3.1. Quantities and Characteristics of the Waste Generated
The UDC has 17 cafeteria-canteen services of which 11 are located on the Elviña-A Zapateira campus, in a radius of less than 1 km, and they cover approximately 80% of the total university service. The estimation study through surveys provided the information of nine of these canteens in Elviña-A Zapateira shown in
Table 1. This data indicates that the generation of organic waste by the university canteens (BWUC) can vary between 6 and 50 kg/day on average for each one of the centers, even though peaks of 100 kg of BWUC could be produced in a single center a day. In this way, 1275 kg of organic waste were estimated to be generated every week in Elviña-A Zapateira in 2009.
The weighing campaign carried out in the PF and the SA during the development of the present study (2011) showed the following results (considering school days, that is, five days a week):
These results show a 32% higher generation in SA and a 23% higher generation in the case of PF, in relation to the estimate made two years earlier. Regarding the type of organic material that constitutes BWUC residues, food leftovers were predominant (44–91% of total BWUC for four centers) while waste from food preparation, squeezed oranges, and coffee grounds were usually present but in variable amounts depending on the center (0–37%). In this way, the processed food always remains as the majority fraction, while other types of materials show a very variable presence. The presence of inappropriate materials such as plastic and metal, mainly bottle stoppers, plastic wrap, sauce, and condiment bags or kitchenware) was measured and was very rare or sporadic. Physical contaminant materials was always far below 1% of raw material.
The chemical composition of the waste materials used for composting is presented in
Table 2, including the BWUC from SF that was not included in the composting program at that time. The content of certain nutrients was variable in BWUC and more reduced in the case of the bulking agent. The bulking material had a relatively high C/N ratio (value of 53), which is due to the presence of branches and leaves between crushed plant remains, along with more woody trunks. Composting of the organic waste generated by university canteens, with C/N ratios of 14–21 (BWUC from SA, PF and SF,
Table 2), need a carbon-rich amendment to balance the C/N ratio and, above all, a structural material that facilitates the aeration of the matrix during the process. The used bulking material provided both the C/N ratio and the structure correction. Lastly,
Table 2 also indicates the content of heavy metals in the initial materials. With the exception of Cd in organic waste from the Science Faculty (SF) canteen, the metal content in all the samples is very low, which allowed us to expect a compost of good chemical quality.
3.2. Composting in the Philology Faculty
The experimental study in the Faculty of Philology lasted for eight months. During this period, three successive composting batches were carried out, in accordance with the process described in
Section 2.
Figure 2 shows the evolution of the second and third batch load as well as the evolution of the temperature, together with the frequency of turns and irrigations during the second batch. After 63 days of loading, the composter of the second batch kept the thermophilic temperature for over an extra month, which, subsequently, dropped from day 120 to values close to those of ambient temperature. The first and third batches showed similar behavior with only minor variations (data not shown).
Table 3 summarizes composting parameters obtained from the second batch in the Philology Faculty. The loading of the second batch in the PF extended for 63 days in which a total of 35 load episodes took place. Including BWUC and bulking agent, a total of 360.5 kg were added, in ratios of 8.5:1 and 1:1 BWUC:GrW in mass and volume, respectively. An average loading rate of 7.7 kg/day was obtained. After the completion of the 63-day loading period, the material continued to be composted for another 60 days. In the overall period, the temperature was above 65 °C for four days, and higher than 55 °C for 18 days. It means that it fulfils the requirement associated with compliance and the regulations of Animal By-Products (ABPs) in the international legislation in terms of community composting. The process also meets the requirements of the regional rules for sewage sludge composting [
33]. The mean thermophilic temperature for the period between the days 11 to 92 was 51.5 ± 9.5 °C (mean ambient temperature of 14.6 ± 3.6 °C). Regarding stability, self-heating assays at day 84 of operation indicated Class II, while a very stable Class V material was obtained on day 129.
