Process Monitoring Techniques Research Articles

A software platform for real-time control and monitoring of a wastewater treatment plant

This paper reports on the design, implementation and real-time operation of advanced process control and monitoring techniques for wastewater treatment plants. The paper presents the development of a software platform and its implementation in a full-scale wastewater treatment plant. The software platform allows the real-time execution of advanced process control techniques as model predictive control, subspace identification, and statistical process analysis using principal component analysis methods. Results obtained by the real-time execution of such algorithms are also presented.

Monitoring and Control of Polymerization Reactors Using NIR Spectroscopy

: The development and use of microprocessor-based devices improved the speed and reliability of transmission and manipulation of data. More recently, some monitoring techniques were greatly improved by the combination of spectroscopic methods and fiber optics technology, which allow for the in situ and in-line acquisition of process data, consequently allowing for reduction of time delays normally involved with sample preparation. Among them, techniques based on the near infrared spectroscopy (NIRS) have certainly become the most important ones. The main objective of this paper is to review the use of NIR spectroscopy as an alternative technique for polymerization process monitoring and control, presenting and discussing successful applications of NIR spectroscopy as a tool for in situ and in-line monitoring and control of solution, suspension, and emulsion polymerization processes.

Determining the best reaction calorimetry technique: theoretical development

Reaction calorimetry is a powerful technique for process monitoring as it allows the estimation of the heat of reaction Q ˙ R , which can be used to determine conversion. In addition, under some circumstances, the overall heat transfer coefficient k, which is useful for maximizing production and for safety analysis, can be estimated. Currently, there is no single technique to estimate both the heat of reaction Q ˙ R and the overall heat transfer coefficient k for the whole range of reactor sizes simultaneously. For small reactors, oscillation calorimetry is a method to estimate Q ˙ R and k simultaneously, whereas the solution of the heat balance equations or state estimators such as an Extended Kalman Filter (EKF), which use the heat balance model, perform better for large reactors. However, the reactor size limit of each technique is unknown, and perhaps more importantly, it is not known if the application ranges of the two techniques overlap. This article addresses this problem by comparing the performance of oscillation calorimetry and heat balance calorimetry using a state estimator in simulation studies for reactors of intermediate size. Additionally, guidelines for the use of oscillation calorimetry are presented.

An Empirical Bayes Process Monitoring Technique for Polytomous Data

AbstractWhen a product item is tested, usually one has more information than just pass or fail. Often there are categories of failure modes. The purpose of this paper is to develop a method to monitor the fractions of the tested items falling into different categories of pass/fail modes. Using the multinomial model with Dirichlet prior, we describe the theory underlying an empirical Bayes approach to monitoring polytomous data generated in manufacturing processes. A pseudo maximum likelihood estimator (PMLE) and the method‐of‐moments estimator (MME) of the hyperparameters of the prior distribution are considered and compared by a simulation study. It is found that the PMLE performs slightly better than the MME. A monitoring scheme based on the marginal distributions of the observed pass/fail fractions is proposed. The average run length behavior of the proposed monitoring scheme is investigated. Finally, an example to illustrate the use of the technique is given. Copyright © 2004 John Wiley & Sons, Ltd.

A Statistical Model for Determining the Through-Transmission Ultrasonic Amplitude Response for Imperfectly Contacting Interfaces during Thermoplastic Fusion Bonding

The focus of this work is to develop a system model for through-transmission ultrasonics (TTU) as a process monitoring technique for nonintrusive, online control of thermoplastic fusion bonding processes. A closed form model is developed to predict the TTU amplitude of a fusion bond using an ultrasonic wave propagation theory with a model for the evolution of intimate contact at the material interfaces. This new approach treats the attenuation of sensor response due to imperfect contact at the bond interface as a statistical averaging of the transmission coefficients for each of the possible sound paths between source and receiver. The results of experimental and model-based parametric studies show that the intimately contacting path dominates the overall signal behavior. An approximation of the full model for an arbitrary number of interfaces based on these results is derived.

