Efficiency of a system for the separate collection of the biowaste from municipal 1 solid waste. A Spanish pilot case study 2

According to EU regulations, member states shall take measures to encourage the recycling of biowaste in a way that fulfils a high level of environmental protection. In Spain, the separate collection of biowaste is only implemented in some regions. For this reason, a pilot scheme based on an information campaign and the location of a specific 35 brown container for biowaste in specific zones of the city was carried out in Castellón de 36 (Spain) over a period of six months. In this period, the collection and composition of the biowaste was monitored in depth with the goal of determining the evolution of the 38 efficiency of the new collection system over time.


54
The European Union (EU) must take taction to guarantee the hierarchy of waste to obtain 55 recycled products of high quality. Therefore, the EU Member States will have to achieve 56 the separation of bio-waste, which will contribute to the protection of the environment  of the total, followed by inert, ash and debris at 21.06%, paper at 8.77%, plastic at 8.18%, 79 glass and ceramics at 4.45% and metals at 2.71% (Majid and Hwee, 2007;Dangi et al., 80 2011; Pirani and Arafat, 2014; Tatano et al., 2017). 81 Experiences carried out in Latvia indicated that 32.9% of recyclable waste is not properly 82 separated, and 29.2% is bio-waste. This implies that almost 60% of the potentially 83 recyclable waste in this country could be used if separated at the source, promoting a 84 circular economy (Kubule et al., 2019). 85 In some German cities, separate collection of household biowaste affects the quality and 86 final composition of the recovered materials. In fact, when the biowaste was separated, 87 the rest of the waste was reduced by up to 30%. Furthermore, the residual waste had less 88 humidity, which improves the efficiency of the incineration plants (Schuch et al., 2017). 89 In the food industry, the efficient separation of bio-waste is even more important due to 90 the large quantities of waste that is produced. Separation at source in the industries and 91 factories is key to improve the quality of the resources obtained from bio-waste. 92 Therefore, on the one hand, education and awareness of society and companies is 93 essential, and on the other hand, investment in support facilities (Wang et al., 2020). 94 Sometimes, when separation is done properly, investments in source separation are very 95 profitable. As an example, in Southampton (UK), separate collection of bio-waste could 96 save the council £690,000 each year, despite having to incur a significant cost on vehicle 97 adaptation and construction of the transfer stations (Bees and Williams, 2017 In Portugal, a comparative study was carried out on the costs of separate collection. It 104 was concluded that the cases of separate collection of bio-waste did not imply an overall 105 increase in costs in the service, they could even decrease them if more than 40% of the 106 population (threshold for the case study) participate in the system (Gomes et al., 2008).

107
On the other hand, the collection system, the levels of separation at source, the urban  The study was divided into five stages: (i) objectives, indicators and scope of the study, 130 (ii) identification of the study area, (iii) definition and dissemination of an information 131 and awareness campaign (information campaign), (iv) experimental design, and (v) data 132 analysis. The aim of the pilot study is to determine the degree of the efficiency of the biowaste 136 separate collection system over time and how the type of separate collection affects it.

137
For this purpose, the following specific objectives have been proposed:

138
• Determine the degree of efficiency of the biowaste collection system.

139
• Determine the degree of the variation in efficiency over time.  In order to determine the degree of efficiency of a collection system, first of all, it is 145 necessary to define a number of indicators. In this work, efficiency is defined in terms of 146 the extent to which clean materials are recovered at source, that is to say, materials  between containers is 100-120 m. Citizens deposit the glass fraction, paper/cardboard 176 fraction and light-packaging fraction in drop-off areas.

177
The frequency of collection of mixed waste (mx) is six days a week. The collection of 178 mixed waste in 2016 was 56,875 tons, which represents a DCRmx of 0.91 kg / inh·day.

179
To carry out the pilot study, it was decided to define three study areas, one for each 180 kerbside system. Table 1 shows the characteristics of each area. The zones have different 181 numbers of inhabitants and containers.
182 Zone 1 is located in the city centre, which corresponds to the old city. It is an area where 185 a low-density residential area is combined with commercial and restaurant areas. Zone 2 186 is located in the northern district of the city. It corresponds to a wide area with a high-187 density residential area with several green areas and little commerce. Zone 3 is located in 188 the west of the city and has similar characteristics to Zone 2. Finally, Zone 4 was also 189 defined, close to Zone 2, where the selective collection of biowaste was not implemented.

190
Samples were taken in this area to determine the composition of the mixed waste 191 container of the current MSW collection system in Castelló de la Plana.  Table 1 shows the number of biowaste containers 218 located in each area.

219
Containers were collected three times a week, since a large volume of biowaste was not 220 expected. Each zone had an independent collection route. Once the collection was 221 complete, the collection truck was weighed at the treatment plant, so daily collection data 222 were available for the three areas. In Zone 1, due to different technical reasons, collection 223 began 6 weeks later than in the rest.  where: 242 n is the number of samples required 243 t α;n-1 is the deviation from the accepted mean value to achieve the desired confidence level ( -244 1), for the "t" distribution. 245 VC (variation coefficient) is the variance that we hope to find in the population (expressed as a 246 decimal) 247 ɛ is the maximum margin of error (expressed as a decimal) 248 To calculate VC, it is necessary to know previously the data about the mean and the standard 249  only the biowaste fraction has been considered, since it is the one of interest for the study. 266 Finally, a margin of error of 10% has been assumed.
267 Table 2 shows the data needed to perform the calculation and the results obtained after  According to Table 2 and Equation 1, it will be necessary to take at least seven samples

294
The average composition for each of the three zones is shown in Table 3. It can be seen 295 that in the three areas, the percentage of biowaste is high, between 79.75% and 82.74%, 296 although there is still around 20% of inappropriate, which will end up being a reject in  The fraction of "fines" is the most abundant in the inappropriate part. This is a fraction 308 with materials less than 10 mm in size, such as dirt, dust, stones, microplastics, metals,

324
Citizens identify dirty food packaging with the fraction of biowaste and, therefore, for 325 future research, citizens should be informed that these materials must be cleaned and 326 placed in the appropriate container. Dirty paper, which is also identified as biowaste, was 327 also found in this fraction but this is not a problem since it is biodegradable. Finally, very 328 little glass was found, except in Zone 1. As Zone 1 is a restaurant and commercial area, 329 wine bottles appeared in some characterizations.

