Sulfur Recovery Research Articles

A temporal convolution network-based just-in-time learning method for industrial quality variable prediction

Real-time acquisition of quality variables is paramount for enhancing control and optimization of industrial processes. Process modeling methods, such as soft sensors, offer a means to predict difficult-to-obtain quality variables using easily measurable process parameters. However, the dynamic nature of industrial processes poses significant challenges to modeling. For instance, conventional models are typically trained offline using historical data, rendering them incapable of adapting to real-time changes in data distribution or environmental conditions. To tackle this challenge, we introduce a novel approach termed the Residual Temporal Attention Temporal Convolution Network (RTA-TCN) and propose a just-in-time learning method based on RTA-TCN for industrial process modeling. The RTA-TCN model incorporates temporal attention into TCN, enabling the integration of previous time-step process variables into the current ones, as well as the fusion of internally relevant features among inputs. Moreover, to prevent the partial loss of original information during feature integration, residual connections are introduced into the temporal attention mechanism. These connections facilitate the retention of original feature information to a maximal extent while integrating relevant features. Consequently, the proposed RTA-TCN demonstrates significant advantages in handling the non-linearity and long-term dynamic dependencies inherent in industrial variables. Additionally, the proposed just-in-time learning method leverages RTA-TCN as a local model and updates it in real-time using online industrial data. This just-in-time learning method enables effective adaptation to varying data distributions and environmental conditions. We validate the performance of our method using two industrial datasets (Debutanizer Column and Sulfur Recovery Unit).

Unveiling the Dual Role of Oxophilic Cr4+ in Cr-Cu2O Nanosheet Arrays for Enhanced Nitrate Electroreduction to Ammonia.

Cuprous oxide (Cu2O)-based catalysts present a promising activity for the electrochemical nitrate (NO3-) reduction to ammonia (eNO3RA), but the electrochemical instability of Cu+ species may lead to an unsatisfactory durability, hindering the exploration of the structure-performance relationship. Herein, we propose an efficient strategy to stabilize Cu+ through the incorporation of Cr4+ into the Cu2O matrix to construct a Cr4+-O-Cu+ network structure. In situ and quasi-in situ characterizations reveal that the Cu+ species are well maintained via the strong Cr4+-O-Cu+ interaction that inhibits the leaching of lattice oxygen. Importantly, in situ generated Cr3+-O-Cu+ from Cr4+-O-Cu+ is identified as a dual-active site for eNO3RA, wherein the Cu+ sites are responsible for the activation of N-containing intermediates, while the assisting Cr3+ centers serve as the electron-proton mediators for rapid water dissociation. Theoretical investigations further demonstrated that the metastable state Cr3+-O-Cu+ favors the conversion from the endoergic hydrogenation of the key *ON intermediate to an exoergic reaction in an ONH pathway, and facilitates the subsequent NH3 desorption with a low energy barrier. The superior eNO3RA with a maximum 91.6% Faradaic efficiency could also be coupled with anodic sulfion oxidation to achieve concurrent NH3 production and sulfur recovery with reduced energy input.

Unveiling the Dual Role of Oxophilic Cr4+ in Cr‐Cu2O Nanosheet Arrays for Enhanced Nitrate Electroreduction to Ammonia

Cuprous oxide (Cu2O)‐based catalysts present a promising activity for the electrochemical nitrate (NO3‐) reduction to ammonia (eNO3RA), but the electrochemical instability of Cu+ species may lead to an unsatisfactory durability, hindering the exploration of the structure‐performance relationship. Herein, we propose an efficient strategy to stabilize Cu+ through the incorporation of Cr4+ into the Cu2O matrix to construct a Cr4+‐O‐Cu+ network structure. In situ and quasi‐in situ characterizations reveal that the Cu+ species are well maintained via the strong Cr4+‐O‐Cu+ interaction that inhibits the leaching of lattice oxygen. Importantly, in situ generated Cr3+‐O‐Cu+ from Cr4+‐O‐Cu+ is identified as a dual‐active site for eNO3RA, wherein the Cu+ sites are responsible for the activation of N‐containing intermediates, while the assisting Cr3+ centers serve as the electron‐proton mediators for rapid water dissociation. Theoretical investigations further demonstrated that the metastable state Cr3+‐O‐Cu+ favors the conversion from the endoergic hydrogenation of the key *ON intermediate to an exoergic reaction in an ONH pathway, and facilitates the subsequent NH3 desorption with a low energy barrier. The superior eNO3RA with a maximum 91.6% Faradaic efficiency could also be coupled with anodic sulfion oxidation to achieve concurrent NH3 production and sulfur recovery with reduced energy input.

