Valorisation of the organic fraction of municipal solid waste

Abstract

The appropriate selection of treatment or final disposal for municipal solid waste generated by a population depends on multiple characteristics; decision-making becomes a challenge for those responsible for its integral management. Therefore, the objective of this research was to evaluate multi-scenarios of valorisation of the organic fraction of municipal solid waste. The methodology used included a sample of 421 houses to determine the generation and composition of the aforementioned municipal solid waste; this data served as input to be able to design four treatment and final disposal scenarios: current situation, composting, anaerobic digestion and landfill. The parameters of humidity, biochemical methane potential and biogas were used. In addition, organic fraction variability and treatment cost were considered. Data analysis included design of experiments with the uncertainty coefficient and predictive equations. The results showed that composting was the scenario that presented the highest coefficient of determination and therefore it would be the best choice to fit the particular conditions of the localities. It is concluded that this applied methodology can be used as a tool in the decision-making process regarding the valorisation of the organic fraction of municipal solid waste and can be replicated at a national and international level.

Keywords

Municipal solid waste organic fraction valorisation multi-scenario evaluation design of experiments

Introduction

The generation of municipal solid waste (MSW) worldwide is approximately 1.3 billion t y^-1 and this value could be increased to 2.2 billion t y^-1 for 2025 (World Bank, 2012). This situation is dangerous, owing to the consequences that could arise from the inadequate waste disposal management, such as pollution of surface and groundwater, soil, air, landscape, disease transmission, generation of biogas, waste burning, drainage obstruction, etc. (Soliz, 2015).

To properly select MSW management for a certain region, it is essential to know its generation and composition. This data is the starting point to begin studies that contribute to establishing municipal public policies of valorisation and treatment alternatives to achieve an integrated waste management. So it is important to review every component of MSW, and especially the organic fraction because it is predominant. A way to achieve this revision is the establishment of multiple scenarios, which simulate alternatives of MSW management (Colón et al., 2015; Fernández-González et al., 2017; Ionescu et al., 2013; Lombardi et al., 2015; López et al., 2015; Massarutto et al., 2011; Oyoo et al., 2014; Parkes et al., 2015; Panepinto et al., 2015; Rajaeifar et al., 2015). These scenarios can be evaluated using different tools, like techniques of advanced design of experiments (DOE), which allows the analysis and optimisation of variables individually or through the combination of effects to obtain optimal operating conditions (Dong and Sartaj, 2016).

Although several authors have used DOE (Babu et al., 2013; Bashir et al., 2010; Dong and Sartaj, 2016; Menon et al., 2017), in this study an analysis of four scenarios with 21 environmental and economic variables was conducted and applied in rural and urban areas to identified the most adequate scenario in the study area regarding to the cost treatment (composting, anaerobic digestion and landfill) of the organic fraction. This included environmental, technical, economic, legal and social aspects with more precision due to the use of the uncertainty coefficient and with the additional benefit of having predictive equations for costs of the treatments according the possible variation in composition and other parameters of the treatments process.

Therefore, the objective of the present research was to evaluate multi-scenarios of valorisation of the organic fraction of MSW using DOE to obtain the best scenario for the study area and predictive equations for economic costs of the treatments aforementioned. Therefore, the amount of new information for decision-making and the analysis proposal in the methodology can be used as a tool of strategic planning for valorisation of the organic fraction of MSW.

Methodology

This research was carried out in two stages as it is shown in Figure 1.

Stage 1. Logistics of field and laboratory activities

Description of the study area

This research was conducted in Mexico, in a region with 5327 km² (SEFIPLAN, 2013). The predominant climate is mild and the annual rainfall is between 1500 and 2000 mm, the average temperature is 12–18 °C (Gutiérrez and Dorantes, 2017). The study population contained 33 municipalities located in the vicinity of the Sierra Madre Oriental. In this area, 1.2 million persons lived in 331,588 households (INEGI, 2015).

According to the number of inhabitants, every municipality can be classified as follows (SEDESOL, 2015):

Stratum 1 or metropolitan (localities with more than 100,000 inhabitants).

Stratum 2 or urban (localities with more than 15,000 and less than 100,000 inhabitants).

Stratum 3 or semiurban (localities with more than 2500 and less than 15,000 inhabitants).

Stratum 4 or rural (localities with less than 2500 inhabitants).

The study area included 17 semiurban, 10 rural, five urban and one metropolitan municipality (INEGI, 2015; SEDESOL, 2017). Regarding their economic activities, there are differences because most of the people can be working in the agricultural sector or in several services with percentages from 3% up to 75%. It was estimated in 2010 that 54.7% of the inhabitants in this area lived in multidimensional poverty; 5% were vulnerable by income; 24.2% vulnerable by social deficiencies; and 16.1% non-poverty and non-vulnerable people (SEFIPLAN, 2013).

To determine the sample size, a multi-stage sampling was carried out using the sampling table MIL STD 105 E (Castillo and De Medina, 2014; De Medina et al., 2013; Gutiérrez, 2010). First, counting the municipalities, then counting the city blocks of every municipality selected and finally counting the houses of the city blocks selected, to obtain the final sample of houses for the study.

In the selected houses, fieldwork was carried out to determine the generation and composition of MSW following the methodology described in the national reference standards; special emphasis was put on the organic fraction (food and garden waste) in the study localities. The collection of MSW was daily and at the same time. The samples were collected in bags and these were transported in a container to preserve their characteristics; the time between sampling and analysis was approximately 30 to 45 minutes (SECOFI, 1985a).

The generation per capita (GPC) of MSW expressed in kg inhab-day^-1, was obtained by dividing the weight of the residue by the number of inhabitants of the household as shown in equation (1) (SECOFI, 1985b).

Generation per capita (GPC) = \frac{kg MSW collected per household}{Number of inhabitants per household}

(1)

Next, the daily GPC average for each of the houses was obtained based on data registered during the sampling week. Based on the average household, the average GPC for the 7 days was obtained, expressed in kg inhab-day^-1 of the MSW of the corresponding locality.

