Large-scale Chemical Plants Research Articles

Recursive identification of kernel-based hammerstein model for online energy efficiency estimation of large-scale chemical plants

Accurate energy efficiency prediction is crucial for decision-making in large-scale chemical production, especially considering energy management and environmental concerns. The complexity of chemical processes, characterized by strong nonlinearities, dynamics, and uncertainties, challenges traditional prediction methods. This paper introduces KHSM (Kernel-based Hammerstein Subspace Model), which models nonlinear static and linear dynamic processes in series by embedding kernel function-based nonlinear mapping into subspace modeling for energy efficiency estimation. KHSM identifies nonlinear process models without explicit nonlinear mapping functions, constructing dynamic models that encompass both energy input and product output variables. An adaptive KHSM variant dynamically updates the energy efficiency model, improving accuracy by iteratively selecting and excluding samples. Comparative analysis shows KHSM outperforms existing methods, achieving a coefficient of determination (R2) closest to 1 and reducing the root-mean-square error (RMSE) by up to an order of magnitude, with the smallest mean and standard deviation. Additionally, modeling efficiency increases by an order of magnitude. The developed model enables ongoing energy efficiency estimation and proactive identification of efficiency trends. Validation through mathematical and real-world ethylene process cases demonstrates KHSM's efficacy and applicability. These estimates facilitate informed adjustments to energy management strategies and optimization of production approaches in energy-intensive chemical processes.

Analysis of the causes of waste heat boiler leakage in large-scale chemical plants

A waste heat boiler is a kind of high-temperature and high-pressure heat exchanger in a large chemical plant, and its safe operation is directly related to the production efficiency and safety risk of a chemical plant. In this paper, the leakage failure analysis of a waste heat boiler in an ethylene plant is carried out, and the analysis methods such as field observation, macro observation of fracture, metallographic analysis, micro observation of fracture, phase analysis, and energy spectrum analysis are adopted respectively. The results show that the direct cause of the leakage of the waste heat boiler is the thinning of the wall thickness of the heat exchange tube caused by alkali corrosion, which leads to the overload and fracture of the heat exchange tube. The pits on the outer surface of some heat exchange tubes are caused by mechanical reasons, which are related to this leakage and closure. The corrosion failure of heat exchange tubes is widespread, which has a great influence on safety production and serious consequences. Through this comprehensive analysis, the failure mechanism is clarified, which can provide a reference for the maintenance of similar equipment. At the same time, preventive measures are put forward from the aspects of technology, structure, and detection to reduce the occurrence of similar accidents.

Is NIMBY inevitable? An empirical exploration of determinants of public attitudes towards unwanted facilities using nationally representative data in China

ABSTRACT ‘Not in my Back Yard’ (NIMBY) protests against the construction of unwanted facilities nearby occur when individuals prioritize personal benefits over broader social benefits, which is an interesting phenomenon in China’s government-regulated market economic system. Although this fascinating phenomenon attracts plenty of research attention, it remains unknown regarding the extent of NIMBY intention and its influencing factors in China. This study describes the NIMBY intention of Chinese residents towards a large-scale chemical plant with low pollution and analyzes its determinants utilizing data from a nationally representative survey in 2017. It shows that 85% of the 38,775 respondents have strong NIMBY intentions. Among all factors, education and urbanization are the two most influential ones and are positively related to NIMBY intention. Besides, women, high-income groups, and people living in areas with strong social networks, high GDP per capita, or high pollution, exhibit strong NIMBY intention. Conversely, trust in the government and willingness to participate in public affairs are negatively correlated with NIMBY intention. Our study indicates that NIMBY intention will inevitably rise with educational attainment and urbanization levels, but the NIMBY dilemma can be alleviated by implementing participatory governance and building trust in local government.

