Content area
The iodine-sulfur (IS) thermochemical cycle process is one method for producing renewable hydrogen fuel using water as a feedstock. To determine the research growth of the IS process, a bibliometric study was conducted using the Web of Science (WOS) database. The aspects of co-occurrence counting, co-authorship analysis (citation and publication), and organization bibliographic coupling were investigated using a well-established bibliometric analysis methodology. The WOS database was used to extract 386 IS article records published between 1980 and 2023 for bibliometric analysis. Significant publication growth has been observed, with China accounting for 41% of all publications. According to research trend analysis, the top ten IS research focuses on the Bunsen reaction mechanism, sulfuric acid decomposition, and catalyst research and application. Based on the analysis, only a limited amount of research has been done on dynamic IS modelling, IS plant simulation, and the design of the process control system for the IS process on an industrial scale. Future research prospects were suggested to encourage ongoing research on the plantwide control (PWC) strategy structure, which consists of advanced dynamic modelling, simulation and control strategies. This would aid in documenting the difficulties encountered in the IS thermochemical plant, thereby accelerating its commercialization.
Introduction
Hydrogen is one of the attractive renewable energy as an alternative to fossil fuels where it is proven with the best energy-weight ratio of 120 to 142 MJ/kg among all conventional fuels [1]. The thermochemical cycle is reported as one of the in-trend methods to produce renewable hydrogen [2]. A report on Solar Thermochemical Cycles for Hydrogen production (STCH) was published in 2011 and presented more than 300 methods of producing renewable hydrogen from numerous types of thermochemical cycles [3]. The extensive STCH report provides crucial insights into the thermochemical cycle processes where IS process is classified as one of the highest efficient processes. In the year of 2010, Zhang and co-workers [4] reported several important factors that can upsurge the potential for scaling up the IS which include all-fluid condition, the purely thermal mode and high thermal efficiency which is also supported by [5]. Furthermore [6] presented that the IS process attains the maximum energy and exergy competence leading to the minimum cost as the IS process can avoid the heavy use of electricity, unlike other types of thermochemical cycles. In addition, it can deal with massive production due to its flexibility to integrate with high-temperature solar systems [7, 8].
The IS plant consists of a number of essential chemical process equipment requiring proper design and reliable models to develop efficient control strategies. The vital process units in the IS plant include reactors and separators. Figure 1 illustrates the three sections of the IS process. Water decomposition occurs through chemical reactions with sulfur dioxide and iodine.
[See PDF for image]
Fig. 1
IS Process Flow Diagram: Section I, Section II, and Section III [9]
Current Research and Development
Bibliometric Study of IS Research
To analyze the development status of the IS process, a bibliometric study was proposed. A bibliometric analysis is a statistical method to identify publications globally for a particular research ground. The bibliometric analysis helps identify the reputable or developing research zones in a specific discipline, thus assessing and enumerating the development of the works of literature published [10].
The established systematic methodology for literature search and bibliometric analysis proposed by [11] was adopted. The keyword for “sulphur (sulfur) and iodine and thermochemical and cycle” was used for literature search in the Web of Science (WOS) database from the year 1980 until 2023. The literature search produced 386 publications with citing articles of 3304 and a cited number of 8059. An enormous number of publications and citations testified to the growth of interest in the IS process. Among the 386 publications, 41% of active publications are from China.
Subsequently, a thorough summary of IS research technology is performed by reviewing and analyzing the related published works of the WOS database. At this point, several tools from VOSviewer © software are utilized which are; term co-occurrence counting co-authorship analysis and organization bibliographic coupling analysis tools:
Analyzing the most popular research keywords based on the term co-occurrence counting tool: Fig. 2 shows the most popular terms presented in IS publications grouped in four different clusters (similarities). Keywords with a minimum of 25 occurrences were considered based on [11]. The terms were then extracted based on relevant scores out of 7273 phrases in the documents using the fully-counting method. Based on the relevance score, it indicates the top 3 popular terms focused on the “catalyst”, “Bunsen reaction” and “sulfuric acid decomposition” studies. The analysis found limited phrases on dynamic modelling, simulation, and process control design of the IS process in the literature. In other words, the research on dynamic modelling, simulation, and control aspects of the IS process is relatively scarce compared to other IS aspects, e.g., experimental work on catalysts.
Analyzing the authorship network using the co-authorship tool: Fig. 3 depicts the co-authorship networks among the researchers in IS research, worldwide. Meanwhile, Table 1 shows the top 10 authors’ collaboration strengths and citation scores. The total link strength column in Table 1 described the overall strength of the co-authorship links of a particular researcher with other researchers. In the ranking of the top most cited articles, three studies were focused on the process thermodynamics, catalyst material and Bunsen reaction.
Identifying the active organization using bibliographic coupling analysis tool: Based on the top 3 publications and citation number (Table 1), most of the studies conducted by established institutes such as the Japan Atomic Energy Agency (JAEA) and Institute of Nuclear and New Energy Technology (INET) are focused on lab and pilot-scale testing.
[See PDF for image]
Fig. 2
The most popular terms presented in IS publications are based on the full-counting method
Table 1. Top 10 authors’ collaboration strengths and score in the IS study.
No. | Author | Total publication | Institution (country) | % of total publications (386) | Total link strength | Citation |
|---|---|---|---|---|---|---|
1 | Wang Laijun | 47 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 12.18 | 146 | 390 |
2 | Zhang Ping | 47 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 12.18 | 146 | 400 |
3 | Chen Songzhe | 42 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 10.88 | 142 | 384 |
4 | Wang Zhihua | 34 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 8.81 | 143 | 509 |
5 | Xu Jingming | 33 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 8.55 | 91 | 338 |
6 | Zhang Yanwei | 33 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 8.55 | 141 | 502 |
7 | Chen Kefa | 32 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 8.29 | 134 | 491 |
8 | Zhou Junhu | 23 | Institute of Nuclear and New Energy Technology, INET of Tsinghua University (China) | 5.96 | 101 | 412 |
9 | Kubo Shinji | 22 | Japan Atomic Energy Agency, JAEA (Japan) | 5.69 | 84 | 369 |
10 | Onuki Kaoru | 19 | Japan Atomic Energy Agency, JAEA (Japan) | 4.92 | 50 | 311 |
(source: WOS, retrieved on 7th February 2023)
[See PDF for image]
Fig. 3
The co-authorship network between the top researchers in the IS study (colour legend represents the average citation per year from 10 to 25 citations)
Concerning IS technology, there is a gap in statistical research trend information especially on control, simulation and modelling-related topics. This gap motivates the rest of the review work aiming at the control, modelling and simulation-based IS technology, as reported in the current literature database using the bibliometric analysis technique. Therefore, in the next section, this review aims to report on available progress in modelling, simulation and control research on the IS process. Specifically, this review paper intends to highlight the real challenges in the dynamic process modelling and control studies of the IS process, which are not discussed elsewhere in the literature. Given the gaps, the current review paper also proposes some possible solutions to reduce those gaps. The details of these studies will be presented in the next section.
