Waste Management 34 (2014) 669–675
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Treatment of the biodegradable fraction of used disposable diapers by co-digestion with waste activated sludge M. Torrijos a,⇑, P. Sousbie a, M. Rouez b, M. Lemunier b, Y. Lessard b, L. Galtier c, A. Simao c, J.P. Steyer a a
INRA, UR0050, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne F-11100, France Suez Environnement CIRSEE, 38 rue du Prés. Wilson, Le Pecq F-78230, France c SITA UP, TOUR CB21, 16 place de l’Iris, PARIS La Défense F-92040, France b
a r t i c l e
i n f o
Article history: Received 8 November 2012 Accepted 22 November 2013 Available online 28 December 2013 Keywords: Anaerobic digestion Co-digestion Diaper Waste activated sludge Pretreatments
a b s t r a c t The results presented in this paper are part of a project aimed at designing an original solution for the treatment of used disposable diapers permitting the recycling of materials and the recovery of energy. Diapers must be collected separately at source and transported to an industrial facility to undergo special treatment which makes it possible to separate plastics and to recover a biodegradable fraction (BFD) made up mainly of cellulose. The methane yield of BFD was measured and found to be 280 ml CH4/g VSfed on average. 150 kg of dry BFD can be retrieved from the treatment of one ton of used disposable diapers, representing an energy potential of about 400 kW h of total energy or 130 kW h of electricity. As the treatment process for used diapers requires very high volumes of water, the setting up of the diaper treatment facility at a wastewater treatment plant already equipped with an anaerobic digester offers the advantages of optimizing water use as well as its further treatment and, also, the anaerobic digestion of BFD. The lab-scale experiments in a SBR showed that BFD co-digestion with sewage sludge (38% BFD and 62% waste activated sludge on volatile solids basis) was feasible. However, special attention should be paid to problems that might arise from the addition of BFD to a digester treating WAS such as insufficient mixing or floating particles leading to the accumulation of untreated solids in the digester. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Disposable diapers are absorbent products for personal hygiene designed to absorb and retain urine and faeces from babies or from adults with incontinence problems. Nowadays, it is estimated that disposable diapers for babies are used in more than 95% of families in advanced economies (EDANA, 2008) and that a child, before being toilet-trained, requires 3796 diapers (EDANA, 2008) with an average of 4.16 a day in the UK for example (Environmental Agency, 2008). The French national survey on household waste conducted in France by ADEME (2010) highlighted that the sanitary textile fraction (including diapers, sanitary towels, tissue papers, cotton, wet wipes, paper towel, paper tablecloths and napkins, etc.) represented 9% by mass of total municipal solid waste, amounting to 34 kg of waste per inhabitant per year. Colón et al. (2010) estimated that the specific generation of waste from disposable diapers in Europe in 2007 was 4,278,461 tonnes which is 1.66% of total municipal waste generated and 3% of the organic fraction. According to Espinosa-Valdemar et al. (2011), diapers accounted for 6% of urban solid waste generated in Mexico in 1997 and in the range 5–15% in 2008. ⇑ Corresponding author. Tel.: +33 4 68 42 51 51; fax: +33 4 68 42 51 60. E-mail address: [email protected] (M. Torrijos). 0956-053X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2013.11.009
An unused disposable diaper generally consists of (EDANA, 2008): (i) a liquid-permeable membrane lining the inside surface, made of non-woven polypropylene (PP) or polyethylene (PE); (ii) a watertight membrane on the outer surface made from PP, PE, starch, woven cloth or rubber; (iii) an absorbent core (pulp fluff) made up of a fibrous material (cellulose, hemp or synthetic materials) enclosed in water-resistant paper; (iv) the absorbent part also contains a super-absorbent polymer material (sodium polyacrylates), which has a high capacity for bonding with water, making it possible to retain urine within the absorbent part. The efficiency of a diaper is highly dependent on its capacity to absorb and retain urine; (v) finally, a diaper also contains minor amounts of tapes, elastics and adhesive material. The typical composition of a disposable baby diaper has been described by EDANA (2011): fluff pulp 36.6%, SAP 30.7%, PP 16%, LDPE 6.2%, tape, elastic and adhesive 10.5%. A used baby diaper will contain excreta made up of faeces and urine. The quantity of excreta per diaper was first evaluated to be about 96 g but the results of a small survey and analysis of bins showed that it was much higher with 192 g of excreta per diaper composed on average of 18% faeces and 82% urine (Environmental Agency, 2008). Close results were reported by Colón et al. (2010) for the average quantity of excreta per diaper (171 g) but with a different distribution (6% faeces and 94% urine). At present, diapers are generally not collected separately and are disposed of as solid municipal waste for further treatment,
670
M. Torrijos et al. / Waste Management 34 (2014) 669–675
mainly by incineration, land filling (Mirabella et al., 2013) and, to a lesser degree, by composting (Colón et al., 2010, 2013) or anaerobic digestion. Used disposable diapers for babies are generated in family houses but also by collective institutions, such as day nurseries or maternity hospitals, which are a strategic key element to introduce separate collection of used diapers and alternatives to the conventional treatments. The quantity of diapers generated by licensed child care arrangements varies a lot among the European countries. Indeed, it was estimated to represent 6.63% of the total used diapers in France but up to 16.36% and 20.72% in Sweden and Denmark, respectively (Colón et al., 2013). A process has been patented for the treatment of absorbent sanitary paper products, including disposable diapers, which makes it possible to separate such products into their different components in a form suitable for recycling or re-use (Conway et al., 1996). The patent describes different possible equipment and designs for the entire treatment-process line. It comprises a first step of soaking in an aqueous solution to remove particulate and/or soluble matter from the products to be treated. The soaking operation is run under conditions conducive to the non-destructive separation of the constituents. In the recommended process set-up, the product obtained is then treated with an aqueous washing solution and a further solution which includes bleaching and antimicrobial agents. Finally, plastics and the cellulosic material are recovered in separate streams. A research programme, based on the Conway patent, has been launched and the results presented in this paper were obtained during this programme. The aim is to divert used diapers from municipal solid waste streams and classic disposal methods such as incineration or land filling by designing an original approach involving material recycling and energy recovery. In this approach, used disposable diapers must be separated at source, collected separately and sent to an industrial center for treatment according to the process patented by Conway et al. (1996). The main steps are: coarse shredding, pulping and separation of plastics, separation of the super-absorbent polymer (SAP) and recovery of the biodegradable fraction of diapers (BFD). As this treatment process requires the use of large volumes of water, the industrial center could be set up at a wastewater treatment plant (WWTP) to optimize water use and further post-treatment. Indeed, the effluent at the outlet of the WWTP could be used in the treatment of the diapers and then recycled back to the start of the wastewater treatment line for pollution removal. Furthermore, the biodegradable fraction of diapers could be co-digested with waste activated sludge for energy recovery in pre-existing anaerobic digesters. Co-digestion would be an advantageous option for anaerobic digesters treating waste activated sludge, particularly in the case of under-loaded digesters, as it should make it possible to increase the anaerobic treatment’s overall methane production. Co-digestion of waste activated sludge has been successfully investigated with several kinds of co-substrate such as fruit and vegetable waste (Rizk et al., 2007, Bouallagui et al., 2009), meat industry waste (Buendía et al., 2009), the organic fraction of municipal waste (Zupancic et al. (2008), Derbal et al. (2009)) and, of course, residues containing fat, oil and grease which are particularly suitable thanks to their high methane potential (Davidsson et al. (2008), Wan et al. (2011), Girault et al. (2012)). The work presented in this paper was part of the research programme aimed at evaluating the technical feasibility of the treatment at an industrial center of used disposable diapers for the recovery of their plastic fraction for recycling as raw material and of their biodegradable fraction for energy production. The experiments focused first, on the assessment of the methane yield of the biodegradable fraction of diapers (BFD); and second, on the co-digestion of the BFD with waste activated sludge in a lab-scale reactor.
