Enrique E. Tarifa

Fault diagnosis, direct graphs, and fuzzy logic

Abstract Fault diagnosis is the problem of finding out the root cause (the fault) of process malfunctions. In this work we present a new model based approach procedure for fault diagnosis in conventional chemical processes. The process model is a Signed Directed Graph (SDG). The SDG is used by a Qualitative Simulator. This stage allows us to know, for each potential fault, the possible process behaviour. This information is compiled into a set of IF-THEN rules. An Expert System evaluates them using information about the actual process state. Fuzzy Logic is used in this evaluation. By the way, the fault whose rule has the highest value of certainty should be the first one considered by the operator. Finally, the additional information provided by the Qualitative Simulation enables the Expert System to explain the diagnostic. Besides, this additional information can be used to improve the diagnostic.

A dynamic simulator for MSF plants

Abstract This work presents a dynamic simulator for MSF desalination plants. It takes into account the heaters and stages dynamic, hydraulic, standard instrumentation and control systems. This simulator was developed to study the effects of faults that may affect a MSF system. In order to extend the results scope, the simulator allows the modification of MSF topology and parameters into a wide range. Indeed, it is possible to change the number of stages belonging to the recovery and rejection sections, controller parameters (set point, integral time and gain), valve size, pump characteristics, seawater conditions, stages and heater dimensions, etc. Since fault simulation is the main simulator goal, the model and its resolution were carefully designed to enhance stability and speed. The user can select the fault to simulate among a set of possible faults (fault in controllers, sensors, pumps, etc.), and can specify the activation time (at which the fault starts), the development time (time elapsed from the fault start up until the fault reaches its maximum magnitude) and the fault magnitude. Thus, it is possible simulate step and ramp perturbations. The simulator was tested with data from real plants and it has shown a good performance. To make the simulator operation easier, it was developed by using a visual language for Windows 95.

Fault diagnosis for a MSF using a SDG and fuzzy logic

Abstract This work outlines the development of a fault diagnostic system for a multi-stage flash (MSF) desalination plant MSF using a real time expert system. This diagnostic system processes the plant data to determine whether the process state is normal or not. In the last case, the diagnostic system determines the cause of the abnormal state. The first step is to determinate the potential faults. This set contains all the faults the diagnostic system should be able to recognize. Then, to improve the diagnostic system performance, a careful selection of the plant sensors that will be supervised by the diagnostic system is done. The knowledge base of the expert system is automatically obtained from a qualitative model of the plant. The qualitative model is a signed directed graph (SDG). The SDG is used by a qualitative simulator to forecast, for each potential fault, the possible qualitative evolutions of the plant. This information is used to generate rules ‘if-then’ to build the knowledge base. During the diagnostic system operation, at each sampling time, the readings of the previously selected sensors are transformed in qualitative values. These values are used by the expert system to evaluate the rules by using fuzzy logic. The result is an index between 0 and 1 for each potential fault. This number represents the certainty about the corresponding fault is affecting the plant. The higher is the value, the higher is the certainty of that affirmation. Finally, a dynamic simulator was used to evaluate the performance of the diagnostic system.

A methodology for fault diagnosis in large chemical processes and an application to a multistage flash desalination process: Part I

Abstract This work presents a new strategy for fault diagnosis in large chemical processes (E.E. Tarifa, Fault diagnosis in complex chemistries plants: plants of large dimensions and batch processes. Ph.D. thesis, Universidad Nacional del Litoral, Santa Fe, 1995). A special decomposition of the plant is made in sectors. Afterwards each sector is studied independently. These steps are carried out in the off-line mode. They produced vital information for the diagnosis system. This system works in the on-line mode and is based on a two-tier strategy. When a fault is produced, the upper level identifies the faulty sector. Then, the lower level carries out an in-depth study that focuses only on the critical sectors to identify the fault. The loss of information produced by the process partition may cause spurious diagnosis. This problem is overcome at the second level using qualitative simulation and fuzzy logic. In the second part of this work, the new methodology is tested to evaluate its performance in practical cases. A multiple stage flash desalination system (MSF) is chosen because it is a complex system, with many recycles and variables to be supervised. The steps for the knowledge base generation and all the blocks included in the diagnosis system are analyzed. Evaluation of the diagnosis performance is carried out using a rigorous dynamic simulator.