IS Pilot Plants
Process Description
One of the important factors that make IS gained attention for hydrogen production at the industrial scale because it does not generate harmful products [5]. The IS process consists of three prevalent reactions: Section I for the Bunsen reaction, Section II for the H2SO4 decomposition to produce sulfur dioxide (SO2) and oxygen (O2), and Section III for the hydrogen iodide (HI) decomposition to hydrogen (H2) and iodine (I2). Equation (1) to (3) show the primary chemical reactions involved [12].
Section I, Bunsen reaction: Exothermic Reaction ∆H = -165 kJ/mol
1
Section II, sulfuric acid decomposition: ∆H = + 173 kJ/mol
2
Section III, hydrogen iodide decomposition: ∆H = + 371 kJ/mol
3
Section I requires a temperature of more than 800oC, and Section III requires a relatively lower temperature of approximately 500oC. Since the temperature requirement for the IS process is very high, only limited heat sources are available to power the process. Two sources currently under intense investigation are nuclear energy and solar energy [13]. Another potential energy source for the IS process is hydro energy; however, this has only received little discussion in the literature.
At present, the prevalent IS research institutions are following the pathway pioneered by the General Atomic (GA) research centre by also proposing heat supply from nuclear power plant facilities to power the industrial scale IS plant [14, 15–16]. Nuclear energy has been the preferred heat source for the IS process ever since the scheme’s introduction by GA over three decades ago [17]. Later on, solar energy gradually received attention as an alternative to nuclear energy. Compared to nuclear energy, recently, solar energy has become the focal point as it is safer and more practical [18].
Pilot Plant
The bibliometric study has revealed several prominent institutes or research centre that is currently actively pursuing IS research on the lab scale and pilot plant scale. As an institution, the JAEA and INET have produced the most publications on IS with high average citations which was presented in Table 1. Formerly known as Japan Atomic Energy Research Institute, the JAEA has been venturing into research and development of the IS process since the early 1990s. It accommodated a specially designed reactor operated at high temperatures, renowned as the biggest pilot plant of a volumetric flow of 25,000 Nm3/d. It was the highest temperature (950oC) reactor operating at JAEA [19]. The goal of JAEA is to seek better components and stability to produce stable oxygen and hydrogen production rates by the H2SO4 and HI decomposers, respectively [15]. The integrated High-Temperature Gas-Cooled Reactor (HTGR) and IS process has become the preferred scheme by the JAEA due to its many advantages on oxygen and hydrogen production rate [20]. In 2019, is reported the duration of the hydrogen production operation successfully extended to 31 h with a rate of 20 NL/h of hydrogen production [21]. Starting from 2016 until the present, JAEA is focusing to produce more than 1000 Nm3/h of hydrogen by using helium-heated, High-Temperature Engineering Test Reactor (HTTR) of plant equipment made of industrial materials.
Besides Japan, the INET a research centre of Tsinghua University China also been active in research on the fundamental development of hydrogen production powered by nuclear since 2005. The INET has focused on studying the behavior of the Bunsen reaction products separation. Since three decades ago, the INET has researched the HI decomposition of Section III using Pt-based catalysts. A proof-of-concept facility reported by [4] demonstrated the constructed lab-scale facility with an integrated nuclear-powered hydrogen production capacity. The facility successfully produced 60 NL/hr of hydrogen in the year 2014 [22].
It is interesting to note that in most publications the mainstream works by the JAEA and INET have only centered on laboratory-based research activities, especially on reaction behaviors [22]. Nevertheless, limited evidence remains on the dynamic (state) modelling, plant simulation and process controller design of the IS process on an industrial scale.
IS Process Constraints
Thermochemical processes not only gained attention in hydrogen production but have also proven to be effective to produce other types of renewable fuels [23]. From the mid-1970s till the mid-1980s, the GA conducted an extensive simulation-based research program on the thermochemical cycle processes, including the IS process [24]. The GA is the first to apply modelling and simulation-based research to develop the IS process. To date, the studies focused on the development of the general first principle model. In such studies, the mainstream objectives are to determine the right process operation settings. According to a report from Perret (2011) [3], the GA successfully modelled the binary parameters for Section II but remained unsuccessful in modelling the ternary parameters for the HI decomposition (Section III) through Aspen Plus simulation. There are a few significant constraints in the IS process that could be of great interest to the dynamic modelling, simulation, and process control design:
The chemical species in Section I include high-concentration acids (H2SO4 and HI) that can form immiscible liquid phases [25]. Iodine and sulfur must be fed to the Bunsen reactor at the proper composition ratio as well. Only 2.69% of water decomposed into hydrogen at 1 pressure and 2500 K, according to [26] analysis of thermodynamics. The percentage decomposition would increase to 25% at a lower pressure of 0.05 bar. Therefore, it is vital to study the optimum operating pressure and temperature. The prediction of the phase states and the determination of the phase compositions are two key components that are necessary for the simulation of Section I [26]. The HI, I2, H2O, and H2SO4 chemical species co-exist in the Bunsen reactor. This phenomenon represents a challenge as an appropriate modelling method shall capture the realistic behaviors of the phase states. The homogeneous phase, the two-liquid phase, and the two-liquid phases with I2 precipitation are the three potential states in the Bunsen reactor. A reliable model should have the ability to predict all possible states under different operating conditions. For example, if operating at the two-phase state is desired, then a precise solution composition calculation is necessary to attain this state. To create the ideal amounts of HI and H2SO4, the feed of SO2 pressure and the H2O feed flow rate must be controlled via a well-designed control system [27].