According to these results, a three-stage process was proposed as indicated in
Figure 3. The complete stabilization of the compost was verified after stage 2 (
Table 3). The process, including the periods of progressive loading and thermophilic composting, requires three to four months to produce stable compost (stages 1 and 2). In this way, stage 3 constitutes an optional stage, in the case of wanting to make vermicompost with the aim of increasing the degree of humification of the material. It may also be considered as an additional maturation stage of interest in the case of sufficient space and if the immediate use of the compost is not necessary.
3.3. Composting in the School of Architecture
3.3.1. Dynamic Composter at Low Loading Rate (Configuration DC1)
Upon receipt of the DC, provided by Plana Compost
®, several preliminary tests were performed to demonstrate the performance of all operating mechanisms, including mixing and ventilation functions. Ventilation was found to be necessary not only for oxygenation, but mainly to remove excess water vapor that otherwise condensed and generated leachate and excess moisture from composting material. In addition, it was considered necessary to shorten the width of the spiral blades to avoid the excessive mixing effect and excess effort required to move the endless spiral of the digester. The DC was then continuously operated for one month (DC1 Configuration, described in
Section 3.3.2). The engine torque was then modified and continuous operation was continued (DC2 Configuration,
Section 3.3.3).
Figure 4 presents the loading rate to the dynamic composter and the evolution of the temperatures in both the dynamic and the static maturation composters. The first experiment of continuous operation of the DC1 (dynamic composter in Configuration 1) allowed us to check different aspects of the operation of the aerobic digester at a low load (days 0-30,
Figure 4A). In the first days, the digester was loaded at a higher rate, reaching about 200 kg of material in total. Above this weight, it was found that the mixing system stopped working due to a lack of sufficient power to move and drag the material by the endless spiral. This forced the removal of material (
Figure 4A) and ensured a limit of 200 kg. The loading rate of 13.5 kg/day on average during the period had to be reduced to no more than 10 kg/day. In these conditions, the operation took place at a temperature in the range of 25–30 °C (
Figure 4B), even though there was an advanced fermentation process and a significant reduction in volatile solids (data not shown). It was generated as leachate at a rate of 1–2 L/day, showing low pH of 6.3 and high electrical conductivity and ammonia content, of 6.3, 12.4 mS/cm, and 442 mg N/L on average, respectively.
In contrast to the temperatures in the range of 25–30 °C for the DC1 (
Figure 4B). The SC (parallel test) reached thermophilic temperatures between 55 and 65 °C from the second day of the operation. This indicates that the low temperatures obtained in the DC1 were consequential for the operation of this unit and not of the characteristics of the waste fed. Likewise, it was found that, in these conditions, once the outlet material of the DC1 has been transferred to a static maturation composter, the material reached and maintained thermophilic temperatures for a period of approximately 15 days.
The low temperatures in this initial phase of the DC1 could be due, in part, to the low load allowed by the mixing system. With a maximum of 200 kg inside, occupying approximately 400 L, the DC1 had unused three-fourths of its volume, which increases the heat losses. On the other hand, for these conditions (maximum load of 200 kg), a maximum retention time of 14 days was obtained and the system could not receive more than 10 kg/day of organic waste.
3.3.2. Static Composter SC1050 in SA
The SC1050 was operated in parallel with the DC1 described in
Section 3.3.1. The planning of the experiment was detailed in
Section 2.4. The planed 1:1 volume ratio of BWUC:GrW resulted in a mass ratio of 3.1. An average loading rate of 13.5 kg/day was obtained, which is the same as for DC1 but higher than that applied in SHC in the PF. Differences in behavior regarding the SC of PF were scarce. After the completion of the 21-day loading period, the material continued to be composted at thermophilic temperature until day 56. During the period of 1 to 56, the mean temperature was 48.0 ± 11.8 °C (mean ambient temperature of 14.4 ± 3.3 °C). In the same period, the temperature was above 60 °C for at least eight days. The temperature progressively decreased staying below 25 °C after three months of the process.
