It is carried out in the range between 750 °C and 1000 °C to obtain heat and electricity and these generation processes can be combined [11]. A typical controlled incineration system (electricity and heat) is composed of a waste storage chamber, an incinerator/furnace, a vapor/generator turbine, a fuel-gas cleaning system and a waste treatment system. The calorific value of waste is an important parameter that greatly contributes to the efficiency of the incineration plant [5].

Incineration is a mature technology, used in several developed countries. France, for instance, widely uses incineration: in 2003, 12.6 million tons of non-hazardous waste were treated at 130 incineration plants. A total of 2.9 TWh of electricity were generated, and 9.1 TWh were consumed in the form of heat by private and public users [12].

China actively promotes the production of energy by incineration. In 2014, the country was building 75 plants to process 110,000 tons per day and have a total installed capacity of 2.2 GW. Germany has an incineration plant, property of the German Cleaning Company, capable of incinerating 520,000 tons per day and generate 188 kWh of electricity every year [13].

Some studies have evaluated the viability of obtaining energy from incineration in countries like Bangladesh, Nigeria, and KSA (Kingdom of Saudi Arabia) [14]–[16]. In [14], the authors present an overview of energy (electricity) generation potential from solid waste in mega-cities of Bangladesh. In [15], the WTE potential of municipal solid waste (MSW) was assessed throughout Nigeria using the population growth rate factor and the boiler, steam and overall efficiencies for calculating the exploitable energy potential. Other authors [16] evaluated the potential contribution of WTE facilities to the total Saudi peak power demand until 2032 by means of a quantitative analysis of six large cities. In that study, the MSW production rate was assumed to be an average 1.4 kg/capita/day. To calculate the total energy content per kilogram of municipal waste, the caloric energy content of the various types of waste and MSW contents were considered.

One of the greatest advantages of this process is that it can treat organic and inorganic waste [17]. Therefore, waste volume can be reduced up to 80%. The plant can be continuously fed, and the treatment is fast. The complexity of the plant is low; it can be installed in urban areas and meet all the technical and environmental regulations.

One of its drawbacks is that it is not viable to build plants to treat a volume lower than 100T of SW per day. In that case, the chemical production of dioxins and slag should be considered. Besides, this technique is not appropriate for SW with high water content [17]–[21].

Also known as biomethanation, this biological conversion technology transforms organic waste into liquid or gaseous fuels by means of biological reagents [18]. This process involves four stages: hydrolysis, acidification, acetogenesis and methanogenesis. It is carried out in a closed container (biogas digester), where bacteria ferment the organic material under oxygen-free conditions to produce biogas. Such biogas can be used in a boiler or alternative engine [25].

In Brazil, anaerobic digestion has been successful in producing electricity in small scale [26]. In Colombia, Chicón project in Chigorodó (Antioquia) was in its implementation stage in 2016. Such project seeks to produce 2 million m3 of biogas and 500 kW of electric power from 15,000 T/year of organic SW [27].

Some authors have evaluated the energy recovery potential of biogas from anaerobic digestion to generate electric or thermal power in Brazil, Tanzania, Algeria, Spain and China [26], [28], [29]. Another work [27] assessed a micro-co-generation unit located in a typical Brazilian agroenergy condominium that uses biogas (produced from waste by anaerobic digestion) as renewable fuel. In [28], the authors presented a strategy to utilize organic solid waste from the city of Dar es Salaam (Tanzania) for producing biogas and, ultimately, generating electric energy. This is called the Taka (waste) Gas Project. Additionally, some actions to implement the project and make it feasible are discussed. Regarding Asia, an analysis of the sustainability of current anaerobic digestion methods in China was introduced in [29]. On the other hand, in [30] the authors focus on the conversion of municipal solid waste to biogas as a local energy supply in urban areas. Three urban models were identified along with a matrix of the typical Organic Fraction of Municipal Solid Waste (OFMSW). In order to analyze energy generation, theoretical production and substrate composition are calculated.

Anaerobic Digestion is profitable and applicable to a production greater than 2 T/day of SW. However, the plant must only be fed the organic fraction, which means waste sorting is necessary. Nevertheless, the process avoids the release of greenhouse gases, its digestate is rich in nutrients and it can be used as organic fertilizer. Additionally, in small-scale plants, the co-digestion of the raw material or SW can be carried out with biosolids. In general, the resulting biogas must be treated for final use. Besides, the complexity of this type of plants is low and they are usually located in rural areas [17]–[21].

