Waste Management 43 (2015) 50–60
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Influence assessment of a lab-scale ripening process on the quality of mechanically–biologically treated MSW for possible recovery Maria Chiara Di Lonardo a,⇑, Erwin Binner b, Francesco Lombardi a a Laboratory of Environmental Engineering, Department of Civil Engineering and Computer Science Engineering, University of Rome ‘‘Tor Vergata’’, Via del Politecnico 1, 00133 Rome, Italy b Institute of Waste Management (ABF-BOKU), University of Natural Resources and Life Sciences Vienna, Muthgasse 107, A-1190 Vienna, Austria
a r t i c l e
i n f o
Article history: Received 26 February 2015 Accepted 22 May 2015
Keywords: Biological stability Heavy metal Leaching behaviour Mechanical–biological treatment Municipal solid waste Ripening process
a b s t r a c t In this study, the influence of an additional ripening process on the quality of mechanically–biologically treated MSW was evaluated in the prospective of recovering the end material, rather than landfilling. The biostabilised waste (BSW) coming from one of the MBT plants of Rome was therefore subjected to a ripening process in slightly aerated lab test cells. An in-depth investigation on the biological reactivity was performed by means of different types of tests (aerobic and anaerobic biological tests, as well as FT-IR spectroscopy method). A physical–chemical characterisation of waste samples progressively taken during the ripening phase was carried out, as well. In addition, the ripened BSW quality was assessed by comparing the characteristics of a compost sampled at the composting plant of Rome which treat source segregated organic wastes. Results showed that the additional ripening process allowed to obtain a better quality of the biostabilised waste, by achieving a much higher biological stability compared to BSW as-received and similar to that of the tested compost. An important finding was the lower heavy metals (Co, Cr, Cu, Ni, Pb and Zn) release in water phase at the end of the ripening compared to the as-received BSW, showing that metals were mainly bound to solid organic matter. As a result, the ripened waste, though not usable in agriculture as found for the compost sample, proved anyhow to be potentially suitable for land reclamation purposes, such as in landfills as cover material or mixed with degraded and contaminated soil for organic matter and nutrients supply and for metals recovery, respectively. In conclusion the study highlights the need to extend and optimise the biological treatment in the MBT facilities and opens the possibility to recover the output waste instead of landfilling. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction In the last two decades, hundreds of mechanical–biological treatment (MBT) facilities were built in Europe to treat the residual municipal solid waste (MSW) from at-source separate collection (Lornage et al., 2007; Barrena et al., 2009; Bayard et al., 2010; Ponsá et al., 2010; Tintner et al., 2010; Di Lonardo et al., 2012). Main goal of MBT is to reduce the environmental impacts and risks to human health related to landfilling, as set by the European Landfill Directive 1999/31/EC (European Commission, 1999). One of the output of the MBT is the biostabilised waste (BSW) produced from the aerobic biological treatment of the organic fraction mechanically separated from the input MSW. In the majority of the MBT facilities, the biological process consists in an active
⇑ Corresponding author. Tel.: +39 06 7259 7497; fax: +39 06 7259 7021. E-mail address: [email protected] (M.C. Di Lonardo). http://dx.doi.org/10.1016/j.wasman.2015.05.028 0956-053X/Ó 2015 Elsevier Ltd. All rights reserved.
biodegradation phase at forced aeration conditions for 20–28 days (Adani et al., 2004a; Di Lonardo et al., 2012; Puyuelo et al. 2010). Differently, in the common practice of the composting processes, the active decomposition phase at forced aeration conditions is followed by a windrow or static pile system at natural aeration conditions for the later stages of decomposition and ripening in order to raise the output compost quality (Richard, 1992). Such a difference in the biological treatment management depends upon the fact that MBT was firstly developed and seen as a MSW pre-treatment for disposal in order to reduce the biodegradable municipal waste going to landfills (Di Lonardo et al., 2012), therefore no high quality of the output was expected. Recently the interest in the possibility to recover the biostabilised waste from MBT is increasing (MacLeod et al., 2008), especially considering the large amounts that are now being produced in efforts to divert waste from landfill (Farrell and Jones, 2010). However, MacLeod et al. (2008) argued that the MBT output must be highly stable so that the users can be confident that
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
contaminants in the material such as heavy metals have reached their maximum concentration. Many studies showed that an efficient reduction of the biological reactivity of mechanically–biologically treated MSW is achieved by means of optimised aerobic biological process lasting at least 8 weeks, including both an active biodegradation phase and a ripening phase (Zach et al., 2000; Francou et al., 2005; Lornage et al., 2007; MacLeod et al., 2008; Barrena et al., 2009; Shao et al., 2009; Bayard et al., 2010; Tintner et al., 2010). In the city of Rome the separate collection of biodegradable municipal waste (BMW) and the related treatment facilities are poorly developed. Data reported in ‘‘Municipal Solid Waste Report 2014’’ (ISPRA, 2014) show that in 2013 a low percentage of BMW, i.e. only 23.9% of the total MSW produced in Rome (corresponding to 1,754,823 Mg), was source segregated. As a result, the residual MSW fed to the MBT plants (about 60% of the total amount of MSW produced in this city) were composed of significant percentages of biodegradable waste, namely 40–50% by weight, as observed by several investigations on material composition performed in the years 2006–2013 (Franzese et al., 2013). In these MBT facilities, about 80% of the biodegradable waste amount, which the input MSW is composed of, is separated through a primary mechanical sieving and then subjected to an aerobic biological treatment at forced aeration conditions for 4 weeks (active biodegradation treatment). Afterwards, the end biostabilised waste is not subjected to a further ripening phase but it is directly landfilled. In previous surveys performed on the biological stability degree, by measuring the dynamic respiration index (DRI) (Di Lonardo et al., 2014), several samples of biostabilised waste coming from the MBT plants of Rome were found highly reactive. Specifically, DRIs were higher than the maximum limit value of 1000 mg O2/kg VS h below which the waste is considered biologically stable (Adani et al., 2004a) and compliant with the requirement set by the Italian landfill regulation for the waste acceptance in landfill (Italian Ministerial Decree, 2010). The high reactivity was mainly related to the relatively short processing period as well as to the significant content of biodegradable waste in residual MSW. Therefore, in this study, the biostabilised waste coming from one of the MBT plants of Rome was subjected to an additional ripening process in slightly aerated lab test cells in order to analyse and evaluate the variation and the increase of the biological stability degree with time of ripening. A physical–chemical characterisation of waste samples progressively taken during the ripening phase was carried out. Furthermore, the ripened BSW quality was also assessed through a comparison with the characteristics of a compost sampled at the composting plant of Rome which treat source segregated organic wastes. The purpose was to evaluate the influence of the additional ripening process on the quality of BSW in the perspective of recovering the end material, rather than landfilling.
