ABSTRACT
Numerical modeling has been used extensively to simulate gas–liquid transfer of sulfur dioxide, assessing how operational parameters affect absorption efficiency in packed or spray columns. Despite individual studies on these contactors, comparative analyses on the same flue gas have been rare. This study uses a numerical model for both packed and spray columns to examine how parameters influence SO2 absorption by sodium sulfite, describing packed and spray columns, is used to investigate the influence of operational parameters on SO2 absorption by sodium sulfite. The model's predictions are validated against experimental data from an industrial pilot plant. Across varying conditions (L/G ratio, temperature, initial SO2 content or initial S(IV) concentration), the packed column achieves higher absorption efficiencies compared to the spray column, with lower assumed energy costs due to a reduced L/G ratio. Temperature proves to be a significant factor, decreasing absorption efficiency by approximately 40% between 40 and 70 °C. SO2 absorption efficiency declines with increasing concentrations of bisulfite and sulfite ions in the absorption solution, dropping to 50% at an S(IV) concentration of 2 kmol m−3 in the liquid phase. Considering the objective of producing a concentrated bisulfite solution and a clean gas, a two-column system is recommended: one for bisulfite solution concentration at acidic pH and the other for gas purification enhancement at basic pH.
HIGHLIGHTS
A process for recovering SO2 from waste incineration flue gas has been numerically modeled.
The model's results were compared with experimental data obtained using an industrial pilot plant.
A comparison of results obtained with a packed and a spray column was carried out.
The impact of industrial operating conditions was tested on both columns.
The packed column provides higher absorption yields for the process studied.
NOMENCLATURE
Concentration of species X in the liquid phase (kmol m−3)
Total interfacial area (m2 m−3)
Wetted interfacial area of packing (m2 m−3)
Column diameter (m)
Diffusion coefficient of species X (m2 s−1)
Droplet diameter in the spray column (m)
Packing ring diameter (m)
Enhancement factor
Standard gravitational acceleration (m s−2)
Flue gas flow rate (m3 s−1)
Column height (m)
Henry's law constant for SO2 (Pa m3 kmol−1)
Equilibrium constant for reaction i (kmol m−3)
Equilibrium constant for water autoprotolysis (kmol2 m−6)
Gas phase mass transfer coefficient for SO2 (m s−1)
Liquid phase mass transfer coefficient for SO2 (m s−1)
Recirculation liquid flow rate (m3 s−1)
Absorption rate of species X (kmol m−2 s−1)
SO2 partial pressure (Pa)
Ideal gas constant (m3 Pa mol−1 K−1)
Cross-sectional area of column (m2)
Total concentration of sulfur ions (SO2,
, and
) (kmol m−3)
Temperature (K)
Droplet velocity in relation to flue gas in the spray column (m s−1)
GREEK LETTERS
LETTERS IN SUBSCRIPT
INTRODUCTION
Sulfur dioxide (SO2) is a significant industrial pollutant requiring specific treatment due to its detrimental effects on the environment and human health. It is a major contributor to air pollution and acid rain (Pandey et al. 2007; Jion et al. 2023). As a result, strict regulations have been implemented to control its emissions. In accordance with the provisions of Directive 2010/75/EU, European regulations stipulate that liquid-fueled power plants with a capacity of more than 300 MW must comply with an SO2 limit of 40 mg/Nm3. In recent years, extensive research has been conducted to enhance desulfurization processes, resulting in a substantial reduction in industrial SO2 emissions worldwide (Li et al. 2022). Waste incineration processes can generate a significant amount of SO2 in flue gases, making the recovery of this pollutant highly beneficial. Several processes have been implemented to recover SO2 in various forms. These include the limestone process to produce gypsum (Lim et al. 2021; Chen et al. 2022), absorption by ammonia to form ammonium sulfate (Gao et al. 2010; Yan et al. 2011) and sodium carbonate processes (Ebrahimi et al. 2003; Lee et al. 2023). The Wellman–Lord process presents an interesting area for further research (Neumann 1991; Shi et al. 2021). This process uses of a sodium sulfite Na2SO3 solution to absorb SO2, which is then converted into a sodium bisulfite NaHSO3 solution. Upon heating, the solution releases a concentrated SO2 gas, which can then be recovered as elemental sulfur through the Claus process (Li et al. 2021).


