3.2. Effect of Main Parameters on SC Calculation
Data at the yearly level can be used to carry out an SC calculation analysis. We identified four levels of the “purity” of the collection (A, B, C, D), and we used five models of SC calculation (C1…C5):
use waste collection data as officially announced by the General Managers of the Province (“original data”)
distribute the multi-material collection in different components such as glass, plastic, metals and scrap based on analysis conducted by the National Consortium for packaging management
eliminate the impurities of the collection
eliminate the waste treatment scraps
The parameters related to the impurities of the collection and the waste treatment scraps can be estimated using a mass balance of selection and treatment plants. For the collection of multi-material packaging and single-material plastic, however, the impurities and scraps can be obtained from the periodic analyses and statistical processing conducted by the National Consortium for Packaging.
The calculation models taken as reference for the application of Formula (1) are the following:
- ➢
C1. Conventional: the numerator (SC) is given by the sum of all fractions separately collected, excluding the unsorted, bulky and street sweeping waste (considered in the denominator as RMSW).
- ➢
C2.
National: in the numerator of the Equation (1), the method adopts Σi SCi. Specifically, the considered streams are the organic fraction (wet + green), packaging (paper, plastic, glass, wood, metal, including fractions resulting from multi-material collection, net of scraps), WEEE from households, bulky to be recovered, used clothing and textile waste and SC of hazardous waste (batteries and accumulators, expired medicine, T/F containers, ink, paint oils, other hazardous waste). The denominator has the same amount reported in the numerator (Σi SCi) to which, at the national level, the unsorted MSW and waste from street sweeping (MWuns), the bulky waste (BW) and the SC scraps (SCS) are added (for example: waste from multi-material collection). Therefore, Formula (1) becomes:
Inert waste is not counted in urban waste and, therefore, it is neither in the numerator nor the denominator of (2), even if arising from household demolition.
- ➢
C3. Provincial: Similar to the conventional model with the difference that in the denominator, the waste from street cleaning is not considered. The obtained value is consequently slightly higher, if compared with the traditional method. The exclusion from the calculation of the street sweeping waste is motivated by climatic variables (mainly snow and ice) that the citizen cannot influence.
Additional assumptions are necessary for managing biostabilisation as landfill pre-treatment and the role of home composting.
It is assumed (C4) that all the unsorted MSW is subjected to a process of biostabilization before being stored permanently in a landfill. In the Province of Trento, biostabilization is required by law when the organic fraction to be landfilled exceeds the limit of 115 kg eq-inh−1 year−1, in order to reduce the possible negative effects on the management of the landfill.
Concerning home composting, when it is sufficiently widespread, it removes a major part of the organic waste from collection service, reducing waste production. This provides an additional model of calculation (C5).
Thus, five different values of percentage of recycling combined with four levels of purity of the collection were calculated, yielding 20 combinations.
3.3. Selective Collection Quality Index
The authors defined a new index to support the %SC to obtain a clear reading of the achieved results. Indeed, in addition to %SC, other indicators need to be integrated in a new index to understand how effective SC is, such as:
collection efficiency for a single fraction
adopted system and method of collection
quality of SC
implementation of the punctual tariff
tourist incidence
The purpose is to define a synthetic index, SCQI (SC quality index) that is fed by the various indicators listed above, in order to identify the best management models. Its definition will require an eight STEP pathway, pointed out as follows.
Regarding the efficiency related to each fraction and the evaluation of the tourist incidence, it is sufficient to apply the respective definitions. The application of the punctual rate is a simple binary yes/no, while it is possible to create an abstraction for classifying the sorting system and method of collection, assigning a value to the single conventional configuration according to a scale of values fixed as reference.
The measure of the quality of the collection is highly conditioned by the scrap of SC, which is the fraction of waste unrelated to the amount of collected fraction. Usually the difference is due to user errors in separating the waste, especially in the case of plastic packaging.
Within multimaterial collection, there are high percentages of scrap, mainly composed of waste not amenable to further recovery.
The high amount of existing plastic for packaging that leads to unify all of the objects made of plastic, facilitates this phenomenon.
It is important to impose continuous control on users to fight the transfer of foreign materials. To this end, we note the importance of the unique user-container relationship to improve the quality of the collection.
The analysis of all (or at least the most important) waste collected fractions would be required to determine the scraps of recycling. However, it is possible to determine with certainty only the impurities in the collection of packaging.
The scrap of the collection can be used as a variable key in assessing the quality of the collection system.
Starting from these considerations, the variables considered in the construction of the SCQI are (STEP 1):
Scrap of SC (S
SC): this is the tons of waste expelled from the SC stream during the selection phase. The data originate from product analysis performed on the collection of multi-material packaging and single plastic materials, which are only systematically available every three months from the National Consortium of Packaging [
22]. For other fractions (i.e., organic, paper and cardboard), few data are available and are more sparse in time; they are assembled directly by the Province of Trento from selected plants.
Gross SC (SCG): this is the tons of waste collected separately, both with curbside systems and at collection centers. The waste fractions are organic, green, paper and paperboard, multi-material, glass, metals, plastics, wood, textiles, WEEE (Waste from Electric and Electronic Equipment), hazardous wastes, and others. It includes residues of waste collection.