The separation of food and garden waste was carried out based on reference standard (SECOFI, 1985c). This standard served to carry out the characterisation of MSW; additionally, an adaptation was made. MSW was then identified and separated into 30 by-products: cotton, cardboard, waxed cardboard, ashes, leather, synthetic fibres, bone, rubber, earthenware and ceramics, wood, construction material, ferrous material, non-ferrous material, paper, wax paper, plastic film, rigid plastic, polyurethane, polystyrene, food waste, garden waste, tetra pack, rags, stained glass, transparent glass, batteries, medicines, disposable diapers, hazardous waste, others.

The percentage of the composition of MSW in each locality was determined by dividing the weight of the considered by-product into the total weight of the sample. This process was performed throughout 7 days of sampling. Then, an average composition per by-product was obtained during the sampling period for all the localities.

Subsequently, the humidity of the sample was determined in the laboratory; the settings for this determination the loss of weight under temperature and time period conditions were considered as per reference standard (SECOFI, 1984). This standard establishes the stove method, which is based on determining weight loss suffered by the MSW sample when subjected to time and temperature conditions established in the standard, taking into consideration that loss of water is originated from water elimination.

Stage 2. Design and evaluation of scenarios

In this stage, four possible scenarios for the valorisation of the organic fraction were designed and evaluated, as follows.

Scenario E1: The treatment or final disposal that each of the localities in the sample currently uses for the organic fraction of their MSW.

Scenario E2: Composting, which was studied through the parameter of humidity that was considered as reported in several studies (Adamcová and Vaverková, 2016; Hannibal et al., 2016; Oviedo et al., 2016); this was determined via laboratory for food and garden waste. Its units were kilogram of water per kilogram of waste.

Scenario E3: Anaerobic digestion, analysed through the parameter of the biochemical methane potential (BMP) according to different authors (Cabbai et al., 2013; Lisboa and Lansing, 2013; Mou et al., 2014), for which the values for food and garden waste were estimated in kilogram of methane per kilogram of waste.

Scenario E4: Landfill, through the parameter of biogas as reported in the literature (Budzianowski, 2016; Fantozzi and Buratti, 2011; Mustafa et al., 2016), for which values for food and garden residue were estimated in kilogram of biogas per kilogram or waste.

The variables that were analysed in this research: (i) the composition of MSW; (ii) the humidity corresponding to composting; (iii) BMP for the anaerobic digestion; (iv) biogas related to the landfill; and (v) economic cost of these treatments.

BMP determination was estimated from the values of the composition of the MSW multiplied by the reference values that were obtained as default values for MSW calculated based on default degradation organic carbon content (DOCC) values of MSW fractions at international level in 2011 (Mou et al., 2014); therefore, the BMP value was obtained for all the localities of the study area as shown in equations (2) and (3):

BMP (food garden) {= composition of food waste * 100 kg CH}_{4} / t

(2)

BMP (garden waste) {= composition garden waste * 133 kg CH}_{4} / t

(3)

For biogas quantification, the first step was the calculation of the conversion factor as shown in equation (4) (Giraldi, 2010):

F = 1 kg CE * \frac{{MWCH}_{4}}{AWC} * FBDG

(4)

where F is the conversion factor; CE is the carbon equivalent; MWCH₄ is the molecular weight methane; AWC is the atomic weight carbon; and FBDG is the factor of biodegradation of MSW in the landfill for food or garden waste (depending on the case).

By multiplying the weight of the food waste in each locality by the value of F (equation (4)) for food waste, a percentage of methane equivalent to 60% of the generated biogas was obtained; considering this data, a correction corresponding to 100% was made. For garden waste, the same procedure was followed, considering that the percentage of methane obtained from biogas was 50% and the adjustment was made for 100%.

With the data obtained from fieldwork, laboratory and literature consulted, the values corresponding to the following variables were identified: (i) composition of food, garden and total waste; (ii) humidity of food and garden waste; (iii) BMP of food and garden waste; (iv) biogas from food and garden waste; (v) economic cost for food, garden and total waste in the cases of composting, anaerobic digestion and landfill. Possible combinations of composition, treatments and operational costs were made for both food and garden waste.

Subsequently, a coding for the variables was designed, which is shown in the results section. Then, the data was introduced in the statistical analysis software JMP version 8.0 to facilitate its analysis as reported in the literature (Abdalla et al., 2012). This software is from the SAS Institute Inc., which is the American company leader in analytical software (SAS, 2018).

Next, a factorial experimental design was carried out, including the analysis of variance (ANOVA). The behaviour of every factor: humidity, BMP, biogas and the percentage of the organic fraction and the economic cost for every treatment aforementioned, was studied. This technique was applied to evaluate the differences of study and it also allowed quantifying the significant parameters. Several experimental trials were made using regression and mathematical models of second-order were obtained, as well as the coefficient of determination (R²), which allowed determining the accuracy of the model. Predictive equations for the economic costs of the treatment in the study were obtained for significant terms. The significance was determined if the probability value (P-value) of the parameter estimates calculated using the software JMP, was lower than the significance level (alpha) 0.10.

Then, the uncertainty coefficient (EPA, 2015), was calculated using the prediction profiler of the software and dividing it between the estimated variable response. Under this criterion, the scenarios were evaluated, considering that the scenario with the minor uncertainty for the cost, offers precision and reliability of the results allowing having the best treatment to fit the particular necessities of the locality.

Results and discussion

The sample size was of eight localities randomly selected and 421 houses were selected in the localities as a sample. The average GPC of MSW obtained for every locality is shown in Figure 2.

Figure 2 shows the locality with the highest generation of MSW was L2, with an average of 0.499 kg inhab-day^-1, while L6 obtained the lowest average with 0.205 kg inhab-day^-1. Localities L1, L3 and L5 were below the average in Mexico, which is 0.408 kg inhab-day^-1; L2 and L8 are above this average; while L5 has practically the same amount (SEDEMA, 2015; SEMARNAT-INECC, 2012).

The GPC for L1, L3 and L5 is similar to the one reported for the average of the study area, which is 0.362 kg inhab-day^-1; localities L6 and L7 are below this value; while the results of L2 and L4 are above it (INEGI, 2016; SEDEMA, 2015).

