Bayesian network-based root cause analysis and fault pathway identification in complex industrial processes

Real-time root cause identification (RCI) of faults or abnormal events in industries gives plant personnel the opportunity to address the faults before they progress and lead to failure. RCI in industrial systems must deal with their complex behavior, variable interactions, corrective actions of control systems and variability in faulty behavior. Bayesian networks (BNs) is a data-driven graph-based method that utilizes multivariate sensor data generated during industrial operations for RCI. Bayesian networks, however, require data discretization if data contains both discrete and continuous variables. Traditional discretization techniques such as equal width (EW) or equal frequency (EF) discretization result in loss of dynamic information and often lead to erroneous RCI. To deal with this limitation, we propose the use of a dynamic discretization technique called Bayesian Blocks (BB) which adapts the bin sizes based on the properties of data itself. In this work, we compare the effectiveness of three discretization techniques, namely EW, EF and BB coupled with Bayesian Networks on generation of fault propagation (causal) maps and root cause identification in complex industrial systems. We demonstrate the performance of the three methods on the industrial benchmark Tennessee-Eastman (TE) process. For two complex faults in the TE process, the BB with BN method successfully diagnosed correct root causes of the faults, and reduced redundancy (up to 50%) and improved the propagation paths in causal maps compared to other two techniques.

Prevalence of SARS-CoV-2 Contamination on Food Plant Surfaces as Determined by Environmental Monitoring

ABSTRACT The SARS-CoV-2 pandemic has presented new challenges to food manufacturers. During the early phase of the pandemic, several large outbreaks of coronavirus disease 2019 (COVID-19) occurred in food manufacturing plants resulting in deaths and economic loss, with approximately 15% of personnel diagnosed as asymptomatic for COVID-19. Spread by asymptomatic and presymptomatic individuals has been implicated in large outbreaks of COVID-19. In March 2020, we assisted in implementation of environmental monitoring programs for SARS-CoV-2 in zones 3 and 4 of 116 food production facilities. All participating facilities had already implemented measures to prevent symptomatic personnel from coming to work. During the study period, from 17 March to 3 September 2020, 1.23% of the 22,643 environmental samples tested positive for SARS-CoV-2, suggesting that infected individuals were actively shedding virus. Virus contamination was commonly found on frequently touched surfaces such as doorknobs, handles, table surfaces, and sanitizer dispensers. Most processing plants managed to control their environmental contamination when they became aware of the positive findings. Comparisons of positive test results for plant personnel and environmental surfaces in one plant revealed a close correlation. Our work illustrates that environmental monitoring for SARS-CoV-2 can be used as a surrogate for identifying the presence of asymptomatic and presymptomatic personnel in workplaces and may aid in controlling infection spread. HIGHLIGHTS

Environmental monitoring shows SARS-CoV-2 contamination of surfaces in food plants

The SARS-CoV-2 pandemic has presented new challenges to food manufacturers. In addition to preventing the spread of microbial contamination of food, with SARS-CoV-2, there is an additional focus on preventing SARS-CoV-2 infections in food plant personnel. During the early phase of the pandemic, several large outbreaks of Covid-19 occurred in food manufacturing plants resulting in deaths and economic loss. In March of 2020, we assisted in implementation of environmental monitoring programs for SARS-CoV-2 in 116 food production facilities. All participating facilities had already implemented measures to prevent symptomatic personnel from coming to work. During the study period, from March 17, 2020, to September 3, 2020, 1.23% of the 22,643 environmental samples tested positive for SARS-CoV-2, suggesting that infected individuals are actively shedding virus. Virus contamination was commonly found on frequently touched surfaces. Most plants managed to control their environmental contamination when they became aware of the positive findings. Comparisons of the personnel test results to environmental contamination in one plant showed a good correlation between the two. Our work illustrates that environmental monitoring for SARS-CoV-2 can be used as a surrogate for identifying the presence of asymptomatic and pre-symptomatic personnel in workplaces and may aid in controlling infection spread.

RESULTS OF MULTI-FACTOR ANALYSIS OF THE CONDITION OF HYDROTECHNICAL CONSTRUCTIONS

Control over the state of the main hydraulic structures of Zhiguli hydroelectric power station is carried out according to observations of the control and measuring equipment installed in the facilities and in their base, as well as the results of systematic inspections and surveys conducted by both the power plant personnel and invited experts. A computer information and diagnostic system for monitoring the condition of facilities is being introduced at the hydroelectric power station. The introduction of this system made it necessary to pay attention to the reliability of the results obtained with the help of instrumentation. The aim of the research is to conduct a multivariate analysis of the state of hydraulic structures Zhigulevskaya HPP on the results of field observations of all devices. It is established that the antifiltration elements of the underground circuit of the HPP building were not effective enough for the specific geological conditions of the base in terms of damping the filtration head. The main pressure drop in the base occurs on the upper tooth of the HPP building, where the maximum filtration gradient is observed, which is very dangerous. In the course of research it was found that several piezometers in the base do not work or their indications cause great doubts, their repair and replacement are required. According to the results of the work, it is concluded that it is necessary to equip the main piezometers of the HPP building with remote water level meters and create an automated system of survey and monitoring of the base state on their basis.