Critical Situations and Prevention of Accidents in Chemico-Technological Systems (Methodological Aspects)

The aim of this work is to study the causes of accidents in chemical processes, develop a methodology for accident prevention via control, and illustrat its realization by examples using a variety of strategies. The general concept of critical situations was introduced systematically covering both emergency and pre-emergency situations. In large-scale chemical plants, examples of accidents are presented. Accident causes as a result of disturbances and control faults in technological processes are analyzed. Approaches for preventing accidents are considered. The revealing of critical situations is presented as a problem of pattern recognition, and the subtasks of the recognition are analyzed. An emergency scale based on the assessment of various states of the chemico-technological process is introduced and applied for distinguishing the different levels of accident. The real obstacles in the prevention of accidents via control are shown and analyzed. Matrices of critical situations with corresponding characteristics are given. The main tasks for the prevention and elimination of critical situations are highlighted and characterized, and our methodology for the realization of these tasks is presented. Practical examples of the prevention of accidents in the industrial ammonia synthesis processes (including approaches and strategies) are demonstrated based on the real-time control of autothermal reactors.

A Cyber-Physical monitoring and diagnosis scheme of energy consumption in Plant-Wide chemical processes

In the large-scale plant-wide chemical process, energy monitoring and diagnosis have a great impact on energy management and sustainable development. Most monitoring and diagnosis methods focused on the construction of the global model with process data, regardless of the meaningful process knowledge and in-depth analysis. Due to the multi-dimensional, correlative, and uncertain characteristics of the collected industrial data, it is laborious to obtain accurate and reliable energy monitoring and diagnosis results. To address these problems, a novel cyber-physical energy monitoring and diagnosis scheme is proposed in this paper. This scheme constructs a cyber-physical model based on process knowledge and data to describe the cause-effect relationships among the energy variables. For energy monitoring, a distributed monitoring approach is developed for energy state estimation based on the constructed monitoring statistics in both variable and residual spaces. For energy diagnosis, a faulty contribution degree index is further addressed and the corresponding root cause location strategy is discussed. The effectiveness and practicality of the proposed scheme are demonstrated via a numerical simulation example and practical ethylene oxide production. Two preset simulated faults and one practical process fault are used and all the abnormal states and root variables are monitored and diagnosed by the proposed scheme. The energy monitoring and diagnosis results provide great support for energy management and development in large-scale chemical plants.

Modeling and Control of Nonlinear Processes Using Sparse Identification: Using Dropout to Handle Noisy Data

Sparse identification of nonlinear dynamics (SINDy) is a recent nonlinear modeling technique that has demonstrated superior performance in modeling complex time-series data in the form of first-order ordinary differential equations (ODEs), which are explicit and continuous in time. However, a crucial step in the SINDy algorithm involves estimating the time derivative of the states from the discrete, measured data. Therefore, the presence of noise can greatly deteriorate the performance if it is not carefully considered and accounted for. In this work, SINDy is used with ensemble learning, where multiple models are identified to improve the overall/final nonlinear model’s performance. Specifically, in the SINDy algorithm, a fraction of the library functions considered for the ODE model representation are randomly dropped out in each submodel to favor model sparsity and stability at the possible risk of lowering the model accuracy. This trade-off is controlled by manipulating the fraction of the library functions dropped out and the total number of models generated, both of which are considered as hyperparameters to be tuned in the proposed algorithm. Data from open-loop simulations of a large-scale chemical plant are generated using the well-known high-fidelity process simulator, Aspen Plus Dynamics, and corrupted with substantial sensor noise to be implemented in the newly proposed algorithm, dropout-SINDy. The dropout-SINDy models obtained from training with the noisy data are then tested in open-loop simulations to demonstrate accurate identification of the steady-state and reasonably close transient behavior under a variety of initial conditions and manipulated input values. Finally, the constructed models are used in a Lyapunov-based model predictive controller to control the large-scale Aspen process, meeting desired closed-loop stability and performance specifications.

Process plant layout optimization considering domino effects and safety devices in a large-scale chemical plant

The layout of a process plant is an essential factor in designing a chemical process which can lead to enhance the performance of the plant as well as reducing economic damages in the case of any incident in the plant. In this study, a mathematical formulation of the plant layout problem considering possible scenarios including fire, explosion, domino effects, and the existence of safety devices is proposed. Benefiting from Bat metaheuristic algorithm at the onset, the mathematical model was validated with a case study from one of our references and was then implemented on a large-scale chemical process plant. The results illustrate that the proposed model can reduce piping costs, land cost, domino hazard index, and the costs associated with domino escalation up to 30%. In addition, Lack of safety devices can increase the piping cost of the plant up to 20%. It was concluded that the proposed model can effectively improve the safety of a plant and reduces its layout costs.

