bernstad et al.pdf

Upload: ligia-dina-solis

Post on 24-Feb-2018

222 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/25/2019 Bernstad et al.pdf

    1/11

    Tank-connected food waste disposer systems Current statusand potential improvements

    A. Bernstad a,, . Davidsson a, J. Tsai a, E. Persson a, M. Bissmont b, J. la Cour Jansen a

    a Water and Environmental Engineering, Department of Chemical Engineering, Lund University, Swedenb VA SYD, Malm Municipality, Sweden

    a r t i c l e i n f o

    Article history:

    Received 1 February 2012Accepted 30 September 2012Available online 31 October 2012

    Keywords:

    Food waste disposersSeparate waste collectionHousehold food wasteAnaerobic digestion

    a b s t r a c t

    An unconventional system for separate collection of food waste was investigated through evaluation ofthree full-scale systems in the city of Malm, Sweden. Ground food waste is led to a separate settling tankwhere food waste sludge is collected regularly with a tank-vehicle. These tank-connected systems can beseen as a promising method for separate collection of food waste from both households and restaurants.Ground food waste collected from these systems is rich in fat and has a high methane potential whencompared to food waste collected in conventional bag systems. The content of heavy metals is low.The concentrations of N-tot and P-tot in sludge collected from sedimentation tanks were on average46.2 and 3.9 g/kg TS, equalling an estimated 0.48 and 0.05 kg N-tot and P-tot respectively per year andhousehold connected to the food waste disposer system. Detergents in low concentrations can resultin increased degradation rates and biogas production, while higher concentrations can result in tempo-rary inhibition of methane production. Concentrations of COD and fat in effluent from full-scale tanksreached an average of 1068 mg/l and 149 mg/l respectively over the five month long evaluation period.Hydrolysis of the ground material is initiated between sludge collection occasions (30 days). Older foodwaste sludge increases the degradation rate and the risks of fugitive emissions of methane from tanksbetween collection occasions. Increased particle size decreases hydrolysis rate and could thus decrease

    losses of carbon and nutrients in the sewerage system, but further studies in full-scale systems areneeded to confirm this.2012 Elsevier Ltd. All rights reserved.

    1. Background

    Source sorting of organic household waste has been proposedwithin the framework of a new European Waste Framework Direc-tive (European Parliament, 2008) and different systems for foodwaste collection are currently being established in various Euro-pean countries. Use of food waste disposers (FWDs) has been sug-gested as a practical way to establish source-separation of foodwaste without increasing transport, avoiding problems related

    to odor and increased need for waste bins amongst others(Marashlian and El-Fadel, 2005). However, there are several ques-tions regarding the effects of FWD connected to the conventionalsewer system. Several potential adverse effects have previouslybeen described.Bolzonella et al. (2003)stated that FWD can causean increased organic load in the biological step at the wastewatertreatment plant (WWTP) and thereby the energy demand forwastewater treatment. Others have raised problems with in-creased oil and grease load at WWTPs and risk of increased H2S-production in sewerage systems, which could result in corrosion

    of cement pipes (Nilsson et al., 1990). The knowledge on the quan-tity and quality of output from FWD actually reaching WWTP isstill incomplete. Considerable removal of dissolved organic matterand proteins in wastewater during sewage transport to WWTP hasbeen seen in previous studies (Raunkjaer et al., 1995). Effects onWWTP processes and potential increased sludge generation dueto FWDs will also to a large extent depend on the WWTP designand whether incoming particulate matter is separated with pri-mary sludge or not.Bolzonella et al. (2003)state that an increased

    carbon concentration in incoming wastewater can improve the C/Nand C/P-ratio in WWTPs and, depending on the process design, re-sult in an improved nutrient removal and reduced requirement forexternal carbon sources.Battistoni et al. (2007)showed that a 41%market penetration of FWD in a small Italian municipality in-creased both COD and N-tot in incoming wastewater, but did notresulted in increased energy use at the local WWTP, where acti-vated sludge processes were used.Evans et al. (2010) andGaliland Yaacov (2001) presented measurements of significant in-creases in biogas production in WWTP-sludge digestion when50% of connected households introduced FWD. The net-effects onsewerage systems and WWTP-processes from FWD installationare still an area of further investigation. These earlier stated

    0956-053X/$ - see front matter 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.wasman.2012.09.022

    Corresponding author. Tel.: +46 222 9557.

    E-mail address: [email protected](A. Bernstad).

    Waste Management 33 (2013) 193203

    Contents lists available atSciVerse ScienceDirect

    Waste Management

    j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / w a s m a n

    http://dx.doi.org/10.1016/j.wasman.2012.09.022mailto:[email protected]://dx.doi.org/10.1016/j.wasman.2012.09.022http://www.sciencedirect.com/science/journal/0956053Xhttp://www.elsevier.com/locate/wasmanhttp://www.elsevier.com/locate/wasmanhttp://www.sciencedirect.com/science/journal/0956053Xhttp://dx.doi.org/10.1016/j.wasman.2012.09.022mailto:[email protected]://dx.doi.org/10.1016/j.wasman.2012.09.022
  • 7/25/2019 Bernstad et al.pdf

    2/11

    problems or questions connected to waste grinders have led tobans on waste grinders for food waste in several areas, such asNew York in combined sewerage systems in the 1970s, Italy in1999 and Raleigh, North Carolina in 2008. These bans have in allcases been lifted after monitoring of the effects to the systems.However, reluctance concerting a more extensive use of food wastedisposers is still seen in many countries.

    1.1. Aim and scope

    In order to exploit the advantages of FWD and at the same timeavoid possible negative effects in sewerage systems and WWTPs,an unconventional disposer system was introduced in two residen-tial areas in the city of Malm, southern Sweden. Another systemwas connected to kitchen plumbing in a restaurant in the samecity. In the present paper, these three full-scale FWD installationsare investigated and potential improvements to the systems arepresented.

    1.2. Description of the systems

    FWDs were installed in kitchen sinks in approximately 60apartments in a residential area in Malm (area A) in 2001. Thekitchen sinks were connected to a pipe system separated fromthe other wastewater system in the building. This wastewaterstream was led through an LPS-unit (Low Pressure Sewer) to a set-tling tank (volume 2.7 m3) divided in sections from whichsuperna-tant is led to the WWTP and the settled waste is collected andtransported for further anaerobic biological treatment. A cuttingpump was used in the LPS-unit. The system in area B was installedin 2007 and was to a large extent a replica of the system in area A,but as the area consisted of 147 apartments in a high rise building(Turning Torso in Malm), there was no need for the LPS and sub-sequently not for the cutting pump used in area A. System C wasinstalled in 2010 and was connected to the food waste disposers

    installed in two restaurants in the same city. Also in this case, nocutting pump was used. The principle of the systems is shown inFig. 1, where the truck for transportation of the settled materialis shown together with the sewer, and the wastewater treatmentplant for handling of the non-settled material.

    The system can be said to be self-regulating in relation to miss-sorting as the disposers in general are sensitive to other materialsthan bio-waste and normally stop if other materials are disposed.However, materials such as soft plastics and paper could probablybe disposed together with food waste without causing severe prob-lems in the disposer and thus enter the settling tanks. Sludge col-lection was performed once a month in system A and B and everysecond week from system C with a tank-vehicle (capacity 12 m3)with a total energy use of 34.0 kW h natural gas/ton collected food

    waste sludge. Information regarding the energy use in LPS-systemcould not be gained from the owner of the system and was there-fore not included in the analysis.