2. Materials and methods 2.1. Biological process in the MBT plants of Rome In the city of Rome four MBT plants are currently in operation with a maximum treatment capacity of 3000 Mg MSW/d. The facilities differ in the type of mechanical pre-treatment employed but the biological process is performed in the same conditions and duration, as hereafter described. BSW used for this investigation was sampled in one of the four MBT facilities whose Fig. 1 shows the flow scheme and the outputs of the treatment. Focusing on the biological process, after metal removal by belt-type
51
electromagnetic separators, the biodegradable fraction (i.e. the undersize flow coming from the primary screening unit at 80 mm) is sent to a biostabilisation basin where an aerobic biodegradation occurs for 4 weeks at forced aeration conditions. Three augers moved by a crane have a dual function: turning over the material in order to keep proper free air space (pores) for satisfying aeration (avoiding the formation of anaerobic conditions, especially at the bottom of the basin) and moving the material along the basin. During turning/moving, water is added to the material by nozzles mounted on the crane, in order to keep the water content favourable for microbial activity. The biostabilised output then is sieved at 20 mm by means of a trommel screen in order to separate an oversize fraction mainly composed of plastics and inert materials from the undersize fraction, namely the final biostabilised waste (BSW) which is daily landfilled. BSW was taken at the outlet of the biodegradation process lasting 4 weeks after the sieving unit at 20 mm, following the procedure laid down in the Italian standard UNI 10802 (2013), and after mixing and quartering, an amount of roughly 30 kg was collected. Furthermore a mature compost sample of roughly 20 kg was collected at the composting plant of Rome which treat the organic fractions of municipal solid waste coming from the at-source separate collection. In this case, beside the 4 weeks of aerobic biodegradation at forced aeration conditions, the treatment includes a ripening phase of up to 6 months, differently from the biological process carried out in the MBT plants. The BSW sample was sent to the laboratory at the Institute of Waste Management of the University of Natural Resources and Life Sciences in Vienna (ABF-BOKU) where it was subjected to a lab scale ripening process and analysed in order to measure the biological reactivity. A further physical–chemical characterisation was instead carried out at the Environmental Laboratory of the University of Rome ‘‘Tor Vergata’’. Differently, all tests performed to characterise the compost sample were conducted at the latter laboratory. 2.2. Lab ripening phase The sampled amount of BSW was subjected to a ripening phase by placing it in a vertical cells system simulating an open windrow configuration (Fig. 2). A relatively low air flow rate was fed to the bottom of the system in order to simulate natural aeration conditions. Each single cell was half-filled with the rotting waste placed on a grid (with mesh opening of 10 mm) to let the air passing through the material homogeneously. A biofilter composed of mature compost was placed at the top of the system to reduce odour emissions. Ambient and rotting material temperature, as well as oxygen and carbon dioxide concentrations in the waste air were daily measured for each single cell in order to monitor the process. Mixing, adjusting in water content and turning of the material were carried out once per week to keep optimised conditions for the ripening process. During mixing of material, samples were taken in order to analyse the biological reactivity and evaluate the changing in reactivity with time. From the measures of temperature, O2 and CO2 concentrations, as well as of the VS (as described in Section 2.4) during the ripening, the actual O2 consumed rates (mg O2/kg VS h) and the actual CO2 released rates (mg CO2/kg VS h) per unit of organic matter (VS) were calculated through the following Eqs. (1) and (2) (Komilis and Kanellos, 2012):
O2
cons
¼
ðO2in O2out Þ Q MWO2 1000 MV VS
ð1Þ
52
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
Fig. 1. Flow scheme of the MBT plant of Rome.