Monitoring of Processes with Multiple Operating Modes through Multiple Principle Component Analysis Models

Because many of the current multivariate statistical process monitoring (MSPM) techniques are based on the assumption that the process has one nominal operating region, the application of these MSPM approaches to an industrial process with multiple operating modes would always trigger continuous warnings even when the process itself is operating under another normal steady-state operating conditions. Adopting principal angles to measure the similarities of any two models, this paper proposes a multiple principal component analysis model based process monitoring methodology. Some popular multivariate statistical measurements such as squared prediction error and its control limit can be incorporated straightforwardly to facilitate process monitoring. The efficiency of the proposed technique is demonstrated through application to the monitoring of the Tennessee−Eastman challenge process and an industrial fluidized catalytic cracking unit. The proposed scheme can significantly reduce the amount of false alarms while tracking the process adjustment.

HACCP with multivariate process monitoring and fault diagnosis techniques: application to a food pasteurization process

Multivariate statistical process monitoring (SPM), and fault detection and diagnosis (FDD) methods are developed to monitor the critical control points (CCPs) in a continuous food pasteurization process. Multivariate SPM techniques effectively use information from all process variables to detect abnormal process behavior. Fault diagnosis techniques isolate the source cause of the deviation in process variable(s). The methods developed are illustrated by implementing them to monitor the critical control points and diagnose causes of abnormal operation of a high temperature short time (HTST) pasteurization pilot plant. The detection power of multivariate SPM and FDD techniques over univariate SPM techniques is shown and their integrated use to ensure the product safety and quality in food processes is demonstrated.

Development of biocalorimetry as a technique for process monitoring and control in technical scale fermentations

Quantitative calorimetry has been developed in a production scale bioreactor with a volume of 100 m 3, using continuous and dynamic methods that require a heat balance be developed around the bioreactor. In order to measure the components of heat balance accurately, the precise measurements of all temperatures and flows were installed in the bioreactor. The on-line measurements of these terms then permitted the on-line calculation of metabolic heat flux from general energy balance in Eq. (1). Good correlation was obtained between oxygen uptake (OUR) and metabolic heat production rates both in continuous and dynamic methods. The metabolic heat production was also estimated from heats of combustion of substrate, biomass and ethanol and compared with experimentally measured values and good correlation was obtained. Biomass concentration was estimated from elemental, electron and heat balances and reasonable estimation have been performed using heat balance calorimetry. Direct relation has been found between cooling water flow rate and metabolic heat flux that simplifies monitoring microbial activity in large-scale fermentations. Dynamic method also allowed quick and practical estimation of overall heat transfer coefficient in the heat exchanger and reduction in the coefficient due to fouling on the surfaces of the plates could easily be identified. The results of this study suggest that quantitative calorimetry in a production scale bioreactor can be simplified by accurately identifying individual heat sources and can be realised using simple instrumentation as compared to the sensors currently used in industry.

Quantitative high-resolution on-line NMR spectroscopy in reaction and process monitoring

On-line nuclear magnetic resonance spectroscopy (on-line NMR) is a powerful technique for reaction and process monitoring. Different set-ups for direct coupling of reaction and separation equipment with on-line NMR spectroscopy are described. NMR spectroscopy can be used to obtain both qualitative and quantitative information from complex reacting multicomponent mixtures for equilibrium or reaction kinetic studies. Commercial NMR probes can be used at pressures up to 35 MPa and temperatures up to 400 K. Applications are presented for studies of equilibria and kinetics of complex formaldehyde-containing mixtures as well as homogeneously and heterogeneously catalyzed esterification kinetics. Direct coupling of a thin-film evaporator is described as an example for the benefits of on-line NMR spectroscopy in process monitoring.

Growth process monitoring and crystalline quality assessment of CuInS(Se) 2 based solar cells by Raman spectroscopy

This work reviews the use of Raman scattering for the crystalline quality assessment of CuInS(Se) 2 thin film chalcopyrite materials for photovoltaic applications. The Raman spectra of these semiconductors show a dominant A 1 mode with spectral features sensitive to the microcrystalline quality of the layers. The spectra are also characterised by additional modes, which have been interpreted and modelled as related to the existence of polymorphic domains with Cu–Au ordering. This interpretation is supported by the detailed microstructural analysis of various specimens, including epi- and polycrystalline layers obtained under different stoichiometric regimes leading to significant differences in final solar cell efficiency. The relative intensity of these additional modes in the spectra is proposed for the monitoring and quality assessment of the absorber layers for photovoltaic devices. Besides, identification of the main vibrational modes related to secondary phases characteristic of the different thin film growth processes allows the use of the technique for process monitoring. Moreover, thanks to its non-destructive character, its implementation can be in principle designed at both in situ and ex situ levels.

