD and varying pro- cessing (loading) rates to produce enough biogas to generate 1 MJ of heat. Waste was assumed to be manually sorted prior to processing. Waste input into the portable AD and was assumed to be monitored for foreign objects and non-organic waste. Heat was assumed to be locally used to maintain AD operational temperature or for water heating and/ or steam generation. The LCA model was developed following the standards from International Organization for Standardization (ISO) 14040 and 14044 (ISO, 2006a, b), using SimaPro 8.2.0 modeling soft- ware (Pré Consultants, Netherlands) and EcoInvent 3.2 (EcoInvent, Switzerland) life cycle inventory database (Wernet et al., 2016) for specific process inventories.

The ASPEN Plus (v8.8; AspenTech, Bedford, Massachusetts) che- mical process optimization software was used to develop a portable AD model and prototype based on laboratory test results (Asato et al., 2017) and to parametrize LCA model inputs and biogas yields. The environmental impacts of AD waste treatment scenarios using varied FW and CB compositions and loading rates were evaluated. The results of these analyses were compared to business-as-usual scenarios avail- able within EcoInvent with two goals in mind, understanding the effects of scale on waste disposal alternatives and the impacts of switching from fossil fuels currently in use to biogas. The biogas generated by the treatment was theoretically used as heat source and compared to con- ventional small-scale heat sources such as coal, natural gas, and diesel. An energy demand versus recovery analysis was conducted to compare efficiency of a small-scale portable AD prototype with conventional large-scale landfill, incineration, and biogasification waste treatment processes available within EcoInvent.

2.1. Portable anaerobic digester

To parametrize the models, characterize varying waste composition and AD settings, and assist in scaling the AD prototype, laboratory tests conducted used a two-stage reactor developed to emulate planned prototype AD processes (Asato et al., 2017). In brief, the first stage was a continuously-stirred tank reactor (CSTR) with a three-liter benchtop fermentor (Applikon Biotechnology B.V.; Delft, Netherlands) and a working volume of one liter. The fermentor included a mechanical impeller, which was set to agitate continuously at 150 rpm. The head plate was fitted with a plastic inlet-outlet tube with an inner diameter of approximately 0.95 cm to accommodate flow of large suspended solids. The CSTR was seeded with five grams of volatile suspended solids (VSS) per liter of sludge. Feeding and effluent removal were performed daily with a peristaltic pump through the inlet-outlet tube. Influent and ef- fluent flow rates were both 0.5 L d−1 to maintain a steady working volume. The influent mixture contained FW and CB diluted in synthetic human wastewater (WW) to achieve the desired concentration. Biogas

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

115

production from the first stage reactor was measured using water dis- placement. The second stage reactor, an upflow anaerobic sludge blanket (UASB), was custom-blown by ChemGlass (Vineland, NJ, USA) and had a working volume of 0.5 L. The UASB reactor was seeded with 10 g VSS L−1 of granular sludge. Influent was fed from the CSTR into the UASB continuously using a peristaltic pump from a continuously mixed hydraulic buffering tank kept on ice to simulate refrigeration, while effluent was removed by overflow and retained in the disposal tank. The AD processes were operated at 35 °C except for the buffering tank which was kept cool to inhibit microbial growth between stages. Biogas generated by the second stage reactor was measured using water displacement. After both reactors were acclimated they were placed in series. Methane content of the biogas was determined using a gas chromatograph (Agilent 6890; Santa Clara, CA) with a flame ionization detector (GC-FID); details of the GC-FID method were reported by Asato et al. (2016). The specific non-combustible gases (likely mostly CO2 with some H2O, NH3, and H2S) generated by the AD process and stages were not analyzed. Non-combustible gases were passed through the boiler after combustion and vented to the atmosphere. Results from the laboratory tests were used to develop and calibrate an ASPEN Plus software model which was then used to construct a portable AD pro- totype.

The experimental scenarios (Table 1) considered in this study tested the sensitivity of two-stage small-scale AD to waste composition and feed rates. The sensitivity factors tested were FW fraction ( fFW ), CB fraction ( fCB), and substrate organic loading rate (OLR) as total che- mical oxygen demand (COD). Percentages of FW and CB inputs were also calculated on a COD basis.

2.2. System boundary and functional unit

All input wastes were considered byproduct wastes of other pro- cesses, meaning that the energy and resources used to produce the food and cardboard were not included in this analysis. This zero-burden approach has been taken because AD is an additional process beyond business-as-usual used to minimize waste disposal and recover energy. The biogas produced was assumed to be used in a boiler to generate heat. The AD processing duration and feedstock quality were varied according to Table 1. Fig. 1 illustrates the system boundaries considered in this study, from “cradle” (waste) to “grave” (combustion and use of biogas).