The reaction between SO2 and sodium sulfite solution is very fast, even instantaneous (Buzek & Jaschik 1995), and gas–liquid transfer limits the reaction. For this reason, contactors with large exchange surfaces are recommended. Packed and spray columns are generally used (Onda et al. 1972). Given their different operating costs, it may be worthwhile to evaluate the performance of these two columns for the process under study, to determine the most suitable technology for industrial-scale application. In this context, an industrial pilot with two absorption columns has been set up on a waste incineration flue gas treatment unit to assess the feasibility of the process and the performance of these two types of contactors. However, various operating conditions (such as temperature or flue gas composition) are imposed by industrial requirements, making it difficult to study the impact of individual parameters. Using modeling as a complement to experimental testing is an effective approach for studying the impact of different parameters on the process.
Numerical methods have been widely used to understand mass transfer of several gases involving complex reversible reactions. In many works, these methods are used to design gas–liquid contactors, including packed columns (Ebrahimi et al. 2003; Liu et al. 2016; Zhu et al. 2021b) and spray columns (Bandyopadhyay & Biswas 2007; Neveux & Le Moullec 2011; Qin et al. 2019). However, in the desulphurization process, these models are only used at basic pH and are not effectively used to study the parameters impacting the process on already dimensioned columns. In addition, few studies compare these two types of contactors for the same gas, and even fewer do so with real flue gas from waste incineration.
The objective of this work is to study the impact of different operating parameters on the absorption yield using a numerical model simulating the gas–liquid transfer during the reaction between sodium sulfite and SO2 at acidic pH. The aim is also to ascertain the most effective column technology for the absorption of SO2 in the desulphurization process implemented using the developed model. The model will first be validated by experimental points carried out on real incineration flue gases. Next, the column profiles of various parameters obtained with the model will be detailed. Finally, the influence of classical industrial operating parameters, such as L/G ratio, temperature, gas and liquid composition, will be assessed based on numerical simulations.
MATERIALS AND METHODS
Description of the pilot plant
Industrial pilot diagram. (1) Absorption column, (2) fan, (3) gas flowmeter, (4) recirculation pump, (5) liquid flowmeters, (6) pH probe, (7) control valve, and (8) gas analyzers.
Industrial pilot diagram. (1) Absorption column, (2) fan, (3) gas flowmeter, (4) recirculation pump, (5) liquid flowmeters, (6) pH probe, (7) control valve, and (8) gas analyzers.
The absorption liquid is injected at the top of the column and continuously recirculated by a pump (4), drawing the liquid from the bottom of the column (Someflu). A liquid flowmeter (5) is placed on the recirculation pipe (Proline Promass F300 from Endress Hauser). A pH probe (6) measures the pH of the absorption solution in real time (Memosens CPS16E from Endress Hauser). The pH can be maintained at a desired value by adding soda via a control valve (7). Another flowmeter (5) is used to record the amount of soda injected. Two gas analyzers (8) are placed at the inlet and outlet of the column to measure the quantities of SO2 present in the flue gases (TESTO 350 from TESTO).
Table 1 outlines all the operating parameters, including those specific to each column. The two absorption columns have been designed to achieve SO2 absorption efficiencies greater than 99% at basic pH. As a result, plastic pall rings with a diameter of 0.0254 m are used as packing, resulting in a wetted interfacial area of 205 m2 m−3. For the spray column, a droplet diameter of approximately 1 mm is achieved using a high flow rate of recirculating liquid (around 8.5 m3 h−1) through a full-cone nozzle (Spraljet HHSJ from Spraying Systems Co.).