There are different factors that could affect the GPC, such as the day of the week, socioeconomic stratum and type of locality (Castillo and De Medina, 2014); particularly small and rural cities have remained stable in their generation of MSW along these last decades, which could explain the values obtained in this research (Jiménez, 2015). Table 1 shows the average composition of MSW obtained in every locality.

Table 1.

Summary of the average composition of MSW for the studied localities in mass-percentage.

Localities
Composition %	L1	L2	L3	L4	L5	L6	L7	L8
Food waste	47.34	43.37	46.45	42.57	51.41	32.86	37.00	38.90
Garden waste	8.49	16.08	11.64	8.76	8.31	9.39	3.00	4.40
Transparent glass	3.00	2.35	2.61	4.12	6.90	3.65	5.16	2.00
Rigid plastic	6.09	6.62	6.90	8.90	6.52	6.29	10.90	5.74
Plastic film	5.40	4.32	5.61	4.62	4.37	5.11	7.24	4.32
Paper	7.19	4.78	6.00	7.21	6.05	3.79	9.29	9.19
Cardboard	2.55	3.44	4.26	2.76	3.16	2.84	5.61	3.64
Non-ferrous materials	1.90	1.62	1.32	2.10	2.50	1.20	2.10	1.52
Tetrapack	1.95	1.96	1.36	1.85	2.20	2.70	2.11	1.54
Disposable diaper	8.08	4.73	5.00	8.00	3.50	1.70	3.53	4.00
Others	8.01	10.73	8.85	9.11	5.08	30.47	14.06	24.75
Sum	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00

In Table 1, it can be observed that the organic fraction (food and garden waste) predominates in all study localities with the following percentages organised from highest to lowest: L5 59.72%, L2 59.45%, L3 58.09%, L1 55.83%, L4 51.33%, L8 43.3%, L6 42.25% and L7 40%. The first four values being higher than the value reported for organic waste in Mexico, which is 52.4% (Jiménez, 2015).

Regarding food waste, it can be seen that L6 has the lowest percentage (32.86%), while L5 has the highest value (51.41%). In all cases, the values found are higher than the national average for this type of waste, which is 25.57% (SEMARNAT-INECC, 2012). This situation is probably owing to the fact that in small rural and urban areas the consumption of unprocessed perishable food is higher than medium or metropolitan cities, where owing to lifestyle, processed foods with longer shelf life and expiration are preferred (De Medina et al., 2016).

The percentages found in this study for garden waste in L1, L4, L5 and L6 are similar to the national indicator, which is 9.38% (SEMARNAT-INECC, 2012). The lowest value obtained in this study was in L7 (3%) and the highest L2 (16.08%), so it is inferred that this variation can be related to the level of socioeconomic development and growth of the locality (De Medina et al., 2016). The results of the four proposed scenarios are shown below.

Scenario E1: Current situation

During this research, the climatic conditions were temperatures between 15 °C to 17 °C, no rain and cloudy days and the sampling was carried out in the winter season from November to February. The localities have a waste management model that includes only two steps: The collection of MSW and its final disposal, except for L3 with vermicomposting; no other place was identified that valorised its organic fraction. This situation can be seen in more detail in Table 2.

Table 2.

Scenario E1: Current situation of the localities regarding their MSW.

Current situation	L1	L2	L3	L4	L5	L6	L7	L8
Final disposal	Landfill	Landfill	Landfill	Landfill	Landfill	Landfill	Landfill	Landfill
Treatment	None	None	Vermicomposting	None	None	None	None	None
Frequency of collection by the public cleaning service	Twice a week	Once a week	Twice a week	Twice a week	Once a week	Once a week	Once a week	Once a week

Scenario E2: Composting

Table 3 shows the humidity results for MSW sample, food and garden waste obtained experimentally.

Table 3.

Results of the humidity test for MSW sample, food and garden waste.

Localities	Total humidity of the MSW sample $\frac{k g o f w a t e r}{k g w a s t e}$	$\frac{k g o f w a t e r}{k g f o o d w a s t e}$	$\frac{k g o f w a t e r}{k g g a r d e n w a s t e}$
L1	0.45162276	0.19321821	0.03465193
L2	0.46055678	0.18422491	0.06830382
L3	0.60145321	0.21965647	0.05457127
L4	0.70629087	0.27911446	0.05743581
L5	0.69548960	0.27320302	0.04416100
L6	0.53426543	0.16600090	0.04745048
L7	0.44491268	0.14343420	0.01162980
L8	0.73569201	0.26296011	0.02974356

As shown in Table 3, the percentages of total humidity of the MSW sample were between 44.49% and 73.56%. These results are according to the literature consulted (Douma et al., 2012; Durán et al., 2013; López et al., 2010).

The localities with the highest percentage of humidity in food waste were L4 (27.91%), L5 (27.32%) and L8 (26.29%). The lowest value was presented in L1 (19.32%). The results of humidity for food waste were lower than those reported by the literature (50%–65%) (Ogunjuyigbe et al., 2017; Rada et al., 2014; Zhou et al., 2018). The results of this research would possibly differ by the type of locality, geographical situation, climate, economic situation and customs. Another factor to consider is if food waste was in contact with other types of residue, like cardboard or paper packaging previous to collection or sampling at the source, this could adsorb part of the food waste’s humidity, reducing the sample’s humidity. It is also worth mentioning food can contain two types of water: free water that can evaporate easily in different ways, and linked water, which is combined chemically with protein (Alkemi, 2018). In the study area, as mentioned previously in the methodology section, most people have limited economic resources. Thus the typical food in the localities does not include great quantities of meat, which is principally composed of protein. So it is likely that, as most of the food waste in the area did not contain this kind of residue, the linked water associated to protein would have been a minimum amount, and the percentages of humidity were therefore low as well.

In the case of garden waste, the highest percentage of humidity was L2 (6.83%) and the lowest L7 (1.16%). These results were lower than those reported by other authors (75%) (Rada et al., 2014). However, other research indicates that the percentage of these residues can vary between 19% and 31% (Ogunjuyigbe et al., 2017). It is likely that the differences can be attributed to the time of year in which the studies were conducted and the type of locality (Ansorena, 2016).