CO2
rel
¼
ðCO2in CO2out Þ Q MWCO2 1000 MV VS
ð2Þ
where (O2in O2out) is the concentration difference between the O2 at the inlet (air concentration equal to 20.8% v/v, taking into account that fed air was not dry) and the measured O2 at the outlet air (% v/v); (CO2out CO2in) is the concentration difference between the CO2 at the inlet (air concentration equal to 0.03% v/v, taking into account that fed air was not dry) and the measured CO2 at the outlet air (% v/v); Q is the air flow rate (l/h) set during the ripening phase; MWO2 is molecular weight of oxygen (31.98 g/mol); MWCO2 is molecular weight of carbon dioxide (12 g/mol); 1000 is the conversion factor (mg/g); MV is the molar volume (l/mol), i.e. MV ¼ MVst TTstt , where MVst = 22.4 l/mol is the molar volume at standard temperature Tst = 273.15 K and standard pressure pst = 1 atm (by assuming that O2 and CO2 are perfect gases) and Tt is the measured temperature at time t during the ripening phase; VS are the volatile solids weekly measured (kg); the VS were assumed constant between two samplings. 2.3. Biological reactivity tests The influence of the ripening process on the biological reactivity of BSW sample was evaluated by performing different types of tests, namely aerobic and anaerobic biological tests, as well as FT-IR spectroscopy method. The aerobic or respiration test measures the cumulative oxygen uptake by microorganisms during a period of 4 days, indicated as RA4 and expressed in mg O2/g DM, according to the Austrian Standard OE NORM S 2027-4 (2012a). The aerobic test was performed by using two static respirometric equipment, namely SapromatÒ and OxiTopÒ (Binner et al., 2012). The anaerobic or incubation test measures the microbial gas generation under
optimised anaerobic conditions during a period of 21 days (Binner and Zach, 1999), indicated as GS21 and expressed in Nl/kg DM (Nl = litres under normalised conditions). The incubation test was conducted according to the specifications of the Austrian Standard OE NORM S 2027-2 (2012b). Furthermore, Fourier Transform Infrared (FTIR) spectroscopy of BSW samples was determined since this method enables to assess the reactivity of MBT waste directly via the chemical composition reflected by the FTIR spectrum (Böhm et al., 2010; Pognani et al., 2010). Infrared spectroscopic investigation was carried out using the attenuated total reflection (ATR) technique (Böhm et al., 2010) by means of Bruker Optics ALPHA FTIR Spectrometer. Prior to the FTIR analysis, samples were air-dried, grinded by means of agate mill and screened through a 0.63 mm sieve in order to obtain pulverised and homogeneous samples. Respiration tests and FTIR spectroscopy were performed for BSW as-received (R0), and for BSW after 2 (R2), 3 (R3), 4 (R4) and 5 (R5) weeks of additional ripening phase. Incubation test, instead, was performed for samples R0, R3 and R5. Respiration and incubation tests were conducted in duplicates. Differently, FTIR-spectroscopy was determined for 5 replicates since the sample amount to analyse was quite low (roughly 10 mg). Then the 5 spectra obtained were averaged for data analysis. As for compost, the biological stability was analysed by determining the dynamic respiration index (DRI), namely the absolute maximum rate of oxygen consumption due to microbial activity (Adani et al., 2004b). DRI was measured and calculated according to the procedure reported in the Italian Standard UNI/TS 11184 (2006) by using a 30 l adiabatic respirometric reactor (Costech International Respirometer 3024). 2.4. Physical–chemical characterisation The samples R0 to R5, as well as the compost sample, were air dried prior to be further analysed to organic matter content, heavy metal total content and release in water phase (leaching test). The organic matter content was determined by measuring the volatile solids (VS) and the total organic carbon (TOC). VS content
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
was determined in duplicates by loss-on-ignition at 550 °C for 8 h (UNI/TS 11184, 2006) on 10 g of sample grinded to size lower than 0.63 mm and pre-dried at 105 °C for 4 h. TOC content was analysed in triplicates by means of Shimadzu SSM-5000A instrument on approximately 0.4 g of sample grinded to size lower than 0.2 mm (EN 13137, 2001). In order to evaluate the loss of organic carbon converted to CO2, the cumulative CO2 and C–CO2 mass was calculated by means of the following equations:
mCO2 ;t
V CO2 ;t Q Dt 24 MWCO2 ¼ MV
mCCO2 ;t
V CO2 ;t Q Dt 24 AWC ¼ MV
ð3Þ
ð4Þ
mCO2 ;cum ¼ RmCO2 ;t
ð5Þ
mCCO2 ;cum ¼ RmC;t
ð6Þ
fibre filters) and (3) final filtration at 0.45 lm (as required by the standard method) by means of syringe filters (Sartorius cellulose acetate filters). The dissolved organic carbon (DOC) (Shimadzu TOC-V CPH/CPN analyser) and heavy metal (Varian ICP-AES) concentrations were measured in the filtered eluates. Heavy metals (HM) and total organic carbon (TOC) percentage release was then calculated by means of the following Eq. (6):
%release ¼ 100
C leached 10 C sol
ð8Þ
where Cleach is the concentration (mg/l) measured in the eluates of the two samples, 10 is the liquid to solid ratio (l/kg DM) and Csol is the total content in the solid materials (mg/kg DM). 3. Results and discussion 3.1. Ripening phase monitoring
where mCO2 ;t and mCCO2 ;t are the masses in grams of carbon dioxide and carbon at time t, respectively; V CO2 ;t is the concentration (% v/v) measured at time t during the lab ripening phase; Q is the air flow rate (l/h) set during the ripening phase; Dt is the time interval between two CO2 measures (d) and 24 is the conversion factor days–hours (h/d); MWCO2 is molecular weight of carbon dioxide (g/mol); AWCCO2 is the atomic weight of carbon (g/mol); MV is the molar volume (l/mol); mCO2 ;cum and mC,cum are the cumulative mass in grams of carbon dioxide, equal to 2417.9 g and the cumulative mass of carbon, equal to 659.4 g. Then, the percentage by weight of carbon loss was calculated by means of Eq. (7), as follows:
% C CO2 ¼ 100
53
mCCO2 ;cum TOCini
ð7Þ
where TOCini is the initial TOC content (measured in the sample R0), expressed in grams. Heavy metal content in solid materials (total content) was determined in triplicates by acid digestion according to the method C indicated in the European standard EN 15411 (2011) with some modifications. Specifically, 3 ml of concentrated HNO3 15.698 N (Sigma Aldrich 30709) and 1 ml of concentrated HCl 12.178 N (Sigma Aldrich 30721) were added to 0.1 g of sample grinded to size lower than 0.5 mm. Then the mixture was put in closed vessels (Parr Instrument Company – model 4744) and kept at 150 °C for approximately 15 h. The obtained solution, after cooling, was firstly filtered by means of Whatman No 41 filter paper so to separate residues and then it was diluted to volume with ultrapure water in 25 ml flask. A final filtration at 0.45 lm (Sartorius cellulose acetate syringe filters) was carried out in order to analyse the dissolved heavy metals in the solution by inductively coupled plasma atomic emission spectrometry (Varian ICP-AES). For the leaching test, air-dried samples were grinded to a particle size lower than 4 mm, as required by the European standard EN 12457-2 (2002). Deionised water was added to each sample at a liquid to solid ratio equal to 10 ml/g and bottles containing the mixture were slowly shaken by means of a rotary tumble for 24 h. Such test was conducted in duplicates for each sample. The obtained eluate, after decanting for 15 min, was firstly analysed by measuring the pH (Eustech Instrument pH 700). Afterwards, three steps of solid–liquid separation were carried out: (1) centrifugation at 12,000 rpm for 15 min (Thermo Scientific SL 16R Centrifuge), (2) vacuum filtration at 0.7 lm (Munktell AB glass