2. Materials and methods 2.1. Separation of the constituents of the diapers at lab-scale For the application of the patented process (Conway et al., 1996) to the separation of the constituents of used disposable diapers, the following process, made up of three major successive steps (see Fig. 1), was set up. (i) First step: coarse shredding to open up the diapers and facilitate subsequent separation of constituents. It should produce plastic pieces large enough in size for them to be removed easily in the second step. Diapers were crushed to a size of approximately 1 cm using a Blik BB 230 crusher equipped with stainless steel rotating blades. (ii) Second step: pulping of the shredded materials and recovery of the plastics. In this step, water was added to reduce dry matter content to a low level (1–2% total solids (TS)) and an adapted mixing was used to permit the ‘‘solubilization’’ of faeces, cellulose and SAP. To prevent the SAP from swelling, a chemical agent was added to the pulping water (dosage used is confidential information). After a few minutes of treatment, the SAP became reticulated and formed small hard particles. At pulper outlet, a screen retained the plastic elements while letting the water containing faeces, cellulose and SAP go through for further treatment. A simple domestic washing machine was used at laboratory scale with a ten-minute washing cycle. Emptying the machine enabled the faeces, cellulose and SAP to be recovered, the plastics remaining in the machine drum. (iii) Third and final step: SAP separation from others components (faeces and cellulose). A simple 100 l tank equipped with a mechanical stiring system was used. Gentle agitation allowed the cellulose to be maintained in suspension whereas the SAP settled to the bottom of the tank where a draining pipe permitted its recovery. The biodegradable solid fraction of diapers (BFD), mainly cellulose, was recovered with a 60 lm screen installed along a water recirculation loop on the top part of the 100 l tank. This BFD fraction had a TS concentration of around 10%. Faeces and urine were eliminated with the water flow. 2.2. Protocol for biochemical methane potential assessment The biochemical methane potential (BMP) was measured in anaerobic flasks. The volume of each flask was 570 ml, with a working volume of 400 ml, the remaining 170 ml volume serving as head space. The following solutions were added to the contents of each flask: (i) 3.4 ml of a solution of NH4Cl at 26.6 g/l, KH2PO4 at 10 g/l, MgCl2 at 3 g/l; (ii) 4 ml of a solution of FeCl2 at 2 g/l, CoCl2 at 0.5 g/l, MnCl2 at 0.1 g/l, NiCl2 at 0.1 g/l, ZnCl2 at 0.05 g/l, H3BO3 at 0.05 g/l, Na2SeO3 at 0.05 g/l, CuCl2 at 0.04 g/l, Na2MoO4 at 0.01 g/ l; (iii) 20.8 ml of a solution of NaHCO3 at 50 g/l. The anaerobic sludge was added to the flasks at a concentration of 5 g VS/l. The initial substrate concentration (S0) over the initial volatile suspended solids (VSS) concentration (X0), i.e. the S0/X0 ratio, was 0.5. Once the flasks were prepared, degasification with nitrogen was carried out to obtain anaerobic conditions and the bottles were closed with red butyl rubber septum-type stoppers which were air tight. Duplicate bottles were incubated at 35 °C. Biogas volume was monitored by the water displacement method. Acidified water (pH = 2) was used to minimize the dissolution of carbon dioxide in water. At each volume measurement, biogas composition was analysed. 2.3. Description of the reactors Reactor experiments were carried out in double-walled glass reactors of 6 l effective volume, maintained at 35 °C by a regulated water bath. Mixing was done by a mechanical stirring system
M. Torrijos et al. / Waste Management 34 (2014) 669–675
671
Fig. 1. Lab-scale process used for the separation of the constituents of used disposable diapers (SAP: super absorbent polymer; BFD: biodegradable fraction of diapers).
made up of a shaft with radial dual impellers. The substrates were added manually through a 3-cm opening located in the top of the reactors. The volume of biogas produced was measured on-line by Milligascounter MGC-1 flow meters (Ritter gas meters) fitted with a 4–20 mA output. The ‘‘Modular SPC’’ software developed at the INRA Narbonne laboratory was used to log the data. The measurement of the volumes of biogas and methane produced was done once a week. The reactors were seeded with anaerobic sludge taken from an industrial-scale anaerobic UASB reactor treating the effluents from a sugar refinery. 2.4. Operation of the reactors 2.4.1. Reactors operated in batch mode The methane yields of the biodegradable fraction of diapers (BFD) and of the waste activated sludge (WAS) were evaluated in 6 l reactors operated in batch mode for about one month. For each substrate, several successive batches were carried out (14 for BFD and 8 for WAS). At the beginning of each batch, 1 g of VS/l was added. At the beginning of the experiment, the S0/X0 ratio was 0.127 g VSadded/g VSSreactor for BFD and 0.07 g VSadded/g VSSreactor for WAS. It was different for the two substrates due to different initial VSS concentrations (8.2 and 13.7 g/l for the reactors fed with BFD and WAS, respectively). The volume of biogas produced over time was monitored on-line, making it possible to measure the total volume of biogas and the kinetics of biogas production for each batch. The biogas production rate (BPR) was maximum at the beginning of a batch and then decreased over time, by the end reaching a low but fairly constant biogas production rate. It was considered that when the BPR was less than 15 ml/h, the biogas production observed was linked only to the endogenous respiration and that the added organic matter had been eliminated. The volume of biogas produced by endogenous respiration was calculated using the biogas production rate measured during the last hours of a batch, when the BPR was less than 15 ml/h, assuming that endogenous respiration was constant throughout the entire duration of a batch. This volume was subtracted from the overall total volume of biogas produced to evaluate the volume of biogas produced exclusively from the added organic matter. At the end of each experiment, the full volume of biogas produced during the last batch was collected into a bag to analyze its average composition in order to measure the volume of methane produced and to calculate the methane potential of the added residue.