These results contrast with those reported in Mexico City for the organic fraction, where the value of 687 g of water per kilogram residue or 68.7% was obtained (González, 2017). However, it should be noted that this last study included remains of paper and non-classifiable organic material within the organic fraction, which could explain the difference in the results found.

Scenario E3: Anaerobic digestion

Table 4 shows the results of the BMP for food and garden waste calculated for the localities.

Table 4.

BMP calculated for the localities.

Localities	BMP kg CH₄ kg^-1 food waste	BMP kg CH₄ kg^-1 garden waste
L1	0.018	0.004
L2	0.021	0.010
L3	0.016	0.005
L4	0.017	0.004
L5	0.018	0.003
L6	0.006	0.002
L7	0.010	0.001
L8	0.016	0.002

In Table 4, L2 has the highest BMP in food (0.021) and garden waste (0.010); while the lower values for food waste were found in L6 (0.006) and in L7 (0.001) for garden waste.

Based on the consulted literature, the predetermined values for BMP of the organic fraction are in the range of 0.0148–0.0445 kg CH₄ kg^-1 residue on a humidity basis (Mou et al., 2014). According to this, the results of food waste coincide because they were between 0.016 and 0.021 kg CH₄ kg^-1, residue except in the case of L6 and L7, which would be below along with garden waste.

The results obtained in this research were lower than those reported for experimental tests for food waste (0.061–0.106 kg CH₄ kg^-1 waste) (Mou et al., 2014). This variation is possibly owing to the lifestyle and economic development of the inhabitants because their research was conducted in a developed country with a high socioeconomic profile for its people and as has been described, the study area of this study included more than 50% of people in multidimensional poverty.

The values found in the localities under study were lower than those indicated in other research (0.322 kg CH₄ kg^-1 waste) (Cabbai et al., 2013), which were probably owing to the fact that they included the organic fraction from shops and restaurants. Likewise, for other specific studies in the organic and food fraction, there were lower results, they were 0.2256 and 0.2919 kg CH₄ kg^-1, respectively (Zhang et al., 2012), these differences are probably owing to the fact that these studies were carried out in a highly commercial port and where the climatic conditions are more favourable for biodegradation.

Scenario E4: Landfill

Table 5 shows the results obtained for biogas calculated from food and garden waste in all localities.

Table 5.

Results of the calculation of biogas generation for food and garden waste.

Localities	Kg biogas kg^-1 food waste	Biogas expressed in kg CH₄ kg^-1 food waste	Kg biogas kg^-1 garden waste	Biogas expressed in kg CH₄ kg^-1 garden waste
L1	0.032	0.019	0.004	0.002
L2	0.038	0.023	0.010	0.005
L3	0.030	0.018	0.005	0.002
L4	0.030	0.018	0.004	0.002
L5	0.032	0.019	0.003	0.001
L6	0.011	0.007	0.002	0.001
L7	0.019	0.011	0.001	0.0005
L8	0.029	0.017	0.002	0.001

Locality L2 presented the highest generation of biogas per kilogram of food waste (0.038) and garden waste (0.010), as can be observed in Table 5. On the other hand, L6 has the lowest biogas generation per kilogram of food waste (0.011) and L8 for garden waste (0.002).

The great majority of researchers at the international level refer to methane generated when conducting biogas studies, so in this section, the discussion of results was made under this premise.

The maximum values obtained from the generation of biogas in this research for food and garden waste were 0.023 and 0.005 kg CH₄ kg^-1 waste, respectively, lower than those reported in the literature indicating 0.251 and 0.026 kg CH₄ kg^-1 waste (Bundhoo et al., 2016). However, this information comes from a metropolis with more than one million inhabitants and the study area is predominantly rural or semi-urban, which presumably had a role in the differences noted.

The biogas results for garden waste obtained in this research were found to be in the range of 0.002 to 0.010 kg CH₄ kg^-1 waste, which were lower than the value reported by the literature of 0.029 kg CH₄ kg^-1 waste (Maalaouf and El-Fadel, 2017). This difference was probably to the fact that straw was found in that study, which was not presented in this study sample, likewise, the authors point out that the waste used had 60% humidity, which is also different in the sample used here, which had maximum humidity values of 27%.

The methane values obtained in this research vary between 0.001 and 0.0019 kg CH₄ kg^-1 waste, which were lower than those indicated in the literature for the organic fraction of 0.223 kg CH₄ kg^-1 waste (Mao et al., 2015). However, it should be noted that those studies refer to the inclusion of rice waste and tea leaves, which appeared in minimal quantities within the waste generated by the study localities.

Regarding to the economic costs for the treatments of this study, Table 6 shows the results of the costs for each of the treatments in dollars per tonne, both in the study area reported by the municipalities and the average value for several countries (De Medina et al., 2017; Milutinović et al., 2014). The different treatment systems would be located in the same site, thus the collection costs in each alternative were not considered, because they would be equal.

Table 6.

Costs of the study treatments.

Treatment	Cost study zone ($ t^-1)	Cost at Latin America and the Caribbean ($ t^-1)	Cost at international level ($ t^-1)
Composting	130.23	20–75	47
Anaerobic digestion	76.10	50–100	85
Landfill	11.92	25–65	72

Note: The table is made from interviews to the authorities responsible of waste management and adapted from ‘Effect of a food waste disposer policy on solid waste and wastewater management with economic implications of environmental externalities’ (Maalaouf and El-Fadel, 2017; Mao et al., 2015; World Bank, 2012).

As can be noted, composting is cheaper in other countries; in the study zone, the cost of the composting process was calculated based on the operating costs of a composting plant in the study area, which included the monthly salary of five workers ($1674.41). Estimating a period of time of the degradation process to obtain the compost of 7 months, considering a total weight of 90 t of treated organic waste, the cost per tonne was $130.23.

The anaerobic digestion and especially landfill are extremely expensive in other countries. It is inferred that these differences are related to the environmental and economic policies that prevail in the different countries and to the type of locality. Likewise, it should be mentioned that in rural areas the design and operation of landfills is not carried out in its entirety as indicated by environmental laws, so it is common for costs to be lower than those reported in industrialised cities.