Operating and technical parameters of absorption columns
Parameters . | Packed column . | Spray column . |
---|---|---|
Range of SO2 at column inlet (%) | 0.1–1.0 | |
Range of sulfite concentration (kmol m−3) | 0–2 | |
Liquid flow (m3 h−1) | 2.5 | 8.5 |
Diameter of the column (m) | 0.45 | 0.32 |
Height of exchange surface (m) | 0.80 | 1.90 |
Drop diameter (mm) | – | 1 |
Packing ring diameter (m) | 0.0254 | – |
Wetted interfacial area (m2 m−3) | 205 | – |
Parameters . | Packed column . | Spray column . |
---|---|---|
Range of SO2 at column inlet (%) | 0.1–1.0 | |
Range of sulfite concentration (kmol m−3) | 0–2 | |
Liquid flow (m3 h−1) | 2.5 | 8.5 |
Diameter of the column (m) | 0.45 | 0.32 |
Height of exchange surface (m) | 0.80 | 1.90 |
Drop diameter (mm) | – | 1 |
Packing ring diameter (m) | 0.0254 | – |
Wetted interfacial area (m2 m−3) | 205 | – |
A test begins by filling the column with water and selecting the pH to regulate the solution. The flue gas is then drawn in by the fan, and SO2 is continuously absorbed by the solution. With time, the sodium bisulfite concentration in the solution increases due to continuous absorption. To evaluate the influence of the S(IV) concentration in the liquid phase on the absorption efficiency of the columns, experiments were carried out with a wide range of concentrations between 0.11 ± 0.02and 1.88 ± 0.14kmol m−3. The SO2 content is highly dependent on incineration conditions and cannot be controlled. On the industrial site, the SO2 content of flue gases varies according to the sulfur content on the incinerated waste and the incineration conditions. Thus, during experiments, the SO2 content may vary from 0.10 ± 0.02 to 0.43 ± 0.06%. Table 2 describes the conditions of the different tests carried out as part of this study.
Pilot-scale test conditions
Column . | [S(IV)] (kmol m−3) . | pH . | Inlet SO2 content (%) . | Temperature (°C) . |
---|---|---|---|---|
Packed column | 1.41 ± 0.11 | 8.5 ± 0.1 | 0.19 ± 0.03 | 68.0 ± 0.5 |
1.88 ± 0.14 | 8.7 ± 0.1 | 0.25 ± 0.04 | 68.50 ± 0.5 | |
0.21 ± 0.02 | 5.3 ± 0.1 | 0.11 ± 0.02 | 68.00 ± 0.5 | |
0.62 ± 0.06 | 8.5 ± 0.1 | 0.34 ± 0.05 | 69.00 ± 0.5 | |
0.15 ± 0.03 | 8.0 ± 0.1 | 0.10 ± 0.02 | 70.00 ± 0.5 | |
0.19 ± 0.03 | 7.0 ± 0.1 | 0.11 ± 0.02 | 70.00 ± 0.5 | |
0.22 ± 0.03 | 6.0 ± 0.1 | 0.29 ± 0.04 | 70.00 ± 0.5 | |
0.23 ± 0.03 | 5.5 ± 0.1 | 0.31 ± 0.04 | 70.00 ± 0.5 | |
Spray column | 0.72 ± 0.06 | 4.8 ± 0.1 | 0.35 ± 0.05 | 68.50 ± 0.5 |
0.79 ± 0.07 | 5.4 ± 0.1 | 0.29 ± 0.04 | 68.50 ± 0.5 | |
0.11 ± 0.02 | 7.9 ± 0.1 | 0.37 ± 0.05 | 72.00 ± 0.5 | |
0.40 ± 0.04 | 6.8 ± 0.1 | 0.43 ± 0.06 | 72.00 ± 0.5 | |
