Prediction of Chemical Composition of Gas Combustion Products from Thermal Waste Conversion

Abstract

The current global energy crisis is driving the need to search for alternative raw materials and fuels that will be able to ensure the continuity of strategic industries, such as the steel industry. A chance to reduce the consumption of traditional fuels (e.g., natural gas) is to utilise the potential of gases from the thermal conversion of waste, and, in particular, pyrolysis gas. Unfortunately, despite its high calorific value, this gas is not always suitable for direct, energy-related use. The limitation is the type of waste subjected to pyrolysis, particularly plastics, rubber and textiles. Due to the above, this article proposes the co-combustion of pyrolysis gas in a ratio of 1:10 with natural gas in a pusher reheating furnace employed to heat the charge before forming. The chemical composition of flue gases generated during the combustion of natural gas alone and co-combustion with pyrolysis gas from various wastes was modelled, namely, two types of refuse-derived fuel (RDF) waste, a mixture of pine chips with polypropylene and a mixture of alder chips with polypropylene. The calculations were performed using Ansys Chemkin-Pro software (ver. 2021 R1). The performed computer simulations showed that the addition of pyrolysis gas for most of the analysed variants did not significantly affect the chemical composition of the flue gases. For the gases from the pyrolysis of biomass waste with the addition of polypropylene (PP), higher concentrations of CO and H₂ and unburned hydrocarbons were observed than for the other mixtures. The reason for the observed differences was explained by conducting a formation path analysis and a sensitivity analysis for the selected combustion products.

1. Introduction

One of the main branches of industry, the steel industry, is of key importance for the economies of all countries; at the same time, it is one of the most energy-intensive and highest emission-producing industries [1,2]. Despite this, its products are among the basic materials used by most sectors of the world economy. The price and quality of these products largely determine the level of competitiveness in world markets [3,4]. In order to meet high economic and ecological requirements, the Polish steel industry is forced to modernise its production processes. It is particularly important to search for new methods of reducing pollutant emissions for all processes in technological chains; this applies primarily to steel mills with a full production cycle [5,6,7]. The currently used iron and steel production technologies, despite many modern investments, are still characterised by a negative impact on the environment, which especially concerns CO₂ emissions and energy consumption [8,9]. Therefore, new solutions are sought, which include innovative technologies developed within international programmes for reducing CO₂ emissions. Noteworthy are the optimisation activities carried out in many global steel mills, which result not only in reduced fuel consumption, but also in reduced pollutant emissions. An example of such a solution is the appropriate selection of combustion process parameters (in particular, the excess air ratio and the temperature of heated air), the use of low-emission burners and the management of so-called production gases in the combustion process, especially those rich in hydrogen [10,11,12,13].

The steel industry, apart from high pollutant emissions, is also characterised by high energy intensity [14,15,16,17]. Considering the fact that improving energy efficiency is one of the priorities of the EU’s energy policy, Poland has also declared a national target in this respect, i.e., 23% by 2030 in relation to the primary energy forecasts from 2007 [18,19]. The potential for improving energy efficiency covers almost the entire Polish economy, but significant benefits are seen in improving energy-intensive processes in steel production.

An opportunity to improve energy efficiency, aimed at reducing fuel consumption, is to utilise the potential of post-process gases, such as coke oven gas and gases from thermal conversion of waste [20,21]. Particular attention should be paid to the gaseous products of the pyrolysis process, namely pyrolysis gas, which currently has no reliable or permanent method of utilisation [22,23,24]. Depending on the material subjected to pyrolysis, pyrolysis gas mainly consists of compounds such as methane, hydrogen, carbon dioxide, carbon monoxide and C2–C6 hydrocarbons [25,26,27]. The yield of pyrolysis gas can even reach over 50% and depends on both the properties of the processed material and the process conditions, in particular the temperature and residence time [28,29,30].

This article presents a promising approach to the energy use of thermal conversion gases, which is based on the possibility of co-combusting pyrolysis gas (from biomass, PP and RDF) with natural gas in a pusher-type metallurgical furnace.

1.1. Energy Potential of Thermal Waste Conversion Products

1.1.1. Gaseous Products

Considering the depletion of fossil fuel resources, all types of waste are becoming an attractive and ecological alternative for obtaining energy. The ongoing energy crisis is deepening as a consequence of the geopolitical situation in Europe and the Middle East. Society and the largest economies of the world face the challenge of creating a new, reliable and stable source of energy that will not exploit the earth’s fossil fuel resources [31]. Numerous scientific publications prove [32,33,34,35,36] that the pyrolysis process, depending on the material subjected to thermal treatment, can be a source of valuable fuels and, in particular, pyrolysis gas [37,38,39,40]. The obtained gaseous products are rich in gases such as H₂, CH₄ and CO (

Table 1. Composition of pyrolysis gas obtained from various wastes (literature data).