    Detergents have earlier been seen disrupting the performanceof anaerobic digesters (Tanaka and Ichikawa, 1993; Gavala et al.,

    2001) and it is known that surfactants can harm methane-formingbacteria through lysis due to the absence of protecting envelopearound many methane-forming bacteria. In normal municipalwastewater, detergents used in households are mixed with a largeamount of water through different flows; showers, toilets andkitchens etc. The tank-system could result in higher concentrationof detergents in ground food waste sludge, as the system separateskitchen sink flows from other wastewater flows. Thus, the effect ofdetergents on biological processes in collected food waste sludgewas of interest to investigate. The following aspects were investi-gated and used as a basis for the evaluation:

    Quantity and characteristics of collected food waste sludge including potential methane production.

    Quantity and characteristics of effluent from collection tanks. Degradation of freshly ground food waste under different

    conditions. Potential fugitive methane production from ground food waste

    under different conditions. Potential inhibition of biological processes due to inhibitory

    substances (detergents).

    2. Methods

    The study combines data from analyses of samples from thefull-scale installations as well as laboratory experiments withfreshly ground food waste. In order to mimic the full-scale systems,a recipe for food waste ground for laboratory experiments weredeveloped based on the ratio of fat, proteins, carbohydrates andVS in full-scale samples (Table 1). The amount of water used forlaboratory grinding was chosen based on a literature reviewof pre-viously assessed water use for food waste grinding, recalculated tothe same unit (litre water/kg food waste) (Table 1).

    Energy use for grinding of food waste with the equipment usedin the full-scale areas was determined based on five measurements

    where 1 kg of fresh food waste was ground per occasion. Eachgrinding session was timed and the effect of the grinder was con-trolled. Energy-use in disposers was determined to 4.4 kW h/tonfood waste. Samples (20 l per occasion) were collected from thethree full-scale suspension-tanks on several occasions during theperiod September 2009 December 2011. Sampling was per-formed once a month, and always from full tanks. Effluent was alsosampled during a period of five months. Samplings were made onsix occasions from the outflow from the suspension tank. Five litresof effluent was collected per occasion. Samples were stored at 6 Cwithin three hours and analyses were made within 24 h. Samplesof sludge from suspension-tanks and effluent were used to deter-mine the following parameters with the following methods:

    TS(% ofwet weight) and VS (% ofTS) in sludge and TS, VS and SS(mg/l) in effluent was determined in triplicates using standardmethods (APHA, 2005, 2540 G).

    pH was determined with an instrument from the manufacturerWTW, model 320.

    Fig. 1. Principles of the food waste disposer and sedimentation tank system in areas AC together with the truck for transportation of the settled material, the sewer and thewastewater treatment plant for handling of the non-settled material.

    194 A. Bernstad et al. / Waste Management 33 (2013) 193203

  • 7/25/2019 Bernstad et al.pdf

    3/11

    Particle size compositions in food waste sludge and effluentwere found from sieving of the material (8, 4, 2, 1 and0.15 mmperforations). In order to decrease the influence of sur-plus water sieves were left running for 10 min, after which theweight of each sieve was recorded. Measurements were per-formed in triplicate.

    Sedimentation was determined using standard methods (SS-EN14702-1:2006). Recordings of settling were made after 1, 2.5, 5,10, 20 and 30 min. Settling velocity was determined as the vol-ume of settled material over 30 min. Determinations weremade in triplicates.

    Volatile fatty acids (VFAs) (propionate and acetate) were mea-sured in food waste sludge and effluent from tank systems aswell as in hydrolyzed food waste samples and samples usedfor determination of spontaneous methane production. Foodwaste sludge samples were centrifuged for five minutes(1250 g, MSE Scientific Instruments, Model GR-8) and all sam-ples (both effluent and sludge) were filtered (Munktell 1002,110 mm). Before analysis, the sample (0.9 ml) was conservedwith phosphoric acid (0.1 ml) and a sample volume of 0.2 mlwas injected into the gas chromatograph (GC) (Agilent 6850Series) equipped with a HP-FFAP column (30 m/0.53 mm/1 lm) at 80130 C (temperature inlet flow = 180 C, oven tem-perature 260 C).

    Heavy metals (Al, Cd, Cr, Fe, Pb and Zn) S, K and P were deter-mined in FWD-sludge and effluent by the laboratory at PlantEcology (Lund University) with use of inductively coupledplasma atomic emission spectroscopy (ICPAES) (PerkinElmermodel OPTIMA 3000 DV).

    TOC and N-tot in FWD-sludge andeffluent was determined withTOC-analyser TOC-VCPH with additional nitrogen analyserTNM-1 from Shimadzu. Determinations after digestion of sam-ples using Swedish Standard SS 02 91 50 for sample preparationof sludge and sediment (20 ml sample to 20 ml distillated waterand 20 ml HNO3and temperature rise through the use of auto-clave (200 kPa, 120 C during 30 min). Proteins, carbohydrates,fat and energy content was determined in dried samples usingN6.25 (Kjeldahl) Mod. NMKL nr 6 Kjeltec, carbohydrates were

    calculated according to SLV FS 1993:21, fat according to SLV VF1980 Lidfett0A.07 and energy content according to SLV FS1993:21.

    Potential unwanted (spontaneous) methane production wasdetermined from 500 ml samples (freshly ground food wastealone and mixed with 10% (VS) up to 4 weeks old food wastesludge from full-scale tanks), stored in 2000 ml glass bottleswith rubber septum lids. Triplicates of the bottles were storedin varying temperatures (6, 20 and 35 C) to represent thefull-scale situation over the year. Gas samples were taken intriplicates from each bottle using syringe with a sampling sizeof 0.2 ml, until no further gas production was detected, usingthe analysis procedure described inHansen et al. (2003). Theaccumulated methane yield was determined as N m3 CH4 gVS

    1,

    using temperature and pressure specified at each analysis occa-sion. No inoculum was used.

    Degradation of ground food waste under different conditionswas assessed through hydrolysis tests. Tests were performedwith ground food waste together with other wastewater flowsand at different temperatures.

    Influence of detergents on degradation of ground food wastesludge was tested through addition of the anionic surfactantSodium dodecyl sulphate (SDS or NaDS) (C12H25SO4Na) (CAS151-21-3, 100%) in concentrations of 0.8, 3.2 and 8.0 g/l freshlyground food waste (1:12 with fresh water). This compound waschosen as it is a relatively common in Swedish hand washingdetergents and was found in a total of 318 products on theSwedish market in 2009. The reported use in these applicationsin 2008 was 67 tons (Kemikalieinspektionen, 2011). Analysesof pH, VFA, COD and NH4 were made on filtered samples(Munktell 1002, 110 mm). Analyses were made after 0, 2 h,24 h, 48 h, 72 h, 96 h, 120 h, 1 week, 2 weeks, 3 weeks and4 weeks in a 28 d long experiment. COD was also analysed insamples with fresh water and detergents only in order to assessthe potential contribution of COD from the detergent itself.

    The potential methane production from ground food waste wasdetermined with the batch tests method described in Hansenet al. (2003). Inoculum was collected from a full scale digestionplant with co-digestion of food waste, manure, slaughter housewaste and industrial wastewater with high organic content.

    Inoculum and collected material were added in a ratio of60:40 in relation to the VS content.

    The effluent flow velocity from full-scale tanks was carried outbased on measurements of the water level after emptying oftanks with precise indications of the time between emptyingand measurement. An average was based on six separate mea-surements, covering different hours of the day and days of theweek.

    3. Results

    3.1. Characteristics of food waste sludge

    TS in sludge collected from areas A and B are in general low andboth TS and VS are higher from the restaurant (C) compared to theresidential areas (Table 2). Although the variations are large be-tween samplings, the differences between areas A, B and C are sig-nificant in all cases (t-test (two tailed)p< 0.01). Also in the case ofpH, significant differences were seen between all areas (t-test (twotailed)p< 0.01).