The decomposition of organic waste in garbage dumps is slightly similar to anaerobic digestion in biogas digesters. Microorganisms living in the organic material, such as residues of food and paper, cause decomposition as well as methane and carbon dioxide release. Landfill gas (LFG) is usually 50% methane and 50% carbon dioxide. Such biogas released by the dumps is gathered and burned to produce electricity. Generally, it is collected by pipes that reach the wells installed inside the landfills [5].

This technology has been successfully applied in countries like Brazil, where a potential of 660 MW from landfills was estimated in 2009. In fact, in 2014, 69 MW were produced from biogas recovered from landfills in São Paulo (11,244,369 inhabitants), Belo Horizonte (2,375,444), Salvador (2,676,606) and Uberlândia (619,536) [31]. Some studies have evaluated the electricity generation potential of landfill biogas [32]–[36]. In [32], the authors estimated the feasibility of LFG in a trigeneration scheme in Hong Kong. In [33], the authors investigated the potential for economically viable electricity generation by means of energy recovery from landfill biogas in Brazil. Moreover, in [34], the authors proposed to feed the national grid with a MSW power plant. In [35], the author presented a feasibility analysis of landfill gas recovery in Mexico. In [36], the authors evaluated the renewable energy potential of MSW and the environmental benefits of carbon reduction in Bangladesh using WTE strategies for urban waste management. The energy potential of different WTE strategies is assessed using a standard energy conversion model and a greenhouse gas (GHG) emissions model. The evaluation was conducted using a first-order decay model. Many of the works above employed the tool LandGEM to evaluate energy production potential.

Producing landfill biogas is a low-cost alternative to generate electric or thermal energy. However, its efficiency is limited to 30 or 40% of the generated gas. Since the natural resources are returned to the soil, swamps might become useful areas. The level of complexity of this kind of plants is low; for that reason, their operation does not require qualified staff. It also presents some disadvantages: it requires large areas to be operated, spontaneous combustion might occur as a result of the accumulation of methane gas, and surface runoff during rains causes the soil and groundwater to be contaminated with lixiviates [17]–[21].

To determine the most suitable technology for each scenario, some aspects should be considered: number of inhabitants, per capita waste production, and waste composition. The following section is a description of the method to select the study cases. Besides, it elaborates on how to select, form a technical point of view, the appropriate conversion technology for each case under analysis. For that purpose, the main advantages and disadvantages of such technologies are included.

To evaluate energy production, three cases were chosen in line with Law 388 of 1997 for Land-use Planning [37]. Section 9 therein defines the plans to be adopted by three groups [38]: Group 1 (G₁), municipalities with less than 30,000 inhabitants; Group 2 (G₂), municipalities with a population between 30,000 and 100,000; and Group 3 (G₃), more than 100,000 inhabitants.

The population forecast for 2015 to 2020 is based on information obtained from the 2005 census (DANE) [40].

To select one municipality per group, the rural/urban ratio (Urban Population - UP) of each population was evaluated. Regarding this parameter, a trend was found in each group: G₁, predominantly rural population (UP>1); G₂, balanced rural and urban population (0.9<UP<1.1); and G₃, predominantly urban population (UP<1).

This classification allows to consider only the municipalities that exhibit said characteristic in each group. Also, it enables to analyze three scenarios with different waste production scales and composition. Consequently, the range of possible conversion technologies for each scenario may vary.

Another selection criterion was the availability of the Urban Solid Waste Management Plans (SWMPs). The municipalities that had not yet developed their SWMP or whose documentation was not publicly accessible were discarded. Besides, if a municipality did not provide sufficient information (physical composition, mass fraction, quantity and features), it was not included.

The information in the SWMPs regarding the amount, physical composition and per capita generation of solid waste was analyzed. By pondering the main advantages and disadvantages of both technologies, it was possible to recommend the most appropriate option (from a technical standpoint) in the three scenarios.

