3.1. Composting performance and compost quality
The performance of the community composting system is represented in Fig. 2, together with the main parameters monitored throughout the process (temperature and interstitial oxygen). The system was fed 3 to 4 days a week with 65 kg of material on average, accounting both for the KW and the bulking agent. After 3 weeks of feeding module 1 was full, with an accumulated mass of 450 kg. The material was then transferred to module 2 (marked as a vertical dotted line in Fig. 2) to promote mixing and aeration, even though only one round of compost was produced due to the school’s calendar. To maintain system operation as in a real scenario, the material was transferred from module 2 to module 3 around day 75. During the composting process there was a 40% volume loss resulting from degradation and water loss, which has not been represented in Fig. 2. The moisture content of the material, although not shown in Fig. 2, was maintained within the recommended range of 40–60% to promote microbial degradation and avoid leachates.
Temperature was monitored daily during the process to ensure the material went through all the composting stages. The temperature first increased from ambient values to 45ºC (mesophilic), reaching maximum degradation and material sanitisation at up to 70ºC (thermophilic), and then descended back to ambient values below 45ºC for the final stabilization of the material (maturation) [26]. The thermophilic phase lasted at least 14 days as required and, after the peak, the temperature descended back to ambient values, two conditions that ensure the sanitisation and stability of the resulting compost [26]. Interstitial oxygen displayed an inverse relationship with temperature, as shown in Fig. 2: at high microbial activity rates temperature increases and oxygen availability decreases, and vice versa. The O2 concentration was maintained over the 10% limit value throughout the process in order to ensure aerobic conditions for microbial activity and avoid undesirable consequences of anaerobic degradation (material rotting, unpleasant odours, etc.) [26]. This was achieved by the addition of pruning waste as bulking agent and the rutinary mixing of the material. The transfers between modules favoured material mixture and aeration, which promoted further degradation and resulted in a slight increase in temperature and decrease in interstitial oxygen.
Several physical-chemical properties were analyzed to characterize the initial feedstock mixture and the final compost (Table 1). The dry matter, organic matter, pH and conductivity values were similar to those found in biowaste composting at domestic and community scale [27]. Respiration indexes are key indicators to ensure the efficiency of the composting process and the stability of the resulting compost [13]. The DRI of the initial mixture was below typical values reported for source-selected OFMSW but similar to OFMSW mixed with green wastes (3–4 and 1–2 g O2 kg− 1 OM h− 1, respectively) [28] because the KW was mixed in a 1:1–1:1.25 volumetric ratio with pruning waste as bulking agent. The DRI decrease from 0.6 to 0.3 g O2 kg− 1 OM h− 1 represents a 50% reduction in respiration activity and evidences the proper stabilization of the biowaste fed to the community composting system through the aerobic degradation of organic matter. The DRI of the final compost is in line with results found in home composting systems, which are below of those of industrial composting plants (0.27 and 1.51 g O2 kg− 1 OM h− 1, respectively) [27], indicating the better efficiency at stabilizing the organic material of the former. Recently, the cumulative respiration activity has been established as the measure to assess the stabilization efficiency of all biostabilized waste destined to be landfilled [20]. The AT4 also displayed a significant 71% decrease from 21.4 to 6.2 g O2 kg− 1 OM, confirming the previous results and the correct stabilization of the biowaste. The AT4 of the final compost is similar to typical results for home composting, which are again below those of industrial composting (8.4 and 113.5 g O2 kg− 1 OM, respectively) [27].
Table 1
Physical-chemical characteristics and static and dynamic respirometry indexes of the raw composting feedstock mixture and the final compost.