    Heavy metals, C, N, K, P and S were analysed in FWD-sludgefrom areas A and B at two occasions (August and October 2010)and area C at one occasion (October 2010) (Table 3). Althoughthe percentage of heavy metals on a DS basis increases after diges-tion, a comparison can be made between results and the SEPAguidance on heavy metals in sludge from WWTP, which is spreadon farmland as well as the certification of digestate from AD-plants

    and sludge from WWTP (SP, 2010; REVAQ, 2011). The results arealso compared to an analysis of food waste ground in laboratory

    Table 1

    Water use in FWD-systems; literature values and used assumption.

    Value Unit Equal to (l/kg food waste) References

    3 l/household, d 7.2 Kppalafrbundet and SRAB (2009)6 l/household, d 14.5 Kppalafrbundet and SRAB (2009)4.3 l/person, d 15.6 Marashlian and El-Fadel (2005)4.5 l/household, d 10.9 Gteborgs Stad (2011)

    19.3 l/kg food waste 19.3 Nilsson et al. (1990)

    12.4 l/kg food waste 12.4 Wainberg et al. (2000)12.0 Used in the present study

    A. Bernstad et al. / Waste Management 33 (2013) 193203 195

  • 7/25/2019 Bernstad et al.pdf

    4/11

    with use of non-ionic water as carrier water (FW fresh, based onrecipe presented inDavidsson et al., 2011).

    3.2. Particle size and sedimentation qualities

    The investigation of the particle size distribution shows a con-sistently higher fraction of larger particles (>4 mm) in sludge col-lected from area B compared to area A, where a cutting pumpwas used in the system (p< 0.5,t-test, 2-tailed) while no signifi-cant differences were found between area C and other areas

    (Fig. 2). The particle size generated was seen to be below the12 mm demanded by the hygienization standards for this type ofbiogas substrate according to the EU ABP-directive (EuropeanParliament, 2009).

    Settling tests with samples from areas A and B showed that anabsolute majority of the particles in the samples sediments within10 min (Fig. 3). No settling test could be done with samples fromarea C, due to the high TS in samples from this area.

    The initial settling velocity during the first 10 min is calculatedas:

    0:001 1000 Vs=10=60 1

    withVs= sludge volume (mm) after 10 min (Table 4).It is seen that the majority of settling particles in the ground

    food waste will settle within 10 min, provided a surface load below35 m/h. The effluent flow velocity was determined based on sixseparate measurements, covering different hours of the day anddays of the week. As no measurements were done over-night, itwas assumed that the measured flow was representative for 12 hof the day (07AM-7PM) and that the flow during the other 12 hwas 5% of this measured flow. This gives a water use of 40.3 l/capi-ta and day. Based on this assumption, the initial settling velocitycould be calculated for the surface area of 2.9 m2 in full-scale tanks,giving a surface load in the settling-tank of 0.11 m/h, thus well be-low the needed 35 m/h.

    3.3. Degradation of ground food waste

    Degradation tests with freshly ground food (alone as well asmixed with different types of other waste flows) show that biolog-ical hydrolysis is taking place in all samples. Increased concentra-tions of dissolved COD and VFA (acetate and propionate) anddecreased TS and VS are seen in all cases, as degradation of VS toa large extent is mirrored in an increase of dissolved COD. These

    processes are all enhanced when food waste was mixed withwastewater (WW), black water (BW) and sludge from full-scaleFWD-tanks (FWS) (Fig. 4). Mixing with these flows also increasedthe concentration of NH4 , which was not seen when food wastewas degraded separately (Table 5). A previous study of the samesystem has also shown that the degradation rate is enhanced byincreasing temperatures (Davidsson et al., 2011).

    3.4. Methane production theoretical and experimental

    In order to determine the theoretical methane production po-tential in the ground food waste, the content of proteins, carbohy-drates and fat was measured. The calculation of the potential wasbased on data fromChristensen et al. (2003)(Table 6).

    Higherfat andlower protein contentis seen insamples from areaC (restaurant). The theoretic methane production per VS is 1635%

    Table 2

    TS, VS, pH and N-tot in full-scale food waste sludge samples as average and standard deviation (SD). Also the average C/N ratio is presented.

    Parameter TS% VS (% of TS) pH N-tot C/N ratio

    Area A B C A B C A B C A B C A B C

    Samples (n) 7 9 3 7 2 3 5 6 4 2 2 2 2 2 2Average 0.4 2.7 9.8 85.9 89.6 97.3 5.65 4.91 4.46 30 46 21 19 12 37SD 0.2 1.5 0.2 4.3 1.1 0.22 0.01 0.20

    Table 3

    Content of metals, K, P and S in FWD-sludge samples analysed at two occasions (mg/kg TS) as well as relevant Swedish regulations and guidelines.

    Area Al Cd Cr Cu Fe K P Pb S Zn Cd/Pa

    A (Aug) 255 0.1 3.1 105 1263 6399 3093 0.9 166 32B (Aug) 691 0.1 8.5 60 437 2242 5275 2.5 94 19A (Oct) 417 0.7 6.2 184 1505 4838 2505 8.8 3690 126 290B (Oct) 141 0.2 1.3 29 227 1426 2549 2.6 1975 53 74C (Oct) 120 1.1 1.2 5.3 424 2506 1833 4.8 769 27 600FW fresh 49 0.1 0.5 4.9 41 8126 2077 0.9 2980 21 48SEPAb 2 60 10 80SPCRc 1 60 10 800REVAQd 34

    a Cd/P quote as mg Cd/kg P.b Swedish Environmental Agency limits for WWTP-sludge use on farmland (SEPA, 1994).c SP (2010)Certification system for digestate from co-digesting biogas plants.d REVAQ (2011). WWTP quality certification system with limit for Cd/P quote in WWTP-sludge use on farmland as average on yearly basis.

    0%

    20%

    40%

    60%

    80%

    100%

    A B C

    Percentoftotalsample

    0-0.15 mm

    1 mm

    2 mm

    4 mm

    8 mm

    Fig. 2. Particle sizes in full-scale samples from areas A, B and C based on sieving of

    three samples from each area. Standard deviationfor particles >4 mm= 6%, 10% and6% in areas A, B and C respectively.

    196 A. Bernstad et al. / Waste Management 33 (2013) 193203

  • 7/25/2019 Bernstad et al.pdf

    5/11

    higher in ground food waste from households (area A and B) com-pared to previous assessments of household food waste collectedin paper bags, due to a larger fraction of carbohydrates in the latter.The reason behind this difference in composition could only partlybe explained by the influence of the collection bags used. Methane

    ml

    0

    200

    400

    600

    800

    1000

    10

    Time (minutes)

    20

    Area A

    30

    Floatinglayer

    Clear

    layer

    Sediment

    ml

    200

    400

    600

    800

    1000

    010 20

    Time (minutes)

    Area B

    30

    Fig. 3. Results from settling tests with samples for area A and B. Average of triplicates.

    Table 4

    Initial settling velocity in food waste sludge from full-scale tanks in areas A and B.

    Area A B

    Initial settling velocity (m/h) 5.2 3.5SD (m/h) 0.4 0.9

    VS(%ofww)

    VFA(mg/l)

    0.00

    0.10

    0.20

    0.30

    0.40

    0.50

    100

    200

    300

    400

    500

    600

    700

    800

    0

    00 10

    10

    Days

    Days

    20

    20

    30

    30

    2000

    2500

    COD

    (mg/l)

    10

    12

    14

    16

    NH4+(mg/l)

    0

    500

    1000

    1500

    0

    2

    4

    6

    8

    0

    0 5

    5

    10

    10

    Days

    Days

    15

    15

    20

    20

    25

    25

    30

    30

    Fig. 4. Development of VS, COD and VFA (acetate + propionate) in hydrolysis experiments with ground food waste (FW1small particles and FW2larger particles) and groundfood waste mixed with wastewater (FW + WW), black water (FW + BW) and sludge from the settling full scale FWD tank (FW + FWS).