Definiteness matters as a discourse cue in L1 and L2 processing of relative clauses

Definiteness matters as a discourse cue in L1 and L2 processing of relative clauses

Layer-wise-residual-driven approach for soft sensing in composite dynamic system based on slow and fast time-varying latent variables

Driven by the requirements for a comprehensive understanding of composite dynamic systems in industrial processes, this paper investigates a new soft sensor for quality prediction based on slow and fast time-varying latent variables extraction using layer-wise residuals. First, the slow feature partial least squares were expanded into long-term dependency by introducing explicit expressions of the potential state of the process into the objective function. Then, the multilayer regression model for exploring composite dynamics driven by layer-wise residuals is developed using a serial structure that can extract both slow and fast time-varying latent variables that are completely orthogonal. Finally, the exponential-weighted partial least squares are proposed for extracting fast time-varying dynamic latent variables by learning the exponential decay properties of the time-series data correlation. Case studies on the industrial debutanizer and sulfur recovery unit show that the prediction accuracy of the proposed approach outperforms traditional methods.

Enhancement and sulfur poisoning mechanism of Ce doped ZrO2 catalyst in COS hydrolysis at low-temperature

In order to achieve efficient and longer-lifespan removal of carbonyl sulfide (COS), a high-performance catalyst for hydrolysis is needed urgently. In this study, Ce doped ZrO2 oxides are used to remove COS with or without H2O. The results showed that the doping of Ce significantly improves the removal of COS by ZrO2 and obtains a sulfur capacity of 155.4 mgS/g at 70 °C, but there is still accompanied by sulfur poisoning. The characterization reveals that CeO2 has a cubic phase, ZrO2 has a mixed phase of tetragonal and monoclinic, and Ce-doped Zr with different ratio of Ce/Zr forms a Ce-Zr solid solution. Ce and Ce-Zr oxides have Ce3+/Ce4+ pairs, adsorbed oxygen species and surface oxygen vacancies, while ZrO2 has adsorbed oxygen and surface oxygen vacancies, the presence of which can promote the dissociation of H2O and surface adsorption of COS. ZrO2 has abundant medium-strong basicity while CeO2 has abundant weak and strong basicity, and different proportions of Ce/Zr adjust the surface basic strength, basic amount and surface redox property of Ce-Zr solid solution, which changes the adsorption and hydrolysis of COS and H2O and related oxidation reactions on Ce and Zr, thereby changing the sulfur poisoning. Ce has abundant –OH groups, which can rapidly adsorb COS, leading to hydrolysis and direct oxidation of COS in the absence of H2O, while the OH groups on ZrO2 promote the hydrolysis of COS to H2S. Although suitable Ce-doped ZrO2 improves the removal efficiency of COS, it still inevitably leads to the sulfur poisoning, but the presence of Zr could change the sulfation degree of CeO2 and the Claus reaction.