| Parameter | Kitchen biowaste + bulking agent mixture | Final compost |
| Dry matter (%) | 42 ± 2 | 53 ± 5 |
| Organic matter (%) | 88 ± 2 | 61 ± 2 |
| pH | 4.2 | 8.1 |
| Electrical conductivity (µS cm− 1) | 17373 | 3203 |
| C/N ratio | – | 11.2 |
| DRI (g O2 kg-1 OM h-1) | 0.6 ± 0.1 | 0.30 ± 0.03 |
| AT4 (g O2 kg-1 OM) | 21.4 ± 0.6 | 6.2 ± 0.3 |
The composition of the final compost was analysed to evaluate its quality. According to the current European legislation on fertilizers [29], the compost obtained in the school is considered Class C because the levels of zinc (906 mg/kg of DM) surpass the limit established for Class B (500 mg kg-1 DM). Otherwise, it would be considered Class A because the remaining heavy metals are all below the limits for the highest quality compost. The accumulation of heavy metals in the soil is a widespread concern, with zinc being one of the main ones. The environmental hazard of metals is linked to their bioavailability, and the composting process is known to increase the complexation of these elements to organic compounds, thus limiting their solubility and potential toxicity [30]. The compost also meets regulatory requirements for levels of pathogenic microorganisms (E. coli and Salmonella spp) and percentage of organic nitrogen (> 85% of total nitrogen) [29]. The fact that most nitrogen is present in stable organic form (2.57 out of 2.83% based on DM) is key to ensure that nitrogen is slowly released into the soil and adsorbed by plants and other organisms, thus avoiding the negative environmental effects of mineral nitrogen leaching [30]. Regarding microorganisms, the absence of E. coli and Salmonella, the two main human pathogenic microorganisms, confirms the correct sanitisation of the material achieved as a result of the temperature increase above 65ºC. However, Clostridium perfringens, another microorganism related to human toxicity, was found in the compost. Spores of C. perfringens are commonly detected after composting, as its spores can survive temperatures up to 70ºC [31]. Although the presence of C. perfringens it is a relevant aspect to consider and is not permitted by other regulations like the hygiene requirements for animal by-products processing [32], it is not the case for fertilizers and, therefore, it would not hinder the use of the compost as such. Last, the compost contains other nutrients that can be released into the soil and become available for the organisms in the short- and long-term, including calcium (3.35%), potassium (2.09%), iron (0.494%), phosphorous (0.475%) and magnesium (0.427%). The full characterisation of the final compost can be found in Table S1 of the supplementary materials.
3.2. GHG emissions
Greenhouse gases emissions were monitored during the complete composting process in a weekly basis. Specifically, CH4 and N2O were the GHG considered due to their non-biogenic origin and their high Global Warming Potential (GWP) compared to that of CO2, 27 and 273 times higher, respectively [33]. Figure 3a shows the evolution of the GHG specific emission rates linked to the material’s temperature evolution along the whole composting process, whereas Fig. 3b shows the GHG cumulative emissions throughout the composting process.
The average CH4 and N2O concentrations measured during the composting process were 6.3 and 3.6 ppmv, respectively, with maximum concentrations of 39.7 and 14.5 ppmv, respectively. In terms of specific emission rates, the average CH4 and N2O reported values were 0.9 and 1.3 mg d− 1 kg− 1 VS, with maximum specific emission rates of 5.7 and 5.9 mg d− 1 kg− 1 VS, respectively. These maximum emission rates were observed after the transfer of the material from module 1 to module 2 (Fig. 3a). CH4 emissions were low throughout the composting process (around 0.9 mg CH4 d− 1 kg− 1 VS), just showing an emission peak at day 32 of operation, which coincided with a mixing operation and a small increase of the material’s temperature. In general, O2 availability within the solid matrix was correctly maintained along the process as it can be observed in Fig. 2. However, this high temperature phase can reduce oxygen solubility within the solid matrix [34], promoting the formation of specific anaerobic spots. Moreover, these anaerobic spots can be facilitated due to an incomplete mixing of the composting material, as mixing operations in community composting can be complicated a complicated task, what ultimately translates into the observed CH4 emission peak episode [15]. Besides, N2O emissions were generally low from the start of the process, possibly due to the inhibition of some N2O generation mechanisms by high temperatures, until the temperature shift from thermophilic to mesophilic conditions observed after day 30 of operation. At this point, a peak of N2O was observed, coinciding with process temperatures below 40 ºC. This fact, together with a possible scarcity in carbon sources availability led to an increase on the denitrifying activity within the solid matrix, thus causing this increased emission of N2O as an intermediate of the NO3− reduction process [35, 36]. Another reported reason described for N2O emission increment is a low moisture content within the solid matrix, what increases free pore space and promotes N2O escape [37].
The emission factors determined for each pollutant measured during the entire composting process are shown in Table 2, expressed in g of pollutant per kg of VS treated. Moreover, a global GHG emission factor is presented, expressed in g of CO2eq per kg of VS treated, considering the specific CH4 and N2O 100-year GWP [33].
Table 2
Emission factors for CH4, N2O, GHG and total VOC determined for the community composting process.
| CH4 (g CH4 kg− 1 VS) | 0.10 |
| N2O (g N2O kg− 1 VS) | 0.14 |
| GHG* (g CO2eq kg− 1 VS) | 41.01 |
| Total VOC (g C-VOC kg− 1 VS) | 1.43 |
*GHG emission factor in CO2eq units considering CH4 and N2O 100-year GWP as 27 and 273 times that of CO2, respectively [33].