    Table 5

    Total change in VS, COD (dissolved) and NH4 over the experiment in samples with ground food waste mixed with wastewater (FW + WW), black water (FW + BW) and sludgefrom the settling full scale FWD tank (FW + FWS). FW 1small particles and FW2larger particles after 30 days.

    Wastewater combination VS reduction (%) COD initial (mg/l) COD final (mg/l) NH4 initial (mg/l) NH

    4 final (mg/l)

    FW + WW 54 155 2097 3.2 10.9FW + BW 48 173 2330 40.4 46.3FW + FWS 38 162 2182 4.0 12.0FW1 27 140 1653 2.0 5.5FW2 22 193 1278 2.0 5.0

    A. Bernstad et al. / Waste Management 33 (2013) 193203 197

  • 7/25/2019 Bernstad et al.pdf

    6/11

    production potential from similar material has previously been as-sessed to 75 N dm3/kg VS (Riber and Christensen, 2006). Previousstudies have suggested an average use of paper bags equal to11.719.5 kg/ton wet food waste or 35% of total TS (Bernstadet al., 2011) in a system similar to the ones described by Davidssonetal.(2007). A recalculationof above presented datafromDavidssonet al. (2007) was made in order to exclude the impact of paper bagsfrom the theoretical methane potential. However, also recalculateddata for the theoretical methane potential show low values whencompared to samples from full-scale areas A and B (Fig. 5).

    Methane production was also determined experimentallythrough batch methane potential (BMP) analyses of freshly groundfood waste samples as well as samples from hydrolysis experi-ments. The methane production per VS was similar in all cases(398405 N dm3 CH4/kg VS). However, when related to the initialVS content (prior to hydrolysis), results suggest a potential de-crease of potential methane production of more than 25% in sam-ples hydrolyzed during 4 weeks and 79% in samples hydrolyzedduring 36 h. Results suggest a decreased methane production po-tential of 526% per ton ground food waste (Fig. 6). Thus, althoughthe methane potential per VS in FWD sludge, the recovered

    methane is potentially decreased through the degradation takingplace before the sludge is collected.

    3.5. Influence of detergents

    Addition of detergents to the ground food waste caused a rapidincrease of dissolved COD, which could only partly be explained bythe COD added through the detergents themselves (Table 7). Thus,the detergents resulted in an increased solubilisation of particulateorganic matter. At the same time, addition of detergents in higherconcentrations resulted in lower concentrations of both VFA andNH4 and a lesser decrease in pH compared to samples were deter-gents not had been added, which was interpreted as inhibition ofbiological hydrolysis (Table 7). Also the methane production perVS increased when lower concentrations of detergents had beenadded to the samples, while addition of detergents at higher con-

    centrations resulted in signs of temporary inhibition of methaneproduction (Fig. 7). After 12 days, a clear difference was seen inreactors without and with the lowest addition of detergent onthe one hand, compared to reactors where detergent had beenadded in higher concentrations on the other (p< 0.01, t-test

    Table 6

    Content of proteins, carbohydrates and fat in full-scale samples in relation to literature values.

    Fraction Unit Area A Area B Area C Householdb CH4potential (N dm3/kgVS)

    Proteinsa % of DS 22.4 29.3 9.8 1018 496Carbohydrates % of DS 27.9 27.2 7.7 1955 415Fat % of DS 42.3 37.5 81.2 1018 1014Ash % of DS 7.4 6.0 1.3 819 0Energy kJ/100 g 2345 2285 3201 19002200

    Theoretical CH4potential N dm3/kgDS 656 638 904 267500Theoretical CH4potential N dm

    3/kgVS 764 713 928 567615

    a As Kjeldahl nitrogen.b Food waste from households with separate collection in paper bags ( Davidsson et al., 2007).

    Fig. 5. Theoretical methane potential based on content of fat, carbohydrates andproteins in areas AC and paper bag collected food waste (PB1-2).

    N

    dm3CH4

    50

    100

    150

    200

    250

    300

    350

    400

    450

    0

    FW

    FW

    (4weeks)

    FW+W

    W(36h)

    FW

    (36h)

    0

    0.05

    0.1

    0.15

    0.2

    0.25

    0.3

    Degradation

    ofVS

    Ndm3 CH4/kg

    VSNdm3 CH4/kg

    food waste

    Fig. 6. Methane potential per kg VS and kg food waste as well as degradation ratioduring hydrolysis according to degradation experiments.

    Table 7

    NH4 , VFA and COD (dissolved) in ground food waste with varying concentrations ofdetergents (mg/l).

    Sample NH4 VFA COD

    Watera Sample

    FW 15 500 1800FW + 0.8 mg/l 30 290 670 5000FW + 3.2 mg/l 2 310 2700 11,000FW+8.0 mg/l 2 310 6700 22,500

    a Potential contribution of COD from detergents alone in fresh water.

    198 A. Bernstad et al. / Waste Management 33 (2013) 193203

  • 7/25/2019 Bernstad et al.pdf

    7/11

    2-tailed). After 30 days, the highest production was seen in reac-tors to which the lowest concentration of detergents had been

    added. However, differences between samples were no longerstatistically significant (p> 0.1,t-test, 2-tailed).

    3.6. Potential fugitive methane emissions from settling tanks

    Results from the degradation tests show that hydrolysis occurin the tanks between sludge collection occasions. Thus there couldalso be a possibility of unwanted emissions of methane from thetank systems between sludge collections. The risk for unwanted(spontaneous) methane production was assessed through simula-tions of the tank process in laboratory scale experiments under dif-ferent conditions. Freshly ground food waste and freshly groundfood waste together with sludge from area B in different ratios

    (10% and 15% of total VS respectively) was assessed (Fig. 8). Thefirst measurement was made nine days after setup. Temperaturemeasurements in settling tanks in full-scale systems A and Bshowed an average temperature of 17.0 C and 19.1 C in the twosystems over the sampling period (SD = 4.4 and 2.3 Crespectively).

    3.7. Composition of effluent

    The composition of effluent from the tank-system in area B isdisplayed inTable 8 as averages based on seven measurementsover a period of five months. Both TS and VS were low. A large partof the solids are non-organic, which could be explained by solved

    salts. The conductivity in effluent was high in comparison to previ-ously presented data for average wastewater and above localguidelines for industrial wastewater released to municipal treat-ment plants (Levlin and Hultman, 2008). pH was in all cases belowneutral; 6.16.6. Analyses of effluent from full-scale installationswere compared with average levels of N-tot, P-tot, CODCr, SS, fatand conductivity in Swedish municipal wastewater and in the in-flow to the WWTP at which the effluent was received. Also nationaland local regulations on these emissions in effluent from WWTPare presented in the comparison (Table 8).

    The concentrations of the metals Al, Cu and Zn were determinedto an average of 0.2, 0.01 and 0.02 mg/l respectively (SD = 0.033,0.003 and 0.009 mg/l respectively). Concentrations of Cd and Pbwere in all cases under detection limits.

    Analyses of nutrients, carbon and metals in food waste sludgeand effluent respectively, together with measurements of the flowin effluent from Area B can be used to estimate the mass-flow ofthese compounds over the system. However, as seen inTables 2,3 and 8, concentration in some cases varied largely between sam-pling occasions food waste sludge as well as of varied largely be-tween sampling occasions. Using minimum and maximum valuesfrom analyses of both sludge and effluent and the flow assumedabove, 4055% of N-tot and C-tot as well as 3862% of P-tot and9599% of K and S in ground food and kitchen waste are trans-ported from the sedimentation tank with effluent. In the case offat, between 30% and 57% is transported to the sewerage systemwith effluent.