Table 7 shows the costs for treatment considering the organic waste generated in every locality and those reported by the municipal authorities.

Table 7.

Economic cost of the different treatments for food and garden waste for localities.

Localities	Organic fraction (kg day^-1)	Composting cost			Anaerobic digestion cost			Landfilling cost
Localities	Organic fraction (kg day^-1)	Food waste ($)	Garden waste ($)	Sum ($)	Food waste ($)	Garden waste ($)	Sum ($)	Food waste ($)	Garden waste ($)	Sum ($)
L1	589.45	1259.53	225.88	1485.41	736.07	132.01	868.08	115.34	20.68	136.02
L2	20.04	36.84	13.66	50.49	21.53	7.98	29.51	3.37	1.25	4.62
L3	109.03	220.07	54.67	274.74	128.61	31.95	160.56	20.15	5.01	25.16
L4	211.62	442.27	91.01	533.28	258.47	53.19	311.65	40.50	8.33	48.83
L5	50.16	108.81	17.59	126.40	63.59	10.28	73.87	9.96	1.61	11.57
L6	13.11	25.69	7.34	33.03	15.01	4.29	19.30	2.35	0.67	3.02
L7	5.86	13.65	1.11	14.76	7.98	0.65	8.62	1.25	0.10	1.35
L8	87.60	198.33	22.43	220.76	115.90	13.11	129.01	18.16	2.05	20.21

In Table 7, the highest and lowest costs were in composting and landfill, respectively. It should be noted that this research only considered the cost of operation and not the income that could be obtained from the sale of the compost.

Data analysis

Table 8 shows the coding for the variables to use in the statistical software JMP version 8.

Table 8.

Codification for the variables in the study.

Coding	Variable
Y	The overall composition of food and garden waste
Y1	The composition of food waste
Y2	The composition of garden waste
X11	The humidity of food waste
X12	The humidity of garden waste
X21	BMP of food waste
X22	BMP of food waste
X31	Biogas of food waste
X32	Biogas of garden waste
C11	Cost of composting food waste
C12	Cost of composting garden waste
C21	Cost of anaerobic digestion for food waste
C22	Cost of anaerobic digestion for garden waste
C31	Cost of the landfill for food waste
C32	Cost of the landfill for garden waste
C1T	The total cost of composting for food and garden waste
C2T	The total cost of anaerobic digestion for food and garden waste
C3T	The total cost of the landfill for food and garden waste

BMP: biochemical methane potential.

The combinations of the variables introduced into the software are illustrated in Table 9.

Table 9.

Combination of encoded variables.

Encoded variable	Versus	Cost
Y	X11	C1T
Y	X12	C1T
Y	X21	C2T
Y	X22	C2T
Y	X31	C3T
Y	X32	C3T
Y1	X11	C11
Y1	X11	C1T
Y1	X21	C21
Y1	X21	C2T
Y1	X31	C31
Y1	X31	C3T
Y2	X12	C12
Y2	X12	C1T
Y2	X22	C22
Y2	X22	C2T
Y2	X32	C32
Y2	X32	C3T

The results showed that only six of the 10 combinations were significant, which are detailed below.

1. The total composition of food and garden waste (Y), humidity of garden waste (X12) and the total cost for composting of food and garden waste (C1T).

The coefficient of determination (Rsquare) was 77.19%, indicating that the deviations between the experimental values and the estimated values are explained by the proposed mathematical model. With the terms and their significant values (Figure 3), the predictive equation was formulated for the total treatment cost using the composting technique (C1T) shown in equation (5):

C 1 T = - 538331.4 - 2171044.8 Y - 831263.4 YX 12 - 2041860 Y^{2}

(5)

2. The composition of food waste (Y1), BMP of food waste (X21) and cost for anaerobic digestion of food waste (C21).

The coefficient of determination was 82.68% (Figure 4) and the predictive equation (6) obtained for the cost of the anaerobic digestion treatment for food waste obtained with the software was the following:

C 21 = 255378.15 - 32477176 X 21 + 170629012 Y 1 X 21 - 0 {. 000000001116 X 21}^{2} {- 3356524 Y 1}^{2}

(6)

3. Composition of food waste (Y1), BMP of food waste (X21) and total cost for anaerobic digestion of food waste and garden (C2T).

The coefficient of determination was 83.11% (Figure 5) and the predictive equation (7) that was obtained for the total cost of the anaerobic digestion treatment was:

C2T = 302255.822 - 38455455 X 21 + 201823797 Y 1 X 21 - 0 {. 0000000013192 X 21}^{2} - 3969677 {. 3 Y 1}^{2}

(7)

4. The composition of food waste (Y1), biogas from food waste (X31) and cost for landfill of food waste (C31).

The coefficient of determination was 82.68% (Figure 6) and with significant terms, the predictive equation (8) was generated for the cost of landfill treatment with respect to the food waste shown below:

C 31 = 40015.658 - 2864300 {. 1 X 31 + 15048498 Y 1 X 31 - 55422314 X 31}^{2} - 525939 {. 75 Y 1}^{2}

(8)

5. The composition of garden waste (Y2), humidity of garden waste (X12) and cost of composting of garden waste (C12).

The coefficient of determination was 92.13% (Figure 7) and the predictive equation (9) obtained for the cost of sanitary landfill treatment with respect to food waste was:

C 12 = - 4884.3779 + 172332 {. 397 Y 2 - 23905125 Y 2 X 12 + 21096651 X 12}^{2} + 5482582 {. 97 Y 2}^{2}

(9)

6. The composition of garden waste (Y2), humidity of garden waste (X12) and the total cost for composting of food and garden waste (C1T).

The coefficient of determination was 94.79% (Figure 8) and the predictive equation (10) obtained for the total cost of the composting treatment was:

C1T = - 30748.881 + 1117485 {. 14 Y 2 - 159301204 Y 2 X 12 + 140200830 X 12}^{2} + 36686733 {. 2 Y 2}^{2}

(10)

With the results shown in the prediction profiler section of the Figures 3, 4, 5, 6, 7 and 8, the uncertainty coefficient was estimated and it is shown in Table 10.

Table 10.

Uncertainty coefficients for the combination of the statistically significant variables of the study.