3.3.3. Dynamic Composter at a Medium Loading Rate (Configuration DC2)
Correcting this initial design of the DC1 would require greater capacity and power to its mixer system. Modifications carried out affected the mixing mechanism and the torque of the engine. These changes allowed to reach about 400–500 kg inside the digester. These conditions were applied from the fortieth day of operation forward (
Figure 4A,B), reaching an average feeding rate of about 20 kg/day. In these conditions, the material inside the digester reached about 800-900 L of volume, which was a reasonable use of the total reactor volume (about 50–60% of it). The generation of leachate was completely avoided after the first week of this period.
Under the new configuration during this second period (days 42 to 107), a total of 1291 kg (2809 L) of BWUC and GrW were fed into the DC2 at a rate of 19.6 kg/day on average. The DC2 achieved a mass reduction for the material in the process of about 35.7%, which generated an outlet rate of 12.6 kg/day that was transferred to the SMC. In this condition, the DC2 reached temperatures in the range of 41–54 °C (47.4 ± 6.3 °C on average), which favored the composting process. The material taken out from the digester, in terms of stability, shows a variable Rottegrade class II-IV.
The material taken out from the two reached thermophilic temperatures of 52–59 °C in the SMC for 2–3 weeks (the batch in
Figure 4C registered 55.4 ± 3.5 °C on average for 10 days), to approach the ambient temperature after four weeks. Various mixing actions and correction of moisture content were done for at least three weeks without a noticeable increase in temperature (
Figure 4C, day 108 onwards). Stable compost of
Rottegrade class IV-V was obtained after 3–4 weeks in the SMC when the DC2 is used as the first stage.
In general, a good oxygenation of the material in composting was observed, both in the DC2 as especially in the SC and SMC (
Figure 5). The lowest oxygen values were recorded on a regular basis in the DC2, especially at the low load stage, which showed values in the range of 8% to 17% oxygen, with an average of 11.4 ± 3.3% (
n = 15) in the period of low load (days 1–21) and 14.1 ± 3.2% (
n = 42) in the period of high load. Static composters showed 19.4 ± 2.2% oxygen on average (
n = 56, SC) and 20.0 ± 0.7% (
n = 46, SMC).
3.3.4. Steady State Operation of DC2 and SMC at the School of Architecture
Subsequently, all the BWUC of the SA was fed into the DC2, which allowed checking its operation with loads of up to 40 kg/day. The mean loading rate in the autumn 2011 (October to December) was 28 kg BWUC/day, showing similar operating characteristics and results. Thus, the treatment capacity of the DC2 was in the range of 13–27 kg BWUC/m3·day (18.5 kg BWUC/m3·day on average). With these high loads, during the autumn, the discharge of the digester was made directly to a big bag in which the maturation continues for a period of 3–5 weeks, and becomes, after this period, stable compost. The use of the big-bag facilitates the draw out of compost and transportation to the area of use as well as traceability of the batches.
Earthworms were applied to some SMC batches after the end of the thermophilic phase, which required a period of several months and periodic supervision to maintain high moisture with the need for watering during the summer months.
Figure 6 summarizes the results obtained in the SA in terms of 3-stages proposed and their duration. As for the PF system based on static composters, the complete stabilization of the compost was verified after stage 2. The process requires 5–8 weeks to produce stable compost (usually class IV). Thus, the time the material must be in process has been reduced to about half the 3–4 months required by the use of static composters, which we can see by comparing
Figure 3;
Figure 6.
3.4. Physical Properties of Material Samples from Dynamic and Static Composting
The physical characteristics of different samples from the dynamic and static composting units are shown in
Table 4. In the initial degradation experiment carried out with the initial configuration of the DC1 and the SC1050, VS content decreased in both units in time because of the effect of degradation. MC was high in the DC1 without watering while it was very low in the SC1050 on day 14 and higher on day 29 due to watering. On the other hand, samples from the steady state operation showed higher VS content and lower MC in the DC2 in comparison to DC1 because of the improved operation of the dynamic composter in configuration 2 and the application of a higher loading rate. The sample from the maturing stage, SMC, had a high MC because of watering to favor the growth of earthworms.