Evaluating and predicting energy efficiency using slow feature partial least squares method for large-scale chemical plants

As demands for energy conversion and management are increasing, accurate evaluation and prediction of energy efficiency are the prerequisites of making production strategies to improve economic benefits in the chemical industry. With the characteristics of slow variations in the chemical process and temporal correlations in the sampling data, it is more difficult for conventional estimation methods to precisely evaluate and infer the energy efficiency. To overcome these difficulties, this paper proposes a novel energy efficiency evaluation and prediction method, named slow feature partial least squares. The proposed method is integrated with slow feature analysis and partial least squares. Slow feature partial least squares can extract dynamic features from temporal behaviors of chemical products and energy media in a supervised manner and construct the model relationship. With the established model, not only are the energy efficiency levels evaluated accurately, but the tendency of energy efficiency is also inferred in advance. The effectiveness and practicality of the slow feature partial least squares method are validated by a mathematical example and an actual ethylene process. The energy efficiency results provide great support for the enhancement of energy efficiency and the development of production strategies in chemical plants with high energy consumption.

Visual Leakage Inspection in Chemical Process Plants Using Thermographic Videos and Motion Pattern Detection.

Liquid leakage from pipelines is a critical issue in large-scale chemical process plants since it can affect the normal operation of the plant and pose unsafe and hazardous situations. Therefore, leakage detection in the early stages can prevent serious damage. Developing a vision-based inspection system by means of IR imaging can be a promising approach for accurate leakage detection. IR cameras can capture the effect of leaking drops if they have higher (or lower) temperature than their surroundings. Since the leaking drops can be observed in an IR video as a repetitive phenomenon with specific patterns, motion pattern detection methods can be utilized for leakage detection. In this paper, an approach based on the Kalman filter is proposed to track the motion of leaking drops and differentiate them from noise. The motion patterns are learned from the training data and applied to the test data to evaluate the accuracy of the method. For this purpose, a laboratory demonstrator plant is assembled to simulate the leakages from pipelines, and to generate training and test videos. The results show that the proposed method can detect the leaking drops by tracking them based on obtained motion patterns. Furthermore, the possibilities and conditions for applying the proposed method in a real industrial chemical plant are discussed at the end.

Forecasting industrial aging processes with machine learning methods

Accurately predicting industrial aging processes makes it possible to schedule maintenance events further in advance, ensuring a cost-efficient and reliable operation of the plant. So far, these degradation processes were usually described by mechanistic or simple empirical prediction models. In this paper, we evaluate a wider range of data-driven models, comparing some traditional stateless models (linear and kernel ridge regression, feed-forward neural networks) to more complex recurrent neural networks (echo state networks and LSTMs). We first examine how much historical data is needed to train each of the models on a synthetic dataset with known dynamics. Next, the models are tested on real-world data from a large scale chemical plant. Our results show that recurrent models produce near perfect predictions when trained on larger datasets, and maintain a good performance even when trained on smaller datasets with domain shifts, while the simpler models only performed comparably on the smaller datasets.

Ineos invests $2 billion in 3 Saudi plants

Ineos will build three large-scale chemical intermediate plants within Project Amiral, an ethylene cracker complex in Jubail, Saudi Arabia, being developed by Saudi Aramco and Total. Costing $2 billion and set to start up in 2025, the facilities that Ineos will build are a 425,000-metric-ton-per-year acrylonitrile plant, a 400,000-metric-ton-per-year linear-α-olefin plant, and what the firm says will be a “world-scale” poly-α-olefin plant. The three plants will be Ineos’s first in Saudi Arabia. Output from the α-olefin plants will be used to make synthetic lubricants. Acrylonitrile is a key building block chemical for making carbon fiber and engineering polymers. “This first investment in the Middle East consolidates our position as the market leader,” says Paul Overment, CEO of Ineos Nitriles. Ineos’s Saudi move follows recent announcements by the privately owned firm that it will invest $3.4 billion in a new ethylene cracker and propane dehydrogenation complex in Antwerp, Belgium, and $1.3