    4. Discussion

    4.1. Losses of food waste to wastewater system

    The FWDs used in the investigated full-scale systems were orig-inally designed to transport ground waste in the sewerage systemto a WWTP, as this would be the normal use of FWDs. Larger par-

    ticles have been assumed to increases the risk for settling of parti-cles in the sewerage systemand subsequent clogging and there hastherefore been a striving towards ever smaller particle sizesamongst FWD producers. This could be assumed to decrease theamount of settled material in tanks. However, both freshly groundfood waste and sludge collected from full-scale systems were seento have good settling-qualities and the low TS and high levels ofdissolved COD in effluent indicate that losses of organic matterfrom the system primary occur through dissolved substances andnon-settling fat. Hydrolysis tests show a slower biological hydroly-sis rate in ground food waste when the particle size was increased.Thus, an increased particle size in ground food waste could poten-tially decrease losses of organic matter and nutrients from the

    Fig. 7. Methane production in ground food waste with different concentrations ofdetergents. FW = Food waste (ground).

    3

    4

    5

    6

    7

    0

    1

    2

    3

    4

    6C 20C 20C 35C 20C

    FW+FWS(10%) FW+FWS(10%) FW+FWS(15%) FW+FWS(10%) FW

    pH

    Ndm3CH4/kgVS

    Temperature and sample

    9 days

    22 days

    35 days

    44 days

    67 days

    70 days

    pH before

    pH after

    Fig. 8. Methane production from laboratory assessment test with samples of freshly ground food waste with and without sludge from area B in different ratios at differenttemperatures. Also pH before and after the test is displayed. FW = Food waste, FWS = Food waste sludge.

    A. Bernstad et al. / Waste Management 33 (2013) 193203 199

  • 7/25/2019 Bernstad et al.pdf

    8/11

    tanks system, mainly due to a slower hydrolysis rate rather thanthrough improved settling qualities. Hydrolysis tests showed thatdegradation of VS and levels of dissolved COD also increased whenground food waste was mixed with other waste flows, includingsludge from full-scale tank systems. Thus, it is not advisable touse the tank system for combined systems for food waste andwastewater/black water and there are reasons to try to diminishthe amount of old food waste sludge left in the sedimentation tankwhen emptied through proper cleansing.

    Accumulation of fat in municipal sewerage systems is anincreasing problem in many Swedish municipalities (Bleckenet al., 2010). Although experiences have suggested that in-sewerbiological processes can acclimate to the change in wastewatercomposition caused by FWDinstallations and thereby reduce prob-lems related to accumulation of organic matter in sewage pipes,there can still be reluctance towards FWD in areas already experi-encing problems with fat-accumulation. This could also be seen asone of the main reasons behind the development of the tank-connected system. Limits for fat in wastewater from commercialentities to the municipal sewerage system varies between Swedish

    municipalities, but commonly range between 50 and 150 mg/l(Blecken et al., 2010). Analyses of effluent from full-scale tank-systems investigated in the present study show that the averageconcentration of fat in effluent is 50% above municipal limits inthe region of application (VA SYD, 2011) equal to 1.8 kg/personand year. Thus, the study indicates that effluent from non-commercial entities can result in considerable contribution of fatto the municipal sewerage system and that the tank-system inits current design not fully solves potential problems related tofat-accumulation.

    The estimated mass-flow of carbon and nutrients over the sys-tem indicate that around 50% of C-tot, N-tot, P-tot and fat is trans-ported from the settling tank with effluent. In the case of K and S,the fraction seems to be even higher. As the major parts of nutri-

    ents can be assumed to be conserved through the anaerobic diges-tion process, it can be assumed to be available as fertilizers ifdigestate is used on farmland. Whether carbon and nutrients arerecovered later in the system will to a large extent depend onthe WWTP design and potential use of WWTP sludge as fertilizer.Independently of this, results show a potential decrease in prob-lems related to for example fat accumulation in sewages whencompared to systems where FWD are connected to sewerage sys-tems directly, as a large part of the fat is collected with the sludge.

    The level of biogas potentially non-realized due to losses of or-ganic matter with effluent canbe assessed using values for theoret-ical methane production per kg COD (0.35 N m3/kg COD) (Spinosaand Vesilind, 2001) in effluent. Assuming a realization of 75% ofthe theoretical methane production and a degradation ratio of

    80% of COD would result in non-recovered methane potential of0.26 N m3 CH4/m3 effluent, or 2.25 N m3 CH4 per year and person

    connected to the FWD tank-system. This can be compared to themethane potential of 3.7 N m3 per person from sludge collectedfrom the systems over the same period, assuming an 80% degrada-tion of VS. This should be seen as an estimate, as no measurementswere made of dissolved methane in effluent, and as the calculationis based on a series of assumptions. However, it is relevant toacknowledge that the current design of the system does not resultin an optimal collection of substrate for methane production fromthe tanks.

    4.2. Fugitive emissions of methane from settling tanks

    Fugitive emissions of methane from tank systems are most un-wanted. Laboratory simulations of fugitive emissions showthat theaddition of older food waste sludge increased the fugitive methaneemissions. Using values from measurements in 20 C, i.e. slightlyabove the average temperature in full-scale systems settling tanksover the sampling period, state a potential production of 0.21.5 N ml CH4/g VS between sludge collection occasions (35 days).Increasing the period between collections to 70 days could in-

    crease the production to 0.43.2 N ml CH4/g VS. Thus, accumulatedmethane production generally increased with increasing time,addition of old food waste sludge and temperature. However, athigher temperatures, the methane production stopped after35 days, probably related to an inhibition as the pH dropped radi-cally. The potential relation between pH in the tank and methaneproduction makes it important to highlight that the method usedfor estimating potential fugitive emissions of methane does nottake the constant inflow of new organic matter and water withhigher pH to the tank into consideration, which could underesti-mate the risks for fugitive methane emissions. Decreasing the cur-rent collection frequency could increase the risk for methaneemissions from settling tanks and is therefore not advisable. Levelsof potential fugitive emissions detected in laboratory experiments

    equals a production of 0.030.23% of the calculated total methaneyield from food waste sludge and can thus be regarded as insignif-icant. However, if the tank not is emptied completely and rinsed,there could be remains of older food waste in the tank. This couldbe seen as an inoculum, increasing the degradation in tank, as seenalso in hydrolysis experiments.

    4.3. Characteristics of food waste sludge

    Previous studies have shown a higher per VS methane produc-tion in batch test AD of food waste collected with food waste grind-ers compared to paper bags (Davidsson et al., 2007). This could beexplained by household behavioural factors and the difficulties tocollect liquid/semi-liquid food waste with high ratio of easily bio-

    degradable carbon such as dairy products, and juice and a de-creased ratio of non-anaerobically biodegradable carbon such as

    Table 8

    Composition in effluent from full-scale tank system in area B (as averages and SD = standard deviation) together with national average, national guidelines and local conditions.

    Flow P-tot (mg/l) N-tot (mg/l) K (mg/l) S (mg/l) TOC (mg/l) COD (mg/l) CODda (mg/l) SSb (mg/l) Fat (mg/l) Cond.c (mS/m)

    Wastewaterd 16 50 530 445 70 93WWTP effluente 0.3 10

    50150 500Inflow WWTPf 5.3 42 559Average effluent 2.2 16.1 19.6 23.3 177.1 1101 575 540 149 772

    SD effluent 0.8 6.4 1.3 30.4 86.7 354 150 120 65 317a Dissolved COD.b Suspended Solids.c Conductivity.d Moderate levels in Sweden (Henze et al., 1995).e National and regional regulations (SFS, 1998; Blecken et al., 2010).f Average in inflow to local WWTP (Sjlunda, Malm) (VA SYD, 2010).