The assessment included pre-treatment and processing of waste (grinding, mixing, and inoculation), consumption of energy and re- sources for treatment, emissions to air, water and soil from the treat- ment process as well as emissions caused by combustion of biogas to produce heat. Construction was excluded from this study because the designed life of the AD unit was 20 years and the construction impacts amortized over this life span would be negligible. Transportation for collection and delivery of waste material to the portable AD was not considered because it was designed for small and/or isolated areas where waste disposal was assumed to be done individually by members

of the community on a localized scale. The portable AD units were designed to be dispersed throughout communities, operate without infrastructure, and be within walking distances of mess halls, disaster relief distribution sites, and residences. The functional unit used was 1 MJ/g-COD-waste substrate; based on one wet basis gram of FW and CB equating to 0.7 ± 0.1 and 1.2 ± 0.2 g-COD with a particulate inert (non-digestible) fraction of COD of 0% and 33%, respectively. The COD was measured via a closed-reflux colorimetric method according to standard methods for each waste category (APHA et al., 2012). A functional unit including energy and COD was selected because it re- flected the quality differences in the waste feedstocks. Materials and processes were assembled based on mass and equated using the same energy generation or potential of one MJ. The AD co-products of me- thane, sludge, treated effluent, and solid organic waste were econom- ically allocated (Table S5). Economic allocation was used because it reflected the expenses and waste disposal volume differences associated with end-of-life.

2.3. Inventory analysis

2.3.1. Waste composition The substrate composition was varied according to Table 1. In the

laboratory, synthetic FW with a C:N ratio of 16.9:1 was prepared to mimic an anticipated or common food mixture with high starch content using (%w/w, wet basis): canned pork and beans (35.6%), potato flakes (7.6%), and white bread (56.9%). Cardboard (CB) with a C:N ratio of 461:1 and a composition (%w/w, dry basis) of 53.8% cellulose, 13.2% hemicellulose, and 22.2% lignin was obtained from local waste (Asato et al., 2016). Synthetic human-derived WW was prepared according to the Organization for Economic Co-operation and Development (OECD/ OCDE, 2001) recipe. The anaerobic inoculum used in this study was obtained from a UASB for treatment of brewery wastewater (New Belgium, Fort Collins, CO), stored at 4 °C, and contained a volatile suspended solids content of 8.1% w/w. Inoculum was acclimated for six months prior to use in AD. Both wastewater inoculum and substrate were assumed to be waste byproducts and no impacts from waste production or consumption were considered.

2.3.2. Anaerobic digestion The total mass balance and energy requirements for AD pre-treat-

ment and operation stages were estimated based on an ASPEN Plus v8.8 software model of a portable two-stage AD experimental prototype and associated laboratory test results. A simplified process flow diagram is shown in Fig. 2.

Biogas production was measured in terms of g COD/day and yield was calculated from the substrate organic loading rate (g COD/L/day). The portable AD design capacity at the time of this analysis could process the organic solid waste typically produced by 70–80 people (∼90 kg COD/day, equating to ∼100 gal of homogenized organic waste/day). The total treatment capacity is dependent on organic waste type and composition. The human-derived wastewater, specifically black water, was assumed to be sourced from on-site from the same population. The prototype AD was designed to optimize CH4 recovery from solid organic waste but also could function in the treatment of human or livestock-sourced wastewater if desired. Specific effluent or outgoing solid waste quality targets were not included in this model, but the AD unit could be adjusted on-site to assist in meeting specific treatment targets. The prototype AD unit was designed to be both portable and scalable, so units can be placed in parallel to accom- modate encampment size or organic waste treatment targets. Yields obtained from the experimental scenarios are summarized in Table 1 and detailed in supplementary material (Table S1). The process re- quired use of sodium carbonate (20% aqueous solution) to maintain the hydraulic buffering tank pH at 6.5, which was used to maintain the pH between 7–8 after the CSTR. Chemical, input waste, product, and by- product quantities are presented in Table 2 for each scenario studied.

Table 1 Scenarios developed based on experimental portable AD conditions and results which were used to calibrate the ASPEN model and parameterize the LCA in- puts.

Scenario OLR (g COD L−1 d−1)

fFW (%) fCB (%) Biogas Yield (%COD)

I 8 35 65 16.6 II 8 65 35 29.1 III 16 65 35 37.4 IV 16 35 65 22.8 V 32 35 65 19.4

Abbreviations: organic loading rate (OLR), chemical oxygen demand (COD), food waste (FW), cardboard (CB), and fraction (f).

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

116

Losses due to process leakages were assumed to be 1.5% of the total biogas volume. Methane yields were between 0.3 and 1.5 g per L of FW- CB-WW mixture into the AD unit.