Energy efficiency evaluation and prediction of large-scale chemical plants using partial least squares analysis integrated with Gaussian process models

In the large-scale chemical plant, energy efficiency evaluation and management plays a crucial role in the sustainable development. However, most data-driven approaches to calculating energy efficiency still focus on the deterministic evaluation regardless of the effects resulting from the data quality. Besides, because of the multi-dimensional, stochastic and uncertain characteristics of the collected industrial data, it is more difficult to accurately and reliably evaluate and predict the energy usage of the complex chemical production process. To solve these limitations, a novel energy efficiency evaluation and prediction method integrated with the Gaussian process (GP) and the partial least squares (PLS) analysis, named GP-PLS method, is proposed in this paper. This integrated method can select the appropriate sampling data, construct the corresponding nonlinear model and indicate the reliability of the model. Based on the constructed models, not only are the accurate energy efficiency evaluation results calculated, but the energy efficiency tendency is also predicted. In addition, according to the reliability of the established model, the informative data are intelligently selected from the historical database to enhance the quality of the model. The effectiveness and practicality of the proposed method are demonstrated through a numerical example and a practical ethylene production. The evaluation results and the prediction values are acquired to decrease the energy consumption, guide the production process and improve the energy efficiency level in the large-scale chemical plant.

A dynamic approach to energy efficiency estimation in the large-scale chemical plant

With the increasing pressures from the energy price and environmental protection, large-scale chemical plants pay more attention to the implementation of energy efficiency estimation to improve its economic benefit and environmental performance. Because of the stochastic and dynamic characteristics of the actual data, traditional estimation methods fail to satisfy the requirement of real-time evaluation. To cope with this limitation, a novel energy efficiency estimation method combining just-in-time (JIT) learning and subspace model identification (SMI) with noise elimination, called e-JITSMI method, is proposed. First, the state space model is constructed to describe the dynamic performance of production processes. This integration method can select the appropriate sampling data, estimate noise effect, and build the corresponding dynamic model. With the built model, not only are the relationships between production and supplied energy built, but the energy efficiency tendency is also predicted at the next moment. In addition, with the arrival of the new sampling data, the dynamic evaluation model is automatically updated. The effectiveness and accuracy of the proposed method are demonstrated through a practical large-scale chemical process. The results present the average accuracy of energy efficiency prediction can reach 88.9% and the tendency of energy efficiency is 100% correct even if the working conditions change.

A Data-Driven Causality Analysis Tool for Fault Diagnosis in Industrial Processes

Data-driven causality analysis is an important step towards fault diagnosis in complex industrial processes. Although many causality analysis tools were developed in different domains, only a few of them are applied in the industry. Accordingly, there is a need to develop a causality analysis tool that serves for fault diagnosis in large-scale chemical plants. This paper develops a decision-support tool to perform causality analysis by extracting useful information from the process historical data. The aim is to help the process operator to understand the underlying systems conditions with minimal efforts and to take appropriate actions in a short response time. The tool is implemented as a graphical user-friendly interface (GUI) that exploits the multivariate time series data and provides the user with stationarity tests and Granger causality analysis. It also offers various visualization charts such as pairwise causality relationships and most importantly the final causal graph. In order to demonstrate the easiness and usability of the developed tool, two different case studies are considered. The first case study is a time-varying simulated model and the second one is the Tennessee Eastman Process as a well-known benchmark. The results show that the cause-and-effect information obtained by the developed tool can assist the user to deeply analyze causal variables and diagnose the corresponding fault with minimal involvement.