Enhanced decomposition of H2S to H2 via iron sulfide nanoparticles: A comprehensive experimental and theoretical evaluation

Hydrogen sulfide (H2S), a common pollutant in both fossil and bio-based fuels, can be transformed into valuable hydrogen (H2) through a two-step thermochemical process known as sulfur looping. This scheme, derived from the chemical looping approach, employs a metal-sulfide-based sulfur carrier as a solid intermediate. Iron sulfide, a benign and abundant material, is a promising candidate as the sulfur carrier; however, it faces challenges related to its poor kinetic activity. In this study, we demonstrate the exceptional performance of nano-sized sulfur carriers for the sulfur looping scheme through comprehensive experimental and theoretical analyses. We synthesized iron sulfide nanoparticles that exhibit a remarkable ∼69% enhancement in reactivity over the bulk-sized sulfur carrier. This enhancement is attributed to the high surface area and the nanoscale effects that enhance intrinsic activity. Furthermore, for the first time, we evaluated the effect of co-feeding carbon dioxide (CO2), a prevalent component in industrial acid gas streams alongside H2S. Our combined thermodynamic, atomistic density functional theory (DFT), and experimental studies indicate that the presence of CO2 enhances the conversion of H2S. Additionally, to gauge the sulfur looping process’s superiority at a larger scale, we conducted comprehensive thermodynamic analyses under industrial acid gas conditions. These analyses reveal an improvement of ∼22 percentage points in energy efficiency and ∼8 percentage points in exergy efficiency over the state-of-the-art Claus process. This study, integrating experiments, atomistic DFT studies, and process simulations, provides valuable guidance in the design of high-performance sulfur carriers for efficient H2S splitting to H2.

The construction of double short-cut sulfur autotrophic denitrification and PN/Anammox partition and its effect on the coupling performance of nitrogen and sulfur removal

How to reduce NO3--N from partial nitrification /Anammox (PN/Anammox) for low carbon emissions, thus achieve deep nitrogen removal, is a significant issue when applying PN/Anammox to treatment inorganic high-ammonia wastewater. This study designed a prepositioned double short-cut sulfur autotrophic denitrification (DSSADN) system integrated with PN/Anammox, achieving simultaneous sulfur recovery and complete autotrophic denitrification. It also examined the impact of reflux changes on the nitrogen and sulfur transformation characteristics in the coupled system when treating ammonia and sulfur-containing wastewater. Results indicate that the coupling of DSSADN zone (zone C1) enhanced TN removal by PN/Anammox, with the TN removal rate (TNRRT), TN removal efficiency (TNRET), and S2- removal rate reaching 0.54 kg/(m³·d), 95.1%, and 100%, respectively. The NO2--N accumulation efficiency (NiAEC1) of DSSADN reached 87.1%, with ΔSO42--S consistently less than 11.8 mg/L, and the S0 accumulation efficiency (S0AEC1) reaching 82.4%. A small amount of O2 carried in the reflux can effectively reduce residual S2- in zone C1, preventing its entry into the PN/Anammox zone (zone C2) and mitigating its toxic effects on Anammox. However, O2 also increases the expression of soxB in zone C1, causing more S2- to be oxidized to SO42-, and reducing S0AEC1. When DOC1 was between 0.05 mg/L and 0.34 mg/L, nirK expression was significantly increased, and the abundance of related functional microorganisms changed, led to more NO3--N being reduced to N2. When DOC1 exceeds 0.34 mg/L, it inhibits the denitrification process, leading to elevated NiAE. Therefore, when regulating reflux in DSSADN+PN/Anammox partitioned integrated coupling system, attention should be given to changes in DOC1, controlling it within 0.05 mg/L.

Coal to Clean: Comparing Advanced Electrodes for Desulfurization and Copper Recovery

This study explores the electrochemical desulfurization of coal and the recovery of copper (Cu) using dimensionally stable anode (DSA) electrodes. Background: The research addresses the need for effective sulfur removal from coal to reduce emissions. Methods: Electrochemical desulfurization was conducted using DSA and graphite electrodes, evaluating parameters like activation energy, desulfurization rate, and energy consumption. Cyclic voltammetry and linear sweep voltammetry were used to study the electrochemical properties. Results: The DSA electrode demonstrated superior performance with higher desulfurization rates, lower activation energy, and better response to temperature increases compared to the graphite electrode. Optimal desulfurization was achieved at 50 °C with the DSA electrode, balancing efficiency and energy consumption. Copper recovery from the solution post-desulfurization was effective, with an 86.34% recovery rate at −0.15 V vs. (Ag|AgCl). The energy consumption for the Cu recovery was calculated to be 10.56 J, and the total cost for recovering 1 ton of Cu was approximately 781.20 €. Conclusions: The study highlights the advantages of DSA electrodes for efficient sulfur removal and metal recovery, promoting cleaner energy production and environmental sustainability. Future research should focus on optimizing electrochemical conditions and scaling up the process for industrial applications.