Fig. 3 shows the setting of the air flow rate and the trends of ambient and rotting material temperature, as well as the CO2 and O2 concentrations measured in the exhaust air during the lab ripening phase. The highest temperatures, 10–15 °C higher than the ambient temperature, were measured in the first week of ripening phase with a maximum value equal to 41.5 °C observed in the 3rd day. During the 2nd week and half of the 3rd week, temperature decreased and kept around 35 °C, being 5 °C higher than the ambient temperature. During the last 10 days, a further decrease of temperature by approaching to ambient temperature was observed and this was a first indication on the biological stabilisation of the material. The maximum oxygen consumption and corresponding carbon dioxide release were observed during the first 2 weeks of the ripening, proving the high reactivity of the BSW as-received. It has to be noticed that between the 3rd and the 5th day, O2 concentration in the exhaust air significantly increased. This was due to the increase of the air flow rate from 20 l/h (set in the first day) to 60 l/h in order to keep O2 concentration in the exhaust air higher
Fig. 2. Lab ripening phase: vertical cells system scheme.
54
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
than 10% (v/v) below which microbial activity may be strongly hampered or inhibited because of oxygen starvation. From day 12 on, O2 concentration kept at 10–11% (v/v) and CO2 at 8–9% (v/v), increasing to 14% (v/v) and decreasing to 6% (v/v), respectively, during the last three days. This indicated again the achievement of biologically stable conditions, therefore the test was stopped after 5 weeks.
3.2. Biological reactivity Fig. 4 shows the actual O2 consumed and CO2 released rates calculated by means of Eqs. (1) and (2). The same trends of O2 consumed and CO2 released rates during the ripening phase were observed with the maximum values found on the 3rd day, when the maximum temperature of the rotting waste was measured as well, indicating the high reactivity of the as-received BSW. Until the 18th day the rates tended to decrease with an oscillating trend, proving that the waste was still reactive, whereas steady rates were observed during the last ten days because of the achievement of biological stabilisation state. In Fig. 4 the maximum limit values identifying a medium biological stability, i.e. 1000 mg O2/kg VS h, and a high biological stability, i.e. 500 mg O2/kg VS h (Adani et al., 2004b) are highlighted. It can be observed that at the beginning of the ripening the actual O2 consumed rates were higher than both stability limits confirming what was found in a previous study (Di Lonardo et al., 2014), namely that BSW coming from the active biodegradation of 4 weeks in the MBT plants of Rome could not be considered biologically stable. Differently, a high biological stability degree was obtained after 15 days of lab ripening phase since the actual O2 consumed rates were lower than
Fig. 3. (a) Setting of the air flow rate and (b) trends of rotting material and ambient temperatures, as well as CO2 and O2 concentrations in waste air during the ripening phase.
500 mg O2/kg VS h. Compost sample showed to be biologically stable with DRI found to be equal to 620 mg O2/kg VS h, therefore comparable to that of BSW after 2 weeks of ripening process. These observations were verified through the measure of the respiration activity (RA4) by means of Sapromat and OxiTop equipment (as described in Section 2.3). Fig. 5 shows the results of the respiration tests for samples R0 to R5. It can be noticed that BSW as-received (R0) was characterised by a significant respiration activity with RA4 values equal to 21.1 mg O2/g DM from Sapromat and equal to 17.0 mg O2/g DM from OxiTop, exceeding the limit value of 7 mg O2/g DM set by the Austrian Landfill Ordinance (2008). During the ripening phase a reduction in respiration activity was observed by reaching RA4 lower than the limit after 4 weeks of treatment. Furthermore, it can be observed that the average cumulative trend of oxygen consumption was quite steep for sample R0 and then progressively less slope, reaching a certain stability shown by very similar trends found for R4 and R5. Comparing RA4 obtained by the two equipment, comparable values were observed for samples R2–R5, whereas for R0, RA4 measured by OxiTop was lower than that measured by Sapromat, i.e. 17.0 and 21.1 mg O2/g DM, respectively, probably because of the dynamic oxygen replacement provided by Sapromat equipment (Binner et al., 2012). The biological reactivity for samples R0, R3 and R5 was also measured through the incubation test measuring the biogas generation (results are shown in Fig. 6). The average gas generation sum within 21 days of test duration was found to be high for R0 exceeding the limit of 20 Nl/kg DM set by the Austrian Landfill Ordinance (2008). Values lower than the limit were obtained at the end of the ripening process (5 weeks), when a significant reduction occurred. It has to be observed that GS21 for repetition 1 of R0 was found to be quite low. This was due to the high reactivity of BSW which caused the formation of acidification conditions, verified by the relatively low pH (
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Influence assessment of a lab-scale ripening process on the quality of mechanically–biologically treated MSW for possible recovery Maria Chiara Di Lonardo a,⇑, Erwin Binner b, Francesco Lombardi a a Laboratory of Environmental Engineering, Department of Civil Engineering and Computer Science Engineering, University of Rome ‘‘Tor Vergata’’, Via del Politecnico 1, 00133 Rome, Italy b Institute of Waste Management (ABF-BOKU), University of Natural Resources and Life Sciences Vienna, Muthgasse 107, A-1190 Vienna, Austria
a r t i c l e
i n f o
Article history: Received 26 February 2015 Accepted 22 May 2015
Keywords: Biological stability Heavy metal Leaching behaviour Mechanical–biological treatment Municipal solid waste Ripening process