2.4.2. Reactor operated in sequencing or SBR mode A 6 l reactor was fed with a mixture made up of 32% BFD and 68% WAS. This ratio is based on the quantity of diapers that could be separately collected in a specific location in France and the quantity of WAS currently generated at the wastewater treatment plant located in the area taken into consideration. To minimize loss of suspended solids during withdrawal, the reactors were operated in sequencing batch reactor (SBR) mode that is to say with cycles made up of the following 4 steps: feed, react, settle and withdrawal. As the reactors were fed each working day (Monday to Friday), there were four 24 h batches and one 3-day batch. Each cycle started with the feeding of the reactor when a known mass of BFD and a known volume of WAS were added manually under a N2 flux. The reactor was then closed so that the react phase could start. The volume of biogas produced was monitored on-line throughout the phase. The duration of the react phase was 22 h for the 24 h batches and 70 h for the 3-day batches. At the end of the react phase, mechanical agitation was stopped for 2 h to allow the sludge to settle. At the end of the decant phase, a volume of supernatant was withdrawn so as to maintain a constant liquid level at the beginning of each cycle. The total duration of the experiment was 118 days. The organic loading rate (OLR) was increased from an initial value of 0.92 g VS/l batch to 1.61 g VS/l batch after 35 days of operation then, finally, to 2.03 g VS/l batch from day 78. The average OLR and HRT per week taking into account 5 days with feeding and 2 days without feeding were 0.66 g VS/l d, 1.15 g VS/l d and 1.45 g VS/l d for OLR and 60 d, 32 d and 27 d for HRT, respectively for the 3 operating conditions. The pH in the reactor was maintained at 7.1 by the addition of 30% soda when needed. The initial solids concentration after inoculation was 10.1 g SS/l and 8.3 g VSS/l. 2.5. Sampling and analysis The soluble fraction was obtained by centrifugation of the samples at 27,000g for 15 min. Soluble COD was measured by a colorimetric method (Jirka and Carter, 1975) using Hach 0–1500 mg/l vials. Other parameters (TS, VS) were measured following Standard Methods, APHA (2012). Volatile fatty acid (VFA) concentration was measured using a gas chromatograph (GC-8000, Fisons Instruments) equipped with a flame ionization detector and an automatic sampler (AS 800, Fisons Instruments). The column used was a semi-capillar Econocap
672
M. Torrijos et al. / Waste Management 34 (2014) 669–675
FFAP (Alltech) column, 15 m length, diameter 0.53 cm, and 1.2 lm Phase EC™ 1000 film. The temperature of the spitless injector was 250 °C, the temperature of the detector 275 °C. The temperature rose from 80 °C to 120 °C in 3 min. The carrier gas was nitrogen (25 kPa). The volume of sample injected was 1 ll. The calibration was made from a mixture of 6 acids (standard solution): acetic (C2), propionic (C3), butyric (C4), isobutyric (iC4), valeric (C5) and isovaleric (iC5) acids at 1 g/l each. The calibration range was 0.25–1 g/l by dilution of the standard solution. The internal standard method (1 g of ethyl-2-butyric acid in 1 l of water acidified with 50 ml of H3PO4) was used to measure total VFA concentration by mixing volumes 1/1 of the internal standard solution and the sample or the standard solution. The margin for error of this measurement was between 2% and 5% with a quantification threshold of 0.1 g/l. The volume of biogas produced was measured by Milligascounter MGC-1 flow meters (Ritter Gas Meters) at 25 °C and local atmospheric pressure. Biogas composition was determined using a gas chromatograph (Shimadzu GC-8A) connected to a C-R8A integrator and equipped with a CTRI Alltech column. The following gases were measured: CO2, H2, O2, N2 and CH4. The column was made up of two concentric columns: the 3.175 mm-diameter inner column, filled with Silicagel, permitted the separation of CO2 from the other gases; the 6.350 mm-diameter outer column, filled with a molecular sieve, separated these other gases. The carrier gas was argon at 2.8 bars. The temperatures were 30 °C for the oven and 100 °C for both the injector and detector. The detection of gaseous compounds was done using a thermal conductivity detector with the intensity of current at 80 mA. The volume of injected biogas was 1 ml. The calibration was done with a standard gas composed of 25% CO2, 5% H2, 2% O2, 10% N2 and 58% CH4. The margin for error of this measurement was 5%. 3. Results and discussion 3.1. Composition of the substrates 3.1.1. BFD A mixture of baby and adult used diapers (50%/50% in mass) was treated according to the protocol described above in Section 2in order to prepare the biodegradable fraction of the diapers (BFD) which was dried after retrieval. The ratio 50% baby/50% adult diapers was chosen according to the potential feedstock of diapers that could be separately collected in a specific location in France. Before its use for laboratory experiments, the BFD was prepared from dried BFD by dilution to a concentration of 10% suspended solids to have the same solids concentration as that of the BFD obtained at the end of industrial-scale treatment. Only one batch of BFD was used for the entire study.
Table 1 Composition of the substrates.