2.3.3. Business-as-usual scenarios To complete the end-of-life analysis, biogas produced from the

anaerobic digestion system was assumed to be directly combusted lo- cally in a boiler to generate heat (Ecoinvent process: heat, central or small-scale | alloc def, U). Net calorific value was assumed to be 22.9 MJ m−3 at standard temperature and pressure (STP) for biogas that contained 63% CH4 (vol.). Emissions from combustion (Table S2) were taken from Fruergaard and Astrup (2011) based on Nielsen et al. (2008). Comparable fossil-fueled, small-scale boilers (5–15 kW) were selected from the Ecoinvent unit process library (Table S3) to evaluate general on-site use differences between biogas and conventional fossil fuels. A specific site or location was not assumed because the unit was designed to be mobile and could be deployed or the products could be used at a variety of locations. As a result, transportation of fossil fuels or biogas was not included in the analyses. This analysis provides a basic indication of conversion life cycle impacts and benefits.

The economies of scale are applicable to biogas generation pro- cesses, thus as a result, portable small-scale biogas generation units (< 100 tonnes organic waste per year) will likely be less efficient than large-scale processes (> 10,000 tonnes organic waste per year). To determine the effect of scaling on waste management, energy balance, and process and system sustainability - the portable AD was compared to standard, full-scale biogas and energy recovery from incineration, biogasification, and landfill waste disposal processes for similarly

mixed proportions of comparably sorted food waste and lignocellulosic- rich wastes documented in Ecoinvent (Table S4). These comparisons were conducted with the intent of gauging the relative impact of waste biogas and/or energy production scale and generation method and waste avoidance between the portable AD unit and business-as-usual

Fig. 1. Life cycle assessment (LCA) system boundary showing energy production from food and cardboard waste treatment by anaerobic digestion.

Fig. 2. Simplified process flow diagramming biogas production through a two-stage CSTR and UASB anaerobic digestion system. Chemicals and input wastes are denoted by white-filled boxes, processes by gray shades, and products/byproduct by black-filled boxes. Material was recirculated through the process until the majority of biodigestable material was removed.

Table 2 Inventory summary for biogas production based on food waste and cardboard anaerobic digestion extracted from ASPEN Plus software model and scaled to match laboratory experiments. Experiments used a 1L CSTR and 0.5L UASB working volume for the reactors.

Material [g/day]

Scenario

I II III IV V

Inputs Food Waste (∼50% water) 4.0 7.5 15.0 8.1 16.2 Cardboard 4.4 2.4 4.8 8.9 17.7 Wastewater 628.4 734.1 1468.1 1256.9 2513.7 Sodium Carbonate 14.0 16.3 32.7 28.0 56.0 Water for Buffering Solution 56.0 65.4 130.7 111.9 223.8 Energy [J/day] 225.5 286.0 570.9 451.0 902.0

Product Methane (63% of biogas) 0.3 0.6 1.5 0.9 1.6

Byproduct Non-digestible Solids (∼90% water) 13.2 15.4 30.9 26.4 52.8 Sludge (∼90% water) 7.1 8.3 16.6 14.2 28.4 Treated Effluent 675.6 789.2 1578.3 1351.3 2702.5 Losses (1.5% loss) 10.6 12.4 24.8 21.2 42.4 Inert Gas (37% of biogas, mostly CO2) 0.2 0.4 0.9 0.5 0.9

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

117

waste-to-energy alternatives.

2.4. Environmental life cycle impact assessment

Environmental impacts were assessed via SimaPro software using the ReCiPe mid-point 1.08 method with a hierarchist perspective and world normalization (Goedkoop et al., 2009) including long-term emissions. Eighteen mid-point metrics from the ReCiPe method in- cluding the 100-year climate change (kg CO2 eq), ozone depletion (kg CFC-11 eq), terrestrial acidification (kg SO2 eq), freshwater eu- trophication (kg P eq), marine eutrophication (kg N eq), human toxicity (kg 1,4-DB eq), photochemical oxidant formation (kg NMVOC), parti- culate matter formation (kg PM10 eq), terrestrial ecotoxicity (kg 1,4-DB eq), marine ecotoxicity (kg 1,4-DB eq), freshwater ecotoxicity (kg 1,4- DB eq), ionizing radiation (kBq U235 eq), agricultural land occupation (m2a), urban land occupation (m2a), natural land transformation (m2), water depletion (m3), metal depletion (kg Fe eq) and fossil fuel deple- tion (kg oil eq) were quantified for each scenario. Energy consumption was computed via SimaPro CED v1.08 (Frischknecht et al., 2003). All biogenically sourced emissions and impacts were included in analyses to provide full accountability. The CED of a product estimates the direct and indirect energy use, including energy required for input materials and processes (Gürzenich et al., 1999; Huijbregts et al., 2006). Impact allocation was calculated based on the economic value of biogas or other energy products and solid waste disposal costs (Table S5). Un- certainty was estimated based on a log-normal distribution from Si- maPro pedigree analysis from a Monte Carlo simulation with a 95% confidence interval and 1000 simulations.