Compound (vol%)	H2	CO	CO2	CH4	C2–C5	Calorific Value (MJ/m3)	Ref.
LDPE:SS	25.57	2.71	4.14	46.31	21.27	43.82	[31]
Tyre	21.50	5.10	26.20	17.30	29.90	30.30	[42]
RDF	29.40	11.30	13.20	32.90	7.20	22.80	[43]
RDF	12.37	29.69	11.89	17.83	25.70	29.94	[44]
Biomass	0.02	6.70	6.80	0.69	0.91	-	[45]

LDPE—low-density polyethylene rejects; SS—sewage sludge.

), which are the basic components of fossil fuels [41].

The literature data indicate that pyrolysis gas may differ significantly in composition and calorific value, depending on the material subjected to pyrolysis. The calorific value of the obtained gas fraction may range from 10 to 15 MJ/m³ for gases from biomass pyrolysis and reach a value of over 40 MJ/m³ for gas from the pyrolysis of various plastics. Considering the calorific value of natural gas, which is about 35 MJ/m³, and the calorific value of coke oven gas, which reaches approximately 17 MJ/m³, it can be stated that gaseous pyrolysis products can be a good alternative to the combustion of natural gas and coke oven gas after meeting appropriate conditions [24,46].

1.1.2. Solid Products

The key parameters that affect the properties of solid pyrolysis products are the process temperature, heating rate, residence time, additives and catalysts, in addition to the composition of the basic raw material that is subjected to pyrolysis. Biochar is very often characterised by a porous structure, high absorption properties and stable chemical properties [47]. Biochar consists mainly of coal, ash, volatile substances and water, with ash constituting the largest part, namely over 80%. Depending on the input material, transition metal compounds may be present in the ash, which have very good absorption capacities [48]. Example compositions of biochar are presented in

Table 2. Examples of compositions of solid fractions from various materials (literature data).

Material (wt%)	C	H	N	S	O	Calorific Value (MJ/kg)	Ref.
Tyre waste	80.82	1.46	0.53	2.41	-	30.0	[42]
Pear wood	84.00	1.91	0.65	0.01	7.26	33.55	[49]
Virginia mallow	49.90	5.69	0.39	0.10	37.42	18.51	[50]

.

As presented in Table 2, the main component of the solid fraction is carbon. Its content ranges from nearly 60% for Virginia mallow to over 80% for pear wood and tyre waste. The conditions of the pyrolysis process have a huge impact on the content of the carbon element in biochar. The higher the temperature, the more carbon is bound in the solid fraction. The carbon content in the solid product is closely related to its calorific value; the higher its content, the higher the calorific value. The calorific value of biochar ranges from 18 to 36 MJ/kg, where, for comparison, the calorific value of hard coal is in the range of 22–24 MJ/kg, the calorific value of brown coal is 7.5–21 MJ/kg and that of eco-pea coal is 24–26 MJ/kg [46,47,51].

1.1.3. Liquid Products

The third product, in addition to the gaseous and solid fractions, produced in the pyrolysis process is pyrolysis oil. It is characterised by a high content of aromatic hydrocarbons. Depending on the material subjected to pyrolysis, the optimum temperature that allows the process to be directed towards the production of pyrolysis oil is about 500 °C [42]. The type of input material has the greatest influence on the chemical composition and calorific value of the oil fraction (

Table 3. Composition of liquid fraction from various wastes (literature data).

Material	Elemental Composition (wt%)						HHV (MJ/kg)	Ref.
	N	C	H	O	Cl	S
PE	0.30	84.28	13.79	1.63	n.d.	n.d.	48.06	[41]
PVC	0.11	56.88	8.76	34.09	0.16	n.d.	18.45
LDPE:SS	0.60	80.40	14.50	n.d.	n.d.	<0.3	37.65	[31]
Olive oil residue	1.64	70.24	8.42	19.70	n.d.	n.d.	32.36	[52]
Waste tyres	0.60	88.10	10.70	0.06	n.d.	0.60	43.30	[53]

PE—polyethylene; PVC—polyvinylchloride; LDPE—low-density polyethylene waste; SS—sewage sludge.

) [22].

As can be seen, the calorific value of all the pyrolysis oils listed in Table 3 differs within a wide range. The lowest value was obtained from PVC, namely 18.45 MJ/kg, while the highest was 48.06 from PE, which exceeds the calorific value of crude oil [31]. The organic fraction of the liquid phase contains chemical substances such as ethylbenzene, styrene and toluene, whose calorific value in the pure form reaches 33–40 MJ/kg, which positively affects the high calorific value of the obtained pyrolysis oils [54].