Total solids Volatile solids VS/TS
WAS 1st batch (%)
WAS 2nd batch (%)
BFD (%)
4.1 3.1 76
4.4 3.2 73
10.9 10.4 95
of primary and secondary sludge and its BMP was fully in agreement with values for this kind of residue as reported in the literature: for example, 146–351 ml CH4/g VSfed reported by Kepp and Solheim (2000) or 206–427 ml CH4/g VSfed reported by Mottet et al. (2010). However, the methane potential per g of raw substrate was much higher for BFD compared to WAS, due to the higher VS content of BFD, with 29 ml of CH4 per g of raw BFD but only 9.3 ml of CH4 per g of raw WAS. In a second stage, the BMPs of different mixtures of BFD and WAS were measured and the results are presented in Fig. 2. The BMPs of the different mixtures ranged between 239 and 340 ml CH4/g VS but without any clear influence on the BMP value of the ratio of the two substrates. This might be explained by the difficulty in obtaining satisfactory reproducibility with small quantities of inhomogeneous substrates. The average BMP of the six mixtures was 283 ± 36 ml CH4/gVSfed which is close to the average value of the BMP of the two substrates alone. This indicates that there was no effect on the BMP, either positive or negative, when the two substrates were mixed. 3.3. Assessment of the methane yield of BFD and WAS in batch reactors Two 6 l reactors were operated in fed-batch mode, one fed the biodegradable fraction of diapers alone, the other with waste activated sludge alone in order to measure the methane yield of each substrate individually. Several batches were done for each substrate (14 for BFD and 8 for WAS) in order to measure the methane yield with acclimatized biomass. Fig. 3 presents a comparison of the evolution over time of the methane yield for BFD and WAS during a batch. For BFD, the same substrate as that used for the measurement of the BMP was used while for WAS a new batch of substrate was used. For BFD, the methane yield was 285 ml CH4/g VSfed, which is very close to the value measured in 550 ml flasks using the BMP protocol (275 ml CH4/g VSfed). However, for WAS, a much higher value was found with 485 ml CH4/g VSfed. The different result obtained for WAS can be explained by the use of a new batch of WAS. The very high methane yield suggests that this WAS
3.1.2. WAS Two batches of waste activated sludge (WAS) were used, the first was used for the BMP tests and the second for the experiments in reactors operated both in batch and SBR modes. The composition of the substrates used in this study is presented in Table 1. 3.2. Assessment of the BMP of BFD, WAS and different mixtures of the two substrates The biochemical methane potentials (BMP) of the biodegradable fraction of the diapers and the waste activated sludge alone were evaluated in a first stage. The BMPs of the two substrates were quite close, with 275 ml CH4/g VSfed for BFD and 300 ml CH4/g VSfed for WAS. The WAS used in this work was a mixture
Fig. 2. Evolution of the biochemical methane potential (BMP) with the increase in the percentage of the biodegradable fraction of diapers in the mixture of BFD and WAS.
400 300 200 BFD
100
WAS
0 0
20
40
60
80
100
120
140
160
180
Volumes of biogas and methane produced per week (L)
Methane yield in NmL/g VS
500
50 45 40
1,60 Biogas OLR
30
1,00
25
0,80
20
0,60
15
0,40
10
0,20
5 0
0,00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
3.4. Results with a SBR fed with a mixture of BFD and WAS A SBR was fed for 118 days with a mixture of 32% BFD and 68% WAS based on TS mass, that is to say 20%/80% on mass basis and 38%/62% on VS basis. The reactor was run at incremental average weekly organic loading rates (OLR) of: 0.66 g VS/l d (5 batches/ week at 0.92 g VS/l batch) for about 5 weeks (days 1–34), then 1.15 g VS/l d (5 batches/week at 1.61 g VS/l batch) for about 6 weeks (days 35–76) and finally 1.45 g VS/l d (5 batches/week at 2.03 g VS/l batch) for the last 6 weeks of the experiment (days 77–118). OLR was not increased beyond 1.45 g VS/l d as this OLR was the maximum OLR planed for application at an industrial scale. The hydraulic retention time was 60 d, 32 d and 27 d at, respectively, OLRs of 0.66, 1.15 and 1.45 g VS/l d. 3.4.1. Volume of gas produced and methane yield The volume of biogas produced during each batch was monitored on-line. As the duration of batch treatment was not constant during the week, with four 24 h batches and one 72 h batch, the volumes of biogas and methane produced were reported per week. Fig. 4 presents the evolution over time of the organic loading rate and of the volumes of biogas and methane produced per week. In week 14, the OLR was lower because the reactor was fed only 4 times. As a consequence, the volumes of biogas and methane were lower compared to the other weeks. Fig. 4 shows clearly that the volumes of biogas and methane produced increased rapidly after each increase in the organic loading rate, with a weekly methane production of 32 l at 0.66 g VS/l d, 57 l at 1.15 g VS/l d and 71 l at 1.45 g VS/l d. These results are plotted in Fig. 5 which again shows clearly that the increase in the volume of methane produced was proportional to the increase in the OLR. Consequently, there was no deterioration of microbial activity even at the highest OLR. The average methane content of the biogas produced was 63 ± 2.9%. The slope of the straight line in Fig. 5 indicates the methane yield of the mixture 32% BFD/68% WAS; it was 411 ml CH4/g VSfed. This value is in total agreement with the expected value calculated using the methane yields measured for each individual substrate in the batch reactor. Indeed, the methane yield was 285 ml CH4/g VSfed for BFD and 485 ml CH4/g VSfed for WAS, and considering the proportion of each component (38% for BFD and 62% for WAS as a percentage of VS), the expected methane yield was 409 ml CH4/g VSfed. This mixture had a much higher methane yield than that measured in the first part of the experiment using the BMP protocol
Time in weeks Fig. 4. Evolution over time of the OLR and of the volumes of biogas and methane produced per week in the SBR fed with a mixture of 32% BFD and 68% WAS (as a percentage of VS).
Volume of methane produced per week (L)
contained a high proportion of primary sludge, probably with a significant quantity of fat and greases.
1,20
35
Time in hours Fig. 3. Evolution over time of the methane yield of the biodegradable fraction of diapers (BFD) and of waste activated sludge (WAS) in a batch reactor, after removal of the endogenous respiration.
1,40
Methane
Organic Loading Rate (g VS/L.d)
673
M. Torrijos et al. / Waste Management 34 (2014) 669–675
35 30 25 20 15 10 5 0 0
0,5
1
1,5
2
Organic Loading Rate (gVS/L.d) Fig. 5. Evolution of the volume of methane produced per week with the OLR in the SBR fed for a mixture of 32% BFD and 68% WAS.