Monitoring transitions in chemical plants using enhanced trend analysis

Chemical plants operate in a multitude of steady states called modes and frequently undergo transitions between them. Transitions usually require extensive operator involvement and hence monitoring the safe execution of transitions is crucial. Existing techniques for process monitoring are currently configured assuming a single mode of operation and hence cannot be directly applied to transitions. In this paper, we use a trend analysis-based technique for monitoring transitions in continuous chemical plants where the real-time evolution of the process variables is abstracted into semi-quantitative trends. These real-time trends are then compared with the dictionary trends, which correspond to those during normal operation of that transition. One challenge is to determine the location in the dictionary with which the real-time trend has to be compared. For this, a dynamic feature synchronization algorithm is proposed which helps in tracking the process during transitions. The details of the algorithm, its implementation as an expert system, and the application to two case studies—a pilot-scale distillation column and a simulation of a fluidized catalytic cracking unit—are presented.

Fixed abrasive diamond wire machining—part I: process monitoring and wire tension force

The process monitoring and mechanics of fixed abrasive diamond wire saw machining are investigated in this study. New techniques to affix diamond particles to a steel wire core have advanced to make this process feasible for the machining of ceramics, wood, and foam materials. Developments in fixed abrasive diamond wire machining are first reviewed. Advantages of using fixed abrasive diamond wire machining are then introduced. The process monitoring and signal processing techniques for measuring the cutting forces, wire speed, down feed rate, and wire bow angle in diamond wire saw machining are developed. The application of a capacitance sensor to measure the wire bow and a procedure to convert the wire bow to vertical cutting force in a rocking motion wire saw machine are developed. The tension force of the wire during cutting is also derived and discussed.

Applications of optical sensing for laser cutting and drilling.

Any reliable automated production system must include process control and monitoring techniques. Two laser processing techniques potentially lending themselves to automation are percussion drilling and cutting. For drilling we investigate the performance of a modification of a nonintrusive optical focus control system we previously developed for laser welding, which exploits the chromatic aberrations of the processing optics to determine focal error. We further developed this focus control system for closed-loop control of laser cutting. We show that an extension of the technique can detect deterioration in cut quality, and we describe practical trials carried out on different materials using both oxygen and nitrogen assist gas. We base our techniques on monitoring the light generated by the process, captured nonintrusively by the effector optics and processed remotely from the workpiece. We describe the relationship between the temporal and the chromatic modulation of the detected light and process quality and show how the information can be used as the basis of a process control system.

Process Monitoring of Fiber-Reinforced Polymer Composites

AbstractThis article presents a review of optical-fiber-based process-monitoring techniques that can be used to track the chemical reactions that take place during the processing of materials, with specific reference to thermosetting resins. The techniques covered include quantitative process-monitoring methods based on near and mid-infrared, Raman, UV–visible, evanescent wave, and fluorescence spectroscopy. The basis for refractive-index-based process monitoring using optical fibers is also presented. The techniques described here can be readily applied to other classes of materials and other areas of interest such as aging and degradation. Recent advances in noncontact process monitoring are also presented.

Control and monitoring of growth of chromium nitride coatings using in-situ spectroscopic ellipsometry

In-situ spectroscopic ellipsometry (SE) in the energy range from 1.5 to 5.5 eV was used to monitor the deposition of chromium nitride coatings deposited by ion-assisted reactive magnetron sputtering. First, a systematic study was conducted on monolithic coatings as a function of each of the following deposition parameters: nitrogen flow rate, bias voltage, and target current. In-situ ellipsometry data were analyzed and the composition of the coatings identified as a function of these parameters. Two additional samples were deposited using arbitrary initial deposition conditions and in-situ ellipsometry was used as a diagnostic tool that quickly indicated the required nitrogen flow rate adjustments to be made in order to produce stoichiometric CrN and Cr 2N coatings. These samples were tested using X-ray diffraction (XRD) and ex-situ SE. Results derived from XRD and SE data correlated well and confirmed the growth of stoichiometric phases. Finally, in-situ ellipsometry was used to monitor the early stages of film growth and graded interfaces in CrN/Cr 2N multilayer structures. The coalescence thickness of CrN on Si and the thickness of the graded region in multilayers were determined. In-situ ellipsometry was shown to be a production worthy technique for process monitoring and control of chromium nitride single and multilayer structures.