SMS Based Remote Monitoring and Control of Industrial Processes using Artificial

This paper presents artificial neural network (ANN) based intelligent control system for effective monitoring and control of industrial processes based on Global System for Mobile Communication network (GSM). This system provides ideal solution for monitoring critical oil/gas plant variables such as temperature, pressure, level and flow- rate. Historical and real time data can be accessed worldwide using the GSM network. The proposed system monitors and controls these variables from the remote location and whenever it crosses the set limit, a microcontroller will sends an SMS to concerned plant personnel(s) mobile phone via GSM network. The concerned personnel can control the system through his mobile phone by sending AT Commands to GSM MODEM and in turn to the microcontroller A Nokia N90 series GSM mobile phone was interfaced with a PC via RS 232 serial port. A PC was used as SMS gateway as well as to create a database to store different records to provide list of changes. An SMS enable was also introduced to enable reporting agent to communicate with the plant. Assembly language, C Language and Visual BASIC were used at different stages to program the necessary hardware. Data obtained from this experiment were analyzed using MATLAB. The performance evaluation and the curve generated with MATLAB shows that using ANN controllers to control industrial Process gives better result/performance than on/off controllers

Complex of measures to reduce noise on the environment and power plant personnel

The article considers the sources of noise that have a negative effect on the electrician of the power plant, as well as on the environment. The analysis showed that noise is the main harmful production factor that adversely affects the working conditions of an electrician. In the course of the study, a set of measures to reduce the impact of industrial noise on the worker and the environment, including the installation of silencers, was proposed. The material used for the manufacture of the silencer, which can significantly reduce the level of industrial noise, was proposed. Recommendations for adjusting the work and rest regime were developed, and more effective personal protective equipment for an electrician was selected..

Predictive Shimming of Frame-Panel Assemblies Using FEM and Laser Scanning

This paper presents a predictive shimming methodology based on finite element analysis modeling (FEM) and laser scanning metrology. The aim of the predictive shimming methodology is to provide the geometry of the necessary shims for an assembly beforehand, so the shims can be manufactured in advance and placed on the exact location reducing ergonomic harmful conditions for plant personnel, improving accuracy, reducing the components assembly time and lowering manufacturing cost. The case study for the predictive shimming method is an aircraft torque box assembly (TBA). The interface of interest is the mating surface of the skin and the rib shear ties. A FEM model of the skin was developed to calculate the skin deformation under prescribed displacements at the shimming locations. The FEM model was successfully verified and validated. Using the FEM deformation data, the geometry of the shims needed for an assembly can be calculated. Laser scanning metrology is used to get a point cloud of the assembly, fine-tuning the shims geometry.

Atributos de los estilos de vida que favorecen la salud en el personal de planta de la Universidad La Salle, Noroeste

En México se vive un aumento de las enfermedades crónicas no trasmisibles que afectan a la población laboral. Objetivo: Analizar las diferencias de los atributos del estilo de vida entre el personal sin riesgo y con riesgo de padecer obesidad, diabetes mellitus tipo 2 e hipertensión arterial. Método: Se estudiaron a 116 integrantes del personal de Universidad La Salle Noroeste (Hombres=67; Mujeres=49). Se evaluaron con el instrumento FANTÁSTICO, mediciones antropométricas y clínicas. Resultados: En la comparación de los grupos con riesgo y sin riesgo, el 30.2% de la población no presentó riesgo cardiovascular (RC), discriminando con el apoyo de la familia. Mientras que las funciones no discriminaron con la glucemia capilar y la presión arterial. Conclusiones: Se sugiere realizar acciones para promover el apoyo familiar, nutrición, el bajo consummo de alcohol y tabaco para generar estilos de vida saludable Palabras clave: estilos de vida; personal de planta; riesgo cardiovascular; hiperglicemia; hipertensión AbstractIn Mexico there is an increase in noncommunicable chronic diseases that affect the working population. Objective: To analyze the differences of the attributes of the lifestyle between the personnel without risk and with risk of obesity, diabetes mellitus type 2 and arterial hypertension. Method: 116 members of the staff of Universidad La Salle Noroeste (Men = 67, Women = 49) were studied. They were evaluated with the FANTASTIC instrument, anthropometric and clinical measurements. Results: In the comparison of the groups of risk and without risk, 30.2% of the population did not present CR, discriminating with the support of the family. While the functions did not discriminate with capillary blood glucose and blood pressure. Conclusions: It is suggested to carry out actions to promote variables such as family support, nutrition, low consumption of alcohol and tobacco to generate healthy lifestylesKeywords: Lifestyles, Plant Personnel, Cardiovascular Risk, Hyperglycemia, Hypertension

Gas Turbine and Combined Cycle Generator Dewpoints: Worse Than Expected

Many hydrogen cooled generators operate with no hydrogen dryer. Plant personnel offer a variety of reasons: • The OEM didn’t supply a dryer to begin with. • Manual Reactivations are too labor intensive. • Don’t think their generator is wet. • Because they have a Gas Turbine (no steam), they don’t think there is anywhere for the hydrogen to pick up moisture. • Don’t think moisture matters. • Purity is good so everything must be fine. • They employ “Bleed and Feed” or “Scavenging” to deal with purity / moisture issues. Generator manufacturers recommend keeping the hydrogen dry. We had a theory that many generators were much wetter than was realized or recommended by the manufacturers. We took portable dewpoint analyzers to the plants to test the actual dewpoint of the hydrogen. This paper presents our results. Over 90% of the generators tested were found to have hydrogen dewpoints higher than the generator manufacturer’s recommendation.