    200 A. Bernstad et al. / Waste Management 33 (2013) 193203

  • 7/25/2019 Bernstad et al.pdf

    9/11

    plastics and lignin in food waste collected through FWG in compar-ison to bag-collection systems. According to previous studies in theUK, dairy products and drinks (not including tea and bottled water)contributes to 12% of the total household food waste generation(WRAP, 2008). The fat content in Swedish and Danish householdfood waste collected in paper bags was according to previousstudies equal to 12.218.1% or 15% of TS (Hansen et al., 2007;Davidsson et al., 2007 respectively), i.e. low compared to the37.542.3% of TS in sludge collected from FWD tank-systemsaccording to the present study. As a consequence, the theoreticalmethane potential per ton VS is 36% higher in FWD-collectedhousehold food waste compared to bag collected. The gastronomicvariations between households and areas will have a large impacton the waste composition and thereby also potentially on themethane production from the ground food waste. A high fat con-tent could cause instability in the digestion process and decreasethe methane yield. However, if ground food waste is mixed withother substrates the risk of such problems will decrease. Groundfood waste could also be used as media for decreasing the DS inseparately collected food waste from areas were paper or plasticbag collection is used (Truedsson, 2010).

    The concentration of Cd varies between 0.1 and 0.7 mg/kg TSin full-scale samples of sludge from households FWD systems.This can be compared to results from analyses of food waste col-lected in paper or plastic bags from three Swedish plants foranaerobic digestion of food waste and other organic materials;0.20.3 mg/kg TS (Truedsson, 2010; Malmqvist, 2011). Due tovarying levels of both Cd and P in analysed samples, the Cd/Pratio varies largely between different areas, but also betweenanalysis occasions and values are in many cases higher thancurrent limitations for spreading of WWTP-sludge on farmland(REVAQ, 2011). In the case of Zn and Cu, measurements showconcentrations above current regulations. This could be relatedto the material composition in the plumbing systems, but furtherstudies are needed to confirm this.

    Thus, based on the present results, no conclusions can be made

    regarding the possibilities of reducing Cd concentrations in sepa-rately collected household food waste. According to previous stud-ies, batteries, pigments in plastics and surface treatment productsare responsible for the largest outflows of Cd in Sweden. Cd is alsofound in artist paint, food, drinking water and as a contaminationin zinc products (Mnsson and Bergbck, 2007). The plausible rea-sons for the in some cases high levels of Cd in food waste sludgefrom tank systems are therefore many. Further investigations andmore samples are needed for increased understanding of thesources to Cd in separately collected food waste from households.

    4.4. Household behaviour

    Previous studies on disposal behaviour in Swedish households

    state source-separation ratios of 2245% in multi-family dwellings(Dahln, 2008; LRV, 2009; Bernstad, 2010) and 7279% in single-family households (Swedish Waste Management Association,2011), in both cases using paper bags for separate collection offood waste. Lack of proper space in kitchen has been singled outas a factor which can decrease the willingness to participate infood waste recycling from the side of the households in areas withpaper bag collection schemes (Bernstad, 2010). Such problems areto a large extent avoided with the use of FWD and the source-separation ratio could therefore be assumed to be higher in house-holds with FWD compared to households with bag collectionschemes. Due to uncertainties in relation to the actual amount ofhouseholds using FWD, number of persons in households andlosses of organic matter with effluent, the source-separation ratio

    in residential areas with FWD (A and B) could not be calculated.However, an estimation was made based on the following assump-

    tions and the below equation (Eq. (2)). The number of householdsconnected to system B was 147, but as a large part of these notwere occupied during the major part of the sampling period, calcu-lations were based on an estimation of 100 occupied households.TSs was determined based on average TS in collected sludge overthe sampling period, while TSfwas based on data fromCarlssonand Uldal (2009)on average TS in separately collected food waste(30%). The amount of collected sludge (

    Vs= 38.4m3 d= 989.2 g/l)

    would thereby reply to 34.2 kg food waste/household and year. Ageneration of 100 kg food waste per person and year (FWtot) wasassumed (Konsumentfreningen Stockholm, 2009). The numberof persons per household in area B was assumed to 1.5.

    Vs d TSs=100 TSf=n=FWtot 100 2

    Vs is volume sludge (m3/year), d is density in collected sludge

    (kg/m3), n is the Number of connected households, TSs is the drysubstance in sludge (% of wet waste), TSfis the dry substance in foodwaste (% of wet waste), FWtot is the total amount of food wastegenerated per person and year

    The source-separation ratio was determined to 23%. The rela-tively low source-separation ratio is in line with waste composi-tion analyses from Swedish areas where FWDs were installed in

    1997, where more food waste was found in residual waste fromhouseholds with FWD compared withhouseholds using bag collec-tion or home composting for selective disposal of food waste (Sura-hammar, 2011). More research is needed in relation to differentfood waste collection schemes and source-separation behaviour.

    4.5. Detergents

    An increased degradation of particulate organic matter inhydrolysis experiments where shown when detergents had beenadded to freshly ground food waste. As one of the primary missionsof tensides in detergents is degradation of fat, it is probable thatthe increased levels of dissolved COD mainly where caused by deg-radation of fat. However, no analyses of hydrolyzed samples were

    made in order to confirm this.In a study by Khalil et al. (1989), methane production was

    halved and the pH dropped from 7.4 to 6.0 over a period of 20 dayswhen anionic detergent (sodium dodecyl-benzene sulphonate SDBS) had been added at concentrations of 2050 mg/l, whileno inhibition was seen when non-ionic detergent (Tergitgol non-yl phenyl polyethylene glycol ether) or soap had been added at thesame concentrations. In the present study, addition of the anaero-bic sulfactant Sodium dodecyl sulphate at a concentration of 8 mg/lcaused temporary inhibition of methane production. As the con-centrations investigated in the present paper where low comparedto the ones assessed byKhalil et al. (1989), it cannot be excludedthat higher concentrations would result in irreversible inhibitionand the results obtained are thereby in line with previous research.

    However, the increased methane production in samples with lowerconcentrations of detergents was less expected. One explanation tothis is that low concentrations of detergents decrease the surfacetension in water and thus increases the availability of substratefor methane producing organisms without causing lysis of micro-organisms (Khalil et al., 1989).

    Inhibition of methane production as a result of spiking with an-ionic detergents (Linear Alkylbenzene Sulphonate, LAS) has previ-ously been related to the fraction of surfactant in the aqueousphase, rather than the total concentration of detergents (Garciaet al., 2006). The reason is according toGarcia et al. (2006)that asorption of anionic detergents to suspended solids reduces theavailability and thereby the toxicity on anaerobic microorganisms.As sorption is promoted by presence of cations, the concentration

    of calcium and magnesium extractable cations could be of impor-tance to inhibitory effects from anionic detergents. Further studies

    A. Bernstad et al. / Waste Management 33 (2013) 193203 201

    http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/25/2019 Bernstad et al.pdf

    10/11

    are thereby needed to explore the relation betweensurfactants andwater hardness in relation to inhibitory effects on methanogenicmicroorganisms. Further studies are also needed to investigatetypical concentrations of different types of tensides in householdkitchen sink effluent.

    4.6. Resource use

    Previous studies have shown that the water use in FWD systemscan vary largely (Table 1).Nilsson et al. (1990)saw a decrease ofwater use of 13 l/day in households after FWD had been installedwhile Lundie and Peters (2005) lift the extra use of water (assumedto 2.9 l/PE, day or 12.4 l/kg food waste) as one of the most negativefactors related to introduction of FWD in a dry country likeAustralia. The data used byLundie and Peters (2005)would replyto less than 2% of the average daily per capita consumption andbe close to irrelevant in a Swedish context (Swedish EnergyAuthority, 2007). Thus, the feasibility of FWD use must be studieswithin the local context.