Instantaneous evaluation of mammalian cell culture growth rates through analysis of the mitotic index

Since a culture increases in cell number when dividing cells separate into two newborn cells, the fraction of mitotic cells in a growing cell population directly reflects the overall growth behavior of a cell culture. To rapidly assess the effects of growth conditions on the fraction of mitotic cells we have employed an antibody specific for the phosphorylated form of histone H3 for the identification of mitotic cells using flow cytometry. The phosphorylation of histone H3 closely correlates with the chromosomal condensation that accompanies the onset of mitosis, and, therefore, it represents a convenient marker for dividing cells. We have optimized the protocol for the staining of mitotic cells for both Chinese hamster ovary and hybridoma cell cultures. Fluorescence micrographs taken of stained cells show that cells in the various stages of mitosis can be detected based on the morphological characteristics of the chromosomes. The variation in the mitotic cell fraction has been determined throughout the batch growth phases of cultures under different growth conditions. The dynamics of the mitotic index show that balanced growth was never truly reached and that the growth rate is in fact quite variable for these cultures since large variations in the mitotic index are observed. In addition, a large increase in the fraction of mitotic cells just prior to the exponential growth phase for all cultures indicates that they are partially synchronized at the exit from the lag phase. According to a two-staged, age structured population balance model, the mitotic index is directly proportional to the growth rate of a culture. The proportionality constant for this case is shown to be the time required for cells to progress through mitosis. This time is believed to be constant for a particular cell line, as shown by experimental data. Thus, growth rates can be determined solely by measurement of the fraction of cells in mitosis. The mitotic index measurements were then used to calculate the growth in cell number of the cultures, and these simulations accurately reflect observed cell counts. Other simulations also show that changes in cell growth can be predicted before they are reflected in the cell count data. This technique can be used as a sensitive indicator of cell growth and could be useful as a process monitoring technique and for developing better feeding strategies for animal cell cultures.

Optical characterization techniques for process monitoring

AbstractThe aim of this special issue is to bring to the attention of the surface science community the potential and application of optical techniques. This paper gives an overview of the major optical characterization techniques and the subsequent papers. Where possible, links with other surface science techniques are discussed and the use of optical methods for process monitoring and control is emphasized. © Crown Copyright 2001. Published by John Wiley & Sons, Ltd.

Optical characterization techniques for process monitoring

The aim of this special issue is to bring to the attention of the surface science community the potential and application of optical techniques. This paper gives an overview of the major optical characterization techniques and the subsequent papers. Where possible, links with other surface science techniques are discussed and the use of optical methods for process monitoring and control is emphasized. © Crown Copyright 2001. Published by John Wiley & Sons, Ltd.

Box–Tidwell transformation based partial least squares regression

Partial least squares (PLS) is a powerful and frequently applied technique for process modelling and monitoring when the data is highly correlated. In this paper, a Box–Tidwell transformation based PLS (BTPLS) algorithm is proposed to address the modelling of non-linear systems. The BTPLS algorithm provides a family of flexible regression models for data fitting, where linear and quadratic PLS are special cases. BTPLS is shown to out-perform quadratic PLS, for non-linear problems, in terms of modelling ability and prediction accuracy, and neural network based PLS algorithms with respect to computational time and model parsimony in terms of the Bayesian information criterion. Linear PLS, quadratic PLS, neural network PLS and BTPLS are compared using a benchmark data set relating to the analysis of cosmetic data, a mathematical simulation and a highly non-linear pH problem. It is shown that the BTPLS algorithm provides a pragmatic compromize between model simplicity and accuracy, and constitutes a complementary modelling technique to both existing linear and non-linear PLS approaches.

Integration of multivariate SPM and FDD by parity space technique for a food pasteurization process

Multivariate statistical process monitoring (MSPM), contribution plots, and parity space fault diagnosis (FD) techniques are used to detect abnormal operation of dynamic processes and diagnose sensor and actuator faults. The methods are illustrated by monitoring the critical control points (CCP) and diagnosing causes of abnormal operation of a pilot pasteurization plant. An empirical model of the process is developed by using subspace state space system identification methods and normal process data. The process data collected under the influence of different magnitude and duration of faults in sensors and actuators are used to validate the MSPM and FD techniques. T 2 and squared prediction error (SPE N ) charts are used as MSPM charts. A parity space technique for dynamic stochastic systems and dynamic trends in contribution plots of T 2 and SPE N statistics are used for FD. The detection and FD by these techniques show significant improvements over univariate methods.