Home composting and full-scale composting processes and their related GHG emissions have been studied at some extent in the past, with several scientific publications reporting specific emission factors for the composting of different types of organic waste and composting systems [8, 13, 38, 39]. However, the assessment of the gaseous emissions related to community composting and its environmental impact present an important scientific gap that must be filled to promote this technological alternative, totally aligned with the current European legislation in terms of biowaste management [40], over the next years. Different works have assessed the GHG emissions and environmental impacts from decentralised composting systems using reported emission factors from other studies (generally home composting works) and Life Cycle Assessment [41, 42], which can be useful tools for their estimation but can also present significant differences with real operating community composting systems. In this work, the GHG emission factor determined for the community composting system studied was 41.01 g of CO2eq kg− 1 VS or 14.94 g of CO2eq kg− 1 of initial mixture (Table 2). When comparing with available scientific literature, the reported GHG emission factors were on the lower end of similar works mainly focused on large composting facilities [42, 43] or home composting systems. González et al. [15] carried out a full emission evaluation (including GHG, ammonia, VOC and odours) of several active community composting sites, but only calculated the emission rates of the three system modules, corresponding roughly to the thermophilic, mesophilic and maturation phases of the composting process. In order to estimate emission factors accurately, more data points must be obtained throughout the process; this makes studies including those values scarcer in the literature, as well as more valuable, especially for community composting systems. Andersen et al. [39] monitored the gaseous emissions of several home composting units treating organic household waste with different organic loads (2.6–3.5 kg per week with high-load phases of 15–25 kg for some units) and mixing habits (weekly, once every six weeks and no mixing) and found that both excessive loads and mixing increased GHG emissions, ranging from 100 to 239 g of CO2eq kg− 1 of waste. The emissions generated in this study were way lower, even more so if we consider that the organic loads were two orders of magnitude higher (250–300 kg of KW per week). The main reason behind the minimised emissions could be the proper structure and porosity of the treated material given by the bulking agent, which was added in much high amounts in the community composting system (1:1–1:1.25 volumetric ratio) compared to the domestic composting study (0.12–0.15 kg per week). Amlinger et al. [43] reviewed the GHG emissions of several biological waste treatments including composting. A multi-family composting system with higher waste organic loads (53 kg per week), although still one order of magnitude lower than the present study, also reported higher GHG emissions (76–187 g of CO2eq kg− 1 of waste). Lower GHG emissions were reported in the case of windrow composting, ranging 14–41 and 9–68 g of CO2eq kg− 1 of waste for biowaste and garden waste, respectively, which are more in line with those of the present study. Last, the Intergovernmental Panel on Climate Change [44] has given a default emission factor for several waste biological treatment, with composting having 4 g CH4 kg− 1 of waste and 0.3 g N2O kg− 1 of waste, accounting for a total of 80 (with a range of 19–379) g of CO2eq kg− 1 of waste. The GHG emissions of the present study are in the lower end of this interval, meaning that the environmental performance of the community composting system resembles that of well-operated full scale composting plants, with the added benefit that decentralised composting does not present the associated environmental impact resulting from the waste collection and transportation needed for centralised composting.
3.3. VOC emissions and characterization
The evolution of the VOC emission along the community composting process and the cumulative VOC emission are shown in Fig. 3a and 3b, respectively. VOC were found to be emitted mainly during the thermophilic phase, especially during the first month of operation, where the most easily biodegradable organic matter is consumed, and volatile compounds are much more easily formed and emitted [15]. The highest total VOC concentration observed was 255 ppmv, whereas VOC average concentration was 29.8 ppmv, corresponding to a maximum and average VOC emission rate of 101.8 and 12.1 mg C-VOC d− 1 kg− 1 VS, respectively. Overall, the total VOC mass emitted along the process was 238 g of C-VOC, what corresponds to a VOC emission factor of 1.43 g C-VOC kg− 1 VS (Table 2). These values can be equivalently expressed as 1.26 g C-VOC kg− 1 DM or 0.53 g C-VOC kg− 1 initial mixture. Similarly to GHG emissions, most of the VOC emissions studies available in the literature are carried out either in laboratory scale reactors [45, 46] or in full-scale waste treatment plants [47, 48]. There is a gap of knowledge regarding the performance of community composting and other middle-sized composting systems, although some works can be found [15]. The total VOC emitted by the community composting system installed in the primary school are slightly above the values found in a similar work [49] conducted in a 50 L pilot scale composter treating OFMSW and wood chips (0.99 g VOC kg− 1 DM). They found that using inert bulking agents such as polyethylene resulted in lower gaseous emissions compared to woody bulking agents, although the quality of the resulting compost was higher in the latter. This is due to the fact that the mixture of food and woody waste has a high biodegradability, as discussed in section 3.1., which enables the maintenance of higher thermophilic temperatures and the occurrence of anoxic zones, both favouring VOC emissions. Cadena et al. [48] developed a systematic methodology to analyse the gaseous emissions of composting plants, and also found lower VOC emissions (0.2 g VOC kg− 1 OFMSW) than the present study. As it happens with GHG, VOC emissions present a high variability among composting processes especially due to the characteristics of the organic waste treated. In addition, VOC emissions have been proven to be highly dependent on temperature and aeration control. Therefore, it is important to notice that community composting systems must be managed properly to avoid undesirable gaseous emissions to ensure not only the comfort of the people nearby the system, but also the environmental sustainability of the process.