Advances in Phosphogypsum Calcination and Decomposition Processes in Circulating Fluidized Beds.

Phosphogypsum (PG) is an industrial hazardous waste product discharged during wet-process phosphoric acid production. Once crystallized, the byproduct PG is filtered and separated from the liquid-phase product and sluiced to the disposal area near the production site for storage, seriously threatening the harmonious symbiosis between humans and nature. Therefore, devising effective solid waste management and cleaner production programs to contain and eliminate PG is of interest to researchers. In this study, the utilization status of PG is comprehensively reviewed, and a feasibility pathway for resourceful recovery of PG is proposed. The key challenges and countermeasures for the high-temperature calcination and decomposition of PG are analyzed and discussed. The visualization analysis based on bibliometrics reveals that the maximum recovery of abundant calcium (as CaO) and sulfur (as SO2) in PG and their utilization for the copreparation of calcium-based materials and sulfuric acid are the most suitable solutions for the large-scale application of PG. Five challenges that restrict the commercial promotion of PG calcination and decomposition processes are perfecting the calcium-sulfur conversion mechanism, establishing a process strengthening strategy, developing value-added technology routes, mastering unit scale-up regularity, and conducting sustainable performance assessment. Industrial applications are expected within 10-15 years.

Global dynamic features and information of adjacent hidden layer enhancement based on autoencoder for industrial process soft sensor application

AbstractThere is a lack of consideration of temporal and spatial correlation in the process variables and adjacent hidden layers correlation in the soft sensor model of stacked autoencoders. To address the issue, a novel global dynamic adjacent layer information enhancement auto encoder (GD‐ALIEAE) method is proposed to improve the poor prediction performance. The gated recurrent unit (GRU) and uniform manifold approximation and projection (UMAP) are applied to the GD‐ALIEAE model for obtaining global dynamic features of the temporal and spatial information of process variables by parallel computation. An adjacent layer information correlation algorithm is proposed to avoid the loss of hidden layers information during the stacking process. The algorithm enhances the features of the low layer through nonlinear mapping, combining the low layer and its adjacent layer as input. The input then is fed to the multi‐head attention mechanism to obtain features that contain adjacent layer correlation. Finally, a prediction model is established through a fully connected layer. Through simulation experiments on two industrial cases of sulphur recovery unit and thermal power plant, and compared with models of stacked autoencoder (SAE), stacked isomorphic autoencoder (SIAE), and target‐related stacked autoencoder (TSAE), the effectiveness of the proposed method was verified.

Highly effective direct conversion of H2S into COx-free H2 and S at low temperature over novel MoxC@ZrO2 microwave catalysts

Direct decomposition of H2S into H2 and S is an attractive alternative to the Claus process, because COx-free H2 and S can be simultaneously recovered from an abundant and toxic waste gas, which has huge social economic and ecological environmental benefits. However, it is still a great challenge to achieve the direct H2S decomposition reaction with high efficiency due to the thermal equilibrium limitation. Herein, a core-shell structured MoxC@ZrO2 was developed as novel microwave catalyst for microwave catalytic H2S decomposition. More specifically, the H2S conversion over MoxC@ZrO2 catalyst is as high as 99.9% at 650 °C, which immensely surpasses the H2S equilibrium conversion (6.1% at 650 °C). This study provides an effective strategy for the design of highly active microwave catalysts, and supplies a strategy for efficiently catalytic decomposition of H2S waste and recovering high-value hydrogen and sulfur at lower reaction temperatures.