Unsorted waste (UNS): this is the tons of rubbish collected by the users with a dedicated system.
Bulky waste (BW): this is the tons of bulky waste collected mainly at the collection center. In some territories, the withdrawal is active on call at the user’s home.
Residents and equivalent population (RES, EQ).
Starting from the shown variables, we can define some derived variables (STEP 2):
Net SC, that is SC without scraps:
Residual MSW, that is MSW not managed by SC:
Percentage of SC (definition of the Province of Trento):
Tourist impact TI, as tourism causes an increase in anthropic pressure:
The SC purity index (PI) has been defined as (STEP 3):
The ratio shows how the collection system is characterized by a little amount of scrap. It has a value between 0 and 1.
The preferable results are characterized by low SSC associated with high SCG. The indicator has a clear physical meaning and it is easy to understand. Low values of the index represent the best class that gets worse with increasing PI.
PI, by itself, does not distinguish between areas characterized by little scrap and an undeveloped SC system with territories with higher scraps, but high levels of SC. To properly judge the effectiveness of a SC system, PI must be associated with other leading indicators, such as the collection rate and the per capita production of waste. We consider it useful to introduce some correction factors for PI in order to obtain a new index that summarizes the effect of the considered variables.
We considered the following factors, with the aim of building a quality index with higher general validity than PI (STEP 4):
- (a)
Level of SC: the territories with a greater percentage of SC generally have a higher amount of scrapes. At constant PI, the territories with lower %
SC must be penalized as a collection system as higher %
SC initially requires a substantial investment in human, organizational and economic resources. Consequently, PI must grow and the correction factor becomes:
with k
%SC ∈ [0, ∞], with extremes at %
SC = 1 (or 100%), and %
SC = 0%.
The infinitive value is an extreme case as it means that there is no SC (not acceptable in EU)
- (b)
Underutilization of SC: SSC is usually a proportion of waste that had to be classfied as undifferentiated. In this context, the approximation is acceptable. A low production of unsorted MSW (UNS) does not automatically show an efficient collection system when associated with high levels of SSC. In this configuration, a transfer of waste from the undifferentiated stream to the collection ones outlines obvious diseconomies of management.
The correction factor is:
with k
UNS ∈ [0, SCG/(SCG + UNS)]
SSC + UNS represents the “real” undifferentiated waste. The areas with low SSC and high UNS have a very low ratio, while the territories with high SSC and low UNS are characterized by a high ratio. The higher the KUW ratio, the more the SC results are underutilized.
- (c)
Migration of bulky waste: the production of bulky waste should be an independent variable of the collection system, reaching standard per capita values. In more advanced collecting systems (high %
SC), there is a migration of waste from undifferentiated to bulky. If strict control is not active at the collecting centers and the punctual tariff is applied, the user is encouraged to deliver part of unsorted MSW to the collecting centers to reduce the variable part of the tariff. The formula of the correction factor is:
with k
BW ∈ [0, 1].
- (d)
Inhabitants and tourist incidence: with all factors constant, the organization of the collection service, in areas with a high number of resident inhabitants and/or a high tourist incidence, is more difficult. The tourist incidence is given by the ratio:
with i = catchment area, j = year
To determine decreasing factors with increasing variable (RES and aIN.T), we calculated the ratio using the maximum value of the data series for the year j, according to the following expressions:
coefficient of permanent residents:
coefficient of tourist incidence:
The linear combination of the two coefficients provides the correction factor k
AB that, in the more severe case (a territory at the same time characterized by maximum permanent residents and highest incidence of tourism: a
RES = 0 and a
T = 0), assumes a unit value. Thus,
Starting from this framework, the authors propose the following formula (STEP 5) for an SC quality index (SCQI), whose calculation pathway can be reconstructed according to the method proposed by Kyriakis et al. [
23]:
A high value of the index corresponds to a low quality. SCQI coincides with the index of purity PI when a territory simultaneously presents all limiting situations. Consequently, SCQI ≥ PI.
The various contributions of the k factors have the same weight in the formula of SCQI. It is considered appropriate to introduce a weight coefficient p [0, 1] for each factor k in order to better calibrate the influence of SCQI based on the historical data available. Therefore, denoting K (capital letter) as the definitive formula of correction factors, the individual expressions become (STEP 6):
To fully assess the influence of the four K contributions in (15), a set of indexes was defined as follows (STEP 7):
It is appropriate to define a limited number of classes (STEP 8) in order to facilitate the reading of PI and SCQI. We have chosen to fix three classes:
Class 1–high
Class 2–average
Class 3–low
Data related to a period of some years can be used in order to define the limit values of PI and the correction factors. For the correction factors, limits can be established in relation to the best and worst performances in the catchment areas, in accordance with the objectives of the planning data. For PI, class limits can be defined by evenly distributing the frequency of the sample. SCQI, contrary to PI, does not have a real physical meaning, but represents an abstraction of the quality of the collection, through the linear combination of dimensionless quantities, based on per capita waste weight and inhabitants.
PI must be analyzed together with other factors that characterize the system in order to have a complete and comprehensive reading of the achieved results. Instead, SCQI does not need to explain the characterizing factors, except to understand the origin of the corrections.