1.29 ± 0.10 | 6.6 ± 0.1 | 0.31 ± 0.04 | 71.50 ± 0.5 | |
1.28 ± 0.10 | 6.2 ± 0.1 | 0.25 ± 0.04 | 72.00 ± 0.5 | |
1.49 ± 0.12 | 5.8 ± 0.1 | 0.33 ± 0.05 | 71.50 ± 0.5 |
Column . | [S(IV)] (kmol m−3) . | pH . | Inlet SO2 content (%) . | Temperature (°C) . |
---|---|---|---|---|
Packed column | 1.41 ± 0.11 | 8.5 ± 0.1 | 0.19 ± 0.03 | 68.0 ± 0.5 |
1.88 ± 0.14 | 8.7 ± 0.1 | 0.25 ± 0.04 | 68.50 ± 0.5 | |
0.21 ± 0.02 | 5.3 ± 0.1 | 0.11 ± 0.02 | 68.00 ± 0.5 | |
0.62 ± 0.06 | 8.5 ± 0.1 | 0.34 ± 0.05 | 69.00 ± 0.5 | |
0.15 ± 0.03 | 8.0 ± 0.1 | 0.10 ± 0.02 | 70.00 ± 0.5 | |
0.19 ± 0.03 | 7.0 ± 0.1 | 0.11 ± 0.02 | 70.00 ± 0.5 | |
0.22 ± 0.03 | 6.0 ± 0.1 | 0.29 ± 0.04 | 70.00 ± 0.5 | |
0.23 ± 0.03 | 5.5 ± 0.1 | 0.31 ± 0.04 | 70.00 ± 0.5 | |
Spray column | 0.72 ± 0.06 | 4.8 ± 0.1 | 0.35 ± 0.05 | 68.50 ± 0.5 |
0.79 ± 0.07 | 5.4 ± 0.1 | 0.29 ± 0.04 | 68.50 ± 0.5 | |
0.11 ± 0.02 | 7.9 ± 0.1 | 0.37 ± 0.05 | 72.00 ± 0.5 | |
0.40 ± 0.04 | 6.8 ± 0.1 | 0.43 ± 0.06 | 72.00 ± 0.5 | |
1.29 ± 0.10 | 6.6 ± 0.1 | 0.31 ± 0.04 | 71.50 ± 0.5 | |
1.28 ± 0.10 | 6.2 ± 0.1 | 0.25 ± 0.04 | 72.00 ± 0.5 | |
1.49 ± 0.12 | 5.8 ± 0.1 | 0.33 ± 0.05 | 71.50 ± 0.5 |
Model description
The developed model aims to simulate the operation of each of the industrial pilot columns described in Section 2.1. The hydrodynamic properties of columns are complex phenomena, making their modeling challenging. To simplify the equations, the following assumptions are made:
Transport and reaction phenomena are assumed to be one-dimensional depending on the height of the column.
Double film theory is used to describe mass transfer.
Plug flow is considered for the liquid and gas phases.
Gas and liquid temperatures and flow rates are assumed to be constant along the column due to the low solute concentration.
The flue gas is assimilated to an SO2/N2 mixture.
Despite the presence of oxygen in the flue gas, the oxidation of S(IV) sulfur ions is not covered in this study.
Chemistry
These reactions are considered instantaneous and only pH dependent.
Reactor model

It is therefore essential to determine the SO2 concentration at the gas–liquid interface to carry out sulfur mass balances along the column.
Film region
As indicated in the calculation assumptions, the double film model is used to represent the diffusion of sulfur species from the gas phase to the liquid phase. In this model, it is assumed that all resistance to mass transfer is localized in a liquid film adjacent to the gas–liquid interface (Lewis & Whitman 1924). It is also assumed that mass transfer on the liquid side takes place solely by molecular diffusion through a liquid film. Beyond this film, compositions are considered uniform.