3. Results and discussion

3.1. Impacts of substrate composition and organic loading rate

Substrates with different ratios of FW and CB yielded variable biogas production (Table 1). Waste compositions with higher propor- tions of CB had slightly lower dry solid waste reductions due to higher percent of non-digestible COD (Table S1), but still yielded a dry solid waste reduction of ∼60% with ∼35% in form of sludge which could be recycled as fertilizer. The resulting life cycle impacts were normalized according to ReCiPe midpoint methodology (Sleeswijk et al., 2008) and compared in Fig. 3. All impact categories presented similar overarching trends where scenarios I and V had a greater impact levels, scenarios II and IV had mid-range impacts, and scenario III had the lowest. Impacts inversely correlated with increasing biogas yield (Table S7). Based on

these results, four life cycle impacts (climate change, eutrophication potential, human toxicity and acidification potential) were selected for discussion within this text. These categories were selected due to their high normalized impact level (Fig. 3) and relevancy to disaster affected areas and military operations and human health impacts. In recent literature meta-analyses, the eutrophication and acidification potential impacts of AD have been generally reported as negative impact cate- gories of concern while climate change and human toxicity impacts have been generally reported as beneficial in comparison to a BAU of landfilling and/or use of fossil fuel (Fan et al., 2017). Acidification potential, was an impact category of particular concern because it re- flects the importance of sulfur dioxide (SO2) emissions as an indirect greenhouse gas and acid rain and smog contributor. Smog formation is a substantial, often regionalized or localized, environmental and human health impact of concern affecting both developed and developing na- tions. Detailed results for all life cycle impact categories are presented in supplementary material (Table S6).

When the experimental results were compared (Fig. 4), scenario III (65% FW, 35% CCB, 16 g/day OLR) performed the best with respect to climate change impact. The climate change impact calculated for sce- nario III was 14% smaller than the worst performer, scenario I (35% FW, 65% CCB, 8 g/day OLR). This peak performance was attributed to an optimal loading rate of FW. While a greater proportion of readily- degradable FW resulted in improved conversion to VFAs and methane, overloading of FW inhibited CH4 production due to accumulation of intermediate metabolites. This contrasts with findings of Di Maria et al. (2016) where higher OLRs were found to be optimal, likely due to VFA conversion differences resulting from dissimilarities in waste feedstock properties (FW and WW composition) and blends (FW-WW vs. FW-CB- WW). In a study evaluating batch dry (total solids content ≥20%) AD of similar FW-CB mixtures, Capson-Tojo et al. (2017) observed com- plete inhibition of methane production and lower yields of fermentation products (including VFAs) at higher FW concentrations. This suggests that even in inhibited states, fermentative efficiency remains sensitive to FW loading and quality. In this study, scenario III performed well because the substrate contained a high proportion (65% of COD) of FW, however not to the point where FW loading negatively impacted fer- mentation. As expected, the climate change impact results directly correlated with biogas generation due to the high greenhouse gas po- tential of methane. The greater the conversion efficiency of CH4 and biogas generation, the lower the resulting climate change impact. Waste feedstock composition, quality, OLR, and blending affect biogas gen- eration and need to be understood to evaluate the full LCA impacts of AD.

Fig. 3. ReCiPe mid-point environmental life cycle impact category results for biogas production obtained after normalization per MJ of heat produced from waste. Impacts highlighted in black box were selected for discussion in this paper. Detailed results (Table S6) and uncertainty analyses (Table S7a–f) for all ReCiPe impacts and scenarios are available in supplementary information.

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

118

Organic loading rate related trends were present in the results. Increasing OLR decreased processing time which lowered the conver- sion efficiency. Lower loading rates increased the time spent in each stage and contact time which increased the conversion efficiency but also increased the processing impact. Scenario III performed slightly better for all environmental life cycle categories analyzed; however, the uncertainty analysis showed that the yield and life cycle impact dif- ferences were not significant between processes with similar OLR or compositions because the variability overlapped for all impacts except for climate change (Fig. 5, Table S7a–f). The composition and OLR of the substrate did not affect statistical significance of the life cycle im- pacts caused by portable AD use for impacts other than climate change. Despite the significance of the climate change impact, the impact range was small (< 10 g CO2-eq/MJ) with the climate change impacts varying by less than 15%. These findings demonstrated that the portable AD system can function as a waste reduction and disposal solution with feedstock and OLR flexibility, however operating at non-optimal con- ditions would increase the impacts of use over the lifespan of the unit. The volume and composition of organic waste from potential temporary encampments will vary and a portable AD should be able to

accommodate these variations without statistically significant impact on overall system sustainability.