Research on a semi-supervised soft sensor modelling method for complex chemical processes based on INGO-VMD-ESN

Abstract The dynamic and non-linear nature of complex chemical processes often leads to low prediction accuracy of key quality variables by traditional soft sensors, thus affecting the overall system control accuracy and operational efficiency. Therefore, this paper proposes a semi-supervised soft sensor modelling method based on improved the northern goshawk optimization (INGO)-variable mode decomposition (VMD)-echo state network (ESN). Firstly, a new semi-supervised fusion method is proposed to address the problem of model training difficulty due to the scarcity of labelled samples and process dynamics, which reconstructs the sample dataset by fusing labelled and unlabelled samples into more representative new samples, improving the model’s generalization ability. Secondly, for the noise interference present in the reconstructed data, the input data is denoised using the VMD method to improve the quality of data. Then, a soft sensor model is built based on ESN. Additionally, the denoising and prediction performance of VMD and ESN is significantly affected by parameters, therefore the paper utilizes the INGO algorithm to achieve parameter rectification for VMD and ESN. Finally, the method is validated based on actual sulphur recovery data from a refinery. The results demonstrate that the method effectively mitigates the impact of dynamics and nonlinearity in the complex chemical process which enhances prediction accuracy.

Recovering element sulfur from sulfur-containing waste through a pressure induced melting-crystallization method: Optimization using response surface methodology

Metallurgical sulfur-containing wastes (MSCWs) is an important secondary resource for recovering element sulfur in chemical and metallurgical industries. However, it is challenging to simultaneously achieve high recovery rate and sulfur purity, due to the multiple impurities combining with sulfur in these wastes. In this paper, we developed a novel method of pressure induced melting-crystallization to efficiently recover sulfur from MSCWs. The effects of condition parameters on the sulfur recovery rate and purity were optimized through response surface methodology. The single objective optimization results showed that the interaction between melting temperature and slurry concentration had a great influence on sulfur purity. The slurry concentration and pressure had great influence on sulfur recovery rate. The results of multi-objective optimization showed that under the conditions of pressure 1.2 MPa, slurry concentration 14.9% and melting temperature 140.9℃, the highest sulfur purity and recovery rate could reach 93.0% and 89.1% respectively. Mechanism investigation revealed that the high efficiency of sulfur recovery was ascribed to recrystallization of molten sulfur under the combined action of pressure and water, forming dense sulfur blocks and separating hydrophilic impurities. This work provides a new way to recover sulfur resources from MSCWs with sulfur content higher than 50%.

Efficient elimination of organic contaminants via a non-radical pathway involving activation of sulfite by synergistic adsorption-catalysis bifunctional chitosan-modified MOF derivative

Usually, Metal-based sulfite activation systems follow a radical generation pathway but are rare in non-radical pathways. Singlet oxygen (1O2) generation-targeted catalyst was synthesized by introducing non-redox metal oxides ZnO into chitosan-modified cobalt-zinc bimetallic oxides (CZ1.0C) via template method. Compared with Co3O4, the catalyst CZ1.0C coupled to the sulfite system (CZ1.0C/SO32-) maintained good adaptability in a common water background and a wide pH range (7–11). In addition, the CZ1.0C/SO32- system showed high efficiency in the degradation of Congo red (kobs = 0.190 min−1) and 100 % sulfite utilization. EPR, probe experiments, and DFT calculations proved the dominant role of 1O2, with a concentration three orders of magnitude higher than that of radicals. The controlled generation of non-radicals was achieved owing to the doping of ZnO and the conversion of lattice oxygen and oxygen vacancies. The catalytic oxidation intermediate HSO5- has also been demonstrated in the system, which achieves non-radical pathway oxidation through electron-mediated interactions. This article aims to develop a new “waste control by waste” strategy, which combines high selective adsorption and catalysis to achieve sulfur resource recovery and water pollutants elimination.