Table 3 shows the characterization of the main VOC quantified from the two different gaseous samples corresponding to the thermophilic and the mesophilic phase of the composting process, respectively.
Table 3
Complete characterisation of VOC detected during the thermophilic and mesophilic phases of the community composting process (ppbv units, symbol - means not measured). DMS: Dimethyl sulphide, DMDS: Dimethyl disulphide.
| VOC | Thermophilic phase | Mesophilic phase |
| Benzene | 29.2 | 5.4 |
| 1-butanol | 91.3 | 29.1 |
| Methylcyclohexane | 0.0 | 0.2 |
| DMDS | 17.2 | 69.9 |
| Pyridine | 32.5 | 0.0 |
| Toluene | 9.4 | 68.6 |
| Ethyl butyrate | 0.4 | 0.0 |
| Tetrachloroethylene | 7.1 | 1.7 |
| Butanoic acid | 16.5 | 0.0 |
| Ethylbenzene | – | 24.8 |
| m-xylene | – | 76.5 |
| 1-hexanol | 90.9 | 0.0 |
| Styrene | 57.1 | 11.4 |
| α-pinene | 1757.0 | 302.8 |
| 2-butoxyethanol | 5.8 | 6.8 |
| Cyclohexanone | 7.5 | 0.0 |
| Decane | 5.6 | 30.5 |
| β-pinene | 3048.1 | 112.6 |
| Benzaldehyde | 93.1 | 1.3 |
| Limonene | 874.9 | 581.9 |
| p-cymene | 236.1 | 51.5 |
| Phenol | 6.5 | 1146.3 |
| Decanal | 15.3 | 0.0 |
| Total | 6401.6 | 2521.3 |
The characterisation of VOC in the thermophilic and mesophilic phases of the composting process showed relevant differences in the emitted compounds depending on the process temperature. A total of 23 VOC were identified in the gaseous samples; similar studies can comprise from 7 to over 60 compounds [45, 50]. It is important to note that the compounds quantified in this analysis only account for a 2.84 and 17% of the total VOC measured in the thermophilic and mesophilic phase, respectively. This phenomenon is quite common when treating highly heterogeneous wastes such as OFMSW or kitchen biowaste and should be considered in order to properly assess VOC emissions using both methods. The most common VOC in the two samples were α-pinene, β-pinene and limonene, with concentrations ranging from 302 to 3048 ppbv, all which are terpenes commonly found in composting systems treating food waste. Limonene has a distinct smell of citrus, and its emission can be attributed to the presence of fruit peels in the kitchen waste, as it was also found in a similar community composting system treating the kitchen waste from a university campus canteen [15]. The presence of α-pinene and β-pinene is commonly related to the degradation of lignocellulosic material such as the pruning waste used as a bulking agent in this study [45, 51]. In the thermophilic phase, relevant amounts of p-cymene (236.1 ppbv), another terpene produced in the degradation of woody-like materials, were also detected, whereas the most abundant VOC in the mesophilic phase was phenol (1146.3 ppbv), which is the monomer resulting from lignin degradation and the main precursor of the humification processes that occur during composting [52].
The gaseous samples were also qualitatively analysed in order to complement the previous VOC quantification with a distribution of the different VOC families emitted in each phase of the composting process (Fig. 5).
The results of the quantitative and qualitative analysis concur that the most abundant VOC emitted in the composting process are terpenes (mainly α-pinene, β-pinene and limonene, as seen previously in Table 3), with a relative abundance of 81 and 39% in the thermophilic and mesophilic phase, respectively. Terpene emissions are associated to the early stages of organic matter degradation and compost immaturity and diminish during the maturation time [53]. This matches the significantly lower amount of terpenes found in the mesophilic phase compared to the thermophilic phase in the quantitative analysis, as well as the decrease in terpene relative abundance in the qualitative analysis, which in turn results in an increase in the relative abundance of other VOC families such as phenols (phenol), aromatic hydrocarbons (benzene, toluene, m-xylene, styrene) and N compounds. The presence of aromatic hydrocarbons, together with some ketones (cyclohexanone), alkanes (decane) and alcohols (1-butanol, 1-hexanol) identified in the quantitative VOC analysis, can be attributed to the degradation of polymers from any plastic packaging present in the composting system such as polyethylene, which are commonly found in composting systems treating food waste, MSW and OFMSW [54]. The absence of N compounds identified in the quantitative analysis traces back to the low percentage of detected VOC from the total VOC measured, which highlights again the importance of using several complementary VOC characterisation methods in order to build a complete picture of VOC emissions during the composting process.