Energy recovery potential (ERP) was calculated following the mathematical models in Section 3.3. They are based on the efficiency of the technology, as well as the mass and the lower calorific value (LCV) of the SW [40], [41]. The mass depends on per capita generation, which is determined by the number of inhabitants in different population scenarios. The model used in [40] for incineration, and given in (1), depends on the LCV of the total generated organic and inorganic SW. The authors in [42] used model (2) for anaerobic digestion. Such model depends on the LCV of methane, the principal component of biogas derived from the fermentation of organic waste. Finally, incineration and anaerobic digestion were recommended, as described in Section 4.2.

Lower Calorific Value (LCV) of waste: The ERP from incineration depends on the LCV of the total waste. After the physical composition of the SW of each scenario is established, the total LCV can be estimated with the weighted LCV per kind of waste and its mass fraction.

The typical LCV of each component of SW was taken from the Guidebook for the application of waste to energy technologies in Latin America and the Caribbean [43]. Such document presents study cases in Buenos Aires (Argentina), Toluca (Mexico) and Valparaíso (Chile), and it suggests that the MSW in these regions presents a similar physical composition [44].

The LCV of each type of waste was compared to other studies conducted in Saudi Arabia and Spain. At the domestic level, they were also contrasted with the LCVs obtained from the chemical characterization by Empresas Varias de Medellín (EEVVM) in 2009 [45]. Those LCVs are presented in Table 1. In this study, the LCVs from Buenos Aires, Toluca and Valparaíso were used for the calculations of the selected scenarios. Table 1 shows that the values reported by EEVVM are even higher to those used in [44]. Consequently, the estimate was made with values below those reported by EEVVM, which reduces uncertainty in the calculation of the ERP.

Table 1. Table 1. LCV (MJ/kg) reference values per type of waste.

Type of waste	Saudi Arabia [41]	Argentina Chile Mexico [44]	EEVVM [46]	Spain [43]
Paper and cardboard	15.82	15.6	23.23	10.05
Assorted foods	5.58	4.6	6.97	2.72
Assorted plastics	32.56	32.4	37.17	35.22
Fabrics	18.84	18.4	18.58	14.35
Wood	15.12	15.4	18.58	13.58

Source: Authors.

Incineration: Equation (1) is the expression to calculate the amount electricity that can be obtained from incineration [41]. An 18% efficiency was applied in this case [45]:

(1)

Anaerobic digestion: This is the preferred process for the organic fraction of MSW, which allows the activity of microbes in presence of humidity. The expression for calculating the electricity generation potential of the total organic fraction of the MSW is given in (2). The efficiency of the process was 26%, which corresponds to a reciprocating internal combustion engine in the literature [42].

(2)

Although anaerobic digestion is carried out under controlled operation conditions, different values of methane generated from the OFSW have been reported. For the present study, 71 Nm³/T was selected, assuming a 55.5% methane in the biogas [46]. The literature reports biogas performance values from 67.5 to 122 Nm³/T of organic fraction of waste [42].

The results obtained with the methods described above suggest selecting a scenario for each group (G₁, G₂ and G₃). In Colombia, 78% of the municipalities are in G₁; 16.4%, in G₂; and the remaining 5.6%, in G₃. The population projections for the selected scenarios were taken from the census conducted by the National Statistics Office (Departamento Nacional de Estadísticas, DANE); they are available until 2020 only.

Scenario 1: G₁ is composed of 870 municipalities with less than 30,000 inhabitants. The analysis of this group revealed that 52% of the municipalities in it (470) have less than 10,000 inhabitants. Therefore, the scope was narrowed down to those 470 municipalities. Subsequently, the UP was evaluated as explained in Section 3. In this group, such index exceeded 1 (predominantly rural); thus the search was reduced to 357 municipalities.

Later, the municipalities of Guayatá, Pinchote and Villa Caro were found to report sufficient information on the characteristics of their SW production in their SWMPs. Therefore, they were preliminarily selected for the evaluation. A detailed revision of the information revealed that Guayatá (UP = 2.94) has a complete SWMP; therefore, it was finally selected as the scenario for G₁ [47].

Fig. 1 shows the projected population of Guayatá for the 2015-2020 period; the rural and urban proportions are differentiated. A slight decrease in both populations can be observed during that period, but it is more noticeable in the rural group. By 2020, the UP will be 2.69. Therefore, Guayatá will maintain its urban-rural ratio over 1.

Fig. 1. Fig. 1. Population of Guayatá projected for the 2015-2020 period. Source: Authors.