(around 283 ml CH4/g VSfed). Again, this is linked to the use of a new batch of WAS in the experiments in the reactor, with a much higher methane yield than that of the first batch (485 ml CH4/g VSfed instead of 300 ml CH4/g VSfed). 3.4.2. Soluble matter and solids concentration in the reactor Soluble COD in the reactor always remained very low, with a maximum of 1.75 g/l at the end of the experiment, and VFA concentration was always very low (
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Treatment of the biodegradable fraction of used disposable diapers by co-digestion with waste activated sludge M. Torrijos a,⇑, P. Sousbie a, M. Rouez b, M. Lemunier b, Y. Lessard b, L. Galtier c, A. Simao c, J.P. Steyer a a
INRA, UR0050, Laboratoire de Biotechnologie de l’Environnement, Avenue des Etangs, Narbonne F-11100, France Suez Environnement CIRSEE, 38 rue du Prés. Wilson, Le Pecq F-78230, France c SITA UP, TOUR CB21, 16 place de l’Iris, PARIS La Défense F-92040, France b
a r t i c l e
i n f o
Article history: Received 8 November 2012 Accepted 22 November 2013 Available online 28 December 2013 Keywords: Anaerobic digestion Co-digestion Diaper Waste activated sludge Pretreatments
a b s t r a c t The results presented in this paper are part of a project aimed at designing an original solution for the treatment of used disposable diapers permitting the recycling of materials and the recovery of energy. Diapers must be collected separately at source and transported to an industrial facility to undergo special treatment which makes it possible to separate plastics and to recover a biodegradable fraction (BFD) made up mainly of cellulose. The methane yield of BFD was measured and found to be 280 ml CH4/g VSfed on average. 150 kg of dry BFD can be retrieved from the treatment of one ton of used disposable diapers, representing an energy potential of about 400 kW h of total energy or 130 kW h of electricity. As the treatment process for used diapers requires very high volumes of water, the setting up of the diaper treatment facility at a wastewater treatment plant already equipped with an anaerobic digester offers the advantages of optimizing water use as well as its further treatment and, also, the anaerobic digestion of BFD. The lab-scale experiments in a SBR showed that BFD co-digestion with sewage sludge (38% BFD and 62% waste activated sludge on volatile solids basis) was feasible. However, special attention should be paid to problems that might arise from the addition of BFD to a digester treating WAS such as insufficient mixing or floating particles leading to the accumulation of untreated solids in the digester. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Disposable diapers are absorbent products for personal hygiene designed to absorb and retain urine and faeces from babies or from adults with incontinence problems. Nowadays, it is estimated that disposable diapers for babies are used in more than 95% of families in advanced economies (EDANA, 2008) and that a child, before being toilet-trained, requires 3796 diapers (EDANA, 2008) with an average of 4.16 a day in the UK for example (Environmental Agency, 2008). The French national survey on household waste conducted in France by ADEME (2010) highlighted that the sanitary textile fraction (including diapers, sanitary towels, tissue papers, cotton, wet wipes, paper towel, paper tablecloths and napkins, etc.) represented 9% by mass of total municipal solid waste, amounting to 34 kg of waste per inhabitant per year. Colón et al. (2010) estimated that the specific generation of waste from disposable diapers in Europe in 2007 was 4,278,461 tonnes which is 1.66% of total municipal waste generated and 3% of the organic fraction. According to Espinosa-Valdemar et al. (2011), diapers accounted for 6% of urban solid waste generated in Mexico in 1997 and in the range 5–15% in 2008. ⇑ Corresponding author. Tel.: +33 4 68 42 51 51; fax: +33 4 68 42 51 60. E-mail address: [email protected] (M. Torrijos). 0956-053X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2013.11.009
An unused disposable diaper generally consists of (EDANA, 2008): (i) a liquid-permeable membrane lining the inside surface, made of non-woven polypropylene (PP) or polyethylene (PE); (ii) a watertight membrane on the outer surface made from PP, PE, starch, woven cloth or rubber; (iii) an absorbent core (pulp fluff) made up of a fibrous material (cellulose, hemp or synthetic materials) enclosed in water-resistant paper; (iv) the absorbent part also contains a super-absorbent polymer material (sodium polyacrylates), which has a high capacity for bonding with water, making it possible to retain urine within the absorbent part. The efficiency of a diaper is highly dependent on its capacity to absorb and retain urine; (v) finally, a diaper also contains minor amounts of tapes, elastics and adhesive material. The typical composition of a disposable baby diaper has been described by EDANA (2011): fluff pulp 36.6%, SAP 30.7%, PP 16%, LDPE 6.2%, tape, elastic and adhesive 10.5%. A used baby diaper will contain excreta made up of faeces and urine. The quantity of excreta per diaper was first evaluated to be about 96 g but the results of a small survey and analysis of bins showed that it was much higher with 192 g of excreta per diaper composed on average of 18% faeces and 82% urine (Environmental Agency, 2008). Close results were reported by Colón et al. (2010) for the average quantity of excreta per diaper (171 g) but with a different distribution (6% faeces and 94% urine). At present, diapers are generally not collected separately and are disposed of as solid municipal waste for further treatment,
670
M. Torrijos et al. / Waste Management 34 (2014) 669–675
mainly by incineration, land filling (Mirabella et al., 2013) and, to a lesser degree, by composting (Colón et al., 2010, 2013) or anaerobic digestion. Used disposable diapers for babies are generated in family houses but also by collective institutions, such as day nurseries or maternity hospitals, which are a strategic key element to introduce separate collection of used diapers and alternatives to the conventional treatments. The quantity of diapers generated by licensed child care arrangements varies a lot among the European countries. Indeed, it was estimated to represent 6.63% of the total used diapers in France but up to 16.36% and 20.72% in Sweden and Denmark, respectively (Colón et al., 2013). A process has been patented for the treatment of absorbent sanitary paper products, including disposable diapers, which makes it possible to separate such products into their different components in a form suitable for recycling or re-use (Conway et al., 1996). The patent describes different possible equipment and designs for the entire treatment-process line. It comprises a first step of soaking in an aqueous solution to remove particulate and/or soluble matter from the products to be treated. The soaking operation is run under conditions conducive to the non-destructive separation of the constituents. In the recommended process set-up, the product obtained is then treated with an aqueous washing solution and a further solution which includes bleaching and antimicrobial agents. Finally, plastics and the cellulosic material are recovered in separate streams. A research programme, based on the Conway patent, has been launched and the results presented in this paper were obtained during this programme. The aim is to divert used diapers from municipal solid waste streams and classic disposal methods such as incineration or land filling by designing an original approach involving material recycling and energy recovery. In this approach, used disposable diapers must be separated at source, collected separately and sent to an industrial center for treatment according to the process patented by Conway et al. (1996). The main steps are: coarse shredding, pulping and separation of plastics, separation of the super-absorbent polymer (SAP) and recovery of the biodegradable fraction of diapers (BFD). As this treatment process requires the use of large volumes of water, the industrial center could be set up at a wastewater treatment plant (WWTP) to optimize water use and further post-treatment. Indeed, the effluent at the outlet of the WWTP could be used in the treatment of the diapers and then recycled back to the start of the wastewater treatment line for pollution removal. Furthermore, the biodegradable fraction of diapers could be co-digested with waste activated sludge for energy recovery in pre-existing anaerobic digesters. Co-digestion would be an advantageous option for anaerobic digesters treating waste activated sludge, particularly in the case of under-loaded digesters, as it should make it possible to increase the anaerobic treatment’s overall methane production. Co-digestion of waste activated sludge has been successfully investigated with several kinds of co-substrate such as fruit and vegetable waste (Rizk et al., 2007, Bouallagui et al., 2009), meat industry waste (Buendía et al., 2009), the organic fraction of municipal waste (Zupancic et al. (2008), Derbal et al. (2009)) and, of course, residues containing fat, oil and grease which are particularly suitable thanks to their high methane potential (Davidsson et al. (2008), Wan et al. (2011), Girault et al. (2012)). The work presented in this paper was part of the research programme aimed at evaluating the technical feasibility of the treatment at an industrial center of used disposable diapers for the recovery of their plastic fraction for recycling as raw material and of their biodegradable fraction for energy production. The experiments focused first, on the assessment of the methane yield of the biodegradable fraction of diapers (BFD); and second, on the co-digestion of the BFD with waste activated sludge in a lab-scale reactor.