1.2. Utilisation of Gases from Thermal Conversion

As shown above in Table 1, Table 2 and Table 3, the energy potential of thermal waste conversion products is very high. In many cases, the calorific value of the obtained products exceeds the calorific values of fossil fuels. The interest in the utilisation of gaseous thermal waste conversion products is evidenced by numerous scientific publications [55,56,57,58]. Hossain et al. [24] conducted studies of pyrolysis gas combustion in spark-ignition engines. The obtained results showed that it is possible to obtain the same power as when operating on natural gas, but the engine operation was less stable. The key parameter may also be the engine failure rate, which cannot be precisely determined after 100 h of engine operation on a given fuel. Liu et al. [23] subjected sewage sludge to pyrolysis. They determined that the optimal conditions for the process in their case were a temperature of 450 °C and residence time of 30 min. Such parameters allowed them to achieve a yield of about 52.36% for biochar and 46.40% for pyrolysis gas. The researchers proposed using the resulting pyrolysis gas for direct combustion, and the resulting flue gases could be utilised to dry wet sewage sludge or generate steam. Mouneir et al. [22] analysed the use of the energy potential of pyrolysis gas obtained from used tyres. The authors proposed utilising the obtained gas to power the pyrolysis process in order to reduce or completely eliminate the need to supply a flammable medium from outside. The researchers found that pyrolysis gas has a huge energy potential, but additional attention should be paid to the quality of the obtained products, the level of sulphur content and detailed monitoring of pollutant emissions. Pessoa Filho et al. [59] aimed to determine the optimal conditions of the pyrolysis process to obtain a high share of pyrolysis gas from the pyrolysis of solid plastic waste, and then simulated the combustion of the gas. The authors proposed a burner concept that had an impact on the reduction of CO_x and C_xH_y emissions. The obtained results showed that the lower calorific values of the received gases were similar to the values of commercial combustible gases. Additionally, the orientation of pyrolysis towards gas production and the selection of appropriate parameters allowed 98.01 vol% pyrolysis gas to be obtained from HDPE pyrolysis. As proven by numerous works [60,61,62,63], the yield of individual products in the pyrolysis and gasification process depends on a number of different parameters, such as the process temperature, pressure, catalysts, residence time and the material subjected to thermal decomposition [59]. The yield of individual pyrolysis products for selected wastes is presented in

Table 4. Yield of pyrolysis products for selected waste.

Material	Product Yield (vol%)			Ref.
	Biochar	Bio-Oil	Gas
Pear wood	30.10	21.10	34.40	[49]
Tyre	44.20	38.30	17.50	[42]
Sewage sludge	54.53	18.62	26.85	[64]
HDPE	0.88	1.11	98.01	[59]

HDPE—high-density polyethylene.

.

In the literature, in addition to the energetic utilisation of pyrolysis gas, one can find publications on the utilisation of synthesis gas from gasification. Nevertheless, it should be emphasised that gas from gasification has a much lower calorific value (4–15 MJ/m³) compared to pyrolysis gas; hence, its energetic utilisation has a much smaller application potential. An analysis of sources in the literature shows that numerous studies are being undertaken in the world on the utilisation of pyrolysis gas but, to date, no technology has been implemented on a large scale. The methods of utilising pyrolysis gas proposed by researchers are mainly employed on a pilot or laboratory scale.

The problem of managing the environmentally burdensome gaseous products of the thermal conversion of selected calorific wastes is of an applied nature, as evidenced by the interest of both waste and metallurgical entrepreneurs, for reasons including the following:

• The possibility of managing environmentally hazardous plastic waste, RDF waste and post-production bio-waste, as well as the gaseous products of the thermal conversion of these wastes;
• The possibility of reducing the use of natural gas to fuel heating furnaces, which will translate into economic benefits.

Therefore, there is an urgent need to develop this research towards selecting the most effective and economically viable solution.

While pyrolytic gas has been studied as an alternative to fossil fuels in various contexts, the specific application within the iron and steel industry, particularly in heating furnaces, remains significantly underexplored. To our knowledge, no prior research has addressed the use of pyrolytic gas for furnace heating in this sector.

Additionally, our research demonstrates how the co-firing of pyrolytic gas with natural gas can lead to substantial natural gas savings. Even a small fraction of pyrolytic gas in the fuel mixture results in significant reductions in natural gas consumption, given the high energy demands of the heating process in steel production. This leads not only to cost reductions but also to lower carbon emissions and improved energy efficiency.

Thus, the incorporation of pyrolytic gas addresses a key gap in both academic research and industrial practice by highlighting the potential economic and environmental benefits of reducing natural gas dependency in steel production.

This paper is organised as follows. Section 2 discusses the materials and methodology used in the research conducted in this paper. Section 3 presents and discusses the results of computer simulations obtained in the study. Section 4 is a discussion of the results and a summary of all the observed observations.

2. Materials and Methods

The work carried out for the purposes of this article was divided into two stages: experimental preliminary studies, performed in a laboratory chamber, and computer simulations performed using the Ansys Chemkin-Pro computer programme (ver. 2021 R1). The scheme of the conducted research work is shown in Figure 1.