Combination of variables	Uncertainty coefficient
X12 / Y / C1T	1.33578676
X21 / Y1 / C21	14.30765420
X21 / Y1 / C2T	3.44516472
X31 / Y1 / C31	3.48679563
X12 / Y2 / C12	1.96700326
X12 / Y2 / C1T	1.67663924

In Table 10 it can be observed that the minor uncertainty coefficient was the first one, corresponding to the composting of garden waste and considering the total cost for the treatment of the organic fraction. This means composting is the alternative with the minor variability to response variables and therefore it can show a major grade of adaptation to the real conditions of the localities in study. This technique also contributes to the sustainability of the region as reported in the literature (Colón et al., 2015; Milutinović et al., 2014; Vucijak et al., 2016; Waqas et al., 2017).

The anaerobic digestion was the scenario with the major uncertainty (second one), so it will be the most unfavourable option for the study area, which is reasonable because, although several studies consider this technique as the best alternative with environmental benefits, it is necessary to have sufficient economic resources to install an anaerobic digestion plant, and it would be infeasible for the study area. However, there are support programmes for developing countries, such as Green Climate Fund (GCF), which belongs to the United Nations Framework Convention on Climate Change (UNFCC), which can contribute up to 1 million USD per country per year. Of this amount, those countries can use up to 300,000 USD per year to favour actions related to the programme. It would be desirable for the municipal governments to initiate the necessary steps to access these resources and thus implement new environmental technologies, such as anaerobic digestion, among many other possibilities (GCF, 2018).

It is worth mentioning that differences reflected in the results of the uncertainty coefficient are probably owing to the several grades of urbanisation of the localities of the study area, which influences in the high variability of their residues.

Conclusions

The generation and composition of MSW were an essential input for the development of this research, as they provided basic information for the evaluation of multi- scenarios.

For the valorisation of organic waste, four scenarios were designed that included the following: (i) Current status of the MSW in the study areas, (ii) Composting through the study of waste humidity, (iii) Anaerobic digestion with BMP; and (iv) Landfill through biogas. In these scenarios, the following findings were presented.

The analysed organic fraction had low percentages of humidity contrasting with what is reported in the literature, most likely as a result of particular situations of each locality under study.

The BMP estimated in this research for all the localities was found within the range reported as predetermined values for the organic fraction; however, the results differed with respect to experimental tests from other studies. This situation was probably related to the fact that the calculation of BMP has not been standardised worldwide, in addition to the peculiarities of waste characterisation and the singular geography of each region.

From the organic waste studied, the amounts of methane obtained from biogas were identified to be lower than reported by various authors, so to have a complete panorama of its generation, it would be convenient to monitor generation for 1 year.

Based on the results, it was concluded that because of the minor uncertainty coefficient, composting for garden waste was the most appropriate scenario for the study area. This means that if this treatment was applied to organic waste, this model would be better adapted to the particular conditions of the localities. It should be noted that the predictive equations can be used to make decisions regarding the treatments that should be chosen by those responsible for waste management.

This document constitutes a valuable tool for the formulation of public policies regarding the process of the valorisation of the organic fraction and can be replicated at a national or international level.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Lorena De Medina-Salas

References

Alkemi

(2018) Laboratory specialized in physico-chemical, microbiological and genetic analysis. Spain. Available at: https://alkemi.es (accessed 25 August 2018).

Ansorena

(2016) El compost de biorresiduos. Normativa, calidad y aplicaciones. Madrid: Editorial Mundi-Prensa.

Abdalla

Ali

SHN

Saad

SAM

et al . (2012) The chemical characteristics of composted and vermicomposted cotton residues case study in Sudan. World Journal of Science, Technology and Sustainable Development 9: 325–336.

Adamcová

Vaverková

(2016) Does composting of biodegradable municipal solid waste on the landfill body make sense? Journal of Ecological Engineering 17: 30–37.

Babu

SGL

Chouksey

Reddy

(2013) Approach for the use of MSW settlement predictions in the assessment of landfill capacity based on reliability analysis. Waste Management 33: 2029–2034.

Bashir

MJK

Aziz

Yusoff

et al . (2010) Application of response surface methodology (RSM) for optimization of ammoniacal nitrogen removal from semi-aerobic landfill leachate using ion exchange resin. Desalination 254: 154–161.

Budzianoswki

(2016) A review of potential innovations for production, conditioning and utilization of biogas with multiple-criteria assessment. Renewable and Sustainable Energy Reviews 54: 1148–1171.

Bundhoo

ZMA

Mauthoor

Romeela

(2016) Potential of biogas production from biomass and waste materials in the small island developing state of Mauritius. Renewable and Sustainable Energy Reviews 56: 1087–1100.

Cabbai

Ballico

Aneggi

et al . (2013) BMP tests of source selected OFMSW to evaluate anaerobic codigestion with sewage sludge. Waste Management 33: 1626–1632.

10.

Castillo

De Medina

(2014) Generación y Composición de Residuos Sólidos Urbanos Domésticos en Localidades Urbanas Pequeñas en el Estado de Veracruz. México. Revista Internacional de Contaminación Ambiental 30: 81–90.

11.

Colón

Cadena

Colazo

et al . (2015) Toward the implementation of new regional biowaste management plans: Environmental assessment of different waste management scenarios in Catalonia. Resources, Conservation and Recycling 95: 143–155.

12.

De Medina

Castillo

Romero

et al . (2013) Analysis of the daily per capita generation of municipal solid waste in the central region of Veracruz México. Using a statistical experimental design. Waste Management 33: 251–254.

13.

De Medina

Castillo

Gutiérrez

(2016) La gestión de los residuos sólidos urbanos en México. In: García

De Medina

(Coords.) Derecho y Gestión de los residuos. México, DF: Ubijus, pp.101–117.

14.

De Medina

Castillo

Giraldi

DMR

et al . (2017) Analysis of economical and environmental costs for the selection of municipal solid waste treatment and disposal scenarios through multicriteria analysis. Sustainability 9: 1758.

15.

Dong

Sartaj

(2016) Statistical analysis and optimization of ammonia removal from landfill leachate by sequential microwave/aeration process using factorial design and response surface methodology. Journal of Environmental Chemical Engineering 4: 100–108.