Modeling & experimental studies on enhancement of H2S conversion using catalytic membrane reactor for hydrogen production

Abstract Acidic gases such as H2S and CO2 are generated in the refineries during coal gasification. These pollutant gases need to be treated before releasing to the atmosphere. Conventionally, H2S gas is treated by Claus process in which H2S is converted to elemental S and H2O, in presence of O2 at ∼1,200 K. In the present energy scenarios, hydrogen has got importance as a source of clean fuel for industrial application. The H2S generated in the refineries can be a potential precursor of hydrogen. If this hydrogen can be obtained by thermal decomposition of H2S, that can increase the overall economic value of the H2S containing streams and the produced hydrogen can be used for hydrocracking and hydro-treating in refineries, as well as to synthesize value-added products for chemical industries. However, H2S thermal decomposition is an endothermic and equilibrium limited reaction (equilibrium conversion is only ∼15 % at 1,000 K), requiring a very high temperature (∼1,500 K) to achieve even a conversion of greater than 40 %. Packed Bed Catalytic Membrane Reactor (PBCMR) for thermal decomposition of H2S can be a potential technology augmentation and process intensification to increase this equilibrium conversion. In this work, modeling and experimental studies of H2S thermal decomposition using in-house designed & developed PBCMR have been carried out. The modeling studies were validated with experimental data. Clay-alumina ceramic tubular membrane (L: 250 mm; OD: 12 mm; thickness: 2 mm) was fabricated using extrusion process having an average membrane pore size of ∼1 μm and with a porosity of ∼20 %. Pt (2 %) coated alumina extrudes were used as catalyst to accelerate the reaction kinetics. Experimental studies showed that H2S to hydrogen conversion of ∼90 % is achieved using PBCMR at ∼1,523 K, compared to only ∼40 % conversion in a conventional packed bed tubular reactor (without membrane). Modelling studies were carried out to study the influence of operating parameters such as, reactor wall temperature, feed temperature, pressure and feed velocity. Studies showed that reactor wall temperature is having the most dominant effect on H2S conversion, which is confirmed by experimental findings. The studies offer useful insights into the application of PBCMR technology for management of waste gas stream containing H2S and recovery of hydrogen.

Tungsten-needle intensifies microwave-sustained plasma accelerating direct H2S conversion to H2

Direct sustainable conversion of hydrogen sulfide (H2S) enables collaborative recovery of H and S resources via a metal-enhanced microwave plasma strategy, avoiding the hydrogen waste in the traditional Claus process. However, the metal size effect on microwave plasma property, the optimal process parameters, and the enhancement mechanism remain unclear in H2S conversion. Herein, the optimal tungsten needle (diameter: 1 mm, length: 60 mm, and tip angle: 10°) is experimentally proven for intensifying microwave discharge in multi-mode cavities. Theoretical calculations and plasma distribution reveal that the optimized tungsten needle achieves the ideal coupling with the microwave field, exhibiting extreme electric field augmentation around the needle tip. Tungsten-needle intensifies microwave-sustained plasma, realizing 40.2 % (90.1 %) conversion of 100 % (10 %) concentration H2S to H2 at a low microwave power of 300 W with a good stability of 30 hrs. Low power, large flow rate, and high H2S concentration are beneficial for improving energy efficiency. The excitation of microwave plasma is accompanied by a massive generation of highly energetic electrons. The direct high-energy electron-H2S collision contributes a lot to H2S splitting, especially for high-concentration H2S. In-situ optical emission spectroscopy confirms the vital S and H radicals in the plasma. The free radical reactions triggered by electron collisions are responsible for the production of H2 and S. This work opens an avenue to sustainable and low-carbon hydrogen production from the direct conversion and utilization of H2S.