    Lundie and Peters (2005)measured the energy need for grind-ing of food waste to 0.02 kW h/kg waste, while Evans et al.

    (2010)assume the yearly energy use to FWD to 23 kW h/house-hold. The current collection scheme of ground food waste resultin an energy use equal to 38 kW h/ton ground food waste, addingenergy use in grinders and fuel use for collection and transporta-tion. This can be compared to the energy use for collection of foodwaste separately collected in paper bags. Use of paper bags, heat-ing/ventilation of facilities where food waste is disposed of byhouseholds and temporarily stored before collection (recyclingbuildings) as well as energy use for collection, transportation andphysical pretreatment of food waste has previously been estimatedto 56 kW h/ton selectively collected food waste (Bernstad and laCour Jansen, 2012). Thus, although there are potentials to reducethe energy consumption in the investigated system radically, prin-cipally through improved collection technique and reduced trans-portation of water, the energy consumption in the current systemis well below the energy use in more conventional systems for sep-arately collected food waste. The energy input can also be com-pared to the potential energy recovery from produced food wastesludge through biogas production. The production equals 80109Nm3 CH4/ton ground food waste, depending on the degrada-tion in the tank before collection (assuming a degradation ratioof 80% of VS in biogas production facilities, based on Davidssonet al., 2007). Thus, the energy consumption in the system equalsbetween 3% and 5% of potential energy recovery. This does not takepotential energy savings related to substitution of chemical fertil-izers with digestate on farmland into consideration.

    4.7. Methods for evaluation

    Tank connected FWD-systems were in the present study inves-tigated through a combination of full-scale sampling and measure-ments and lab scale studies. Although laboratory studies wereaimed to mimic full-scale systems, this might not have succeededin all cases. As an example, potential fugitive emissions of methanebetween emptying of tanks assessed in laboratory did not includethe effect of the constant inflow of new organic material and waterto the tank, which might have decreased the risk for a pH drop andthus increased the risk for methane production in the tank.

    5. Conclusions

    The unconventional food waste disposal systems investigated in

    the present paper can be seen as a promising method for separatecollection of food and kitchen waste from both households and

    restaurants. The sludge collected from the systems is rich in fatand has a high methane potential when compared to food wastecollected in more conventional food waste collection systems. Thecontent of heavy metals is low. However, the Cd/P quote varieslargely and can exceed Swedish national regulations for use ofbio-fertilizers from an anaerobic digestion on farmland. Detergentsin low concentrations in food waste disposersludge canresult in in-creased degradations rates and biogas production, while higherconcentrations can result in temporary inhibition of methane pro-duction from ground food waste. The concentrations of N-tot andP-tot in sludge collected from sedimentationtanks were on average46.2 and 3.9 g/kg TS, equalling an estimated 0.48 and 0.05 kg N-totand P-tot respectively per year and household connected to thefood waste disposer system. The concentration of fat in collectedfood waste sludge was high, and the losses of fat to the seweragesystem are thus vastly decreased through the use of settling tanks.The average concentration of COD in effluent from tanks exceeded1000 mg/l, of whichmorethan 50%was dissolved. Losses to effluentrepresent more than 50% of C-tot, N-tot and P-tot in ground foodwaste. Results indicate that a hydrolysis takes place in the tank,and that this is induced in the presence of older food waste sludge.Increased particle size decreases the hydrolysis rate and could thusdecreaselosses of carbon and nutrients to the sewerage system, butfurther studies in full-scale systems are needed to confirm this.Increasedcollection frequency could increase the amount of carbonand nutrients recovered through the system through sludgecollection, but would also increase costs and energy use.

    Acknowledgements

    The authors would like to warmly thank the VA SYD SolidWaste Management Department, in particular Henrik Aspegrenand Roland Nilsson, for all their support and help during the workpresented in this paper. The author would also like to thank HamseKjerstadius for his help with collection of material used for parts ofthe laboratory work presented in the paper.

    References

    APHA, 2005. Standard Methods for the Examination of Water and Wastewater,21th ed. American Public Health Association, Washington DC.

    Battistoni, P., Fatone, F., Passacantando, D., Bolzonella, D., 2007. Application of foodwaste disposers and alternate cycles process in small-decentralized towns: acase study. Water Res. 41 (4), 893903.

    Bernstad, A., 2010. Environmental Evaluation of Solid Household WasteManagement the Augustenborg Ecocity Example. Licentiate thesis, Waterand Environmental Engineering, Lund University, Sweden.

    Bernstad, A., la Cour Jansen, J., 2012. Separate Collection of Household Food Wastefor Anaerobic Degradation - Comparison of Different Techniques from aSystems Perspective. Waste Manage. 32 (5), 806815.

    Bernstad, A., la Cour Jansen, J., Aspegren, H., 2011. A life-cycle approach tomanagement of organic household food waste A swedish case study. WasteManage. (Oxford) 31, 18791896.

    Blecken, G., Viklander, M., Svensson, G., Hedstrm, A., 2010. Fett I avloppsnt

    kartlggning och tgrdsfrslag (Fat in wastewater sewerage systems mapping and suggestions for improvements). Report 2010-03. Svenskt Vattenoch Utveckling. (in Swedish).

    Bolzonella, D., Pavan, P., Battistoni, P., Cecchi, F., 2003. The under sink garbagegrinder: a friendly technology for the environment. Environ. Technol. 24 (3),349359.

    Carlsson, M., Uldal, M., 2009. Substrathandbok fr biogasproduktion (Substrates forbiogas production). Svenskt Gastekniskt Center (SGC). Stockholm, Sweden (inSwedish).

    Christensen, T.H., la Cour Jansen, J., Jrgensen, O., 2003. Datarapport omsammanstning og biogaspotentiale i organisk dagrenovation (Data report oncomposition and biogas potential in organic household waste generation).Miljprojekt Nr. 815, Miljstyrelsen (in Danish).

    Dahln, L., 2008. Household Waste Collection Factors and Variations. Doctoralthesis. 2008:33. Lule Technical University. Department of Civil, Mining andEnvironmental Engineering. Division of Waste Science and Technology.

    Davidsson, ., Gruvberger, C., Christensen, T.H., Lund Hansen, T., Jansen, J., 2007.Methane yieldin source-sorted organic fraction of municipal solid waste. WasteManage. (Oxford) 27, 406414.

    Davidsson, ., Pettersson, F., Bernstad, A., 2011. Frstudie av olika system frmatavfallsutsortering med avfallskvarnar (Investigationof different systems for

    202 A. Bernstad et al. / Waste Management 33 (2013) 193203

  • 7/25/2019 Bernstad et al.pdf

    11/11

    separate collection of food waste using food waste disposers). Report 231.Svenskt Gastekniskt Centrum (SGC), Stockholm, Sweden.

    EuropeanParliament, 2008. EU Waste Framework Directive Directive 2008/98/EC oftheEuropean Parliament and of theCouncilof 19 November2008on WasteandRepealing Certain Directives. European Comission, Brussels, Belgium.

    European Parliament, 2009. Regulation (ec) no 1069/2009 of the EuropeanParliament and of the COUNCIL of 21 October 2009 Laying down Health Rulesas Regards Animal by-Products and Derived Products not Intended for HumanConsumption and Repealing Regulation (EC) No 1774/2002.

    Evans, T.D., Andersson, P., Wievegg, ., Carlsson, I., 2010. Surahammar a case

    study of the impacts of installing food waste disposers in fifty percent ofhouseholds. Water Environ. J. 24, 309319.