Application of formal verification and falsification to large-scale chemical plant automation systems

In this paper, we apply formal verification and falsification of temporal logic specifications to analyze chemical plant automation systems. We present new results, obtained by applying a recently-developed approach to handle combined invariance and reachability requirements. In addition, we develop a set of tests that can be generated automatically for a given control system, some of which have the same form as those in the existing literature, and some of which combine invariance and reachability, to which we apply the new approach mentioned previously. In both cases, we work with abstractions of the automation systems in order to apply symbolic model checking to industrial-scale problems. We demonstrate the results using a series of small illustrative examples, and also report results from an industrial case study. The methods that we apply are implemented in a pair of open-source software tools, which we describe briefly.

Prototype observation and influencing factors of environmental vibration induced by flood discharge

Due to a wide range of field vibration problems caused by flood discharge at the Xiangjiaba Hydropower Station, vibration characteristics and influencing factors were investigated based on prototype observation. The results indicate that field vibrations caused by flood discharge have distinctive characteristics of constancy, low frequency, small amplitude, and randomness with impact, which significantly differ from the common high-frequency vibration characteristics. Field vibrations have a main frequency of about 0.5–3.0 Hz and the characteristics of long propagation distance and large-scale impact. The vibration of a stilling basin slab runs mainly in the vertical direction. The vibration response of the guide wall perpendicular to the flow is significantly stronger than it is in other directions and decreases linearly downstream along the guide wall. The vibration response of the underground turbine floor is mainly caused by the load of unit operation. Urban environmental vibration has particular distribution characteristics and change patterns, and is greatly affected by discharge, scheduling modes, and geological conditions. Along with the increase of the height of residential buildings, vibration responses show a significant amplification effect. The horizontal and vertical vibrations of the 7th floor are, respectively, about 6 times and 1.5 times stronger than the corresponding vibrations of the 1st floor. The vibration of a large-scale chemical plant presents the combined action of flood discharge and working machines. Meanwhile, it is very difficult to reduce the low-frequency environmental vibrations. Optimization of the discharge scheduling mode is one of the effective measures of reducing the flow impact loads at present. Choosing reasonable dam sites is crucial.

HAZOP analysis based on sensitivity evaluation

Hazard and operability (HAZOP) analysis approach based on sensitivity evaluation is presented in this paper to ensure production safety in complex and large-scale chemical plants. In this approach, the sensitivity evaluation is introduced into HAZOP deviation analysis to measure the effect degree caused by each cause on the corresponding deviation. A quantitative sensitivity evaluation model of deviation cause is first established, in which the departure degree, time duration to reach the maximum deviation, and stability degree of the target process variable are determined as evaluation factors. The sensitivity evaluation of each deviation cause is performed based on the dynamic process simulation. The multiple deviation causes that correspond to the deviation are ranked based on sensitivity index, thereby identifying the significant effect causes on deviation from multiple causes; this identification will assist plant operators and safety management staff on their on-line process monitoring and fault diagnosis. Finally, the proposed approach is applied to the depropanization unit in a gas fractionation plant. The sensitivity evaluation was carried out for 10 deviation causes that correspond to the deviation “higher overhead pressure”, in which three and seven sensitive and non-sensitive causes were identified, respectively. The result of the case study demonstrates that the proposed approach can improve the readability and guidance of conventional HAZOP report, which is consistent with the actual production conditions.

Revitalization of Rural Economies though the Restructuring the Self-sufficient Realm