Units in Section II must operate at temperatures exceeding the critical temperature of water (374oC), which is roughly 800oC. According to reports, the IS plant’s sulfuric acid decomposer unit required the maximum temperature to function [28]. Because the H2SO4 decomposition is an endothermic reaction, it is vital to provide sufficient high-temperature heating medium to the unit, which poses a significant challenge. A higher temperature often favors a better decomposition efficiency. However, the higher the temperature, the more energy consumption to heat the heating medium. To avoid a potentially harmful process runaway reaction, the reactor’s temperature must stay below 1140oC [29]. It is possible to save energy consumption if the operating reactor temperature can shift to a lower value that keeps the operational efficiency at the same value as when operating at relatively higher temperatures. Such reduced operating temperature, if attainable, can also lead to better safety conditions. The high operating temperature issue remains a real challenge to designing Section II effectively. Furthermore, the system’s ability to separate water and sulfuric acid before they enter the H2SO4 decomposer should be adequate. Because the IS process is a cycle, failing to remove water reduces cycle efficiency, lowering hydrogen production efficiency [29].
Multiple liquid phases, azeotropes, and solid precipitation are some of the complicated nonlinear behaviors of the HI-I2-H2O system that Section III frequently meets [17, 30, 31–32]. Comparing Section I and II to Section III, the Section III’s current understanding from a thermodynamic standpoint is still insufficient. Non-ideal solutions and incompletely immiscible systems are formed by the binary mixture of HI-H2O and the ternary mixture of HI-I2-H2O in Section III. It is challenging to model and predict the thermodynamic behavior of immiscible systems. Additionally, a high concentration of HI is prevented by the production of an azeotrope in the binary mixture of HI-H2O. Low levels of HI cause HI to slowly and incompletely decompose into H2 and I2. Due to the enormous calorific capacity of the HI mixture, which contains a significant amount of iodine and water, this issue results in high energy expenditure. The JAEA developed an Electrodialysis (EED) technology to address the problem [33]. A few years earlier, the GA had recommended an extractive distillation technique that uses phosphoric acids (H3PO4) as a medium to distil HI solution above the azeotrope [34]. However, the H3PO4 itself should be recoverable in a separate cycle, which often consumes a large amount of energy. Before the GA recommendation, researchers at the Aachen University of Germany proposed a reactive distillation column in which the concentration and decomposition processes take place concurrently [35]. A problem may arise if the HI content of the liquid phase increases as it ascends the distillation column, where more HI molecules are likely to leak. As a result, the vapor phase HI content in the upper region of the column rises above than that in the lower. Furthermore, due to the inherent characteristics of the distillation column, the temperature is lower in the upper distillation region. The thermal decomposition of HI occurs significantly only when the operating temperature exceeds the minimum value of 330oC [35]. This requirement causes a high temperature throughout the column, posing another challenge. The difficulty is maintaining a two-phase flow for distillation under such extreme conditions, where the system pressure must be sufficiently high. The environment inside the reactive distillation column becomes extremely corrosive as a result of the high pressure and temperature.
The actual implementation of the IS process is significantly more complex than just showing how its chemical processes work. Numerous recycling streams are steadfastly emphasized throughout the entire IS process. As a result, it is challenging to assess the dependability of a proposed flowsheet using solely experimentation tools. Following the present discussion, previous studies have demonstrated that the cause of the lack of dynamic modelling, simulation and control research of the IS is comparatively owed to the complex nature of the IS process itself. In other words, the vital parts that constitute the IS plant remain under extensive research, i.e., immature status and incomplete information make it hard to develop a rigorous dynamic model for simulation and control studies. Consistent with the literature, it is endorsed that there is abundant room for further IS research in the modelling, simulation and process control cluster.
Dynamic Modelling, Simulation and Control System for the IS Process
This section presented the research and prospects of the control system, simulation and dynamic modelling for the IS process.
Control System for the IS Process
It is reported that the hydrogen production processes under research worldwide are growing with the challenge not only of developing a safe process at a commercial scale but including efficient storage and transportation [36]. Therefore, safety is of paramount importance to ensure efficient IS process operation at the industrial scale. As mentioned previously, JAEA and INET have dedicated significant research and development efforts, particularly on reaction mechanisms to ensure a stable IS process. Control systems play an essential role in ensuring the stable, safe and consistent operation of a chemical plant. Consequently, it is imperative to conduct a thorough analysis of how the consistency of the industrial-scale IS hydrogen production facility is affected when control system components fail [20]. An appropriate control system is required to deal with complex dynamics such as inverse response, dead time, process disruptions, and nonlinearities. The control system should also be able to deal with changing operational conditions as a result of unmeasured disturbances.
Currently, the conventional controller is the most widely used to control industrial processes. Proportional-Integral-Derivative (PID) controller tuning methods have evolved over the last few decades. Some researchers have also investigated the use of PID controllers in various thermochemical cycle processes. Sack et al. (2012) [37] created a Proportional-Integral (PI) controller with multiple input-single outputs (MISO) to control a receiver-reactor for thermochemical water splitting integrated with a solar tower system. Researchers [16] observed that the process restrictions, such as the requirement for a minimum operating temperature for HI conversion, made the hydrogen iodide decomposition reactor in Section III susceptible to overheating. In a work by [20], it is shown that the proposed control scheme to control the heat source generation of the plant was able to compensate for the disturbances. However, its validation remains necessary in further work. Other researchers used a network control method to create a different control system [38]. The proposed method showed the ability to handle the challenge arising from the multivariable interactions in the controlled system. Nevertheless, there have been no process verification and validation studies to ensure the method’s dependability.
The PID-based controller’s ability to effectively regulate the entire plant may be considerably diminished by the many limitations in the IS process. Additionally, the design of the control system is frequently severely hampered by process nonlinearity. Loads of recycling lines, non-stationary behavior in the systems, and the time delay of the sensors are the additional obstacles [39]. In general, the application of PID in local areas of the plant has been prompted by the high expense of modelling and the simplicity of controller design requiring little need for models. However, when used on the complete plantwide system, PID controllers will perform relatively poorly. Additionally, model-free approaches, which frequently lack comprehensive information on a particular process, may result in a low-performance controller [40].
Because of the limited dynamic models and complex system behaviors (e.g., multiple constraints and nonlinearity), the efficient design of a control system for the IS process under realistic conditions remains a big challenge. Due to these issues, there has been an increase in interest in studying sophisticated control techniques, notably model-based controllers to govern the process [40, 41]. The Internal Model Control (IMC) is one of the well-known model-based control systems as well as the Generic Model Control (GMC) and the Model Predictive Control (MPC) [40, 42, 43, 44–45].
Dynamic Modelling for the IS Process
Numerical solutions are capable of producing reliable designs through accurate prediction of the flow field details while having relatively low costs as compared to traditional theoretical methods and experiments. This is especially useful in measuring working conditions that are hard to measure in experiments and has been implemented in the study of heat transfer, species transport, and multiphase flow [46]. Earlier modelling research on the thermochemical cycle process has centralized on the cycle efficiency enhancements particularly, the analysis of static process behavior [25, 47, 48–49]. Very few efforts have been made to develop and analyze the process dynamics and modelling of the whole IS plant.