Scenario 2: G₂ is composed of 183 municipalities. Most of them are concentrated in the range between 30,000 and 50,000 inhabitants (117 municipalities). Out of these, 15 present a UP between 0.9 and 1.1 (i.e., a proportional relationship between rural and urban population).

In this group, official information was only found for the Municipality of Andes. Their waste management plan describes the physical composition, mass fraction and SW generated per capita [48]. Andes’ UP is 1.02. By 2020, this figure will come down to 0.96. This decrease indicates a more significant growth of the urban area.

The demographic growth in Andes projected for the 2015-2020 period is shown in Fig. 2. Although the urban population is expected to grow significantly (almost 1,700 inhabitants), the rural share will stay steady with about 220 new inhabitants. This is a positive indicator because the projections of production of solid waste (therefore, electric energy) rise.

Fig. 2. Fig. 2. Population of Andes projected for the 2015-2020 period. Source: Authors.

Scenario 3: G₃ is composed of 62 municipalities. Out of these, 60 present a UP below 1 (predominantly urban). Most populations in this group were found to be in the range bet ween 100,000 and one million inhabitants.

In this subgroup, the required information on physical composition, mass fraction and SW generated per capita was obtained from the Municipalities of Pasto and Pereira. Finally Pasto, with a 0.2 UP, was chosen because it had a complete SWMP [50]. The other municipalities with a UP below 1 provided little information or, in many cases, their SWMP was not officially published.

The demographic growth in Pasto projected for the 2015-2020 period is shown in Fig. 3. This municipality presents an increase of about 26,000 inhabitants in the urban area. In contrast, the rural component does not exhibit a significant expansion; this is, its rural population remains constant. By 2020, their UP will be 0.19, which indicates a slight growth of the urban area.

Fig. 3. Fig. 3. Population of Pasto projected for the 2015-2020 period. Source: Authors.

Fig. 4 shows the geographic location of Guayatá, Andes and Pasto (Scenarios 1, 2 and 3, respectively) in the map of Colombia.

Fig. 4. Fig. 4. Geographic location of the municipalities selected as Scenarios for groups G1 (Guayatá), G2 (Andes) and G3 (Pasto). Source: Authors.

LCV of waste: Based on the physical composition of the SW, the total LCV can be estimated from the weighted LCV of the mass fractions of each type of waste. The typical LCV of each component of the SW (reference LCV) was taken from the Guidebook for the application of waste to energy technologies in Latin America and the Caribbean [43], [44]. The composition and LCV of the waste generated in the three scenarios can be observed in Table 2.

Table 2. Table 2. Total LCV, Reference LCV and MF in three scenarios.

	Reference LCV [MJ/kg] [44]	MF [%]
		Guayatá[48]	Andes [49]	Pasto [50]
Paper and cardboard	15.6	12.4	7.94	8.31
Assorted food waste	4.6	51.4	60.7	70
Assorted plastics	32.4	12.7	2.16	8.57
Fabrics	18.4	0.7	__	1.41
Wood	15.4	1.2	__	0.73
Total LCV		8.73	4.73	7.66

Source: Authors.

The recoverable fraction and per capita waste generation reported in their solid waste management plans can be observed in Table 3. The Municipalities of Pasto and Andes lack per capita production indicators for rural areas. In the waste management plans of Guayatá, Sabaneta and Medellín, these values are 0.3, 0.28 and 0.27 kg/inhab-day, respectively. Since these numbers are similar in municipalities with different populations, the average among the reported values was taken: 0.28.

Table 3. Table 3. Per capita waste generation [kg/inhab-day] in each scenario.

	Guayatá	Andes	Pasto
Urban	0.48	0.48	0.55
Rural	0.30	0.28	0.28
Recoverable mass [%]	78.40	70.81	89.02

Source: Authors.

This section presents the selection of the most adequate technology for each scenario. In general, incineration (thermal conversion) and anaerobic digestion (bioconversion) were found to be the most appropriate options considering their advantages and disadvantages.

Thermal conversion:

Since gasification has rarely been implemented for processing MSW at the international level and due to its complexity, it was not taken into account in this evaluation.

Incineration is a widely used technology in SW urban processing and its level of complexity is low. As a result, it was considered to be applicable to all three scenarios.

However, inconveniences arise from incineration when it is applied to Scenario 1 Guayatá (5,126 inhabitants) because of its low total waste production (1.74 T/day). Furthermore, the operating and maintenance costs are high for a small power station of this type [20].