16.

Douma

Guy

Serge

et al . (2012) Composition of municipal solid waste (MSW) generated by the city of Chlef (Algeria). Energy Procedia 18: 762–771.

17.

Durán

Garcés

Velasco

et al . (2013) Revista internacional de contaminación ambiental 29: 47–57.

18.

EPA (2015) Measurement Uncertainty. Environmental Protection Agency. Available at: https://www.epa.gov/sites/production/files/2015-05/documents/402-b-04-001c-19-final.pdf (accessed 11 March 2018).

19.

Fantozzi

Buratti

(2011) Anaerobic digestion of mechanically treated OFMSW: Experimental data on biogas/methane production and residues characterization. Bioresource Technology 102: 8885–8892.

20.

Fernández-González

Grindlay

Serrano-Bernado

et al . (2017) Economic and environmental review of waste-to-energy systems for municipal solid waste management in medium and small municipalities. Waste Management 67: 360–374.

21.

GCF (2018) Readiness and Preparatory Support Guidebook. Green Climate Fund.

22.

Giraldi

DMR

(2010) Evaluación del potencial de mitigación de gases de efecto invernadero en la gestión de residuos. Doctoral Thesis. Universidad Politécnica de Cataluña.

23.

González

(2017) Determinación de las sustancias que permiten la producción de biogás a partir de residuos sólidos urbanos municipales. Master Thesis. Universidad Nacional Autónoma de México.

24.

Gutiérrez

(2010) Calidad y productividad. México, DF: McGraw Hill.

25.

Gutiérrez

Dorantes

(2017) Especies forestales de uso tradicional del estado de Veracruz: Zona Centro-Xalapa. CONAFOR – CONACYT–UV. Available at: http://www.verarboles.com/ZCentroXalapa/centroxalapa.html (accessed 16 January 2018).

26.

Hanníbal

Rafaela

Guevara

et al . (2016) Obtención de compost a partir de residuos sólidos orgánicos generados en el mercado mayorista del Cantón Riobamba. European Scientific Journal 12: 76–94.

27.

INEGI (2015) Encuesta Intercensal 2015. Instituto Nacional de Estadística y Geografía. Available at: http://www.beta.inegi.org.mx/proyectos/enchogares/especiales/intercensal/ (accessed 15 January 2017).

28.

INEGI (2016) Estadísticas a propósito del día mundial del medio ambiente. Instituto Nacional de Estadística y Geografía. Available at: http://www.inegi.org.mx/saladeprensa/aproposito/2016/ambiente2016_0.pdf (accessed 5 March 2017).

29.

Ionescu

Rada

Ragazzi

et al . (2013) Integrated municipal solid waste scenario model using advanced pretreatment and waste to energy processes. Energy Conversion and Management 76: 1083–1092.

30.

Jiménez

MNM

(2015) La gestión integral de residuos sólidos urbanos en México: entre la intención y la realidad. Letras Verdes. Revista Latinoamericana de Estudios Socioambientales 17: 29–56.

31.

Lisboa

Lansing

(2013) Characterizing food waste substrates for co-digestion through biochemical methane potencial (BMP) experiments. Waste Management 33: 2664–2669.

32.

Lombardi

Carnevale

Corti

(2015) Comparison of different biological treatment scenarios for the organic fraction of municipal solid waste. International Journal of Environmental Science and Technology 12: 1–14.

33.

López

Passeggi

Borzacconi

(2015) Variable kinetic approach to modelling an industrial waste anaerobic digester. Biochemical Engineering Journal 96: 7–13.

34.

López

Soliva

Martínez-Farré

et al . (2010) Evaluation of MSW organic fraction for composting: Separate collection or mechanical sorting. Resources, Conservation and Recycling 54: 222–228.

35.

Maalaouf

El-Fadel

(2017) Effect of a food waste disposer policy on solid waste and wastewater management with economic implications of environmental externalities. Waste Management 69: 455–462.

36.

Mao

Feng

Wang

et al . (2015) Review on research achievements of biogas from anaerobic digestion. Renewable and Sustainable Energy Reviews 45: 540–555.

37.

Massarutto

De Carli

Graffi

(2011) Material and energy recovery in integrated waste management systems: A life-cycle costing approach. Waste Management 31: 2102–2111.

38.

Menon

Wang

Giannis

(2017) Optimization of micronutrient supplement for enhancing biogas production from food waste in two-phase thermophilic anaerobic digestion. Waste Management 59: 465–475.

39.

Milutinović

Stefanović

Dassisti

et al . (2014) Multi-criteria analysis as a tool for sustainability assessment of a waste management model. Energy 74: 190–201.

40.

Mou

Scheutz

Kjedsen

(2014) Evaluating the biochemical methane potential (BMP) of low-organic waste at Danish landfills. Waste Management 34: 2251-2259.

41.

Mustafa

Calay

Román

(2016) Biogas from organic waste – a case study. Procedia Engineering 146: 310–317.

42.

Ogunjuyigbe

ASO

Ayodele

Alao

(2017) Electricity generation from municipal solid waste in some selected cities of Nigeria: An assessment of feasibility, potential and technologies. Renewable and Sustainable Energy Reviews 80: 149–162.

43.

Oviedo

OER

Domínguez

Torres

et al . (2016) A qualitative model to evaluate biowaste composting management systems using causal diagrams: A case study in Colombia. Journal of Cleaner Production 133: 201–211.

44.

Oyoo

Leemans

Mol

APJ

(2014) Comparison of environmental performance for different waste management scenarios in East Africa: The case of Kampala City, Uganda. Habitat International 44: 349–357.

45.

Panepinto

Blengini

Genon

(2015) Economic and environmental comparison between two scenarios of waste management: MBT vs Thermal treatment. Resources, Conservation and Recycling 97: 16–23.

46.

Parkes

Lettieri

Bogle

(2015) Life cycle assessment of integrated waste management systems for alternative legacy scenarios of the London Olympic Park. Waste Management 40: 157–166.

47.