330
If the composition of the biowaste container is compared with that of mixed waste (Table   331 4), a clear difference is observed in all the fractions, mainly in the biowaste with a notable 332 increase, which is why the separate collection system has been successful. Something   Table 3 shows that there are no significant differences in the biowaste percentages among the 349 three zones, and therefore in their QCRbw. To determine whether it was true, it was necessary to 350 demonstrate it statistically. For this reason, an Analysis of Variance was carried out. In this case, 351 it was verified that normality can't be assumed for the biowaste fraction. Consequently, the 352 Krustal Wallis test was used with a confidence level of 95% (α = 0.05).

353
After comparing the means, a p-value of 0.814 (p-value > 0.05) was obtained, so it could 354 be stated with 95% confidence that there are no statistically significant differences in the

359
The variation in the QCRbw of the biowaste container over time can be seen in Fig. 2. In 360 all zones the QCRbw increased as the pilot study progressed. Zone 1 underwent the 361 greatest increase, going from 70% at the beginning of the experiment to 90% at the end.

362
In Zones 2 and 3, the progress was smoother because the initial QCRbw data were higher. In the three zones, QCRbw values of around 90% are reached at the end of the experiment.

367
This increase may be mainly due to citizens' learning and familiarization with this new 368 separate waste collection, which implies an increase in their collaboration. Throughout 369 the experiment there was no reinforcement of information, so it follows that citizens 370 themselves voluntarily improved the quality of the waste in the container. Therefore, it is 371 assumed that this is the value that can be reached after its definitive implantation in the 372 town and that, with a continuous awareness campaign, it could be maintained over time.  Table 4 shows the average composition of the mixed waste container for zones 1, 2, 3 and 376 4. The characterization was carried out in the last month of the experiment, when the 377 QCRbw of the selective collection was found to be higher.

378
According to the data in Table 4, despite the fact that selective collection of light-379 packaging, paper/cardboard and glass is carried out in the city, significant percentages of 380 these materials appear in the container in the four zones. It is also important to highlight 381 that there are apparently no differences between the four zones. confidence level of 95% (α = 0.05) was used. We decided to employ this non-parametric 388 test because the number of mixed waste composition data for each zone is small (three 389 data).

390
From the results, it can be stated (with 95% confidence) that there are no significant 391 differences in the percentages of biowaste in the mixed waste container between the four 392 zones, since the p-value obtained is 0.7793 (p-value > 0.05). This is due to the fact that 393 only a small part of the biowaste is diverted from the mixed waste container to separate 394 biowaste collection, as indicated by the SRbw values shown in Table 5. Therefore, the  to collect it door-to-door. This system proved to be the best solution for the high public From an analysis of the previous data, it can also be concluded that the collection system 446 affects its performance, since each system defines the separation between containers 447 differently and, therefore, its CRbw. Therefore, in the pilot study that was carried out, it 448 22 has been possible to show that, under the same characteristics of the population, the 449 collection system with a lower CRbw has a higher SRbw.

450
To analyse the variation in SRbw throughout the experiment, the SRbw values of the three 451 zones have been represented on a graph (Fig. 3).   Regarding QCRbw the study allowed the following issues to be determined:

478
• In the three areas, the QCRbw obtained is the same from the statistical point of 479 view. Therefore, it has been shown that the citizens in the three areas separated 480 their waste at source in a similar way.

481
• The percentage of inappropriate material is 20%, consisting mainly of recyclable 482 materials. To increase the QCRbw, more information should be provided to 483 citizens on how to manage food-containing packaging. 484 • It has been verified that in the three zones the QCRbw increased as the pilot study 485 progressed, finally reaching 90%. This increase may be due to citizen learning 486 and familiarization with this new selective waste collection. Therefore, it is hoped 487 24 that this is the value that can be reached after it is implemented definitively in the 488 city.

489
• Regarding the mixed waste container, the four zones present QCRbw values that 490 are not statistically different. This fact shows that there has been no significant 491 transfer of biowaste to the brown container.

492
On the other hand, regarding the amounts of waste collected separately in the brown 493 container, this study allowed the following issues to be determined:

494
• Zone 1, with more commercial activity and older citizens, presents the worst 495 results in DCRgbw and DCRnbw, despite having the best CRbw.

496
• On analysing the DCRgbw and DCRnbw from Zones 2 and 3 it has been shown that 497 their value increases when CRbw decreases and, therefore, to have better results it 498 will be necessary to decrease the CRbw, that is, to reduce the distance from the 499 citizen to the container.

500
• The SRbw increases in the three zones over time, although in different ways, which 501 means that citizens' collaboration has increased. 502 • The collection system affects the SRbw, since each system defines the separation 503 between containers differently and, therefore, its CRbw will be different.

504
The increase in SRbw and QCRbw over time makes the last month of the experiment the 505 most efficient. Thus, it can be said that there is a positive evolution of the experiment, 506 which encourages implementation of the system throughout the city.

507
Finally, the proposed methodology and its results in the pilot study can be useful at the 508 international level when implementing a separate biowaste collection system in a city 509 with similar characteristics.