Introduction

 

The generation of urban wastes has increased tremendously in recent years, due to the rising population and industrialization around the world. Accumulation in landfills is a major issue, due to the limited number of available sites, the high cost, and the environmental restrictions (Bondarczuk et al., 2016). Thus, thermal treatment of these wastes for energy production, destroying pathogens and other pollutants, seems an attractive solution, in view also of global crisis. It has been estimated that urban wastes could contribute to energy consumption from 1 to 3 EJ/y worldwide (Vamvuka, 2009).

The majority of waste-to-energy systems are based on combustion. However, air pollution from incineration plants was the driving force for the development of gasification technologies, which offer the advantages of higher efficiency, reduced emissions, and greater flexibility in terms of scaling, feedstocks, and end products (Vamvuka & Sfakiotakis, 2015; Hla et al., 2016). A two-stage process in fixed bed appliances, where the first step is pyrolysis of wastes and the second step reaction of char with carbon dioxide as the gasifying agent, presents several benefits. Bio-oil and gases produced during the first step can be used as sources of chemicals and energy (Policella et al., 2019), leaving a highly reactive bio-carbon to be converted to fuel gas free of tar (Vamvuka & Sfakiotakis, 2015). Furthermore, the use of the principal constituent of industrial flue gases causing global warming, as a post-treatment technique, gives a potential solution to the greenhouse gas effect.

The reaction of char with carbon dioxide, the so-called Boudouard reaction, is a slow process proceeding at high temperatures. For the design of gasifiers, kinetic analysis is required. Nth order kinetics has been commonly applied for particle sizes below 1mm when the reaction is chemically controlled (Lahijani et al., 2015; Sadhwani et al., 2016). Various models have been employed, such as iso-conversional models (Zhao et al., 2017; He et al., 2021), the volume reactor model (Xu et al., 2016; Diedhiou et al., 2019), the shrinking core model (Sadhwani et al., 2016;  Xu et al., 2016; Diedhiou et al., 2019) and the random pore model (Xu et al., 2016; Diedhiou et al., 2019; He et al., 2021). Isoconversional methods introduce approximations, whereas the random pore model has been proven quite successful, however, it is very complex.

The reactivity of char in the above models accounts for temperature, partial pressure of reactant, and structural characteristics. Other factors found to affect reactivity are the conditions of the pyrolysis step (Min et al., 2011; Li et al., 2017) and the composition of fuel ash. Inherent mineral matter, such as K, Na, Ca, and Fe compounds, have been reported (Bouraoui et al., 2016; Naidu et al., 2016) to improve gasification reactivity. Thus, carbonates, oxides, or chlorides of these metals were used as effective external catalysts (Vamvuka et al., 2012; Lahijani et al., 2015; Sadhwani et al., 2016).

Urban solid waste streams are highly heterogeneous, making their thermal treatment difficult (Mohammadi et al., 2020). Their high moisture content and ash content, lowering the calorific value, and the presence of hazardous species (Vamvuka et al., 2018) may pose serious limitations in energy enterprises. Therefore, blending with higher quality woody residues is often mandatory, to accelerate their promotion for energy recovery. There is insufficient data throughout the literature on the gasification behavior of urban wastes or their kinetics, while there is a lack of information on their co-gasification with woody agricultural or forest residues. In a previous study (Kwon et al., 2010) the gasification of municipal solid wastes (MSW) under various concentrations of carbon dioxide was investigated and the effects of waste-to-polymer ratios on product yields were explored. Some authors (Arafat & Jijakli, 2013) developed a simple equilibrium model to examine the effects of operating temperature and different MSW compositions on the performance of the process, without considering reaction kinetics, whereas others (Hla et al., 2016) calculated intrinsic reaction rates. More recently, the characteristics of hydrogen and syngas formation were determined for MSW (Zheng et al., 2018) at different carbon dioxide/steam ratios and gasification temperatures.