Nevertheless, it is applicable to Andes and Pasto, which have 45,184 and 440,040 inhabitants, respectively. Currently, these municipalities produce 17.62 and 225 T of waste/day. The study in [19] revealed that, above 100 T/day, incineration can be implemented by means of a circulating fluidized bed. This technology is already being commercialized and can be adapted to SW with low calorific value.

Based on the above, we can conclude that incineration is an alternative for municipalities in G₂ and G₃. However, there might be issues with technology transference and scalability in Guayatá (G1). The daily waste production capacity in that population is way below the one reported by other processes worldwide [19], [50]–[53]. Although Andes produces a low amount of waste with 45,184 inhabitants, its population and waste generation are expected to grow.

Biological conversion technologies: Even though landfill gas offers advantages such as the low cost of investment and collected waste, the latter must be properly stored and covered, thus generating additional expenses [24]. Besides, the generated biogas’ recovery rate might be less efficient, as in the case of anaerobic digestion [54].

Anaerobic digestion was selected because it is carried out under controlled temperature, humidity, pH and oxygen-free conditions, in digester tanks [25].

This technology is applicable to the three scenarios because the produced waste is organic (see Table 2). In addition, there are other well-known technologies that can be implemented to generate electric energy on a small scale (30 kW) [55]. Anaerobic digestion offers positive environmental benefits, such as controlling the emission of greenhouse gases.

In conclusion, anaerobic digestion is technically viable for the three communities under study because it can be implemented on small and large scales. This fact facilitates its acquisition for producing electrical energy.

Once the technologies were selected for each scenario, the electrical energy potential that can be recovered from them was estimated. The results are presented below.

Incineration: To estimate the ERP from incineration, the LCVs of the SW generated in Scenarios 2 and 3 were calculated (Table 2). Furthermore, the LCVs reported in [44] were considered for each type of waste. Such LCVs have been employed in studies on other cities in Latin America, as mentioned in Section 4.1.

Most waste is organic biodegradable material, followed by plastics or paper and cardboard. Fig. 5 presents the daily electrical energy production from incineration in Andes, which applies the model described in Equation (1). It can be observed that, by 2020, up to 4.34 MWh/day might be obtained.

Fig. 5. Fig. 5. Electrical energy production from incineration in Andes. Source: Authors.

Fig. 6 shows the electrical energy production in Pasto based on the same model. It can be seen that, by 2020, it would be possible to recover 90.41 MWh/day. This city has a greater energy recovery potential because of its larger population. Energy production is directly proportional to the number of inhabitants, and it increases or decreases according to the projected population growth.

Fig. 6. Fig. 6. Electrical energy production from incineration in Pasto. Source: Authors.

In the case of Guayatá, the energy production potential from incineration was not evaluated (Section 4.2).

Anaerobic digestion: Fig. 7, 8 and 9 detail the projections of electrical energy generated from anaerobic digestion for the 2015-2020 period. Generation in the Municipality of Guayatá will be low and range between 0.12 and 0.14 MWh/day during the 2015-2020 period, as it can be seen in Fig. 7.

Fig, 7. Fig, 7. Electrical energy production from anaerobic digestion in Guayatá. Source: Authors.

Fig. 8. Fig. 8. Electrical energy production from anaerobic digestion in Andes. Source: Authors.

Fig. 9. Fig. 9. Electrical energy production from anaerobic digestion in Pasto. Source: Authors.

Fig. 8 and 9 suggest that, by 2020, energy generation in Andes and Pasto will increase every year up to 1.23 and 18.25 MWh/day, respectively.

The LCV of the biogas used to evaluate the mathematical models was 5.97 kWh/m³ (21,51 MJ/m³), which corresponds to the study by [56] (See Section 3.3).

Based on these results, by 2020, a total 2,829,000 kWh/month would be obtained from incineration in Pasto and Andes. If a four-person household is assumed to consume 145KWh/month, the average demand of 19,510 households could be met.

On the other hand, if energy was obtained from anaerobic digestion, a total 579,000 kWh/month could be recovered in the three communities. This supply could satisfy the average demand of 3,900 4-person households.

These estimates can illustrate the potential and impact of waste to energy technologies on such municipalities. Additionally, environmental and waste management benefits should be considered before evaluating this type of technologies.