Simulation Works
The modelling of the main reactor, cooling systems, heating equipment, and separators is part of the systematic system of IS process development via simulation. Ying et al. consider and simulate an alternative method for HI decomposition via electrolysis [13]. Based on the theoretical and experimental analysis, the HI electrolysis model with a proton exchange membrane electrolytic cell was developed. The flowsheet was then simulated in Aspen Plus using a user-defined electrolytic module to integrate the HI electrolysis in the IS. The IS process is simulated using the ELECNRTL electrolyte model. The flowsheet is set up to produce 10 L/h of H2. With proper internal heat exchange, the proposed flowsheet has a theoretical thermal efficiency of 25% to 42% and a thermal efficiency of 33.3%. Sensitivity analysis was then performed on the HI decomposition and H2SO4 decomposition sections to determine the main factors influencing overall performance. In the HI section, increasing the ratio from 0.9 to 0.98 increases overall thermal efficiency from 32.1 to 33.9%. Meanwhile, the distillate-to-feed rate ratio, the plate number of the distillation column, and the H2SO4 conversion ratio are the most important influencing factors for the H2SO4 decomposition section. [50].
Similarly, Wang and co-workers [2] also simulated two improved flowsheets for the IS coupled with a VHTR via Aspen Plus. The two proposed configurations for the IS-based hydrogen and electricity cogeneration system are the series connection system (SCS) and the parallel connection system (PCS). The simulation shows that 99% of the energy consumption and over half of the heat consumption can be attributed to the H2SO4 and HI sections, while 80% of the electricity consumption is used by the EED cell for the production of over-azeotropic HI solution. After implementing an internal heat exchange network to recover the heat released, the thermal efficiency was increased from 17.7% to an ideal thermal efficiency of 43.3%. The PSO algorithm is then used to optimise the system, and the performance of the two proposed systems is compared under different operating conditions. In a nutshell, as the demand for hydrogen rises, the total efficiencies of both systems fall, with PCS falling faster than SCS. If the hydrogen production capacity is low, PCS performs better meanwhile SCS performs better when the production capacity is high.
Other than simulating the flowsheets, simulation studies have also been carried out to investigate the feasibility of improved equipment to carry out specific reactions in different sections of the SI cycle. For example, [13], a simulation study was carried out for the Bunsen reaction using an electrochemical cell (EC) to improve the efficiency of the IS process. The simulation results show that the outlet HI concentration (6.40 to 6.78 mol/L) fails to reach the azeotropic concentration (7.46 mol/L), and it is noted that the over-azeotropic state can be achieved if there was a higher initial HI concentration, longer channel length, or smaller value of sulfuric acid. Besides, the effect of current density, temperature and flow rate on the reaction rate is also investigated. On the other hand, Gao et al. (2021) [51] conducted a simulation study on the structural design of the catalytic zone for the H2SO4 bayonet heat exchanger decomposer. Three optimized structures were proposed for the catalytic zone, with different dimensions and the number of inner tubes (single or double tubes). ANSYS Fluent is used to solve the species transport model and ICEM is used to generate the tetrahedral grid for the geometric model of the reactor. It is shown that double inner tubes with a bigger cross-sectional annulus area perform better. This design gives a higher flow velocity and a larger catalyst volume, which promotes higher heat transfer and decomposition rates [51].
Control System for IS-related unit Operations
The IS process is being researched worldwide, and different flowsheets are being developed for each section in the IS to improve its efficiency and optimize the cost [52]. In the flowsheet developed by GA, a bayonet acid decomposer is used, while the flowsheet studied by Korea Atomic Energy Research Institute (KAERI) uses a shell & tube model to simulate the sulfuric acid decomposer, and a catalyst-packed tube and shell to simulate the SO3 decomposer. GA uses a newly designed decomposer which integrates the function of the sulfuric acid vaporizer, superheater and decomposer into one single unit [53]. This integration helps to reduce the number of unit operations. In the GA flowsheet, only two units were used, which are the concentrator and the bayonet acid decomposer. In the KAERI flowsheet, four units were used, which are the concentrator, vaporizer, and two decomposers. By integrating the evaporator with the decomposer, the energy demand can be reduced [54]. In these flowsheets, there are two types of major equipment to be controlled which are continuously stirred tank reactor (CSTR) and tubular reactor. Due to limited work available, the applications of a process controller in other similar process equipment will be referred to illustrate on IS process. The following sections review the process controller research on those two types of equipment in various chemical processes.
Continuous Stirred Tank Reactor (CSTR)
The design of the Bunsen reactor of the IS process can be depicted by a CSTR in actual and practical applications. Inevitably, CSTR comprises a wide application in the chemical industry. CSTR can be divided into two types; with a jacket (exothermic or endothermic) and without a jacket. A study by [55] on a simple reflux ratio control of a CSTR with one recycle line was carried out to investigate whether the self-optimizing control scheme for the recycling process is relevant to different cases or processes. They chose an MPC for the supervisory layer and used a self-optimizing control structure. When compared, the MPC and PI both perform well in controlling the nominated layer. They reached a new finding where the scheme is proven applicable to processes with a bounded and unbounded process. These findings made the MPC scheme particularly a suitable controller to control a Bunsen reaction in the IS process which consists a number of constraints to tolerate. The research will have been more relevant if it addresses more than one recycles line which is a typical case in the current industrial plant.
In the search to overcome the inadequacy of the conventional controller in controlling bioprocesses, [56] created an NMPC controller to control a bioethanol production integrated CSTR system. The CSTR was designed to be controlled by a NARX-based NMPC. In the setpoint tracking, disturbance rejections, and robustness tests, the NMPC outperformed the LMPC (using a state-space model) and conventional PID controller. The model’s accuracy is critical in controlling the controller’s action performance. Even though the NARX model is straightforward in its implementation in the MPC, there is still room for improvement. Expanding the NARX model study by simplifying the NARX model structure shall become more reliable for a plantwide control application, especially for developing an IS commercial plant.