In the first stage, in the experimental part, the scope of the research included determining the temperature profile and concentration of selected pollutants, namely NO, during the combustion of natural gas in the heating chamber. In the second stage, computer simulations were carried out by means of the Ansys Chemkin-Pro programme (ver. 2021 R1), within which the chemical composition of the following products was modelled:

• Natural gas combustion.
• Co-combustion of natural gas with pyrolysis gases from the thermal conversion of various wastes.

The calculations were performed for the heating zone of a pusher furnace for hot forming, located in one of the steelworks in Poland, in the Heavy Plate Rolling Mill. The results obtained from the modelling were verified by comparing the calculated NO concentrations during the combustion of natural gas with the results obtained under experimental conditions.

2.1. Material

Four gases were selected for computer simulations from the thermal conversion of two types of RDF and a mixture of alder chips with polypropylene and a mixture of pine chips and polypropylene. The main reason for choosing this waste combination is the local circumstances and waste availability. Both pine and alder waste are sawmill and furniture industry by-products. They can be successfully used for energy purposes. The problem, however, is with polypropylene waste, which is not recyclable due to contamination. The properties of plastics and their high calorific value limit their thermal conversion. RDF in Poland does not have a permanent customer, which means that huge quantities are stored. The mixture of the above-mentioned wastes proposed in the article, among other reasons due to its calorific value, enables them to be thermally converted and the products of this process are suitable for further use. Moreover, the addition of polypropylene waste to pine and alder waste allows one to minimise the effects of incorrect segregation and managing problematic waste.

2.2. Modelling Procedure in Ansys Chemkin PRO (Ver. 2021 R1)

Computer calculations were carried out using the licenced CHEMKIN-PRO software (ver. 2021 R1). A detailed GRI-Mech 3.0 chemical mechanism developed by the University of California, enriched with C1-C6 hydrocarbons, including kinetic data of more than 60 compounds and nearly 350 chemical reactions, and thermodynamic and transport data, was implemented for the calculations. In addition to the elements and compounds involved in the combustion process, the kinetic data file contained a set of reactions together with values to calculate reaction rate constants from the Arrhenius Equation (1) [65]:

$$k=A{T}^b\exp \left(-{\displaystyle \frac{E}{RT}}\right)$$

(1)

where

A—pre-exponential multiplier (pre-exponential constant).

b—temperature exponent.

E—activation energy, J/mol.

R—universal gas constant = 8.314 J/(mol K).

T—absolute temperature in the reaction zone, K.

The programme used for modelling is used for the chemical analysis of phenomena and processes occurring during the thermal conversion of fuels and waste, and the model adopted for the calculations does not take into account the phenomenon of diffusion. At the core of the model, there is a chemical mechanism that has been implemented from The Creck Modelling Group. The chemical mechanism is based on the Arrhenius reaction rate constant equation.

For the calculations, a model was used based on the condition of ideal mixing of the reactants, the so-called “Perfectly Stirred Reactor”, and a model with a free-spreading flame reactor, “The Freely Propagating Flame Reactor”. The first model was used for combustion calculations in the burner, while the second model was used in the test chamber. Based on the results from the preliminary tests carried out in the laboratory chamber, the necessary data were obtained to formulate the boundary conditions in the modelling procedure. The obtained results provided information concerning temperature distribution and flow parameters, such as pressure and amount of media fed.

Selecting the reactor was also based on the experience of other researchers who have successfully used this type of reactor for modelling thermal conversion processes of solid fuels and waste. It is noteworthy, as pointed out in their paper by Che et al. [66], that great advances in the numerical simulation of thermal conversion technology for solid fuels and waste have been made by the Huazhong University of Science and Technology State Key Laboratory of Coal Combustion. They used the HSC chemistry package and the PSR model to simulate the thermodynamic equilibrium and dynamic equilibrium of the palm oil waste pyrolysis process, obtaining very good conformity with the experimental results.

The chemical mechanism adopted for the calculations has been repeatedly used by Ranzi, Faravelli and Frassoldati [67,68,69,70], as well as by the authors of this paper, to model the chemical kinetics of combustion, pyrolysis and gasification of various biomass species. It is noteworthy that the mechanism was validated with the results obtained from the experiments, which increases the reliability of the results obtained from the calculations. For the calculations, the assumption was made that combustion takes place in a reactor with perfect mixing of the reactants, i.e., a Perfectly Stirred Reactor (PSR). A model of the conducted calculations is shown in Figure 3.

2.3. Preliminary Research

The preliminary tests were carried out on the experimental stand presented in Figure 2. The main element of the test stand was a high-temperature, cylindrical heating chamber (1), with a diameter of 0.12 m and a length of 3 m, made of a ceramic pipe. In order to reduce heat losses, thermal insulation made of fibrous materials (2) was made along the entire length of the chamber. A 15 kW gas swirl burner (5) was located at the inlet to the chamber. Seven holes (3) were placed along the entire length of the chamber (

Table 5. Locations of measurement points.