3.2. AD recovery energy as a substitute fuel

The primary uses of biogas within temporary or isolated encamp- ments would likely be heating or fueling a small electrical generator. For military and natural disaster relief operations, the most common business-as-usual scenario would be small-scale, on-site use of diesel. However, other fossil fuels could be locally used depending on avail- ability. To further explore the life cycle impact differences between biogas and fossil fuels in this context, the peak performance scenario III for the portable AD biogas production system was compared with comparable small-scale coal, natural gas, and diesel fueled boilers for heat. The systems for generating heat from these fuel types can vary greatly in efficiency, this analysis was intended to provide a broad brushstroke overview of potential benefits and impacts. Fig. 6 sum- marizes the impacts analyzed for obtaining 1 MJ of energy as heat. Results show combusted biogas obtained from portable AD of organic waste overall had lower life cycle impact than standard fossil fuel

Fig. 4. Comparison of life cycle assessment (LCA) results of biogas production and com- bustion for portable anaerobic digestion treat- ment scenarios for (a) climate change, (b) eu- trophication, (c) human toxicity, and (d) acidification potential. Error bars shown were calculated using a Monte Carlo simulation with 1000 model iterations performed within SimaPro. Uncertainty analyses presented an overlap in results for impacts other than cli- mate change potential; indicating, that al- though trends can easily be identified, the dif- ferences in these results were not significant.

Fig. 5. Variability of biogas yield and combustion impacts as percent weight conversion (COD basis) for selected environmental life cycle impact categories as calculated via SimaPro Monte Carlo analysis. Vertical error bars represent the standard deviations in the experimental data on biogas yields based on weight percentage of waste COD converted and collected as biogas at each stage. Horizontal error bars represent the standard deviations of the selected ReCiPe Monte Carlo results. In this figure, higher biogas yields correspond with lower environmental impacts. However, these differences were not significant due to substantial overlap of results for impacts other than climate change potential.

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

119

heating options. This result was attributable to lower emissions asso- ciated with biogas combustion. Coal combustion had the highest life cycle impact in all categories analyzed and had a 68% greater impact than biogas combustion likely due in part to the inefficiency of coal use in small-scale processes. Burning coal for heat resulted in climate change impacts of 0.18 kg CO2 eq/MJ compared to 0.08, 0.08, and 0.06 kg CO2 eq/MJ respectively for diesel, natural gas, and scenario III biogas use. Biogas combustion impacts were like those of natural gas. Both natural gas and biogas are predominantly CH4, so they have si- milar combustion characteristics. The 5–12% relative difference be- tween the natural gas and biogas values was attributable to fossil versus biological fuel sourcing. The portable AD unit yielded better life cycle impacts than natural gas in all categories apart from acidification po- tential due to the unavoidable production of indigestible solids and effluent from the waste feedstock. All fossil fuel resources are non-re- newable, whereas methane generated from organic waste is a renew- able resource that also provided a volumetric dry solid waste reduction of ∼60% with a substantial 17% energy recovery rate. Meaning, ap- proximately one fifth of the waste energy potential available was con- verted to a ready-to-use renewable energy resources.

Use of on-site generated fuel from waste would avoid fuel transport, however because this analysis was not site specific, fossil fuel transport was not included. Remote and hard-to-access locations often require greater transport distances and inefficient modes of travel which further decreases the sustainability of all activities. For every km a tonne of fuel is transported, between 0.016 to 0.9 kg CO2 (USEPA, 2015) of emissions occur depending on mode of transport. The emissions from transport of fuels can exceed the emissions from combustion, especially if trans- ported via air more than 4000 km. As a result, on-site generation of biogas in remote or hard-to-access areas with limited fuel resources can be substantially more sustainable than transport of fuels for heating or small-scale electrical generation and would provide the additional benefits of organic waste reduction and treatment.

3.3. Waste treatment energy demand and waste avoidance

The results presented here refer to the comparison of the portable AD performance with three full-scale treatment processes (landfill, in- cineration, and biogasification) (Schmidt, 2010) commonly used for food and paper/cardboard waste. These types of facilities can vary in efficiency, benefits, and impacts due to size, age, waste composition, cost, services, infrastructure, and many other differences. This com- parison provides a synopsis of the basic impact of scale. Fig. 7 re- presents the amount of energy demanded to treat 1 kg of waste (65% FW, 35% CCB, 16 g/day OLR) for those treatments and the biogas en- ergy recovered. Transportation was excluded from all the processes to make them comparable to the portable AD scenario. Results show that portable AD requires less energy to treat 1 kg of solid organic waste than landfill treatment, however, it requires more energy than estab- lished full-scale incineration and biogasification processes. The biogas energy recovery potential of these processes was not substantially

affected by scale. However, the full-scale facilities analyzed were more efficient with regards to processing due to the economies of scale. Larger processes have an inherent operational efficiency advantage due to dispersal of costs over greater volumes and can leverage external infrastructure efficiencies. The portable AD was designed to be small- scale and operate with minimal external infrastructure dependencies. As a result, this process had to be largely self-contained and any ex- ternal infrastructure use had to incorporate inefficiencies present and any associated impacts. Within remote or hard-to-access locations, waste disposal alternatives are limited. The dry solid waste reduction and avoidance of ∼60% which a portable AD presents a substantial benefit in these situations. Use of digestate was not specified in the model, however the sludge can be locally used as fertilizer or processed waste can be dried and landfilled. The portable AD unit concentrates solids and treats wastewater by reducing organic content and removing solids. Liquid effluent leaving the portable AD is ∼7% greater in vo- lume but with a reduced organic and inorganic content. Analytical re- sults demonstrated that a portable AD can offer a small-scale alternative to landfilling with minimal infrastructure dependencies and energy recoveries comparable to larger-scale processes.