a b s t r a c t In this study, the influence of an additional ripening process on the quality of mechanically–biologically treated MSW was evaluated in the prospective of recovering the end material, rather than landfilling. The biostabilised waste (BSW) coming from one of the MBT plants of Rome was therefore subjected to a ripening process in slightly aerated lab test cells. An in-depth investigation on the biological reactivity was performed by means of different types of tests (aerobic and anaerobic biological tests, as well as FT-IR spectroscopy method). A physical–chemical characterisation of waste samples progressively taken during the ripening phase was carried out, as well. In addition, the ripened BSW quality was assessed by comparing the characteristics of a compost sampled at the composting plant of Rome which treat source segregated organic wastes. Results showed that the additional ripening process allowed to obtain a better quality of the biostabilised waste, by achieving a much higher biological stability compared to BSW as-received and similar to that of the tested compost. An important finding was the lower heavy metals (Co, Cr, Cu, Ni, Pb and Zn) release in water phase at the end of the ripening compared to the as-received BSW, showing that metals were mainly bound to solid organic matter. As a result, the ripened waste, though not usable in agriculture as found for the compost sample, proved anyhow to be potentially suitable for land reclamation purposes, such as in landfills as cover material or mixed with degraded and contaminated soil for organic matter and nutrients supply and for metals recovery, respectively. In conclusion the study highlights the need to extend and optimise the biological treatment in the MBT facilities and opens the possibility to recover the output waste instead of landfilling. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction In the last two decades, hundreds of mechanical–biological treatment (MBT) facilities were built in Europe to treat the residual municipal solid waste (MSW) from at-source separate collection (Lornage et al., 2007; Barrena et al., 2009; Bayard et al., 2010; Ponsá et al., 2010; Tintner et al., 2010; Di Lonardo et al., 2012). Main goal of MBT is to reduce the environmental impacts and risks to human health related to landfilling, as set by the European Landfill Directive 1999/31/EC (European Commission, 1999). One of the output of the MBT is the biostabilised waste (BSW) produced from the aerobic biological treatment of the organic fraction mechanically separated from the input MSW. In the majority of the MBT facilities, the biological process consists in an active
⇑ Corresponding author. Tel.: +39 06 7259 7497; fax: +39 06 7259 7021. E-mail address: [email protected] (M.C. Di Lonardo). http://dx.doi.org/10.1016/j.wasman.2015.05.028 0956-053X/Ó 2015 Elsevier Ltd. All rights reserved.
biodegradation phase at forced aeration conditions for 20–28 days (Adani et al., 2004a; Di Lonardo et al., 2012; Puyuelo et al. 2010). Differently, in the common practice of the composting processes, the active decomposition phase at forced aeration conditions is followed by a windrow or static pile system at natural aeration conditions for the later stages of decomposition and ripening in order to raise the output compost quality (Richard, 1992). Such a difference in the biological treatment management depends upon the fact that MBT was firstly developed and seen as a MSW pre-treatment for disposal in order to reduce the biodegradable municipal waste going to landfills (Di Lonardo et al., 2012), therefore no high quality of the output was expected. Recently the interest in the possibility to recover the biostabilised waste from MBT is increasing (MacLeod et al., 2008), especially considering the large amounts that are now being produced in efforts to divert waste from landfill (Farrell and Jones, 2010). However, MacLeod et al. (2008) argued that the MBT output must be highly stable so that the users can be confident that
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
contaminants in the material such as heavy metals have reached their maximum concentration. Many studies showed that an efficient reduction of the biological reactivity of mechanically–biologically treated MSW is achieved by means of optimised aerobic biological process lasting at least 8 weeks, including both an active biodegradation phase and a ripening phase (Zach et al., 2000; Francou et al., 2005; Lornage et al., 2007; MacLeod et al., 2008; Barrena et al., 2009; Shao et al., 2009; Bayard et al., 2010; Tintner et al., 2010). In the city of Rome the separate collection of biodegradable municipal waste (BMW) and the related treatment facilities are poorly developed. Data reported in ‘‘Municipal Solid Waste Report 2014’’ (ISPRA, 2014) show that in 2013 a low percentage of BMW, i.e. only 23.9% of the total MSW produced in Rome (corresponding to 1,754,823 Mg), was source segregated. As a result, the residual MSW fed to the MBT plants (about 60% of the total amount of MSW produced in this city) were composed of significant percentages of biodegradable waste, namely 40–50% by weight, as observed by several investigations on material composition performed in the years 2006–2013 (Franzese et al., 2013). In these MBT facilities, about 80% of the biodegradable waste amount, which the input MSW is composed of, is separated through a primary mechanical sieving and then subjected to an aerobic biological treatment at forced aeration conditions for 4 weeks (active biodegradation treatment). Afterwards, the end biostabilised waste is not subjected to a further ripening phase but it is directly landfilled. In previous surveys performed on the biological stability degree, by measuring the dynamic respiration index (DRI) (Di Lonardo et al., 2014), several samples of biostabilised waste coming from the MBT plants of Rome were found highly reactive. Specifically, DRIs were higher than the maximum limit value of 1000 mg O2/kg VS h below which the waste is considered biologically stable (Adani et al., 2004a) and compliant with the requirement set by the Italian landfill regulation for the waste acceptance in landfill (Italian Ministerial Decree, 2010). The high reactivity was mainly related to the relatively short processing period as well as to the significant content of biodegradable waste in residual MSW. Therefore, in this study, the biostabilised waste coming from one of the MBT plants of Rome was subjected to an additional ripening process in slightly aerated lab test cells in order to analyse and evaluate the variation and the increase of the biological stability degree with time of ripening. A physical–chemical characterisation of waste samples progressively taken during the ripening phase was carried out. Furthermore, the ripened BSW quality was also assessed through a comparison with the characteristics of a compost sampled at the composting plant of Rome which treat source segregated organic wastes. The purpose was to evaluate the influence of the additional ripening process on the quality of BSW in the perspective of recovering the end material, rather than landfilling.