Regarding the evaluation of physical properties of compost samples (
Table 4), the values of VS and MC should be considered as independent parameters while some of the parameters in
Table 4 are mathematically related, as indicated by Equations (1) to (5) (
Section 2.4). A first look at
Table 4 as well as the linear correlation graphs indicated that the DC1 sample from day 29 frequently appeared as an anomalous value. DC1 on day 29 was considered to be in a steady state for the conditions in Configuration 1, as 29 days was approximately twice the estimated retention time. On the contrary, DC1 on day 14 was still in an early evolutionary situation.
Thus, the correlation between variables was performed with all data and excludes the DC1
29 sample in order to better verify the differences. The results are shown in
Table 5. The number of significant correlations (
p < 0.05) was 12 without the DC1
29 sample and 11 including it. The main differences were found for the parameters BD
dry, BD
wet, ϕ, and AC. Significant correlations of DB
dry with FAS or with MC without DC1
29 sample as well as for ϕ with PD and with VS was lost when this sample was included. The opposite occurred for the correlation of ϕ with DB
dry, DB
wet, and AC.
Cases of nonlinear correlation were obtained for some parameters. In fact, BDwet and WHC increased and AC decreased in a linear manner with MC up to 65% and then continued to vary while MC asymptotically approached the limit value of 70%. Thus, other factors than MC determined the behavior of these parameters. The same behavior was observed for these variables against FAS for which the asymptote was the lower value of 28% FAS. This was due to the strict linear relation between FAS and MC in the operating conditions.
In the initial configuration DC1, the material had the largest bulk densities, BD
dry and BD
wet, and WHC, while showing low AC and FAS values. Among these parameters, only low FAS was explained by the high MC, while the others were on the asymptote for MC (i.e., the zone near 70% MC or water saturation zone in which these parameters change without following MC). Configuration DC2 improved these properties that approached those obtained for static composters (SC1050 and SMC). DC1 samples showed a greater compaction of the material, which was indicated by lower contraction capacity, and lower AC values, giving the material a mushy appearance that disappeared in configuration 2 (DC2). In fact, with the change in configuration, the AC increased from values below 35 to values above this threshold, which was indicated as convenient for composting [
34] and similar to that of static composters (
Table 4).
All the samples analyzed had high or very high ϕ values (
Table 4). Compost samples can be classified as high or having very high porosity, depending on whether ϕ is higher than 80% or 90%, respectively [
34]. ϕ values obtained for the samples SC and SMC are higher than those of the DC1 and DC2 samples. In general, the ϕ increased throughout the composting process, as the VS decreased. However, for DC1 between 14 and 29 days, ϕ decreased rather than increased. In this period, the process of biological degradation that decreased VS (from 80.4% to 77.2%) did not increase porosity, perhaps due to a faster and more intense physical phenomenon that led to increased dry density and reduced particle size [
35]. The data suggests that mechanical mixing in the DC accelerated substrate hydrolysis, reduced the particle size, and shortened the composting time. This occurred in both DC1 and DC2 configurations. If a medium or high loading rate was applied, which required configuration DC2, the heat generated counterbalanced the energy losses and the DC reached thermophilic temperatures. However, if low loading rate or low mass content in the digester occurred (configuration DC1), the temperature would remain low, in the mesophilic range.
Cv does not correlate with any of the other variables (R2 < 0.11), and there is no difference between DC and SC samples, even though the values increase with composting time in both cases. It was in the 5% to 14% range, so it may be considered low but appropriate.