Fig. 6. Major impact categories normalized by the highest value per category to compare heat obtained from small-scale biogas, standard coal, natural gas, and diesel fuel use. In this basic comparison, on-site biogas production and use was shown to be comparable to that of natural gas and generally lower in impact than that of coal or diesel for small-scale heat generation. This rudimentary analysis assumed that these en- ergy resources were locally available.

Fig. 7. Bar chart representing energy necessary to treat 1 kg of waste (filled bar using scale on the left) plus the energy recovered (hollow bar using scale on the right) from different treatment processes, including portable anaerobic diges- tion (AD). In this elementary analysis, large-scale municipal landfill, incinera- tion, and biogasification/AD facilities with energy recovery are compared to a portable small-scale AD to better understand the impacts of scale. This chart shows that energy recovery potential is generally comparable between large- and small-scale operations. However, large-scale energy recovery operations likely have the benefit of the economies of scale.

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

120

4. Conclusion

Portable AD units can be used to sustainably produce renewable energy as biogas from organic waste in remote or hard-to-access areas. The climate change impact was the only environmental impact assessed that was significantly affected by organic waste composition or feed rate. This impact was due to the direct correlation between the capture of methane as biogas and its climate change potential. The overall sustainability of the portable AD was found to be generally agnostic to food or paper/cardboard mixture ratios or OLRs. The life cycle impacts of biogas use were like that of natural gas and overall less than fossil fuel use. The portable, small-scale AD units were determined to be a reliable, sustainable source of renewable energy with energy recovery potentials comparable to full-scale conventional waste-to-energy op- erations.

Acknowledgement

This work was supported by the Air Force Civil Engineer Center (AFCEC) under Contract No. FA4819-14-C-004.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in the online version, at doi: https://doi.org/10.1016/j.resconrec.2018.08. 008.

References

Ahamed, A., Yin, K., Ng, B.J.H., Ren, F., Chang, V.W.C., Wang, J.Y., 2016. Life cycle assessment of the present and proposed food waste management technologies from environmental and economic impact perspectives. J. Clean. Prod. 131, 607–614. https://doi.org/10.1016/j.jclepro.2016.04.127.

APHA, AWWA, WEF, 2012. Standard Methods for the Examination of Water and Wastewater, 22nd edition. American Public Health Association (APHA), American Water Works Association (AWWA), and Water Environment Federation (WEF), Washington, D.C.

Arafat, H.A., Jijakli, K., Ahsan, A., 2015. Environmental performance and energy re- covery potential of five processes for municipal solid waste treatment. J. Clean. Prod. 105, 233–240. https://doi.org/10.1016/j.jclepro.2013.11.071.

Asato, C.M., Gonzalez-Estrella, J., Jerke, A.C., Bang, S.S., Stone, J.J., Gilcrease, P.C., 2016. Batch anaerobic digestion of synthetic military base food waste and cardboard mixtures. Bioresour. Technol. 216, 894–903. https://doi.org/10.1016/j.biortech. 2016.06.033.

Asato, C.M., Gonzalez-Estrella, J., Skillings, J., Vargas Castaño, A., Stone, J.J., Gilcrease, P.C., 2017. Upublished Experimental Test Results.

Bernstad Saraiva Schott, A., Wenzel, H., la Cour Jansen, J., 2016. Identification of deci- sive factors for greenhouse gas emissions in comparative life cycle assessments of food waste management – an analytical review. J. Clean. Prod. 119, 13–24. https:// doi.org/10.1016/j.jclepro.2016.01.079.

Capson-Tojo, G., Trably, E., Rouez, M., Crest, M., Steyer, J.-P., Delgenès, J.-P., Escudié, R., 2017. Dry anaerobic digestion of food waste and cardboard at different substrate loads, solid contents and co-digestion proportions. Bioresour. Technol. 233, 166–175. https://doi.org/10.1016/j.biortech.2017.02.126.

Chiu, S.L.H., Lo, I.M.C., 2018. Identifying key process parameters for uncertainty pro- pagation in environmental life cycle assessment for sewage sludge and food waste treatment. J. Clean. Prod. 174, 966–976. https://doi.org/10.1016/j.jclepro.2017.10. 164.

Cho, J.K., Park, S.C., Chang, H.N., 1995. Biochemical methane potential and solid state anaerobic digestion of Korean food wastes. Bioresour. Technol. 52 (3), 245–253. https://doi.org/10.1016/0960-8524(95)00031-9.