2. Materials and methods 2.1. Separation of the constituents of the diapers at lab-scale For the application of the patented process (Conway et al., 1996) to the separation of the constituents of used disposable diapers, the following process, made up of three major successive steps (see Fig. 1), was set up. (i) First step: coarse shredding to open up the diapers and facilitate subsequent separation of constituents. It should produce plastic pieces large enough in size for them to be removed easily in the second step. Diapers were crushed to a size of approximately 1 cm using a Blik BB 230 crusher equipped with stainless steel rotating blades. (ii) Second step: pulping of the shredded materials and recovery of the plastics. In this step, water was added to reduce dry matter content to a low level (1–2% total solids (TS)) and an adapted mixing was used to permit the ‘‘solubilization’’ of faeces, cellulose and SAP. To prevent the SAP from swelling, a chemical agent was added to the pulping water (dosage used is confidential information). After a few minutes of treatment, the SAP became reticulated and formed small hard particles. At pulper outlet, a screen retained the plastic elements while letting the water containing faeces, cellulose and SAP go through for further treatment. A simple domestic washing machine was used at laboratory scale with a ten-minute washing cycle. Emptying the machine enabled the faeces, cellulose and SAP to be recovered, the plastics remaining in the machine drum. (iii) Third and final step: SAP separation from others components (faeces and cellulose). A simple 100 l tank equipped with a mechanical stiring system was used. Gentle agitation allowed the cellulose to be maintained in suspension whereas the SAP settled to the bottom of the tank where a draining pipe permitted its recovery. The biodegradable solid fraction of diapers (BFD), mainly cellulose, was recovered with a 60 lm screen installed along a water recirculation loop on the top part of the 100 l tank. This BFD fraction had a TS concentration of around 10%. Faeces and urine were eliminated with the water flow. 2.2. Protocol for biochemical methane potential assessment The biochemical methane potential (BMP) was measured in anaerobic flasks. The volume of each flask was 570 ml, with a working volume of 400 ml, the remaining 170 ml volume serving as head space. The following solutions were added to the contents of each flask: (i) 3.4 ml of a solution of NH4Cl at 26.6 g/l, KH2PO4 at 10 g/l, MgCl2 at 3 g/l; (ii) 4 ml of a solution of FeCl2 at 2 g/l, CoCl2 at 0.5 g/l, MnCl2 at 0.1 g/l, NiCl2 at 0.1 g/l, ZnCl2 at 0.05 g/l, H3BO3 at 0.05 g/l, Na2SeO3 at 0.05 g/l, CuCl2 at 0.04 g/l, Na2MoO4 at 0.01 g/ l; (iii) 20.8 ml of a solution of NaHCO3 at 50 g/l. The anaerobic sludge was added to the flasks at a concentration of 5 g VS/l. The initial substrate concentration (S0) over the initial volatile suspended solids (VSS) concentration (X0), i.e. the S0/X0 ratio, was 0.5. Once the flasks were prepared, degasification with nitrogen was carried out to obtain anaerobic conditions and the bottles were closed with red butyl rubber septum-type stoppers which were air tight. Duplicate bottles were incubated at 35 °C. Biogas volume was monitored by the water displacement method. Acidified water (pH = 2) was used to minimize the dissolution of carbon dioxide in water. At each volume measurement, biogas composition was analysed. 2.3. Description of the reactors Reactor experiments were carried out in double-walled glass reactors of 6 l effective volume, maintained at 35 °C by a regulated water bath. Mixing was done by a mechanical stirring system
M. Torrijos et al. / Waste Management 34 (2014) 669–675
671
Fig. 1. Lab-scale process used for the separation of the constituents of used disposable diapers (SAP: super absorbent polymer; BFD: biodegradable fraction of diapers).
made up of a shaft with radial dual impellers. The substrates were added manually through a 3-cm opening located in the top of the reactors. The volume of biogas produced was measured on-line by Milligascounter MGC-1 flow meters (Ritter gas meters) fitted with a 4–20 mA output. The ‘‘Modular SPC’’ software developed at the INRA Narbonne laboratory was used to log the data. The measurement of the volumes of biogas and methane produced was done once a week. The reactors were seeded with anaerobic sludge taken from an industrial-scale anaerobic UASB reactor treating the effluents from a sugar refinery. 2.4. Operation of the reactors 2.4.1. Reactors operated in batch mode The methane yields of the biodegradable fraction of diapers (BFD) and of the waste activated sludge (WAS) were evaluated in 6 l reactors operated in batch mode for about one month. For each substrate, several successive batches were carried out (14 for BFD and 8 for WAS). At the beginning of each batch, 1 g of VS/l was added. At the beginning of the experiment, the S0/X0 ratio was 0.127 g VSadded/g VSSreactor for BFD and 0.07 g VSadded/g VSSreactor for WAS. It was different for the two substrates due to different initial VSS concentrations (8.2 and 13.7 g/l for the reactors fed with BFD and WAS, respectively). The volume of biogas produced over time was monitored on-line, making it possible to measure the total volume of biogas and the kinetics of biogas production for each batch. The biogas production rate (BPR) was maximum at the beginning of a batch and then decreased over time, by the end reaching a low but fairly constant biogas production rate. It was considered that when the BPR was less than 15 ml/h, the biogas production observed was linked only to the endogenous respiration and that the added organic matter had been eliminated. The volume of biogas produced by endogenous respiration was calculated using the biogas production rate measured during the last hours of a batch, when the BPR was less than 15 ml/h, assuming that endogenous respiration was constant throughout the entire duration of a batch. This volume was subtracted from the overall total volume of biogas produced to evaluate the volume of biogas produced exclusively from the added organic matter. At the end of each experiment, the full volume of biogas produced during the last batch was collected into a bag to analyze its average composition in order to measure the volume of methane produced and to calculate the methane potential of the added residue.