In 2017, a Neural Network-Based NMPC has been developed by [57] to control a CSTR. The performance of CSTR was evaluated using a deep learning neural network MPC. To design the weight of the neural network model, a hybrid particle swarm optimization-gravitational search is used. It was discovered that the controller had a lower ISE than other types of Neural Network-Based MPC. However, it seems that the ISE results of the other three controllers (including a PID controller) which were used as a comparison in controlling the particular CSTR were less than 0.5. This value is already considered a very minimum error in the process control field. This indicates that the PID controller was already performing very well in controlling the process. Hence, there is no need for the author to propose or use any other controller, moreover, a modified one is undoubtedly more complex than the existing PID controller.
Another deep learning model that was successfully applied in computer vision is the deep belief network (DBN). Its main strength is its strong feature-learning capability in detecting nonlinear dynamics between given data pairs. Wang and colleagues [2] created a DL-based MPC to control a second-order CSTR, which employs a growing DBN (GDBN) model as a predictive model for the MPC. The novel DeepMPC was compared to other MPCs in their study, including differential recurrent NN-based MPC (DRNN-MPC), adaptive NN-based MPC, and others. The DeepMPC outperforms the other MPCs in terms of maximum overshoot, rising time, averaged value, and variance of root-mean-square error (RMSE) in a tracking process. But, it has a slower rising time than the DRNN-MPC. It was remarked that this may be due to the improper assignment of hyperparameters [2].
Kumar and co-workers [58] proposed an internal model control (IMC) PID controller to control the reactor temperature by altering the jacket feed flow rate and the controller was proven to be a success after testing on a non-linear bioreactor process model under set-point change and disturbance rejection [57]. The approach displayed improved closed-loop performance in settling time and IAE compared to the previously published method.
Another paper [59] proposed a novel (least squares support vector machines and Laguerre filters) LSSVM-L Hammerstein model structure for CSTR servo and regulatory control. The goal is to control reactant temperature and product concentration by varying the coolant flow rate while taking into account the effect of load changes on feed flow rate and feed temperature. The proposed model provides a good model fit by implementing the optimized tuned parameters, with its variance accounted for (VAF), MSE, and RMSE being around 98%, 0.1208, and 0.3475, respectively. The proposed NMPC with the LSSVM-L Hammerstein model and LMPC with the Laguerre model were then compared. When disturbances were introduced, the proposed NMPC has lower ISE and IAE performance indices, indicating that it has better regulatory control than the LMPC. The NMPC also has a slightly better total variation (TV) performance than LMPC, where it has a smoother control signal and utilization to adjust the manipulated variable [59].
Besides controlling temperature, the control of pH in a CSTR is also investigated. A region-less explicit MPC-based reference governor was designed to improve the performance of a PI controller in controlling the pH of the CSTR by manipulating the pump voltage. The region-less explicit approach is used to help reduce the memory requirements and make it less computationally demanding. The proposed control method has mitigated the overshoot of the controlled variable, thus saving medium and energy consumption by about 2%, as compared to the conventional PI controller [60]. Meanwhile, in [61], a conventional robust MPC was designed to control the pH by manipulating the volumetric feed flow rates of the acid and base. A real-time compensation of the CSTR’s asymmetric behavior was proposed to allow the switching between two appropriate control setups according to real-time operating conditions. This implementation has improved the performance of quality criterion ISI and ISDI, indicating a reduction in production and maintenance costs. Further studies can be conducted to compare this control method with other conventional control methods.
Tubular Reactor
A tubular flow reactor is a vessel with continuous flow. The flow is usually steady and configured so that the chemical conversion and other dependent variables are functions of position within the reactor rather than time. Tubular reactors are commonly used for gaseous reactions, but they can also be used for some liquid-phase reactions [62]. Tubular flow reactors are mostly found in the chemical industry and wastewater treatment plants. Because of the high time delay in these types of reactors, controlling output variables is extremely difficult.
Song and co-workers [63] proposed an improved event-triggered controller to control the temperature and concentration profiles of a chemical tubular reactor subjected to singular perturbations. To reduce network communication load, they also proposed an event-triggered mechanism (ETM) for PDE systems based on spatiotemporally sampled data. SPPDE is the abbreviation for the system, which is described using partial differential equations (PDE) with singular perturbations. An asynchronous fuzzy point controller was designed using T-S fuzzy rules, and a point-control approach was used to reduce controller design costs. The temperature and concentration profiles are well-controlled, according to the closed-loop simulation results. A non-linear state feedback control was designed by [64] for a tubular reactor, where the distributed jacket temperature is used as a control input. Using the availability function and reduced availability function as Lyapunov candidate functions, four non-linear feedback control laws are designed. Closed-loop simulation results show that the proposed feedback controls with reduced availability function improve the performance of the closed-loop system in terms of input control amplitude and settling time. Besides, the proposed control law also gives a robust performance when perturbations of the inlet species’ temperature are introduced.
A feedforward-output feedback (FF-OF) robust controller was designed by [7] for a jacketed tubular reactor to control the product gas composition by regulating the cooling jacket temperature while handling the disturbance in feed temperature. A feedforward (FF) concentration regulatory controller and a temperature tracking controller were designed and combined to yield the composition-temperature cascade controller. Then, the controller is combined with a geometric observer to produce the robust FF-OF geometric controller. According to the simulation results, the designed controller with a dynamic setpoint compensation based on the measured feed temperature disturbance performed better than its PI controller counterpart. The proposed controller can give a robust performance, reduce the variability of the exit composition by 95% as compared to the conventional PI controller, and maintain the temperature of the hot spot within bounds [7].
Zhang and colleagues [65] also developed a state-feedback model predictive controller and a Luenberger observer-based MPC for a jacketed tubular reactor with a non - stationary jacket temperature. The research looked at a reversible exothermic reaction (A B), which can be described by nonlinear coupled 4 × 4 hyperbolic PDEs. The Cayley-Tustin method is used to linearize and discretize the reactor’s nonlinear continuous-time model. Based on simulation case studies, the recommended MPC designs converged to the stable condition quicker than the open-loop profile while satisfying the system’s physical constraints. In presence of disturbances, the proposed MPCs provide good control performance. In terms of settling time and absolute error, the state-feedback MPC performs slightly better than the observer-based MPC [65]. In addition, [66] designed a similar observer-based MPC for an axial dispersion tubular reactor with recycling. The system was able to achieve stability within the defined constraints while rejecting disturbances by regulating the cooling jacket temperature.
Franco-de los Reyes and colleagues produced a model-based MIMO controller for a multi-jacket exothermic tubular reactor to regulate the jacket temperature control inputs based on the estimated temperature-concentration state profile. Incorporating a pointwise measurement injection observer (PWO) with a MIMO SF passive controller yields potential the for better controller design. The proposed controller’s performance is compared to that of an adaptive controller in single-jacket and two-jacket cases. The simulation results show that the suggested single-jacket controller performs the best [67].