Di Maria, F., Micale, C., Contini, S., 2016. Energetic and environmental sustainability of the co-digetstion of sludge with bio-waste in a life cycle perspective. Appl. Energy 171, 67–76. https://doi.org/10.1016/j.apenergy.2016.03.036.

Edwards, J., Othman, M., Crossin, E., Burn, S., 2017. Anaerobic co-digestion of municipal food waste and sewage sludge: a comparative life cycle assessment in the context of a waste service provision. Bioresour. Technol. 223, 237–249. https://doi.org/10.1016/ j.biortech.2016.10.044.

El-Fadel, M., Findikakis, A.N., Leckie, J.O., 1997. Environmental impacts of solid waste landfilling. J. Environ. Manage. 50 (1), 1–25. https://doi.org/10.1006/jema.1995. 0131.

Emberton, J.R., Parker, A., 1987. The problems associated with building on landfill sites. Waste Manage. Res. 5 (4), 473–482. https://doi.org/10.1177/ 0734242X8700500161.

Evangelisti, S., Lettieri, P., Borello, D., Clift, R., 2014. Life cycle assessment of energy from waste via anaerobic digestion: a UK case study. Waste Manage. 34 (1), 226–237. https://doi.org/10.1016/j.wasman.2013.09.013.

Fan, Y.V., Lee, C.T., Klemeš, J.J., 2017. The update of anaerobic digestion and the en- vironmental impact assessments research. Chem. Eng. Trans. 57, 7–12. https://doi.

org/10.3303/CET1757002. Finkbeiner, M., Inaba, A., Tan, R., Christiansen, K., Klüppel, H.-J., 2006. The new in-

ternational standards for life Cycle Assessment: ISO 14040 and ISO 14044. Int. J. Life Cycle Assess. 11 (2), 80–85. https://doi.org/10.1065/lca2006.02.002.

Frischknecht, R., Jungbluth, N., Althaus, H.-J., Doka, G., Dones, R., Hischier, R., Hellweg, S., Nemecek, T., Rebitzer, G., Spielmann, M., 2003. Implementation of Life Cycle Impact Assessment Methods. Final Report EcoInvent 2000. Swiss Centre for LCI, Dübendorf, CH. www.ecoinvent.ch.

Fruergaard, T., Astrup, T., 2011. Optimal utilization of waste-to-energy in an LCA per- spective. Waste Manage. 31 (3), 572–582. https://doi.org/10.1016/j.wasman.2010. 09.009.

Goedkoop, M.J., Heijungs, R., Huijbregts, M., Schryver, A.D., Struijs, J., Zelm, Rv., 2009. ReCiPe 2008: A Life Cycle Impact Assessment Method Which Comprises Harmonised Category Indicators at the Midpoint and the Endpoint Level. first edition. Report I: Characterisation.

Guinee, J.B., 2002. Handbook on life cycle assessment operational guide to the ISO standards. Int. J. Life Cycle Assess. 7 (5), 311–313. https://doi.org/10.1007/ BF02978897.

Gürzenich, D., Mathur, J., Bansal, N.K., Wagner, H.-J., 1999. Cumulative energy demand for selected renewable energy technologies. Int. J. Life Cycle Assess. 4 (3), 143–149. https://doi.org/10.1007/BF02979448.

Hagos, K., Zong, J., Li, D., Liu, C., Lu, X., 2017. Anaerobic co-digestion process for biogas production: progress, challenges and perspectives. Renew. Sustain. Energy Rev. 76, 1485–1496. https://doi.org/10.1016/j.rser.2016.11.184.

Hawkes, D., 1980. Factors affecting net energy production from mesophilic anaerobic digestion, anaerobic digestion. In: Stafford, D.A., Wheatley, B.I., Hughes, D.E. (Eds.), Proceedings of the First International Symposium on Anaerobic Digestion, held at University College, Cardiff, Wales, September 1979. Applied Science Publishers, London.

Huijbregts, M.A.J., Rombouts, L.J.A., Hellweg, S., Frischknecht, R., Hendriks, A.J., van de Meent, D., Ragas, A.M.J., Reijnders, L., Struijs, J., 2006. Is cumulative fossil energy demand a useful indicator for the environmental performance of products? Environ. Sci. Technol. 40 (3), 641–648. https://doi.org/10.1021/es051689g.

ISO, 2006. ISO 14040: 2006 Environmental management – Life cycle assessment – Principles and framework, in: (ISO), I.S.O. (Ed.) 13.020.60 ISO 14040: 2006. https:// www.iso.org/standard/37456.html.

ISO, 2006. ISO 14044: 2006 Environmental management – Life cycle assessment – Requirements and guidelines, in: (ISO), I.S.O. (Ed.) 13.020.60 ISO 14044: 2006. https://www.iso.org/standard/38498.html.