2. Materials and methods 2.1. Biological process in the MBT plants of Rome In the city of Rome four MBT plants are currently in operation with a maximum treatment capacity of 3000 Mg MSW/d. The facilities differ in the type of mechanical pre-treatment employed but the biological process is performed in the same conditions and duration, as hereafter described. BSW used for this investigation was sampled in one of the four MBT facilities whose Fig. 1 shows the flow scheme and the outputs of the treatment. Focusing on the biological process, after metal removal by belt-type
51
electromagnetic separators, the biodegradable fraction (i.e. the undersize flow coming from the primary screening unit at 80 mm) is sent to a biostabilisation basin where an aerobic biodegradation occurs for 4 weeks at forced aeration conditions. Three augers moved by a crane have a dual function: turning over the material in order to keep proper free air space (pores) for satisfying aeration (avoiding the formation of anaerobic conditions, especially at the bottom of the basin) and moving the material along the basin. During turning/moving, water is added to the material by nozzles mounted on the crane, in order to keep the water content favourable for microbial activity. The biostabilised output then is sieved at 20 mm by means of a trommel screen in order to separate an oversize fraction mainly composed of plastics and inert materials from the undersize fraction, namely the final biostabilised waste (BSW) which is daily landfilled. BSW was taken at the outlet of the biodegradation process lasting 4 weeks after the sieving unit at 20 mm, following the procedure laid down in the Italian standard UNI 10802 (2013), and after mixing and quartering, an amount of roughly 30 kg was collected. Furthermore a mature compost sample of roughly 20 kg was collected at the composting plant of Rome which treat the organic fractions of municipal solid waste coming from the at-source separate collection. In this case, beside the 4 weeks of aerobic biodegradation at forced aeration conditions, the treatment includes a ripening phase of up to 6 months, differently from the biological process carried out in the MBT plants. The BSW sample was sent to the laboratory at the Institute of Waste Management of the University of Natural Resources and Life Sciences in Vienna (ABF-BOKU) where it was subjected to a lab scale ripening process and analysed in order to measure the biological reactivity. A further physical–chemical characterisation was instead carried out at the Environmental Laboratory of the University of Rome ‘‘Tor Vergata’’. Differently, all tests performed to characterise the compost sample were conducted at the latter laboratory. 2.2. Lab ripening phase The sampled amount of BSW was subjected to a ripening phase by placing it in a vertical cells system simulating an open windrow configuration (Fig. 2). A relatively low air flow rate was fed to the bottom of the system in order to simulate natural aeration conditions. Each single cell was half-filled with the rotting waste placed on a grid (with mesh opening of 10 mm) to let the air passing through the material homogeneously. A biofilter composed of mature compost was placed at the top of the system to reduce odour emissions. Ambient and rotting material temperature, as well as oxygen and carbon dioxide concentrations in the waste air were daily measured for each single cell in order to monitor the process. Mixing, adjusting in water content and turning of the material were carried out once per week to keep optimised conditions for the ripening process. During mixing of material, samples were taken in order to analyse the biological reactivity and evaluate the changing in reactivity with time. From the measures of temperature, O2 and CO2 concentrations, as well as of the VS (as described in Section 2.4) during the ripening, the actual O2 consumed rates (mg O2/kg VS h) and the actual CO2 released rates (mg CO2/kg VS h) per unit of organic matter (VS) were calculated through the following Eqs. (1) and (2) (Komilis and Kanellos, 2012):
O2
cons
¼
ðO2in O2out Þ Q MWO2 1000 MV VS
ð1Þ
52
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
Fig. 1. Flow scheme of the MBT plant of Rome.