Measurement Point Number	T1	T2	T3	T4	T5	T6	T7
Distance from burner outlet (m)	0.29	0.65	0.97	1.42	1.91	2.51	2.91

), used to measure temperature by means of a PtRh10-Pt thermocouple (6) and the composition of flue gases employing a VARIO PLUS analyser (7). The gas and air volume flow was regulated utilising valves with rotameters (4). Air was supplied from a compressor and natural gas from the grid.

The locations of the measurement points are presented in Table 5.

Temperature measurements at individual points along the length of the heating chamber were taken using a 0.5 mm thick PtRh10-t jacketed thermocouple, covered in quartz tubing, placed perpendicular to the flame axis. A BRYMEN 805 digital multimeter was used to read the temperature. The investigations were carried out for the excess air ratio λ = 1.05. The gas and air flows were kept constant and amounted to, for air, $\dot{V}$_A = 0.00428 m³/s and, for gas, $\dot{V}$_NG = 0.00045 m³/s.

2.4. Computer Simulations

Computer simulations were performed using ANSYS Chemkin-Pro software (ver. 2021 R1), version 2021R1 (licence within the Pioneer network). According to the literature data, this is one of the most commonly employed programmes, next to ASPEN PLUS, for modelling chemical phenomena [71,72,73,74]. Based on the experience of other researchers [75,76,77,78,79], a modified chemical mechanism was implemented for the calculations, taking into account C1–C5 hydrocarbons, including 360 reactions as well as 61 compounds and chemical elements, with kinetic and thermodynamic data. Two types of reactors were adopted in the modelling procedure. The first, a Perfectly Stirred Reactor (PSR), was the equivalent of gas combustion in a burner, and a plug flow reactor (PFR) had the gas combustion process take place in a heating chamber with a constant temperature profile (Figure 3).

The input data file also included experimentally determined values (

Table 6. Input data in modelling procedure for experimental chamber.

Parameter	Value **
Air flow (m3/s)	0.00428
Gas flow (m3/s)	0.00045
Air temperature (K)	293
Gas temperature (K)	293
Adiabatic temperature * (K)	1980
Chamber length (m)	3
Chamber dimeter (m)	0.12
Residence time (ms)	0.15
Pressure (atm)	1

* Adiabatic temperature was calculated for natural gas using Chemkin-Pro software (ver. 2021 R1), Equilibrio model; ** modelling input data taken from the experimental study.

), namely the following:

• Mass flow of reactants (air and gas).
• Chemical composition of reagents.
• Temperature of supplied media.
• Pressure.
• Adiabatic combustion temperature.
• Temperature profile in the combustion chamber.
• Chamber dimensions (diameter, length).
• Residence time of the reagents in the highest temperature zone.

Basic computer simulations were performed for the heating zone of a pusher furnace (Figure 4).

The following gases from waste pyrolysis were analysed:

• RDF (P1, P2) [25,44].
• a mixture of alder chips with polypropylene (P3) [82].
• a mixture of pine chips with polypropylene (P4) [82].

The streams of the analysed gases (${\dot{V}}_{PG}$) were calculated from Relationship (2), using the conversion factor of natural gas to another combustible gas (3):

$${\dot{V}}_{PG}={\dot{V}}_{NG}\times n$$

(2)

$$n={\displaystyle \frac{{LHV}_{NG}}{{LHV}_{PG}}}$$

(3)

where

LHV_NG—calorific value of natural gas, MJ/m³.

LHV_PG—calorific value of pyrolysis gas, MJ/m³.

The calculated flow rates of the analysed gases and their chemical composition are presented in

Table 7. Flow rates and chemical composition of analysed gases.

Gas	Natural Gas—NG	RDF1 —P1	RDF2 —P2	AW + PP * —P3	PW + PP * —P4
LHV (MJ/m3)	34.4	29.94	24.2	21.67	24.2
n	1	1.145	1.410	1.587	1.421
${\dot{V}}_{NG}$ (m3/s)	0.3669	0.3302	0.3302	0.3302	0.3302
${\dot{V}}_{PG}$ (m3/s)	-	0.0420	0.0517	0.0582	0.0521
Gas composition (vol%)
H2	-	12.37	13.2	8.28	7.94
CO2	0.315	11.89	21.1	32.65	29.44
CO	-	29.69	19.8	27.07	26.43
CH4	96.648	17.83	18.9	16.47	23.15
C2H4	-	13.84	7.1	1.6	1.47
C2H6	1.836	4.03	2.3	3.08	2.88
C3H6	-	7.34	-	5.34	5.56
C3H8	-	0.43	6.1	1.15	1.08
C4H10	-	0.05	3.7	0.41	0.47
C5H12	-	0.01	-	1.41	0.28
N2	1.201	-	-	-	-
Ref.	[83]	[44]	[25]	[82]	[82]

* Ratio in both mixtures was 70% wood chips and 30% polypropylene.

, while the remaining parameters utilised in the calculations for the reheating furnace are presented in

Table 8. Input data in modelling procedure for reheating furnace.