Mao, C., Feng, Y., Wang, X., Ren, G., 2015. Review on research achievements of biogas from anaerobic digestion. Renew. Sustain. Energy Rev. 45, 540–555. https://doi.org/ 10.1016/j.rser.2015.02.032.

Medina, V.F., Waisner, S.A., 2011. Military solid and hazardous wastes assessment of issues at military facilities and base camps. Waste: A Handbook for Management. p. 357.

Nielsen, M., Illerup, J.B., Birr-Petersen, K., 2008. Revised Emission Factors for Gas Engines Including Start/Stop Emissions. Sub-report 3 NERI Technical Report No. 672. National Environmental Research Institute (NERI), Denmark p. 72.

OECD/OCDE, 2001. OECD Guideline for the Testing of Chemicals: Simulation Test - Aerobic Sewage Treatment. Organization for Economic Co-operation and Development, Paris, France, pp. 1–50.

Opatokun, S.A., Lopez-Sabiron, A.M., Ferreira, G., Strezov, V., 2017. Life cycle analysis of energy production from food waste through anaerobic digestion, pyrolysis and in- tegrated energy system. Sustainability 9 (1804), 1–15. https://doi.org/10.3390/ su9101804.

Pierie, F., Benders, R.M.J., Bekkering, J., van Gemert, W.J.T., Moll, H.C., 2016. Lessons from spatial and environmental assessment of energy potentials for anaerobic di- gestion production systems applied to the Netherlands. Appl. Energy 176, 233–244. https://doi.org/10.1016/j.apenergy.2016.05.055.

Pohland, F.G., Ghosh, S., 1971. Developments in anaerobic stabilization of organic wastes - the two-phase concept. Environ. Lett. 1 (4), 255–266. https://doi.org/10.1080/ 00139307109434990.

Schmidt, J.H., 2010. Database Manual EU & DK Input Output Database. 2.0 LCA con- sultants - SimaPro 7.

Sleeswijk, A.W., van Oers, L.F.C.M., Guinée, J.B., Struijs, J., Huijbregts, M.A.J., 2008. Normalisation in product life cycle assessment: an LCA of the global and European economic systems in the year 2000. Sci. Total Environ. 390 (1), 227–240. https://doi. org/10.1016/j.scitotenv.2007.09.040.

Tagliaferri, C., Evangelisti, S., Clift, R., Lettieri, P., Chapman, C., Taylor, R., 2016. Life cycle assessment of conventional and advanced two-stage energy-from-waste tech- nologies for methane production. J. Clean. Prod. 129, 144–158. https://doi.org/10. 1016/j.jclepro.2016.04.092.

USEPA, 2015. Emission Factors for Greenhouse Gas Inventories. Center for Corporate Climate Leadership, USEPA p. 5.

USEPA, 2016. Advancing Sustainable Materials Management: 2014 Facts Sheet. Verma, S., 2002. Anaerobic Digestion of Biodegradable Organics in Municipal Solid

Wastes. Columbia University. Wernet, G., Bauer, C., Steubing, B., Reinhard, J., Moreno-Ruiz, E., Weidema, B., 2016.

The ecoinvent database version 3 (Part I): overview and methodology. Int. J. Life Cycle Assess. 21 (9), 1218–1230. https://doi.org/10.1007/s11367-016-1087-8.

Xiao, B., Quin, Y., Zhang, W., Wu, J., Qiang, H., Liu, J., Li, Y.-Y., 2018. Temperature- phased anaerobic digestion of food waste: a comparison with single-stage digestions based on performance and energy balance. Bioresour. Technol. 248, 826–834. https://doi.org/10.1016/j.biortech.2017.10.084.

Xu, C., Shi, W., Hong, J., Zhang, F., Chen, W., 2015. Life cycle assessment of food waste- based biogas generation. Renew. Sustain. Energy Rev. 49, 169–177. https://doi.org/ 10.1016/j.rser.2015.04.164.

Zhang, R., El-Mashad, H.M., Hartman, K., Wang, F., Liu, G., Choate, C., Gamble, P., 2007. Characterization of food waste as feedstock for anaerobic digestion. Bioresour. Technol. 98 (4), 929–935. https://doi.org/10.1016/j.biortech.2006.02.039.

C. Isola et al. Resources, Conservation & Recycling 139 (2018) 114–121

121

  • Life cycle assessment of portable two-stage anaerobic digestion of mixed food waste and cardboard
    • Introduction
    • Methods
      • Portable anaerobic digester
      • System boundary and functional unit
      • Inventory analysis
        • Waste composition
        • Anaerobic digestion
        • Business-as-usual scenarios
      • Environmental life cycle impact assessment
    • Results and discussion
      • Impacts of substrate composition and organic loading rate
      • AD recovery energy as a substitute fuel
      • Waste treatment energy demand and waste avoidance
    • Conclusion
    • Acknowledgement
    • Supplementary data
    • References