2.4.2. Reactor operated in sequencing or SBR mode A 6 l reactor was fed with a mixture made up of 32% BFD and 68% WAS. This ratio is based on the quantity of diapers that could be separately collected in a specific location in France and the quantity of WAS currently generated at the wastewater treatment plant located in the area taken into consideration. To minimize loss of suspended solids during withdrawal, the reactors were operated in sequencing batch reactor (SBR) mode that is to say with cycles made up of the following 4 steps: feed, react, settle and withdrawal. As the reactors were fed each working day (Monday to Friday), there were four 24 h batches and one 3-day batch. Each cycle started with the feeding of the reactor when a known mass of BFD and a known volume of WAS were added manually under a N2 flux. The reactor was then closed so that the react phase could start. The volume of biogas produced was monitored on-line throughout the phase. The duration of the react phase was 22 h for the 24 h batches and 70 h for the 3-day batches. At the end of the react phase, mechanical agitation was stopped for 2 h to allow the sludge to settle. At the end of the decant phase, a volume of supernatant was withdrawn so as to maintain a constant liquid level at the beginning of each cycle. The total duration of the experiment was 118 days. The organic loading rate (OLR) was increased from an initial value of 0.92 g VS/l batch to 1.61 g VS/l batch after 35 days of operation then, finally, to 2.03 g VS/l batch from day 78. The average OLR and HRT per week taking into account 5 days with feeding and 2 days without feeding were 0.66 g VS/l d, 1.15 g VS/l d and 1.45 g VS/l d for OLR and 60 d, 32 d and 27 d for HRT, respectively for the 3 operating conditions. The pH in the reactor was maintained at 7.1 by the addition of 30% soda when needed. The initial solids concentration after inoculation was 10.1 g SS/l and 8.3 g VSS/l. 2.5. Sampling and analysis The soluble fraction was obtained by centrifugation of the samples at 27,000g for 15 min. Soluble COD was measured by a colorimetric method (Jirka and Carter, 1975) using Hach 0–1500 mg/l vials. Other parameters (TS, VS) were measured following Standard Methods, APHA (2012). Volatile fatty acid (VFA) concentration was measured using a gas chromatograph (GC-8000, Fisons Instruments) equipped with a flame ionization detector and an automatic sampler (AS 800, Fisons Instruments). The column used was a semi-capillar Econocap
672
M. Torrijos et al. / Waste Management 34 (2014) 669–675
FFAP (Alltech) column, 15 m length, diameter 0.53 cm, and 1.2 lm Phase EC™ 1000 film. The temperature of the spitless injector was 250 °C, the temperature of the detector 275 °C. The temperature rose from 80 °C to 120 °C in 3 min. The carrier gas was nitrogen (25 kPa). The volume of sample injected was 1 ll. The calibration was made from a mixture of 6 acids (standard solution): acetic (C2), propionic (C3), butyric (C4), isobutyric (iC4), valeric (C5) and isovaleric (iC5) acids at 1 g/l each. The calibration range was 0.25–1 g/l by dilution of the standard solution. The internal standard method (1 g of ethyl-2-butyric acid in 1 l of water acidified with 50 ml of H3PO4) was used to measure total VFA concentration by mixing volumes 1/1 of the internal standard solution and the sample or the standard solution. The margin for error of this measurement was between 2% and 5% with a quantification threshold of 0.1 g/l. The volume of biogas produced was measured by Milligascounter MGC-1 flow meters (Ritter Gas Meters) at 25 °C and local atmospheric pressure. Biogas composition was determined using a gas chromatograph (Shimadzu GC-8A) connected to a C-R8A integrator and equipped with a CTRI Alltech column. The following gases were measured: CO2, H2, O2, N2 and CH4. The column was made up of two concentric columns: the 3.175 mm-diameter inner column, filled with Silicagel, permitted the separation of CO2 from the other gases; the 6.350 mm-diameter outer column, filled with a molecular sieve, separated these other gases. The carrier gas was argon at 2.8 bars. The temperatures were 30 °C for the oven and 100 °C for both the injector and detector. The detection of gaseous compounds was done using a thermal conductivity detector with the intensity of current at 80 mA. The volume of injected biogas was 1 ml. The calibration was done with a standard gas composed of 25% CO2, 5% H2, 2% O2, 10% N2 and 58% CH4. The margin for error of this measurement was 5%. 3. Results and discussion 3.1. Composition of the substrates 3.1.1. BFD A mixture of baby and adult used diapers (50%/50% in mass) was treated according to the protocol described above in Section 2in order to prepare the biodegradable fraction of the diapers (BFD) which was dried after retrieval. The ratio 50% baby/50% adult diapers was chosen according to the potential feedstock of diapers that could be separately collected in a specific location in France. Before its use for laboratory experiments, the BFD was prepared from dried BFD by dilution to a concentration of 10% suspended solids to have the same solids concentration as that of the BFD obtained at the end of industrial-scale treatment. Only one batch of BFD was used for the entire study.
Table 1 Composition of the substrates.