According to the above discussion, the present study aimed to investigate the valorization of refused-derived fuel (RDF), produced from urban solid waste management units, for energetic uses, through its co-gasification with agro- or forest wastes and carbon dioxide agents. (Shihab et al., 2019). The thermal behavior, the reactivity, the conversion, and the cold gas efficiency were evaluated based on the structural characteristics and inherent minerals of the materials, as well as on the basis of alkali catalysts and bio-oils admixed. Kinetic analysis of the co-gasification process was performed by developing an independent parallel reactions model.

MATERIALS AND METHODS

Selected waste materials

Aiming to improve the gasification reactivity of the municipal solid waste material of this study, a variety of biomass fuels, abundantly found in Greece and Mediterranean countries in general, were selected. These were Sunflower or Helianthus annuus L. (SUN) a short rotation coppice crop provided by Pettas A.E. Company, one agro-industrial waste Cotton stems (CST) provided by a cotton ginning factory, as well as a forest waste Pine cones (PCO), collected from a local forest in Crete island. Refused Derived Fuel (RDF) was obtained from the solid waste management company of Chania, Crete, named DEDISA.

After air drying, all samples were riffled and ground in a cutting mill to a particle size below 1 mm. Representative samples were characterized in terms of proximate and ultimate analysis, following European standards CEN/TC335. Structural characteristics were determined according to the BET method, using an analyzer model Nova 2200 of Quantachrome. Prior to tests, materials were out-gassed overnight at 110 °C, under vacuum.

Mineral species were identified by an X-ray fluorescence spectrometer, model S2 Ranger/EDS of Bruker AXS.

Preparation of biochars

Raw materials and mixtures of agricultural and forestry wastes with RDF at mass ratios 70:30 were devolatilized in a fixed bed unit (Teftiki, 2021).  The reactor was charged with about 15 g of sample, flushed with nitrogen (flow rate 200 mL/min) for 30 min at room temperature, and heated up to 600^oC, with a heating rate of 10^oC/min. Volatiles were passed through isopropanol, ice-cooled baths, to collect liquid fractions, from which bio-oil was collected after centrifugation at 6000 rpm for 20 min from the top layer. Following a retention time of 30 min at 600^oC, the reactor was cooled under nitrogen and the solid biochar was weighed and characterized with the same methods as for raw fuels. Also, each biochar, before being used in the gasification tests, was ground to a particle size below 200 μm, to conduct the experiments under a kinetic regime.

Gasification tests

Gasification experiments were carried out in a TG/DTG thermobalance of Perkin Elmer, with temperature precision ±2 ºC and sensitivity <5 μg. About 15 mg of each biochar was gasified in a carbon dioxide atmosphere of flow rate 35 mL/min up to 950°C, with a heating rate of 10^oC/min, until a constant weight was reached. Two to three replicates confirmed the high reproducibility of the results, expressed by the relative standard deviation RSD reported below.

External catalysts were purchased by Sigma-Aldrich company and consisted of pure carbonates of calcium, potassium, and lithium. Calcium carbonate was calcined at 1000°C, to avoid interference with DTG curves at high temperatures, due to its decomposition above 700°C. Catalyst loading was performed according to the incipient wetness method and ranged from 0.1 to 0.5 gMe/gC in char (Vamvuka et al., 2012; Sadhwani et al., 2016). For mixtures of biochars with the bio-oils, a mass ratio of 1:1 was chosen.

From processing of weight loss (TG) and rate of weight loss (DTG) curves as a function of temperature, the initial (T_i) and peak inflection (T_max) temperatures, as well as the peak reaction rate (R_max) were determined, whereas gasification reactivity was defined according to the equation:

 

Rf=Rmax/(Tmax)x100 (min-1/°C)

(1)

 

Gasification efficiency was expressed by:

(2)

 

where m_CO and m_bioch are the masses of product gas CO and initial biochar, respectively.