Parameter	Value *
Air temperature (K)	623
Gas temperature (K)	293
Combustion chamber temperature (K)	1550
Chamber length (m)	8.1
Cross section (m2)	9.75
Residence time (ms)	1
Pressure (atm)	1
Air flow ${\dot{V}}_{Air}$ (m3/s)	35.583

* Input data for modelling was provided by the local steelworks.

. The mixed gas was burned at a ratio of 1:10 to natural gas.

3. Results

3.1. Combustion of Natural Gas in a Laboratory Chamber

The measured values of NO concentration and temperature in the laboratory chamber are summarised in

Table 9. Temperature and NO concentration distribution along length of experimental chamber.

Distance from Burner Outlet (m)	0.29	0.65	0.97	1.42	1.91	2.51	2.91	Measurement Uncertainty
NO Concentration (ppm)	135	134	137	135	135	134	136	±5 ppm
Temperature (K)	1583	1488	1413	1302	1194	1101	996	± 1 °C

.

The experimentally determined temperature profile (Table 9) was implemented in computer simulations and then the results for the NO concentration obtained from the calculations and the experiment in the laboratory chamber were compared (Figure 5).

As can be seen, the obtained simulation results are very similar to the results obtained in the laboratory experiment, which allows us to assume that the model adopted for calculations correctly reflects the conditions prevailing during the combustion of natural gas in the chamber. The above statement was the basis for conducting the simulation of the co-combustion process of natural gas with gases from waste pyrolysis in a metallurgical reheating furnace.

3.2. Co-Combustion of Gases from Waste Pyrolysis in Reheating Furnace

Calculations were performed both for the co-combustion of the pyrolysis gases listed in Table 9 with natural gas and for natural gas alone. The results obtained for the main combustion products at the end of the heating zone are presented in Figure 6, while Figure 7 and Figure 8 additionally show the change in the concentration of NO, CO, CO₂, H₂ and unburned hydrocarbons as a function of the zone length.

The performed computer simulations (Figure 6 and Figure 7 show that the chemical composition of the flue gases from the co-combustion of pyrolysis gases is qualitatively similar to the composition of the flue gases obtained from the combustion of natural gas alone. Quantitatively, however, differences are observed for the CO and H₂ concentrations, which are related to the composition of the pyrolysis gas. The smaller the share of CO and H₂ in the composition of the pyrolysis gas, the lower the concentration of these compounds is in the combustion products. The highest concentration of CO and H₂ in the flue gases was observed for the co-combustion of natural gas with gas from the pyrolysis of alder chips with polypropylene (P3) and amounted to 6.67% for H₂ and 6.46% for CO, respectively. The lowest concentrations were recorded for the gas from the pyrolysis of RDF (P1) and amounted to 0.35% for H₂ and 0.49% for CO, respectively. The fraction of polypropylene in the P3 and P4 mixtures was 30%.

The concentrations of the analysed compounds, both from the combustion of natural gas alone and from co-combustion with pyrolysis gases along the length of the heating zone, have a linear course for most of them, namely CO, CO₂ and H₂. A different situation is observed for nitrogen oxides: NO, NO₂ and nitrous oxide, N₂O. The highest concentrations occur in the burner zone, and then they drop rapidly, which is related to the high temperature in the PSR, responsible for the formation of the so-called thermal nitrogen oxides. For all the analysed variants, the lowest NO concentration is observed for the P3 gas and it is 12 ppm, while the highest is for natural gas alone and it is 133 ppm. The highest NO₂ concentration is for natural gas and it is 14 ppm; the lowest is for the P3 gas—0.03 ppm. The N₂O concentration is also the highest for natural gas and it is 0.02 ppm, whereas the lowest concentration reaches 0.001 ppm for the P3 gas. In combustion products, in addition to a significant share of compounds such as H₂, O₂, H₂O, CO₂, CO and N₂, there are also compounds whose share is trace, included in Figure 10 in the “other” category. These include, among others, the above-mentioned NO, NO₂ and N₂O. The concentration of other compounds is the highest for the P3 gas and is 0.07%, while the lowest is for NG and P4—0.02%.

As for the concentration of unburned hydrocarbons (Figure 8), the highest values are observed during the combustion of P3 gas and amount to about 500 ppm. Such a high value may result from the higher concentration of C5 hydrocarbons in the fuel mixture compared to the other combusted mixtures. On the other hand, the lowest trace concentration was observed during the combustion of NG and mixtures with the P1 and P2 gases. From the technological point of view, unburned hydrocarbons and CO and H₂ formed during the combustion of mixtures with natural gas will burn out in the furnace equalisation zone due to the excess air ratio prevailing there (λ = 1.07).

In order to explain the mechanisms responsible for the formation of the analysed compounds, a formation pathway analysis and sensitivity analysis were performed in the Ansys Chemkin-Pro programme (ver. 2021 R1).