CO2
rel
¼
ðCO2in CO2out Þ Q MWCO2 1000 MV VS
ð2Þ
where (O2in O2out) is the concentration difference between the O2 at the inlet (air concentration equal to 20.8% v/v, taking into account that fed air was not dry) and the measured O2 at the outlet air (% v/v); (CO2out CO2in) is the concentration difference between the CO2 at the inlet (air concentration equal to 0.03% v/v, taking into account that fed air was not dry) and the measured CO2 at the outlet air (% v/v); Q is the air flow rate (l/h) set during the ripening phase; MWO2 is molecular weight of oxygen (31.98 g/mol); MWCO2 is molecular weight of carbon dioxide (12 g/mol); 1000 is the conversion factor (mg/g); MV is the molar volume (l/mol), i.e. MV ¼ MVst TTstt , where MVst = 22.4 l/mol is the molar volume at standard temperature Tst = 273.15 K and standard pressure pst = 1 atm (by assuming that O2 and CO2 are perfect gases) and Tt is the measured temperature at time t during the ripening phase; VS are the volatile solids weekly measured (kg); the VS were assumed constant between two samplings. 2.3. Biological reactivity tests The influence of the ripening process on the biological reactivity of BSW sample was evaluated by performing different types of tests, namely aerobic and anaerobic biological tests, as well as FT-IR spectroscopy method. The aerobic or respiration test measures the cumulative oxygen uptake by microorganisms during a period of 4 days, indicated as RA4 and expressed in mg O2/g DM, according to the Austrian Standard OE NORM S 2027-4 (2012a). The aerobic test was performed by using two static respirometric equipment, namely SapromatÒ and OxiTopÒ (Binner et al., 2012). The anaerobic or incubation test measures the microbial gas generation under
optimised anaerobic conditions during a period of 21 days (Binner and Zach, 1999), indicated as GS21 and expressed in Nl/kg DM (Nl = litres under normalised conditions). The incubation test was conducted according to the specifications of the Austrian Standard OE NORM S 2027-2 (2012b). Furthermore, Fourier Transform Infrared (FTIR) spectroscopy of BSW samples was determined since this method enables to assess the reactivity of MBT waste directly via the chemical composition reflected by the FTIR spectrum (Böhm et al., 2010; Pognani et al., 2010). Infrared spectroscopic investigation was carried out using the attenuated total reflection (ATR) technique (Böhm et al., 2010) by means of Bruker Optics ALPHA FTIR Spectrometer. Prior to the FTIR analysis, samples were air-dried, grinded by means of agate mill and screened through a 0.63 mm sieve in order to obtain pulverised and homogeneous samples. Respiration tests and FTIR spectroscopy were performed for BSW as-received (R0), and for BSW after 2 (R2), 3 (R3), 4 (R4) and 5 (R5) weeks of additional ripening phase. Incubation test, instead, was performed for samples R0, R3 and R5. Respiration and incubation tests were conducted in duplicates. Differently, FTIR-spectroscopy was determined for 5 replicates since the sample amount to analyse was quite low (roughly 10 mg). Then the 5 spectra obtained were averaged for data analysis. As for compost, the biological stability was analysed by determining the dynamic respiration index (DRI), namely the absolute maximum rate of oxygen consumption due to microbial activity (Adani et al., 2004b). DRI was measured and calculated according to the procedure reported in the Italian Standard UNI/TS 11184 (2006) by using a 30 l adiabatic respirometric reactor (Costech International Respirometer 3024). 2.4. Physical–chemical characterisation The samples R0 to R5, as well as the compost sample, were air dried prior to be further analysed to organic matter content, heavy metal total content and release in water phase (leaching test). The organic matter content was determined by measuring the volatile solids (VS) and the total organic carbon (TOC). VS content
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
was determined in duplicates by loss-on-ignition at 550 °C for 8 h (UNI/TS 11184, 2006) on 10 g of sample grinded to size lower than 0.63 mm and pre-dried at 105 °C for 4 h. TOC content was analysed in triplicates by means of Shimadzu SSM-5000A instrument on approximately 0.4 g of sample grinded to size lower than 0.2 mm (EN 13137, 2001). In order to evaluate the loss of organic carbon converted to CO2, the cumulative CO2 and C–CO2 mass was calculated by means of the following equations:
mCO2 ;t
V CO2 ;t Q Dt 24 MWCO2 ¼ MV
mCCO2 ;t
V CO2 ;t Q Dt 24 AWC ¼ MV
ð3Þ
ð4Þ
mCO2 ;cum ¼ RmCO2 ;t
ð5Þ
mCCO2 ;cum ¼ RmC;t
ð6Þ
fibre filters) and (3) final filtration at 0.45 lm (as required by the standard method) by means of syringe filters (Sartorius cellulose acetate filters). The dissolved organic carbon (DOC) (Shimadzu TOC-V CPH/CPN analyser) and heavy metal (Varian ICP-AES) concentrations were measured in the filtered eluates. Heavy metals (HM) and total organic carbon (TOC) percentage release was then calculated by means of the following Eq. (6):
%release ¼ 100
C leached 10 C sol
ð8Þ
where Cleach is the concentration (mg/l) measured in the eluates of the two samples, 10 is the liquid to solid ratio (l/kg DM) and Csol is the total content in the solid materials (mg/kg DM). 3. Results and discussion 3.1. Ripening phase monitoring
where mCO2 ;t and mCCO2 ;t are the masses in grams of carbon dioxide and carbon at time t, respectively; V CO2 ;t is the concentration (% v/v) measured at time t during the lab ripening phase; Q is the air flow rate (l/h) set during the ripening phase; Dt is the time interval between two CO2 measures (d) and 24 is the conversion factor days–hours (h/d); MWCO2 is molecular weight of carbon dioxide (g/mol); AWCCO2 is the atomic weight of carbon (g/mol); MV is the molar volume (l/mol); mCO2 ;cum and mC,cum are the cumulative mass in grams of carbon dioxide, equal to 2417.9 g and the cumulative mass of carbon, equal to 659.4 g. Then, the percentage by weight of carbon loss was calculated by means of Eq. (7), as follows:
% C CO2 ¼ 100
53
mCCO2 ;cum TOCini
ð7Þ