Conclusively, the process control research and development for CTSR and tubular has evolved significantly. For some processes PID is still workable but for the others, the advanced process control shows better regulatory control and robustness. As both conventional and advanced process control use different plant instrumentation and facilities hence it is important to study the process control and venture the area thoroughly prior to IS process commercialization.
Plantwide Control (PWC) for the IS Process
The previous section presented the research and prospects of the control system, simulation and dynamic modelling for the IS process. It can be seen in those research, the control system and dynamic modelling of IS process design have been done disparately. In other words, those works are confined to particular sections or equipment of the thermochemical cycle process, i.e., not on the entire plantwide scale. Hence, a necessity to propose an effective dynamic model and process control system for the plant scale to assess the feasibility of the designed flowsheet is indispensable.
There are numerous measurements and control loops in the IS process. As a result, a PWC strategy is the best candidate for controlling the entire plant system. Since Buckley’s pioneering work four decades ago, the PWC design has sparked a lot of research interest in the process control community (1964). The PWC is defined as an overall plant control philosophy that emphasises architectural decisions.
Integrated system design is required to address significant operational difficulties in all three sections of the IS process. Thus far, the PWC integrated plant research project of the IS process has received little attention from the scientific community. Furthermore, there has been no report on the entire industrial-scale IS process. The main objective of the PWC structure analysis can be to achieve a flow sheet design that is both financially viable and dynamically controllable. There has been no integrated study of the IS process to date. Overall, the PWC strategies enable the incorporation of advanced and traditional approaches into a sensible approach for the sophisticated and novel IS process.
In the PWC strategy, one can adopt both decentralized PID and MPC schemes, or any integration of conventional and advanced control systems. A decentralised PID control scheme, on the other hand, is incapable of preventing the reactor’s input constraints from being violated. It is worth noting that the PWC will effectively handle all of the previously mentioned challenges.
Conclusion
A bibliometric study for the IS process has been performed. It was found that the most studied research area of IS is on the Bunsen reaction mechanism, sulfuric acid decomposition, and catalyst application led by INET in China and JAEA research institutions in Japan. The contribution of this study has been to confirm the research trend of IS process. The findings presented here shed new light on the scarcity of work on dynamic modelling, simulation-based, and process control research on the IS process. Following that, an evaluation of the research and prospects for dynamic modelling, simulation, and process control for the IS process was performed. Some IS process control studies concentrated on enhancing cycle efficiency, particularly static characteristics, with few efforts made to explore process dynamics and modelling of the whole IS system. More effort should be made to improve the competency of the IS process so that it can be economically available during the commercialization stage. To resolve the previously mentioned technical challenges, a PWC strategy comprised of advanced dynamic modelling and control strategies is recommended to control the entire industrial IS plant. In a nutshell, the IS process possessed a bright future breakthrough via the route of dynamic modelling, simulation and process control research henceforth successfully achieving the industrialization of sustainable hydrogen production.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
1. Rasul, MG; Hazrat, MA; Sattar, MA; Jahirul, MI; Shearer, MJ. The future of hydrogen: Challenges on production, storage and applications. Energy Convers Manag; 2022; 272, 116326. [DOI: https://dx.doi.org/10.1016/j.enconman.2022.116326]
2. Wang, G; Jia, QS; Qiao, J; Bi, J; Zhou, M. Deep learning-based Model Predictive Control for continuous Stirred-Tank Reactor System. IEEE Trans Neural Networks Learn Syst; 2021; 32,
3. Perret R (2011) “Solar Thermochemical Hydrogen Production Research (STCH) Thermochemical Cycle Selection and Investment Priority,” Sandia Rep, no. May, pp. 1–117,
4. Zhang, P; Chen, SZ; Wang, LJ; Yao, TY; Xu, JM. Study on a lab-scale hydrogen production by closed cycle thermo-chemical iodine-sulfur process. Int J Hydrogen Energy; 2010; 35, pp. 10166-10172. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2010.07.150]
5. Madadi Avargani, V; Zendehboudi, S; Cata, NM; Saady,; Dusseault, MB. A comprehensive review on hydrogen production and utilization in North America: prospects and challenges. Energy Convers Manag; 2022; 269, 115927. [DOI: https://dx.doi.org/10.1016/j.enconman.2022.115927]
6. Mehrpooya M, Habibi R (2020) A review on hydrogen production thermochemical water-splitting cycles. J Clean Prod 275. https://doi.org/10.1016/j.jclepro.2020.123836
7. Nájera, I; Álvarez, J; Baratti, R; Gutiérrez, C. Control of an exothermic packed-bed tubular reactor. IFAC-PapersOnLine; 2016; 49,
8. Sattler C, Roeb M, Agrafiotis C, Thomey D (2017) “Solar hydrogen production via sulphur based thermochemical water-splitting,” Sol Energy, vol. 56,
9. Sakaba, N et al. Hydrogen production by thermochemical water-splitting IS process utilizing heat from high-temperature reactor HTTR. Whec 16; 2006; no June, pp. 1-11.