Dual temporal attention mechanism-based convolutional LSTM model for industrial dynamic soft sensor

Abstract Deep learning is an appropriate methodology for modeling complex industrial data in the field of soft sensors, owing to its powerful feature representation capability. Given the nonlinear and dynamic nature of the process industry, the key challenge for soft sensor technology is to effectively mine dynamic information from long sequences and accurately extract features of relevance to quality. A dual temporal attention mechanism-based convolutional long short-term memory network (DTA-ConvLSTM) under an encoder-decoder framework is proposed as a soft sensor model to acquire quality-relevant dynamic features from serial data. Considering different influences of process variables for prediction at multiple time steps and various locations, ConvLSTM and temporal self-attention mechanism are utilized as the encoder to adaptively fuse spatiotemporal features and capture long-term dynamic properties of process in order to capture the trends of industrial variables. Furthermore, a quality-driven temporal attention mechanism is employed throughout the decoding process to dynamically select relevant features to more accurately track quality changes. The encoder-decoder model meticulously analyses the interactions between process and quality variables by incorporating dual-sequence dynamic information to improve the prediction performance. The validity and superiority of the DTA-ConvLSTM model was validated on two industrial case studies of the debutanizer column and sulfur recovery unit. Compared to the traditional LSTM model, the proposed model demonstrated a substantial improvement with the accuracy R2 up to 97.3% and 94.9% and the root mean square error reducing to 0.122 and 0.022.

Innovative process for sulphur recovery from waste incineration flue gases: production of marketable sodium bisulphite solution

ABSTRACT This study presents an innovative process for recovering sulphur from hazardous waste incineration flue gases, designed to produce a marketable sodium bisulphite solution while ensuring complete SO2 removal. This new process is characterized by a double absorption strategy at two different pH levels. The first step, at an acidic pH, generates the desired bisulphite solution, while the second step, at a basic pH, produces the sulphite solution for recycling into the first step and ensures total SO2 removal. The process’s performance and feasibility were evaluated on a laboratory scale using a batch reactor with synthetic gas. The parametric study focused on the initial sulphite concentration in the absorption solution and the reactor temperature. A removal efficiency exceeding 95% was achieved across all initial sulphite concentrations and temperature ranges, when the pH was maintained above 6. At pH 5, where bisulphites are the predominant sulphur species, the removal efficiency remained substantial at approximately 70%. The oxidation of sulphites/bisulphites by oxygen in the flue gases was minimal, with less than 5% conversion to sulphate. Additionally, pH-controlled experiments were conducted to optimize plant start-up procedures. For the basic reactor, starting with water and adjusting the pH to 8 during SO2 absorption effectively minimized sodium hydroxide consumption. In contrast, for the acidic reactor at pH 5, initiating the process with a concentrated sulphite solution resulted in more stable absorption rates. These findings underscore the process's potential for efficient sulphur recovery and highlight the importance of pH management in optimizing operational stability and chemical consumption.

Soft Sensors for Industrial Processes Using Multi-Step-Ahead Hankel Dynamic Mode Decomposition with Control

In this paper, a novel data-driven approach for the development of soft sensors (SSs) for multi-step-ahead prediction of industrial process variables is proposed. This method is based on the recent developments in Koopman operator theory and dynamic mode decomposition (DMD). It is derived from Hankel DMD with control (HDMDc) to deal with highly nonlinear dynamics using augmented linear models, exploiting input and output regressors. The proposed multi-step-ahead HDMDc (MSA-HDMDc) is designed to perform multi-step prediction and capture complex dynamics with a linear approximation for a highly nonlinear system. This enables the construction of SSs capable of estimating the output of a process over a long period of time and/or using the developed SSs for model predictive control purposes. Hyperparameter tuning and model order reduction are specifically designed to perform multi-step-ahead predictions. Two real-world case studies consisting of a sulfur recovery unit and a debutanizer column, which are widely used as benchmarks in the SS field, are used to validate the proposed methodology. Data covering multiple system operating points are used for identification. The proposed MSA-HDMDc outperforms currently adopted methods in the SSs domain, such as autoregressive models with exogenous inputs and finite impulse response models, and proves to be robust to the variability of systems operating points.