where TOCini is the initial TOC content (measured in the sample R0), expressed in grams. Heavy metal content in solid materials (total content) was determined in triplicates by acid digestion according to the method C indicated in the European standard EN 15411 (2011) with some modifications. Specifically, 3 ml of concentrated HNO3 15.698 N (Sigma Aldrich 30709) and 1 ml of concentrated HCl 12.178 N (Sigma Aldrich 30721) were added to 0.1 g of sample grinded to size lower than 0.5 mm. Then the mixture was put in closed vessels (Parr Instrument Company – model 4744) and kept at 150 °C for approximately 15 h. The obtained solution, after cooling, was firstly filtered by means of Whatman No 41 filter paper so to separate residues and then it was diluted to volume with ultrapure water in 25 ml flask. A final filtration at 0.45 lm (Sartorius cellulose acetate syringe filters) was carried out in order to analyse the dissolved heavy metals in the solution by inductively coupled plasma atomic emission spectrometry (Varian ICP-AES). For the leaching test, air-dried samples were grinded to a particle size lower than 4 mm, as required by the European standard EN 12457-2 (2002). Deionised water was added to each sample at a liquid to solid ratio equal to 10 ml/g and bottles containing the mixture were slowly shaken by means of a rotary tumble for 24 h. Such test was conducted in duplicates for each sample. The obtained eluate, after decanting for 15 min, was firstly analysed by measuring the pH (Eustech Instrument pH 700). Afterwards, three steps of solid–liquid separation were carried out: (1) centrifugation at 12,000 rpm for 15 min (Thermo Scientific SL 16R Centrifuge), (2) vacuum filtration at 0.7 lm (Munktell AB glass
Fig. 3 shows the setting of the air flow rate and the trends of ambient and rotting material temperature, as well as the CO2 and O2 concentrations measured in the exhaust air during the lab ripening phase. The highest temperatures, 10–15 °C higher than the ambient temperature, were measured in the first week of ripening phase with a maximum value equal to 41.5 °C observed in the 3rd day. During the 2nd week and half of the 3rd week, temperature decreased and kept around 35 °C, being 5 °C higher than the ambient temperature. During the last 10 days, a further decrease of temperature by approaching to ambient temperature was observed and this was a first indication on the biological stabilisation of the material. The maximum oxygen consumption and corresponding carbon dioxide release were observed during the first 2 weeks of the ripening, proving the high reactivity of the BSW as-received. It has to be noticed that between the 3rd and the 5th day, O2 concentration in the exhaust air significantly increased. This was due to the increase of the air flow rate from 20 l/h (set in the first day) to 60 l/h in order to keep O2 concentration in the exhaust air higher
Fig. 2. Lab ripening phase: vertical cells system scheme.
54
M.C. Di Lonardo et al. / Waste Management 43 (2015) 50–60
than 10% (v/v) below which microbial activity may be strongly hampered or inhibited because of oxygen starvation. From day 12 on, O2 concentration kept at 10–11% (v/v) and CO2 at 8–9% (v/v), increasing to 14% (v/v) and decreasing to 6% (v/v), respectively, during the last three days. This indicated again the achievement of biologically stable conditions, therefore the test was stopped after 5 weeks.
3.2. Biological reactivity Fig. 4 shows the actual O2 consumed and CO2 released rates calculated by means of Eqs. (1) and (2). The same trends of O2 consumed and CO2 released rates during the ripening phase were observed with the maximum values found on the 3rd day, when the maximum temperature of the rotting waste was measured as well, indicating the high reactivity of the as-received BSW. Until the 18th day the rates tended to decrease with an oscillating trend, proving that the waste was still reactive, whereas steady rates were observed during the last ten days because of the achievement of biological stabilisation state. In Fig. 4 the maximum limit values identifying a medium biological stability, i.e. 1000 mg O2/kg VS h, and a high biological stability, i.e. 500 mg O2/kg VS h (Adani et al., 2004b) are highlighted. It can be observed that at the beginning of the ripening the actual O2 consumed rates were higher than both stability limits confirming what was found in a previous study (Di Lonardo et al., 2014), namely that BSW coming from the active biodegradation of 4 weeks in the MBT plants of Rome could not be considered biologically stable. Differently, a high biological stability degree was obtained after 15 days of lab ripening phase since the actual O2 consumed rates were lower than
Fig. 3. (a) Setting of the air flow rate and (b) trends of rotting material and ambient temperatures, as well as CO2 and O2 concentrations in waste air during the ripening phase.
500 mg O2/kg VS h. Compost sample showed to be biologically stable with DRI found to be equal to 620 mg O2/kg VS h, therefore comparable to that of BSW after 2 weeks of ripening process. These observations were verified through the measure of the respiration activity (RA4) by means of Sapromat and OxiTop equipment (as described in Section 2.3). Fig. 5 shows the results of the respiration tests for samples R0 to R5. It can be noticed that BSW as-received (R0) was characterised by a significant respiration activity with RA4 values equal to 21.1 mg O2/g DM from Sapromat and equal to 17.0 mg O2/g DM from OxiTop, exceeding the limit value of 7 mg O2/g DM set by the Austrian Landfill Ordinance (2008). During the ripening phase a reduction in respiration activity was observed by reaching RA4 lower than the limit after 4 weeks of treatment. Furthermore, it can be observed that the average cumulative trend of oxygen consumption was quite steep for sample R0 and then progressively less slope, reaching a certain stability shown by very similar trends found for R4 and R5. Comparing RA4 obtained by the two equipment, comparable values were observed for samples R2–R5, whereas for R0, RA4 measured by OxiTop was lower than that measured by Sapromat, i.e. 17.0 and 21.1 mg O2/g DM, respectively, probably because of the dynamic oxygen replacement provided by Sapromat equipment (Binner et al., 2012). The biological reactivity for samples R0, R3 and R5 was also measured through the incubation test measuring the biogas generation (results are shown in Fig. 6). The average gas generation sum within 21 days of test duration was found to be high for R0 exceeding the limit of 20 Nl/kg DM set by the Austrian Landfill Ordinance (2008). Values lower than the limit were obtained at the end of the ripening process (5 weeks), when a significant reduction occurred. It has to be observed that GS21 for repetition 1 of R0 was found to be quite low. This was due to the high reactivity of BSW which caused the formation of acidification conditions, verified by the relatively low pH (