    Galil, N., Yaacov, L., 2001. Analysis of sludge management parameters resultingfrom the use of domestic garbage disposers. Water Sci. Technol. 44, 2734.

    Garcia, M.T., Campos, E., Dalmau, M., Illn, M., Snchez-Leal, J., 2006. Inhibition ofbiogas production by alkyl benzene sulfonates (LAS) in a screening test foranaerobic biodegradability. Biodegradation 17, 3946.

    Gavala, H.N., Haagensen, F., Mogensen, A.S., Angelidaki, I., Ahring, B.K., 2001. Linearalkylbenzene sulfonates: inhibition effect on acetate and propionatedegradation during anaerobic digestion. In: Proceedings from the 4thInternational Symposium for Environmental, Biotechnology, pp. 452455.

    Gteborgs Stad, 2011. Om kksavfallskvarnar tillts i Gteborg - Tekniska ochekonomiska konsekvenser. (Allowing food waste disposers in Gothenburg Technical and economic consequences). CIT Urban Water Management AB,Stockholm, Sweden (in Swedish).

    Hansen, T.L., EjbyeSchmidt, J., Angelidaki,I., Marca, E.,la Cour Jansen, J., Mosbk, H.,Christensen, T., 2003. Method for determination of methane potentials of solidorganic waste. Waste Manage. 24, 393400.

    Hansen, T., la Cour Jansen, J., Davidsson, ., Christensen, T.H., 2007. Effects ofpretreatment technologies on quantity and quality of source-sorted municipalorganic waste for biogas recovery. Waste Manage. 27 (3), 398406.

    Henze, M., Harremos, P., la Cour Jansen, J., Arvin, E., 1995. Wastewater Treatment:Biological and Chemical Processes. Springer Verlag, Heidelberg, Berlin,Germany.

    Kppalafrbundet and SRAB, 2009. Biologisk behandling av organiskt matavfallmed hjlp av avfallskvarnar (BOA) (Biological treatment of organic food wastewith food waste disposers). Kppalafrbundet and SRAB, Stockholm, Sweden(in Swedish).

    Kemikalieinspektionen, 2011. Use of Tensides en Sweden 2008.Kemikalieinspektionen, Stockholm, Sweden. .

    Khalil, E.F., Whitmore, T.N., Gamal-El-Din, H., El-Bassel, A., Lloyd, D., 1989. Theeffects of detergents on anaerobic digestion. Appl. Microbiol. Biotechnol. 29,517522.

    Konumentfreningen Stockholm, 2009. Rapport frn en slaskhink (Report from adustbin). Stockholm, Sweden, 2009 (in Swedish).

    Levlin, E., Hultman, B., 2008. Konduktivitetsmtningar som mt- och kontrollmetodvid kommunala avloppsanlggningar/Conductivity measurements as measure

    and control method in municipal wastewater treatment plants. Svenskt VattenUtveckling, Report No. 200804, Stockholm (in Swedish).

    LRV, 2009. Insamling av matavfall frn flerfamiljshus erfarenheter frn Lund(Separate collection of food waset from multi-family dwellings). LundsRenhllningsverk, SYSAV, Lund University, Lund, Sweden (in Swedish).

    Lundie, S., Peters, G.M., 2005. Life cycle assessment of food waste managementoptions. J. Clean. Prod. 13 (3), 275286.

    Mnsson, N., Bergbck, B., 2007. Bly, cadmium och kvicksilver Flden och lager IStockholms technosfr (Lead, Cadmium and Mercury Flows and layers in theStockholm technospere). Stockholms Stad (in Swedish).

    Malmqvist, L., 2011. Utvrdering av frbehandlingsprocessen pbiogasanlggningen vid Kungsngens grd i Uppsala/Evaluation of thepretreatment process at the biogas facility at Kungsngen Farm in Uppsala.Master Thesis, Water and Environmental Engineering, Lund University, Lund,Sweden (in Swedish).

    Marashlian, N., El-Fadel, M., 2005. The effect of food waste disposers on municipalwaste and wastewater management. Waste Manage. Res. 23 (20), 2031.

    Nilsson, P., Hallin, P.-O., Johansson, J., Karln, L., Lilja, G., Pettersson, B. ., PetterssonJ., 1990. Kllsortering med avfallskvarnar: En fallstudie i Staffanstorp (Source-separation with food waste disposers: a case study from Staffanstorp). Report

    REFORSK FoU number 54, Lund, Sweden (in Swedish).Raunkjaer, K., Hvitved-Jacobsen, T., Nielsen, P.H., 1995. Transformation of organic

    matter in a gravity sewer. Water Environ. Res. 67, 181188.REVAQ. 2011. Regler fr certifieringssystemet REVAQ tervunnen vxtnring.

    (Rules for REVAQ certification system for recycled nutrients). Edition 2.1,2011-02-07 (in Swedish).

    Riber, C., Christensen, T.H., 2006. Method for fractional solid waste sampling andchemical analysis. Int. J. Environ. Anal. Chem. 87 (5), 321336.

    SEPA. 1994. Swedish Environmental Agency Limits for WWTP-Sludge Use onFarmland. SNFS 1994:2, MS: 72.

    SFS. 1998. SFS 1998:9441. Guidelines on Effluent from Wastewater TreatmentPlants. Svensk frfattningssamling (SFS). Stockholm, Sweden.

    SP. 2010. Certifieringsregler fr Biogdsel (Certification rules for digestate). SPCR120, SP, Sveriges Tekniska Forskninginstitut, Bors, Sweden (in Swedish).

    Spinosa, L., Vesilind, P.A. (Eds.), 2001. Sludge into Biosolids: Processing, Disposaland Utilization. International Water Association (IWA) Publication, 2001. ISBN:9781900222082. IWA Publishing, London, UK.

    Surahammar, 2011. Waste Composition Analyses Performed in SurahammarMunicipality, 2011.

    Swedish Energy Authority, 2007. Energy Performance Contracting en modell frminskad energianvndning och miljpverkan/Energy Performance Contracting a model for decreased energy use and environmental impacts. Report No2007:35, Eskilstuna, Sweden (in Sweden).

    Swedish Waste Management Association, 2011. Nationell kartlggning avplockanalyser av hushllens krl- och sckavfall/National maping of wastecomposition analyses of household bin and bag waste. Report 2011:04, Malm,Sweden (in Swedish).

    Tanaka, S., Ichikawa, T., 1993. Effects of photolytic pretreatment on biodegradationand detoxification of surfactants in anaerobic digestion. Water Sci. Technol. 28(7), 103110.

    Truedsson, C., 2010. Positiva och negativa miljaspekter genom optimisering av.frbehandling av matavfall vid Sysav Biotech/Positive and NegativeEnvironmental Aspects through Optimization of Food Waste PretreatmentatSysav Biotech. Master Thesis, Water and Environmental Engineering, LundUniversity, Lund, Sweden (in Swedish).

    VA SYD, 2010. Environmental Report, Sjlunda Wastewater Treatment Plant.VA SYD, 2011. Tillggsbestmmelser till ABVA (Allmnna bestmmelser fr

    brukande av den allmnna vatten- och avloppsanlggningen) (Localguidelines for emissions to municipal wastewater sewerage systems) (inSwedish). .

    Wainberg, R., Nielsen, J., Lundie, S., Peters, G., Ashbolt, N., Russel, D., Jankelson, C.,2000. Assessment of Food Disposal Options in Multi-Unit Dwellings in Sydney.Report from the Co-operative Research Centre for Waste Management &Pollution Control (CRC), University of Western Sydney. Report 2883R.

    WRAP, 2008. The Food we Waste. Report RBC405-0010, second ed. WRAP, London,UK. ISBN 1-84405-383-0.

    A. Bernstad et al. / Waste Management 33 (2013) 193203 203