Concentration profile of the various compounds in the liquid film

Boundary conditions
The equations associated with acid–base equilibria (Equations (12)–(14)) and the total sulfur balance (Equation (15)) are also applied at the gas–liquid interface.
For efficient solution of this system, the liquid film is discretized into n elements of length . The finite-difference method is used to transform differential equations into algebraic equations. By combining Equations (10)–(15) in the liquid film, a system of
equations with
unknowns, representing the concentration of each chemical species at each layer
of the film, is obtained. This system is solved by a Newton-based method on MatLab® using boundary conditions at the gas–liquid interface and bulk liquid. It enables determining the SO2 absorption rate with Equation (5).
Calculation methods
The absorption columns studied operate in counter current. The gas phase enters at the bottom of the column with a known composition and exits from the top. The liquid phase follows the opposite path, with a known composition at the column inlet. The objective is to determine the composition of the gas phase at the top of the column and the composition of the liquid phase at the bottom under different operating conditions.

The calculation starts at the top of the column with the known liquid composition. For each section , the SO2 absorption rate is calculated by recovering the SO2 concentration at the gas–liquid interface using the double film model (see Section 2.2.3). Since the partial pressure at the top of the column is initially unknown, an assumption must be made. The model then calculates the evolution of pressure and concentration in the liquid phase using Equations (7) and (8). Here, only the total sulfur ion concentration is determined. A small additional algorithm is used to determine the precise quantity of species in solution, using electroneutrality.

Parameter determinations
Physicochemical parameters
In this study, we simplified the system as much as possible by assuming the flue gas to be nitrogen, supplemented by SO2 contents close to those observed on an industrial scale, between 0.1 and 1.0%. SO2 is absorbed by a liquid solution of sodium sulfite Na2SO3 at concentrations of between 0.1 and 2.0 kmol m−3. This range is chosen to be similar to that used in pilot-scale experimental studies. The molar mass, density and viscosity of both the gas and liquid phases are calculated as functions of SO2 content and S(IV) ion concentration, respectively, using ProSim's Simulis® software.
The effect of temperature is considered in the calculation of the Henry's law constant of SO2 and the equilibrium constants of the reactions using empirical correlations (Ebrahimi et al. 2003). In addition, the diffusion coefficients of each species in the liquid phase are calculated at the studied temperature using the Stokes–Einstein equation (Bravo et al. 2002) with the diffusion coefficients at 25 °C determined by Haynes (2014). The diffusion coefficient of SO2 in the gas phase is determined by the empirical equation given by Reid et al. (1988).
Mass transfer coefficients and interface areas
RESULTS AND DISCUSSION
Model validation
Comparison between experimental absorption efficiencies and those defined by the numerical model for the two columns.
Comparison between experimental absorption efficiencies and those defined by the numerical model for the two columns.
Parameter profiles along the columns
Parameters profiles along the packed column. (a) and (d) Evolution of SO2 content and the absorption efficiency. (b) and (e) Evolution of sulfur species concentrations in the liquid phase and pH. (c) and (f) Evolution of SO2 absorption rate. (a)–(c): initial pH of 8.5; (d)–(f) initial pH of 5.0. T = 50 °C; initial S(IV) = 0.5 kmol m−3 of S(IV) at the inlet; 1.0% of SO2 at the inlet.
Parameters profiles along the packed column. (a) and (d) Evolution of SO2 content and the absorption efficiency. (b) and (e) Evolution of sulfur species concentrations in the liquid phase and pH. (c) and (f) Evolution of SO2 absorption rate. (a)–(c): initial pH of 8.5; (d)–(f) initial pH of 5.0. T = 50 °C; initial S(IV) = 0.5 kmol m−3 of S(IV) at the inlet; 1.0% of SO2 at the inlet.