Rada

Ragazzi

Villotti

et al . (2014) Sewage sludge drying by energy recovery from OFMSW composting: Preliminary feasibility evaluation. Waste Management 34: 859–866.

48.

Rajaeifar

Tabatabaei

Ghanavati

et al . (2015). Comparative life cycle assessment of different municipal solid waste management scenarios in Iran. Renewable and Sustainable Energy Reviews 51: 886–898.

49.

SAS (2018) Analytics Software & Solutions. SAS Institute Inc. Available at: https://www.sas.com/en_us/home.html (accessed 11 March 2017).

50.

SECOFI (1985a) Norma Mexicana NMX-AA-015-1985 Protección al ambiente-contaminación del suelo-residuos sólidos municipales-muestreo-método de cuarteo. Secretaria de Comercio y Fomento Industrial. Diario Oficial de la Federación. Last published reform 6 November 1992. Available at: http://biblioteca.semarnat.gob.mx/janium/Documentos/Ciga/agenda/DOFsr/NMX-AA-015-1985.pdf (accessed 10 August 2018).

51.

SECOFI (1985b) Norma Mexicana NMX-AA-061-1985 Protección al ambiente-contaminación del suelo-residuos sólidos municipales-determinación de la generación. Secretaria de Comercio y Fomento Industrial. Diario Oficial de la Federación. Last published reform 6 November 1992. Available at: http://biblioteca.semarnat.gob.mx/janium/Documentos/Ciga/agenda/DOFsr/DO3433.pdf (accessed 10 August 2018).

52.

SECOFI (1985c) NMX-AA-022-1985 Protección al ambiente-contaminación del suelo-residuos sólidos municipales-selección y cuantificación de subproductos. Last published reform 6 November 1992. Available at: http://biblioteca.semarnat.gob.mx/janium/Documentos/Ciga/agenda/DOFsr/NMX-AA-022-1985.pdf (accessed 10 August 2018).

53.

SECOFI (1984) Norma Mexicana NMX-AA-016-1984 Protección al ambiente-Contaminación del suelo-residuos sólidos municipales-determinación de Humedad. Secretaria de Comercio y Fomento Industrial. Diario Oficial de la Federación. Last published reform 14 December 1984. Available at: http://biblioteca.semarnat.gob.mx/janium/Documentos/Ciga/agenda/DOFsr/NMX-AA-016-1984.pdf (accessed 10 August 2018).

54.

SEDESOL (2015) Modulo Único de Clasificación (MUC) formulación del MUC. Lineamientos de evaluación de condiciones socioeconómicas de los hogares, anexo B. México, DF: Subsecretaría de planeación, evaluación y desarrollo regional. Dirección general de geoestadística y padrones de beneficiarios.

55.

SEDESOL (2017) Catálogo de localidades. Veracruz. Unidad de microrregiones. Secretaría de Desarrollo Social. Available at: http://www.microrregiones.gob.mx/catloc/Default.aspx?tipo=clave&campo=mun&valor=30 (accessed 17 January 2017).

56.

SEFIPLAN (2013) Programas Regionales Veracruzanos. Programa Región Capital 2013-2016. Veracruz: Secretaría de Finanzas y Planeación.

57.

SEDEMA (2015) Inventario de Residuos Sólidos CDMX-2014. Ciudad de México: Secretaría del Medio Ambiente del Gobierno de la Ciudad de México.

58.

SEMARNAT-INECC (2012) Diagnóstico básico para la gestión integral de los residuos. Instituto Nacional de Ecología y Cambio Climático. México, DF: Instituto Nacional de Ecología y Cambio Climático-Secretaría de Medio Ambiente y Recursos Naturales.

59.

Soliz

TMF

(2015) Political ecology and critical geography of waste in Ecuador. Letras verdes. Revista Latinoamericana de estudios Socioambientales 17: 4–28.

60.

Vucijak

Kurtagic

Silajdzic

(2016) Multicriteria decision making in selecting best solid waste management scenario: A municipal case study from Bosnia and Herzegovina. Journal of Cleaner Production 130: 166–174.

61.

Waqas

Nizami

Aburiazaiza

et al . (2017) Optimization of food waste compost with the use of biochar. Journal of Environmental Management. Epub ahead of print 18 January 2018. DOI: 10.1016/j.jenvman.2017.06.015.

62.

World Bank (2012) What a Waste. A Global Review of Solid Waste Management. Washington, DC: World Bank.

63.

Zhang

Banks

Heaven

(2012) Anaerobic digestion of two biodegradable municipal waste streams. Journal of Environmental Management 104: 166–174.

64.

Zhou

Selvam

Wong

JWC

(2018) Chinese medicinal herbal residues as a bulking agent for food waste composting. Bioresource Technology 249: 182–188.

Encoded variable	Versus	Cost
Y	X11	C1T
Y	X12	C1T
Y	X21	C2T
Y	X22	C2T
Y	X31	C3T
Y	X32	C3T
Y1	X11	C11
Y1	X11	C1T
Y1	X21	C21
Y1	X21	C2T
Y1	X31	C31
Y1	X31	C3T
Y2	X12	C12
Y2	X12	C1T
Y2	X22	C22
Y2	X22	C2T
Y2	X32	C32
Y2	X32	C3T

Encoded variable	Versus	Cost
Y	X11	C1T
Y	X12	C1T
Y	X21	C2T
Y	X22	C2T
Y	X31	C3T
Y	X32	C3T
Y1	X11	C11
Y1	X11	C1T
Y1	X21	C21
Y1	X21	C2T
Y1	X31	C31
Y1	X31	C3T
Y2	X12	C12
Y2	X12	C1T
Y2	X22	C22
Y2	X22	C2T
Y2	X32	C32
Y2	X32	C3T

Encoded variable	Versus	Cost
Y	X11	C1T
Y	X12	C1T
Y	X21	C2T
Y	X22	C2T
Y	X31	C3T
Y	X32	C3T
Y1	X11	C11
Y1	X11	C1T
Y1	X21	C21
Y1	X21	C2T
Y1	X31	C31
Y1	X31	C3T
Y2	X12	C12
Y2	X12	C1T
Y2	X22	C22
Y2	X22	C2T
Y2	X32	C32
Y2	X32	C3T