3.3. Measurement Error Analysis

The analysis of measurement errors introduced in the course of experimental tests includes systematic errors, resulting from the design of the measurement apparatus, as well as the measurement methodology. The measurement of nitric oxide concentration carried out with the Testo 360 analyser is subject to an error, the magnitude of which depends on its accuracy class [84,85]:

$$kl={\displaystyle \frac{{∆x}_{max}}{Z}}·100\%$$

(4)

From the given instrument class, the absolute error value can be calculated:

$${∆\mathit{x}}_{\mathit{m}\mathit{a}\mathit{x}}={\displaystyle \frac{\mathit{k}\mathit{l}·\mathit{Z}}{\mathbf{100}}}$$

(5)

The accuracy of the exhaust gas analyser, for the final value of the measuring range, i.e., 200 ppm, according to the verification certificate, was 3.8%.

The actual temperature t at each point of the flame was calculated according to Formula (6) derived from the weld energy balance [84]:

$$\mathit{t}={\mathit{t}}_{\mathit{s}}=\mathit{a}·{\mathit{\epsilon }}_{\mathit{s}}·{\mathit{C}}_{\mathit{c}}·{\mathit{d}}_{\mathit{s}}⌊{{\left({\displaystyle \frac{{\mathit{T}}_{\mathit{s}}}{\mathbf{100}}}\right)}^4-\left({\displaystyle \frac{{\mathit{T}}_{\mathit{w}}}{\mathbf{100}}}\right)}^4⌋$$

(6)

where

t_s, T_s—weld temperature, °C; K.

a = 5—experimental constant, m-K/W.

C_c = 5.67—blackbody radiation constant (Stefan Boltzman constant), W/m²·K.

d_s—thermocouple weld diameter, m.

ε_s—emissivity of the thermocouple weld; for the PtRh-Pt thermocouple used in the measurements ε_s = 0.000106 t_s + 0.0383.

T_w = 293—ambient temperature, K.

The maximum temperature difference between the measured and calculated temperatures was 13 °C.

The maximum concentration difference between the measured and calculated value with respect to the standard oxygen concentration was 20 ppm, and the minimum was 2.5 ppm. The characteristics of parameters for the used measuring devices are shown in

Table 10. Characteristics of measurement parameters.

Measured Quantity	Measurement Range	Accuracy	Resolution
VARIO PLUS analyser
NO, ppm	0 ÷ 5000	±5 ppmv ±5% (<1000 ppmv) ±10% (>1000 ppmv)	1.00
PtRh10-Pt thermocouple
Temperature, °C	Up to 1800	±1	0.10

.

3.4. Formation Path Analysis and Sensitivity Analysis

As mentioned earlier, owing to the complexity of the kinetics of chemical reactions occurring during the co-combustion of pyrolysis gases, the NO formation paths (Figure 9) and reactions involved in the formation of compounds such as NO, NO₂, N₂O, CO, CO₂ and H₂ were analysed. Sensitivity analyses were also performed for all the studied gas mixtures. The results of the analyses are presented in Figure 10, Figure 11 and Figure 12.

By analysing the formation paths, it can be seen that the mechanism of nitrogen oxide formation during the combustion of mixtures of natural gas with the P2, P3 and P4 gases is much more complex than that for the combustion of natural gas alone and with the P1 gas. The NO formation pathway is the same for NG and P1, but differences are observed in the rate values, as presented in Figure 10a,b. In NO formation, not only CH radicals participate, but also HNO, NCO, NH, HCN, NH₂, HNCO and H₂O. In order to explain the reason for the observed differences, a sensitivity analysis was performed (Figure 10) [86].

The dominant reactions responsible for the formation of NO (Figure 10) during the combustion of natural gas are Reactions (7)–(9):

N + OH ↔ NO + H

(7)

HNO + H ↔ H₂ + NO

(8)

NO₂ + H ↔ NO + OH

(9)

the reactions that are mainly responsible for the decomposition of NO are Reactions (10) and (11):

NO + O + M ↔ NO₂ + M

(10)

H + NO + M ↔ HNO + M

(11)

During the combustion of gas mixtures P1, P2 and P4, Reactions (7) and (8) are also dominant. In the case of the combustion of gas mixture P3, the main reaction of NO formation is Reaction (8), which has the largest share of NO formation for this mixture. The most important reactions responsible for NO decomposition during the combustion of gas mixture P1 are also Reactions (11) and (10), just as in the case of natural gas. During the combustion of mixture P2, the reactions responsible for NO decomposition are (12), (13) and (14):

C + NO ↔ CO + N

(12)

C + NO ↔ CN + O

(13)

CH + NO ↔ HCN + O

(14)

For the P3 mixture, this is only Reaction (15):

CH₂ + NO ↔ H + HNCO

(15)

And for the P4 mixture, Reactions (12)–(14) are exactly the same as for the P2 mixture.

By analysing the rate of N₂O formation (Figure 11) for all the considered variants, it can be seen that the dominant N₂O formation reaction for NG and P1 is Reaction (16):

N₂O (+M) ↔ N₂ + O (+M)

(16)

For mixtures P2, P3 and P4, there are three main reactions: (16)–(18).