At a pH of 8.5, almost all incoming SO2 is absorbed along the column, with the SO2 content at the outlet being about 0.006%, giving an absorption efficiency of 99.4%. This indicates the column's high absorption capacity as shown in Figure 4(a). Most of the SO2 is absorbed in the first 0.3 m of the column, achieving an absorption efficiency of around 85% in this section. Due to SO2 absorption, the pH decreases from 8.5 to 7.3. The concentration profile of the species in solution clearly shows that SO2 reacts with sulfite ions to form bisulfite ions
via reaction (1) (see Figure 4(b)). At the bottom of the column, the solution initially containing 0.5 kmol·m−3 sulfite is found to contain 0.22 kmol·m−3 bisulfite and 0.38 kmol·m−3 sulfite. As the gas ascends the column, the SO2 absorption rate decreases, reflecting the evolution of SO2 content in the gas phase shown in Figure 4(c). The SO2 absorption rate is directly proportional to the SO2 concentration in the gas, indicating that mass transfer is controlled by the gas side, consistent with findings from other studies (Ebrahimi et al. 2003; Zhu et al. 2021b). The enhancement factor is very high throughout the column, ranging from 100 to 680. This shows that SO2 absorption is greatly enhanced by the reaction with sulfite ions.
When the pH of the initial solution is 5.0, the trends are relatively different. First, the low pH results in a small amount of sulfite ions, which is not sufficient to absorb all the SO2 present in the gas (see Figure 4(e)). The quantity of bisulfite ions does increase as the liquid passes through the column, but to a lesser extent than at basic pH: it rises from 0.50 to 0.58 kmol·m−3 at a pH of 5, whereas at a pH of 8.5 it rises from 0.03 to 0.22 kmol·m−3. Despite this, absorption efficiency remains high, reaching a value of 84% under these conditions, as shown in Figure 4(d). The absorption rate (Figure 4(f)) is about 2.5 times lower than that obtained at a pH of 8.5. This is clearly linked to the lower absorption efficiency observed. In addition, the absorption efficiency also decreases as the gas rises in the column, since the SO2 content also decreases. The calculated enhancement factor is very low, ranging between 0.15 and 0.30, meaning that the reaction no longer favors SO2 absorption at this pH.
The profiles associated with the spray column, not shown here, are similar to those of the packed column, resulting in the same analysis. The values obtained differ, with an absorption efficiency of approximately 99% at a pH of 8.5 and around 70% at a pH of 5 under the same conditions as those presented in Figure 4.
With these results, it seems possible to obtain a concentrated sodium bisulfite solution at a pH of 5 through SO2 absorption. Therefore, the following study will focus on the influence of operating conditions at acidic pH, as this is the most suitable range for generating a sodium bisulfite solution.
Influence of the L/G ratio and the temperature
The L/G ratio is a crucial parameter for gas–liquid contactors. These ratios generally differ between packed and spray columns. It is often higher for spray columns, as the recirculating liquid flow rate must be substantial to ensure a small drop size at the nozzle outlet (Qin et al. 2019). While the gas flow rate is generally driven by the industrial unit, the recirculating liquid flow rate is a variable factor that can be adapted by modifying the pump size and power, among other methods. Therefore, it is important to study the impact of the L/G ratio to determine the optimal value for maximizing absorption efficiency.
(a) Evolution of absorption efficiency evaluated by the model as a function of L/G ratios. (b) Evolution of absorption efficiency evaluated by the model as a function of temperature. 1.0% of SO2 in the inlet; 0.5 kmol m−3 of S(IV) at a pH of 5.0 in the inlet.
(a) Evolution of absorption efficiency evaluated by the model as a function of L/G ratios. (b) Evolution of absorption efficiency evaluated by the model as a function of temperature. 1.0% of SO2 in the inlet; 0.5 kmol m−3 of S(IV) at a pH of 5.0 in the inlet.