The analyzed composts show high or very high values of AC, except for the DC1
29 sample. None of the samples presents risk of asphyxiation, in concordance with oxygen profiles (
Figure 5). For the DC2 conditions, AC and WHC parameters appeared equilibrated and similar to that of values for SC samples.
FAS correlated strictly with MC (R
2 = 0.997). The values of FAS obtained for the SC samples were above the optimal range (30–35%) proposed for composting [
28,
36]. This was due to the low MC caused by the high temperatures and intense evaporation that would require more intensive watering practices. Given the high oxygen concentration in SC material during the thermophilic phase as well as the high AC values, all these data suggest that a finer bulking material could be used. A finer bulking material will reduce the natural ventilation, which, in turn, would reduce water evaporation. However, there would be an increased risk of anaerobic conditions, which is a potential but absent problem under the conditions applied.
For the case of the samples DC1, the values of FAS are slightly below the optimum range with the values obtained around 28%. This fact, associated with a high MC and low AC, could lead to anaerobic conditions that slow down the composting process [
37]. However, oxygen content (
Figure 5) indicated no oxygen limitation and it indicated that the bulking agent material used was appropriate for good oxygen diffusion in the operating conditions.
3.5. Compost Quality and Overall Evaluation
Table 6 shows the characteristics of the final samples from two SA and PF compost batches. The compost has a high fertilizer power with 2.5–3.6% nitrogen content and C/N ratios between 11–15. The lowest C/N ratio and higher content in N of the PF sample is partly explained because a lower proportion of the bulking agent material was used. These C/N ratios, particularly, PF compost indicate a good compost maturity as well as a good N conservation [
38,
39]. The nutrient content was in the range of values previously reported for domestic composting programs [
40]. Phosphorus content is similar or somewhat higher than that found in other industrial composts and in home composts. The values of K and Mg in
Table 6 were lower than those reported by Vázquez and Soto [
40] for home composts while the values for N and Ca were higher.
In relation with the chemical quality, heavy metal concentrations are generally low, although Cd is close to the limit value of class A (
Figure 7). Cadmium is present in food waste, so a process with a low proportion of bulking agent material or with a very advanced degradation of organic matter can lead to values that exceed the limit of class A. In addition, there has been no
Salmonella and there were 321 cfu/g of fecal coliforms, which would be compatible with class A of the US legislation on biosolids EPA 40 CFR Part 503 [
41].
The high quality of the obtained compost indicated that the separation of organic waste at the source in these university canteens was implemented with very good results. The presence of inappropriate materials (or physical contaminant materials such as plastic and metal) was very rare or sporadic, which was always far below 1% of raw material. The obtained results also emphasize the satisfactory participation of the different agents involved, from the staff of the cafeteria services, to the gardening company. The total of the organic waste generated in these two centers, just over 5000 kg in 2011, together with approximately another 2000 kg of crushed plant remains, has been transformed into compost. The obtained compost was used as fertilizer in the university vegetable gardens cultivated by students and staff.
Once the system had been implemented, the personnel in charge of the supervision of the equipment and composting areas dedicated 1 h a week to each composting area. This included students with environmental scholarship, combined later with job placement workers provided by a charitable non-governmental organization. The composting process in static composters has been evaluated as simpler for reduced amounts of organic waste generated by university canteens, up to 15–20 kg/day. In this range of operations, in the following years, new composting areas were installed in which static composters of different volumes were used, which varied from 340 to 1400 L. At the end of 2019, nine composting areas were in operation on the UDC campuses, which treated approximately 80% of the BWUC generated [
42]. For higher generation rates, it was considered convenient to use a mechanical dynamic composter that performs the work of mixing and turning of the material during the first high rate composting phase. The dynamic composter available in the SA facilitated and accelerated the process.
As suggested by Valentukevičienė et al. [
43], these on-campus composting systems are also a valuable opportunity for student participation in research and internship activities through teamwork with projects in various study subjects. In practice, they have provided research practices for undergraduate and doctoral students [
44].