NH + NO ↔ N₂O + H

(17)

NCO + NO ↔ N₂O + CO

(18)

In turn, the dominant reaction responsible for the decomposition of N₂O in the case of all the combusted gases and mixtures is Reaction (19):

N₂O + H ↔ N₂ + OH

(19)

By observing the NO₂ formation rate (Figure 12) for all the considered variants, it can be seen that the dominant NO₂ formation reactions for all the combusted fuels are Reactions (20) and (21):

NO + O + M ↔ NO₂ + M

(20)

HO₂ + NO ↔ NO₂ + OH

(21)

On the other hand, NO₂ decomposition is most influenced by Reaction (22), which is responsible for NO₂ decomposition during the combustion of all the gases:

NO₂ + H ↔ NO + OH

(22)

Formation pathways with interpretations of the results for compounds such as CO, CO₂ and H₂ can be found in the Supplementary Materials to complement the above analysis.

As can be seen in Figure 10, Figure 11 and Figure 12, the rate of formation of the selected compounds for the same reactions varies depending on the fuel composition. Temperature does not affect the rate of formation of compounds in this case because it is constant along the entire length of the heating chamber (1550 K).

4. Conclusions

The issue of managing the environmentally harmful gaseous products of the thermal conversion of selected calorific wastes discussed in this article addresses the problems faced by both the waste industry and highly energy-intensive industries, such as the steel industry. The results presented in this paper clearly indicate that it is possible to safely manage environmentally harmful gaseous products from the pyrolysis process of selected wastes for firing metallurgical reheating furnaces, both in terms of technology and the environment. The detailed analysis of the chemical mechanisms describing the process of the co-combustion of pyrolysis gases with natural gas in the heating zone of the pusher furnace provides comprehensive knowledge on combustion kinetics and environmental effects, namely the following:

• The performed computer simulations show that the chemical composition of the flue gases resulting from the co-combustion of pyrolysis gases is qualitatively similar to the composition of flue gases obtained from the combustion of natural gas alone.
• The results most similar to the combustion of natural gas alone, in terms of the composition of the flue gases, were obtained for the P1 mixture.
• The greatest influence on the differences in the composition of the resulting flue gases is the composition of the pyrolysis gas co-fired with natural gas.
• The addition of gas from the pyrolysis of biomass waste with the addition of PP leads to a higher concentration of CO and H₂ as well as unburned hydrocarbons in the flue gases than for the other mixtures.
• The smaller the share of CO and H₂ in the composition of the pyrolysis gas, the lower the concentration of these compounds in the combustion products is.
• The highest concentration of CO and H₂ in the flue gases was observed for the co-combustion of natural gas with gas from the pyrolysis of alder chips with polypropylene (P3), while the lowest was for the gas from the pyrolysis of RDF (P1), which is related to the addition of polypropylene in the P3 and P4 mixtures at a level of 30%.
• By analysing the formation paths, it can be seen that the mechanism of nitrogen oxide formation during the combustion of mixtures of natural gas with the P2, P3 and P4 gases is much more complex than for the combustion of natural gas alone and with P1 gas. In NO formation, not only do CH radicals participate, but so do HNO, NCO, NH, HCN, NH₂, HNCO and H₂O.

The study also showed a number of environmental benefits:

• The analysis has shown that the addition of pyrolysis gas does not increase nitrogen oxide emissions, which will not result in increased emission charges.
• Managing unused waste, including agricultural waste, will have a positive impact on the waste market in Poland. Large companies (e.g., steel mills) using pyrolysis to meet their own energy needs will become regular consumers of this waste.
• We will see the formation of an outlet market for RDF waste, which, in Poland, due to legal conditions, is stored without thermal use.

We will see the dissemination of new solutions concerning thermal waste treatment technologies into the existing waste management system. In conclusion, it should be emphasised that knowledge of the chemical mechanisms presented in the article, in the application context, may enable appropriate actions to be taken aimed at managing gases after the thermal conversion of waste, while maintaining correct operation of the heating chamber and minimising pollution. By analysing the obtained results, it can be clearly stated that the most desirable gases, potentially used for co-combustion in the steel industry, will be gases from RDF pyrolysis because the obtained flue gases are very similar in composition to the flue gases from pure natural gas, which will not translate into an increase in emission fees, and will contribute to a significant reduction in the operating costs of individual furnaces in the steel industry. This study constitutes an introduction into further simulations and pilot-scale studies to confirm the feasibility of effectively replacing natural gas with other alternative fuels. The complexity of the pyrolysis process, making it difficult to describe mathematically, is undoubtedly a disadvantage. However, the selection of boundary conditions and laboratory tests will be able to validate the validity of the computational model and narrow down the parameters that should be used at the laboratory and pilot scales. Moreover, the focus should be on selecting process parameters so that no additional costs are incurred due to excess air emissions.