In the literature, other authors reported on SO2 absorption in both types of columns, but with different operating conditions. Qin et al. (2019) investigated SO2 absorption by limestone in a pilot-scale spray column. For an SO2 content of 0.8% and an absorption solution at a pH of 6, they observed high efficiencies that increased slightly from 93.5 to 98.5% when the L/G ratio was raised from 20 to 30. This increase is less significant than that seen in Figure 5(a) within the same range of L/G ratio, which may be due to the higher pH of their solution (6 against 5 in the present study), being more favorable to SO2 uptake. Wang et al. (2011) studied SO2 absorption using a sulfite solution in a packed column on a laboratory scale. With a lower SO2 content (around 0.3%) and a solution at a pH of 6, they reported a 20% increase in efficiency when the L/G ratio was increased from 1 to 4. The authors attributed this to the slower decline of the solution pH at a higher L/G ratio, which allows a greater quantity of sulfite ions in the solution to absorb SO2 along the column. Increasing the L/G ratio by increasing the liquid recirculation flow rate will, however, result in higher energy consumption. Therefore, increasing this ratio does not seem necessary for the packed column, since the impact on absorption efficiency is relatively minor.
The temperature of the incineration flue gases from the industrial unit under study is around 90 °C. The use of a heat exchanger upstream of the process makes it possible to lower this temperature, thereby enhancing the absorption phenomenon, which is typically improved at lower temperatures (Roustan 1997). Studying the influence of temperature using the model can therefore be insightful. Figure 5(b) illustrates the evolution of absorption yields as a function of temperature for the two columns under the same conditions as in Figure 5(a). Absorption efficiency of both columns decreases significantly as temperatures rise over a range of tens of Celsius degrees. For the packed column, the yield decreases from 91 to 54% when the temperature rises from 40 to 70 °C, while for the spray column, efficiency drops from 72 to 31%. First, absorption is an exothermic phenomenon and the increase in temperature reduces its effectiveness. In addition, as the temperature decreases, SO2 solubility increases, allowing more SO2 to dissolve in the liquid and react with sulfite ions, thereby enhancing absorption efficiency. Finally, the pKa of acid–base pairs are temperature dependent. As the temperature decreases, these pKa values also decrease, resulting in a higher concentration of sulfite ions in the solution that are capable of absorbing SO2 at the same pH.
This result has been widely observed in both column types (Ebrahimi et al. 2003; Liu et al. 2016; Zhao et al. 2021). For instance, Zhu et al. (2021b) reported a yield reduction of over 40% in their packed column for a temperature increase from 25 to 80 °C (with SO2 content of 1.0% absorbed by a solution at a pH of 6). To achieve optimal absorption efficiency, the gas temperature should be minimized as much as possible, regardless of the gas–liquid contactor technology used. However, reducing temperature requires heat exchange using cooling water, which can result in high energy consumption to operate the heat exchanger. Therefore, it is essential to find a balance between maximizing absorption efficiency through temperature reduction and minimizing excessive energy consumption.
Influence of sulfur species concentration in the liquid and gas phases
(a) Evolution of absorption efficiency evaluated by the model as a function of the initial partial pressure of SO2. (b) Evolution of absorption efficiency evaluated by the model as a function of the initial S(IV) concentration. T = 50 °C; L/G for packed column = 4.5; L/G for spray column = 12.5; pH = 5.0.
(a) Evolution of absorption efficiency evaluated by the model as a function of the initial partial pressure of SO2. (b) Evolution of absorption efficiency evaluated by the model as a function of the initial S(IV) concentration. T = 50 °C; L/G for packed column = 4.5; L/G for spray column = 12.5; pH = 5.0.
Existing studies generally show a different effect for a packed column, with a decrease in absorption efficiency as the inlet SO2 content increases. Wang et al. (2011) suggest that higher SO2 content in the gas leads to a more rapid decrease in the absorption solution pH along the column, reducing absorption efficiency. These studies typically use lower-concentration absorption solutions (<0.2 kmol m−3). In our case, the solution is sufficiently concentrated, providing a significant buffering effect that likely prevents a fast pH drop. The higher the inlet SO2 content, the greater the amount of SO2 absorbed, and therefore the shorter the time needed to reach a high concentration of sodium bisulfite.