Total solids Volatile solids VS/TS
WAS 1st batch (%)
WAS 2nd batch (%)
BFD (%)
4.1 3.1 76
4.4 3.2 73
10.9 10.4 95
of primary and secondary sludge and its BMP was fully in agreement with values for this kind of residue as reported in the literature: for example, 146–351 ml CH4/g VSfed reported by Kepp and Solheim (2000) or 206–427 ml CH4/g VSfed reported by Mottet et al. (2010). However, the methane potential per g of raw substrate was much higher for BFD compared to WAS, due to the higher VS content of BFD, with 29 ml of CH4 per g of raw BFD but only 9.3 ml of CH4 per g of raw WAS. In a second stage, the BMPs of different mixtures of BFD and WAS were measured and the results are presented in Fig. 2. The BMPs of the different mixtures ranged between 239 and 340 ml CH4/g VS but without any clear influence on the BMP value of the ratio of the two substrates. This might be explained by the difficulty in obtaining satisfactory reproducibility with small quantities of inhomogeneous substrates. The average BMP of the six mixtures was 283 ± 36 ml CH4/gVSfed which is close to the average value of the BMP of the two substrates alone. This indicates that there was no effect on the BMP, either positive or negative, when the two substrates were mixed. 3.3. Assessment of the methane yield of BFD and WAS in batch reactors Two 6 l reactors were operated in fed-batch mode, one fed the biodegradable fraction of diapers alone, the other with waste activated sludge alone in order to measure the methane yield of each substrate individually. Several batches were done for each substrate (14 for BFD and 8 for WAS) in order to measure the methane yield with acclimatized biomass. Fig. 3 presents a comparison of the evolution over time of the methane yield for BFD and WAS during a batch. For BFD, the same substrate as that used for the measurement of the BMP was used while for WAS a new batch of substrate was used. For BFD, the methane yield was 285 ml CH4/g VSfed, which is very close to the value measured in 550 ml flasks using the BMP protocol (275 ml CH4/g VSfed). However, for WAS, a much higher value was found with 485 ml CH4/g VSfed. The different result obtained for WAS can be explained by the use of a new batch of WAS. The very high methane yield suggests that this WAS
3.1.2. WAS Two batches of waste activated sludge (WAS) were used, the first was used for the BMP tests and the second for the experiments in reactors operated both in batch and SBR modes. The composition of the substrates used in this study is presented in Table 1. 3.2. Assessment of the BMP of BFD, WAS and different mixtures of the two substrates The biochemical methane potentials (BMP) of the biodegradable fraction of the diapers and the waste activated sludge alone were evaluated in a first stage. The BMPs of the two substrates were quite close, with 275 ml CH4/g VSfed for BFD and 300 ml CH4/g VSfed for WAS. The WAS used in this work was a mixture
Fig. 2. Evolution of the biochemical methane potential (BMP) with the increase in the percentage of the biodegradable fraction of diapers in the mixture of BFD and WAS.
400 300 200 BFD
100
WAS
0 0
20
40
60
80
100
120
140
160
180
Volumes of biogas and methane produced per week (L)
Methane yield in NmL/g VS
500
50 45 40
1,60 Biogas OLR
30
1,00
25
0,80
20
0,60
15
0,40
10
0,20
5 0
0,00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
3.4. Results with a SBR fed with a mixture of BFD and WAS A SBR was fed for 118 days with a mixture of 32% BFD and 68% WAS based on TS mass, that is to say 20%/80% on mass basis and 38%/62% on VS basis. The reactor was run at incremental average weekly organic loading rates (OLR) of: 0.66 g VS/l d (5 batches/ week at 0.92 g VS/l batch) for about 5 weeks (days 1–34), then 1.15 g VS/l d (5 batches/week at 1.61 g VS/l batch) for about 6 weeks (days 35–76) and finally 1.45 g VS/l d (5 batches/week at 2.03 g VS/l batch) for the last 6 weeks of the experiment (days 77–118). OLR was not increased beyond 1.45 g VS/l d as this OLR was the maximum OLR planed for application at an industrial scale. The hydraulic retention time was 60 d, 32 d and 27 d at, respectively, OLRs of 0.66, 1.15 and 1.45 g VS/l d. 3.4.1. Volume of gas produced and methane yield The volume of biogas produced during each batch was monitored on-line. As the duration of batch treatment was not constant during the week, with four 24 h batches and one 72 h batch, the volumes of biogas and methane produced were reported per week. Fig. 4 presents the evolution over time of the organic loading rate and of the volumes of biogas and methane produced per week. In week 14, the OLR was lower because the reactor was fed only 4 times. As a consequence, the volumes of biogas and methane were lower compared to the other weeks. Fig. 4 shows clearly that the volumes of biogas and methane produced increased rapidly after each increase in the organic loading rate, with a weekly methane production of 32 l at 0.66 g VS/l d, 57 l at 1.15 g VS/l d and 71 l at 1.45 g VS/l d. These results are plotted in Fig. 5 which again shows clearly that the increase in the volume of methane produced was proportional to the increase in the OLR. Consequently, there was no deterioration of microbial activity even at the highest OLR. The average methane content of the biogas produced was 63 ± 2.9%. The slope of the straight line in Fig. 5 indicates the methane yield of the mixture 32% BFD/68% WAS; it was 411 ml CH4/g VSfed. This value is in total agreement with the expected value calculated using the methane yields measured for each individual substrate in the batch reactor. Indeed, the methane yield was 285 ml CH4/g VSfed for BFD and 485 ml CH4/g VSfed for WAS, and considering the proportion of each component (38% for BFD and 62% for WAS as a percentage of VS), the expected methane yield was 409 ml CH4/g VSfed. This mixture had a much higher methane yield than that measured in the first part of the experiment using the BMP protocol
Time in weeks Fig. 4. Evolution over time of the OLR and of the volumes of biogas and methane produced per week in the SBR fed with a mixture of 32% BFD and 68% WAS (as a percentage of VS).
Volume of methane produced per week (L)
contained a high proportion of primary sludge, probably with a significant quantity of fat and greases.
1,20
35
Time in hours Fig. 3. Evolution over time of the methane yield of the biodegradable fraction of diapers (BFD) and of waste activated sludge (WAS) in a batch reactor, after removal of the endogenous respiration.
1,40
Methane
Organic Loading Rate (g VS/L.d)
673
M. Torrijos et al. / Waste Management 34 (2014) 669–675
35 30 25 20 15 10 5 0 0
0,5
1
1,5
2
Organic Loading Rate (gVS/L.d) Fig. 5. Evolution of the volume of methane produced per week with the OLR in the SBR fed for a mixture of 32% BFD and 68% WAS.
(around 283 ml CH4/g VSfed). Again, this is linked to the use of a new batch of WAS in the experiments in the reactor, with a much higher methane yield than that of the first batch (485 ml CH4/g VSfed instead of 300 ml CH4/g VSfed). 3.4.2. Soluble matter and solids concentration in the reactor Soluble COD in the reactor always remained very low, with a maximum of 1.75 g/l at the end of the experiment, and VFA concentration was always very low (