10. Ismail SA, Ang WL, Mohammad AW (2020) “Electro-Fenton technology for wastewater treatment: A bibliometric analysis of current research trends, future perspectives and energy consumption analysis,” J. Water Process Eng, vol. 40, no. December p. 101952, 2021, doi: https://doi.org/10.1016/j.jwpe.2021.101952
11. Sillero, L; Sganzerla, WG; Forster-Carneiro, T; Solera, R; Perez, M. A bibliometric analysis of the hydrogen production from dark fermentation. Int J Hydrogen Energy; 2022; 47,
12. Kubo, S et al. A demonstration study on a closed-cycle hydrogen production by the thermochemical water-splitting iodine - sulfur process. Nucl Eng Des; 2004; 233, pp. 1-3. [DOI: https://dx.doi.org/10.1016/j.nucengdes.2004.08.025]
13. Ying, Z; Wang, Y; Zheng, X; Geng, Z; Dou, B; Cui, G. Experimental study and development of an improved sulfur–iodine cycle integrated with HI electrolysis for hydrogen production. Int J Hydrogen Energy; 2020; 45,
14. Cho, WC; Park, CS; Kang, KS; Kim, CH; Bae, KK. Conceptual design of sulfur-iodine hydrogen production cycle of Korea Institute of Energy Research. Nucl Eng Des; 2009; 239,
15. Kasahara, S et al. Current R&D status of thermochemical water splitting iodine-sulfur process in Japan Atomic Energy Agency. Int J Hydrogen Energy; 2017; 42, pp. 13477-13485. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2017.02.163]
16. Wang L, Zhang P, Chen S, Xu J (May 2014) Overview of the development of catalysts for HI decomposition in the iodine–sulfur thermochemical cycle at INET. Nucl Eng Des 271:60–63. https://doi.org/10.1016/j.nucengdes.2013.11.010
17. Kang, KS; Kim, CH; Cho, WC; Jeong, SU; Park, CS; Bae, KK. Demonstration of the HI decomposition section embedded with electrodialysis stack in the sulfur-iodine thermochemical cycle for hydrogen production. Nucl Eng Des; 2013; 256, pp. 67-74. [DOI: https://dx.doi.org/10.1016/j.nucengdes.2012.12.005]
18. Gunawan TA, Cavana M, Leone P, Monaghan RFD, July (2022) 116398 doi: https://doi.org/10.1016/j.enconman.2022.116398
19. Sakaba N, Sato H, Ohashi H, Nishihara T, Kunitomi K (2012) “Development Scenario of the Iodine-Sulphur Hydrogen Production Process to be Coupled with VHTR System as a Conventional Chemical Plant Development Scenario of the Iodine-Sulphur Hydrogen Production Process to be C,” J. Nucl. Sci. Technol, vol. 45, No 9, no. August pp. 962–969, 2012
20. Sato, H; Kubo, S; Yan, XL; Tachibana, Y; Kato, Y. Control strategies for transients of hydrogen production plant in VHTR cogeneration systems. Prog Nucl Energy; 2011; 53,
21. Noguchi, SKH; Takegami, H; Kamiji, Y; Tanaka, N; Iwatsuki, J; Kasahara, S. R&D status of hydrogen production test using IS process test facility made of industrial structural material in JAEA. Int J Hydrogen Energy; 2019; 44, pp. 12583-12592. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2018.11.121]
22. Xu, L; Wang, L; Zhang, P; Hu, S; Li, D; Chen, S. INET’s catalysts study for HI decomposition in the thermochemical water-splitting iodine???Sulfur process for hydrogen production. Int J Hydrogen Energy; 2017; 42,
23. Ong, HC; Chen, WH; Singh, Y; Gan, YY; Chen, CY; Show, PL. A state-of-the-art review on thermochemical conversion of biomass for biofuel production: a TG-FTIR approach. Energy Convers Manag; 2020; 209, 112634. [DOI: https://dx.doi.org/10.1016/j.enconman.2020.112634]
24. Funk, JE. Thermochemical hydrogen production: past and present. Int J Hydrogen Energy; 2001; 26,
25. Paul MM, Brown LC, Technology A, Park TC (2003) “Thermodynamics of the Sulfur-Iodine cycle for Thermochemical Hydrogen production,” Outlook, no. 1,
26. Baykara, SZ. Hydrogen production by direct solar thermal decomposition of water, possibilities for improvement of process efficiency. Int J Hydrogen Energy; 2004; 29, pp. 1451-1458. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2004.02.014]
27. Mohd, N; Nandong, J. Multi-scale Control of Bunsen Section in Iodine-Sulphur Thermochemical cycle process. Chem Prod Process Model; 2017; [DOI: https://dx.doi.org/10.1515/cppm-2017-0036]
28. Kim, J et al. A study on the dynamic behavior of a sulfur trioxide decomposer for a nuclear hydrogen production. Int J Hydrogen Energy; 2008; 33, pp. 7361-7370. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2008.10.006]
29. Lee, BJ; No, HC; Yoon, HJ; Jin, HG; Kim, YS; Lee, JI. Development of a flowsheet for iodine-sulfur thermo-chemical cycle based on optimized Bunsen reaction. Int J Hydrogen Energy; 2009; 34,
30. Guo, H; Zhang, P; Chen, S; Wang, L; Xu, J. Comparison of various circuit designs in the HI decomposition section of the iodine-sulfur process. Int J Hydrogen Energy; 2012; 37,
31. Yoon, HJ; Kim, SJ; Cheon No, H; Lee, BJ; Kim, ES. A thermo-physical model for hydrogen-iodide vapor-liquid equilibrium and decomposition behavior in the iodine-sulfur thermo-chemical water splitting cycle. Int J Hydrogen Energy; 2008; 33,
32. Choi, JY; Kim, YS; Sah, I; No, HC; Jang, C. Corrosion resistances of alloys in high temperature hydrogen iodide gas environment for sulfur-iodine thermochemical cycle. Int J Hydrogen Energy; 2014; 39,
33. Kubo S et al (2012) “R&D status on thermochemical IS process for hydrogen production at JAEA,” Energy Procedia, vol. 29, no. Extension 3793, pp. 308–317, doi: https://doi.org/10.1016/j.egypro.2012.09.037
34. Murphy, JE; O’Connell, JP. Process simulations of HI decomposition via reactive distillation in the sulfur-iodine cycle for hydrogen manufacture. Int J Hydrogen Energy; 2012; 37,
35. Vitart, X; Duigou, aL; Carles, P. Hydrogen production using the sulfur-iodine cycle coupled to a VHTR: an overview. Energy Convers Manag; 2006; 47, pp. 2740-2747. [DOI: https://dx.doi.org/10.1016/j.enconman.2006.02.010]
36. Lim, D; Kim, A; Cheon, S; Byun, M; Lim, H. Life cycle techno-economic and carbon footprint analysis of H2 production via NH3 decomposition: a Case study for the Republic of Korea. Energy Convers Manag; 2021; 250, 114881. [DOI: https://dx.doi.org/10.1016/j.enconman.2021.114881]