The concentration of S(IV) in the solution is another crucial factor to consider. The process aims to increase the bisulfite solution's concentration over time. Therefore, studying the influence of S(IV) concentration on absorption efficiency at acidic pH is important. Figure 6(b) shows the evolution of absorption yield as a function of S(IV) concentration for both columns. The results indicate that S(IV) concentration significantly affects absorption yields, with rather similar evolutions for both columns. Up to 0.25 kmol m−3, absorption efficiency slightly increases for both columns, from 86 to 91% for the packed column and from 52 to 64% for the spray column. Absorption efficiency peaks at this concentration under the given conditions. Beyond this concentration, efficiency decreases with increasing S(IV) concentration, with a drastic decrease for the packed column, which drops to 46% at 2.0 kmol m−3.
The influence of S(IV) concentration on absorption efficiency has been investigated in other studies, which generally indicates that increasing this concentration negatively affects absorption efficiency, as observed in this study. According to Zhu et al. (2021b), this is due to the accumulation of SO2 in the liquid phase, which reduces the absorption rate of gaseous SO2 as described in Equation (5). This subsequently decreases the overall absorption efficiency. The effect is less pronounced for the spray column because the higher liquid flow rate compared to the packed column results in a smaller pH decrease along the column. Thus, increasing sulfite concentration over time reduces absorption efficiency, which in turn increases the time required to reach a high sodium bisulfite concentration.
CONCLUSIONS
As part of the development of the circular economy, a project was undertaken to treat hazardous waste incineration flue gases with the dual objective of eliminating their SO2 content and recovering sulfur in the form of a marketable bisulfite solution. For this purpose, a pilot plant was set up on an industrial site, featuring two absorption columns: a packed one and a spray one. The present study aimed to develop a numerical model that describes the operation of both columns and to investigate the influence of key operating parameters.
Comparison of simulations with experimental results from the industrial pilot plant shows that the developed model accurately represents the studied systems. While SO2 uptake is optimal at basic pH (100% at a pH of 8 with the packed column), this study focused on uptake at a pH of 5, which favors the production of a bisulfite solution. Whatever the operating parameters (L/G ratio, temperature, initial SO2 content or initial S(IV) concentration), the packed column achieves higher absorption yields (85–90%) than the spray column, with an assumed lower energy cost thanks to a significantly lower required L/G ratio (<10). Temperature is a significant influencing parameter, reducing absorption efficiency by around 40% between 40 and 70 °C. Therefore, from an industrial perspective, it is important to find a compromise between heat exchanger cost and expected performance. Furthermore, SO2 absorption efficiency decreases as the concentration of bisulfite and sulfite ions in the absorption solution increases, reaching 50% for an S(IV) concentration of 2 kmol m−3 in the liquid phase. To produce a concentrated bisulfite solution, a two-column system should be considered: one column to concentrate the bisulfite solution and the other to complement the gas purification. The results provide valuable insights into the key parameters influencing the desulphurization process, which can serve as a basis for scaling up to industrial applications. Future work will focus on validating the model under real operating conditions and assessing its adaptability to full-scale industrial units.
It would be interesting to extend this model into a dynamic one to accurately assess the influence of operating parameters on absorption performance over time. This will allow a more accurate comparison of the effectiveness of the model with scale tests. Finally, the model presented in this study is developed based on a simplified gas mixture (SO2 and N2). However, hazardous waste incineration flue gases contain other compounds, such as CO2, which can also be absorbed by the bisulfite solution. Therefore, the model will have to be enriched to include undesirable compounds that could alter absorption efficiency.
ACKNOWLEDGEMENTS
The project has received funding from the European Union's Horizon 2020 Research and Innovation Program under grant agreement no. 859318.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.