 

The cold gas efficiency was calculated as follows:

(3)

 

where LHV_CO and LHV_bioch are the lower heating values of CO gas and biochar in MJ/kg, respectively.

Mathematical modeling of kinetics

The independent parallel reactions model (IPR) was developed for the kinetic analysis of the carbon dioxide gasification of the materials studied. The decomposition of the various pseudo-components was described by parallel reactions so that the overall conversion was the sum of all partial values for k reactions:

i=1,2,3…k

(4)

 

where c_i is the contribution factor of each pseudo-component.

 

The conversion rate of each reaction i was expressed as follows:

 

(5)

 

where α_i the degree of conversion, t the time, A_i the pre-exponential factor, Ei the activation energy, R the universal gas constant, T the absolute temperature, P_CO2  the partial pressure of carbon dioxide, ν the reaction order and f(α_ι) a function of fractional conversion reflecting the change of reactivity:

f(αi)=(1-αi)ni

(6)

 

 where n_i is the reaction order with respect to char conversion. 

The optimal parameters, that minimized the difference between experimental and calculated rate, were derived from the development of an algorithm in Matlab^® code (gradient-based minimization function fmincon):

 

(7)

 

 where (dα_ι/dt)_exp and (dα_ι/dt)_cal are the experimental and calculated conversion rates, respectively, and N is the number of experimental points.

 

The deviation between experimental and calculated DTG curves by the model was determined by the following equation:

 

(8)

 

where N is the number of data points and Z is the number of parameters employed in the model.

         

 

a)

b)

c)

d)

Figure 4. Modeling of gasification rate of (a) SUN (b) CST (c) PCO and (d) RDF biochars

Kinetic analysis was also performed for the gasification of biochars studied in presence of the most effective catalyst CaO as shown above, at a low percentage of 10% wt, as well as for the gasification of biochar/bio-oil blends at ratios 1:1. The kinetic parameters are included in Table 3. As can be noticed, both addition of external catalysts or bio-oil resulted in a significant reduction in activation energy, which for RDF material was 38%. Activation energy values of raw biochars were lowered from 148-252 kJ/mol to 138-201 kJ/mol.  On the other hand, reaction orders of raw biochars were increased from 0.3-0.5 to 0.6-1, revealing a greater dependence of gasification rate on conversion when external catalysts were used. Current data are in line with literature for chemically controlled gasification processes (Sadhwani et al., 2016; Edreis et al., 2018; He et al., 2021).

CONCLUSION

Conversion of RDF material to carbon monoxide at 950°C was 55%, while that of agricultural and forest wastes was 81-95%. Blending RDF with these wastes at mass ratios 30:70 increased its conversion up to 86.4%. Peak inflection temperature occurred between 890°C and 940°C. Gasification reactivity followed the order CST>SUN>PCO>RDF. In the case of CST biochar, it was correlated to its inherent alkali metals.

The addition of alkali compounds as external catalysts enhanced conversion at low loadings up to 100% and the process, in general, took place at much lower temperatures. The catalytic activity followed the order: CaO> Li₂CO₃>K₂CO₃. The addition of bio-oil at ratios 1:1 increased the conversion up to values 85-92 %. The gasification efficiency of RDF material blended with the woody fuels and 10% wt CaO as catalyst reached values 85-90%.

The kinetic model applied was proven very successful, with deviation values below 5 %. In presence of 10% wt, CaO the activation energy values of biochars were lowered from 148-252 kJ/mol to 138-201 kJ/mol, implying a better performance.

ACKNOWLEDGMENTS: The authors kindly thank the laboratories of Hydrocarbons Chemistry and Technology, Geochemistry and Ore Processing and Beneficiation of the Technical University of Crete for the ultimate, ash and BET analyses.

CONFLICT OF INTEREST: None

FINANCIAL SUPPORT: None

ETHICS STATEMENT: None