According to rising social and environmental concerns, the number of small-scale oil mills growing and milling rapeseed has increased over the past decade. This study was undertaken to clarify small-scale rapeseed oil mill production, which is under pressure from mass-produced vegetable oil, and propose the application of this community-based system to developing nations. We analyzed all domestic rapeseed oil mills between 2009 and 2011, and discussed the transition from traditional mills to newly established town office-sponsored mills and its effect on the rural economy. We clarified that small-scale oil mills in rural areas have survived through strong self-sufficient production and have restructured into self-managing local communal businesses. This restructuring of local rapeseed production has had a positive effect on the local community and has revitalized the rural economy. Thus, we concluded that it may be suitable for microcredit in developing nations. Discipline: Agricultural Economics Additional key words: community mill, inducing production effect, local rapeseed production, restoring local production Present address: 2 College of Bioresource Sciences, Nihon University (Fujisawa, Kanagawa 252-0880, Japan) *corresponding author: nonaka@affrc.go.jp +81-19-643-3494 Received 16 November 2012; accepted 13 March 2015. Introduction Nowadays, vegetable oil, one of the most essential foodstuffs, is manufactured in large-scale chemical plants from imported oilseed grown in a few exporting nations. Japan used to grow and mill rapeseed in rural areas; however, imported oilseed now dominates the domestic market and Japanese production is close to zero. Recent discussions on the growing use of bio-diesels, however, have raised public concern about domestic rapeseed cultivation, and the number of rural areas growing rapeseed has increased since 2000. Such regional areas have established small-scale oil mills to process rapeseed that farmers in their neighborhood grow, and this growth in rapeseed production and milling activity (rapeseed production hereafter) has revitalized rural economies. Kawate et al.(2006), Goto et al. (2008) (2009), Ono et al. (2011), Ono and Nonaka (2011), Hirano (2008), Watanabe (2010), and Nonaka (2007) (2008) have all analyzed new rapeseed oil milling in rural areas based on field research and reported that increasing small-scale milling is established by local government and supported by local society. However, previous studies on this growth in small-scale rapeseed milling are based on specific case studies, and their structure and benefit to rural economy remain unclear. Major oil mills, such as the Honen Corporation, started mass production in large-scale chemical plants using Chinese soybean in the late 1920s, and they sold soy oil to Japan and other overseas markets. However, domestic rapeseed oil production survived until Japan’s high economic growth era (mid-1950s to 1973). Most farmers grew rapeseed as a second crop in paddy fields, while small-scale oil mills in towns processed the rapeseed, a system that resisted pressure from mass-produced goods. Although the growth in new small-scale rapeseed oil production is attracting research interest as a method of promoting small-scale business in rural areas, there are differences and similarities between old small-scale rapeseed production and new ventures in this regard. Sasama (1981) analyzed the history of Japanese vegetable oil milling, but not small-scale rapeseed oil milling specifically. Nonaka

A Hybrid Process Monitoring and Fault Diagnosis Approach for Chemical Plants

Given their potentially enormous risk, process monitoring and fault diagnosis for chemical plants have recently been the focus of many studies. Based on hazard and operability (HAZOP) analysis, kernel principal component analysis (KPCA), wavelet neural network (WNN), and fault tree analysis (FTA), a hybrid process monitoring and fault diagnosis approach is proposed in this study. HAZOP analysis helps identify the fault modes and determine process variables monitored. The KPCA model is then constructed to reduce monitoring variable dimensionality. Meanwhile, the fault features of the monitoring variables are extracted, so then process monitoring can be performed with the squared prediction error (SPE) statistics of KPCA. Then, multiple WNN models are designed through the use of low-dimensional sample data preprocessed by KPCA as the training and test samples to detect the fault mode online. Finally, FTA approach is introduced to further locate the fault root causes of the fault mode. The proposed approach is applied to process monitoring and fault diagnosis in a depropanizer unit. Case study results indicate that this approach can be applicable to process monitoring and diagnosis in large-scale chemical plants. Accordingly, the approach can serve as an early and reliable basis for technicians’ and operators’ safety management decision-making.

Modeling and simulation of supersonic gas separations

Supersonic separation is an operation potentially applicable to gas processing in large scale chemical plants. Supersonic flows are achieved in properly sized converging–diverging nozzles, whose performance is simulated in this paper. The overall simulation procedure consists of different types of thermodynamic calculations, including the determination of phase equilibrium conditions and of thermodynamic sound speed in non-ideal single-phase and multiple-phase systems. The thermodynamic properties of the fluid are evaluated using the Peng–Robinson equation of state, but the procedure is general and compatible with other thermodynamic models. Despite being less detailed than computational fluid dynamics (CFD) simulations previously reported in the literature, the results are in very good agreement with CFD. The proposed procedure constitutes a tool for the preliminary design of supersonic separators that is conceptually simple and accurate, with rigorous evaluations of the thermodynamic properties.