37. Säck, JP; Roeb, M; Sattler, C; Pitz-Paal, R; Heinzel, a. Development of a system model for a hydrogen production process on a solar tower. Sol Energy; 2012; 86,
38. Al-Dabbagh, AW; Lu, L. Dynamic flowgraph modeling of process and control systems of a nuclear-based hydrogen production plant. Int J Hydrogen Energy; 2010; 35, pp. 9569-9580. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2010.06.059]
39. Rodriguez-Toral, MA; Morton, W; Mitchell, DR. Using new packages for modelling, equation oriented simulation and optimization of a cogeneration plant. Comput Chem Eng; 2000; 24,
40. Ashoori, A; Moshiri, B; Khaki-sedigh, A; Reza, M. Optimal control of a nonlinear fed-batch fermentation process using model predictive approach. J Process Control; 2009; 19,
41. Costa AC, Meleiro LAC, Filho RM (2002) Non-linear predicti v e control of an extracti v e alcoholic fermentation process. 38:743–750
42. Deshpande, AP; Patwardhan, SC; Narasimhan, SS. Intelligent state estimation for fault tolerant nonlinear predictive control. J Process Control; 2009; 19,
43. Henson MA (2003) Dynamic modeling and control of yeast cell populations in continuous biochemical reactors. 27:1185–1199. https://doi.org/10.1016/S0098-1354(03)00046-2
44. Zhu G, Henson MA, Ogunnaike BA (2000) A hybrid model predictive control strategy for nonlinear plant-wide control. 10:449–458
45. Garcia, CE; Prett, DM; Morari, M. Model predictive control: theory and practice- a survey. Automatica no; 1989; 25, pp. 335-348. [DOI: https://dx.doi.org/10.1016/0005-1098(89)90002-2]
46. Sun, JWQ; Gao, Q; Zhang, P; Peng, W; Chen, S; Zhao, G. Numerical study of heat transfer and sulfuric acid decomposition in the process of hydrogen production. Int J Energy Res; 2019; 43, pp. 5969-5982. [DOI: https://dx.doi.org/10.1002/er.4714]
47. Kasahara, S; Kubo, S; Hino, R; Onuki, K; Nomura, M; Nakao, SI. Flowsheet study of the thermochemical water-splitting iodine-sulfur process for effective hydrogen production. Int J Hydrogen Energy; 2007; 32,
48. Lee, BJ; Cheon, H; Joon Yoon, NOH; Jun Kim, S; Kim, ES. An optimal operating window for the Bunsen process in the I-S thermochemical cycle. Int J Hydrogen Energy; 2008; 33,
49. Goldstein, S; Borgard, JM; Vitart, X. Upper bound and best estimate of the efficiency of the iodine sulphur cycle. Int J Hydrogen Energy; 2005; 30,
50. Zhang, Q et al. Establishment and assessment of a novel bioethanol and efficient biogas coupling fermentation system integrated with the pretreatment of a cellulolytic microbial consortium. J Clean Prod; 2014; 83, pp. 142-150. [DOI: https://dx.doi.org/10.1016/j.jclepro.2014.07.076]
51. Gao, Q; Zhang, P; Peng, W; Chen, S; Zhao, G. Structural design simulation of bayonet heat exchanger for sulfuric acid decomposition. Energies; 2021; 14,
52. Brown LC, Besenbruch GE, Lentsch RD, Schultz KR, Funk JF, Pickard PS (2003) “High Efficiency Generation of Hydrogen Fuels using Nuclear Power,”
53. Russ B (2009) “Sulfur Iodine Process Summary For the Hydrogen Technology Down-Selection: Process Performance Package,” no. June, p. INL/EXT-12-25773,
54. Nagarajan, V; Chen, Y; Hung, T-C; Wang, Q; Ponyavin, V. Numerical modeling of bayonet-type heat exchanger and decomposer for the decomposition of sulfuric acid to sulfur dioxide. Heat Transf Eng; 2014; 35, pp. 6-8. [DOI: https://dx.doi.org/10.1080/01457632.2013.837644]
55. Saka, H.; Naka, Y.
56. Mohd, N; Aziz, N. Performance and robustness evaluation of nonlinear autoregressive with exogenous input Model Predictive Control in controlling industrial fermentation process. J Clean Prod; 2016; 136, pp. 42-50. [DOI: https://dx.doi.org/10.1016/j.jclepro.2016.06.191]
57. Deepa, SN; Baranilingesan, I. Optimized deep learning neural network predictive controller for continuous stirred tank reactor. Comput Electr Eng; 2017; 0, pp. 1-16. [DOI: https://dx.doi.org/10.1016/j.compeleceng.2017.07.004]
58. Kumar M, Prasad D, Giri BS, Singh RS (2019) “Temperature control of fermentation bioreactor for ethanol production using IMC-PID controller,” Biotechnol Rep, vol. 22, no. e00319.,
59. Naregalkar, A; Durairaj, S. A novel LSSVM-L Hammerstein model structure for system identification and nonlinear model predictive control of CSTR servo and regulatory control. Chem Prod Process Model; 2021; [DOI: https://dx.doi.org/10.1515/cppm-2021-0020]
60. Holaza, J; Klaučo, M; Drgoňa, J; Oravec, J; Kvasnica, M; Fikar, M. MPC-based reference governor control of a continuous stirred-tank reactor. Comput Chem Eng; 2018; 108, pp. 289-299. [DOI: https://dx.doi.org/10.1016/j.compchemeng.2017.09.020]
61. Oravec, J; Horváthová, M; Bakošová, M. Multivariable robust MPC design for neutralisation plant: experimental analysis. Eur J Control; 2021; 58, pp. 289-300.4217839 [DOI: https://dx.doi.org/10.1016/j.ejcon.2020.07.012]
62. Sinnot RK (2009) Coulson and Richardson’s Chemical Engineering, 6th edn. Butterworth-Heinemann.
63. Song, X; Zhang, Q; Song, S; Ahn, CK. Improved event-triggered control for a chemical tubular reactor with singular perturbations. J Process Control; 2022; 112, pp. 49-56. [DOI: https://dx.doi.org/10.1016/j.jprocont.2022.02.007]
64. Zhou, W; Hamroun, B; Le Gorrec, Y; Couenne, F. A thermodynamic approach to the stabilization of tubular reactors. J Process Control; 2021; 108, pp. 98-111. [DOI: https://dx.doi.org/10.1016/j.jprocont.2021.11.006]
65. Zhang, L; Xie, J; Koch, CR; Dubljevic, S. Model Predictive Control of Jacket Tubular reactors with a reversible exothermic reaction. Ind Eng Chem Res; 2020; 59,
66. Khatibi, S; Ozorio Cassol, G; Dubljevic, S. Model predictive control of a non-isothermal axial dispersion tubular reactor with recycle. Comput Chem Eng; 2021; 145, 107159. [DOI: https://dx.doi.org/10.1016/j.compchemeng.2020.107159]
67. Franco-de los Reyes, HA; Schaum, A; Meurer, T; Alvarez, J. Stabilization of an unstable tubular reactor by nonlinear passive output feedback control. J Process Control; 2020; 93, pp. 83-96. [DOI: https://dx.doi.org/10.1016/j.jprocont.2020.07.005]
© The Author(s) under exclusive licence to International Center for Numerical Methods in Engineering (CIMNE) 2023.