Fault diagnosis for a MSF using neural networks

Abstract This work outlines the development of a fault diagnostic system for a multi-stage flash (MSF) desalination plant using artificial neural networks (ANNs). This diagnostic system processes the plant data to determine whether the process state is normal or not. In the last case, the diagnostic system determines the cause of the abnormal process state. The diagnostic system has an ANN for each potential fault. Every ANN processes the plant data looking for symptoms of their respective faults. At a given time, the result reported by an ANN is an index between 0 and 1. This number represents the certainty about the corresponding fault is affecting the plant. The higher is the value, the higher is the certainty of the affirmation. The structure of each ANN is simpler than those reported in the bibliography; however, the performance is better. These results are obtained due to a careful selection of the diagnostic system output and the use of a special training method. That training method calculates an appropriate value for the output of each ANN instead of setting it at 0 or 1 only. The new value of the output does not depend on the fault that causes the inputs but it does only on the degree of matching between the observed evolution and the expected one for the fault corresponding to each ANN. Finally, a dynamic simulator was used to evaluate the performance of the diagnostic system.

A fault diagnosis prototype for a bioreactor for bioinsecticide production

Abstract The objective of this work is to develop an algorithm for fault diagnosis in a process of animal cell cultivation, for bioinsecticide production. Generally, these processes are batch processes. It is a fact that the diagnosis for a batch process involves a division of the process evolution (time horizon) into partial processes, which are defined as pseudocontinuous blocks. Therefore, a PCB represents the evolution of the system in a time interval where it has a qualitative behavior similar to a continuous one. Thus, each PCB, in which the process is divided, can be handled in a conventional way (like continuous processes). The process model, for each PCB, is a Signed Directed Graph (SDG). To achieve generality and to allow the computational implementation, the modular approach was used in the synthesis of the bioreactor digraph. After that, the SDGs were used to carry out qualitative simulations of faults. The achieved results are the fault patterns. A special fault symptom dictionary —SM—has been adopted as data base organization for fault patterns storage. An effective algorithm is presented for the searching process of fault patterns. The system studied, as a particular application, is a bioreactor for cell cultivation for bioinsecticide production. During this work, we concentrate on the SDG construction, and 3btaining real fault patterns by the elimination of spurious patterns. The algorithm has proved to be effective in both senses, resolution and accuracy, to diagnose different kinds of simulated faults.

A methodology for fault diagnosis in batch reactors

Abstract The goal of a fault diagnostic system is to detect any incipient abnormal situation in a plant operation and to find out the fault that causes that situation, e.g. a blocked valve or a failed pump. In such a task, time is highly restricted and the system must be capable to help the operator in correcting the abnormal situation as quickly as possible, in order to achieve a stable and secure operation of the plant. There are a few previous works about diagnosis for batch processes. The main problem is that batch processes evolve along the normal operation. Tarifa and Scenna (1995) presented a method destined only to batch bioreactors. In this work, a more general method is presented and it is applied on a batch reactor.

Coupling Solvent Extraction Units to Cyclic Adsorption Units

The possibility of regenerating the solvent of extraction units by cyclic adsorption was analyzed. This combination seems convenient when extraction is performed with a high solvent-to-impurity ratio, making other choices of solvent regeneration, typically distillation, unattractive. To our knowledge, the proposed regeneration scheme has not been considered before in the open literature. Basic relations were developed for continuous and discontinuous extraction/adsorption combinations. One example, deacidification of plant oil with alcohol, was studied in detail using separate experiments for measuring process parameters and simulation for predicting performance at different conditions. An activated carbon adsorbent was regenerated by thermal swing, making cyclic operation possible. When extracting the acid with methanol in a spray column, feed = 4 L min−1, solvent = 80 L min−1, feed impurity level 140 mmol L−1, and extract concentration 7.6 mmol L−1, the raffinate reaches a purity of 1.2 mmol L−1, the solvent being regenerated cyclically in the adsorber (364 kg) to an average of 0.7 mmol L−1. Regeneration of the solvent by cyclic adsorption had a low heat duty. Values of 174 kJ per litre of solvent compared well with the high values for vaporization of the whole extract phase (1011 kJ L−1).

A New Method for Estimating the EFV Distillation Curve

Abstract This work presents a new method to estimate the equilibrium flash vaporization (EFV) distillation curve from standard laboratory analytical assay procedures. In this method, experimental data are utilized by a chemical process simulator to obtain the EFV curve. Any simulator able to model petroleum can be used for this task. To evaluate the performance of the proposed method, several types of petroleum were analyzed experimentally in this work. The obtained curves were compared with those produced by other methods showing a good match. The main advantage of the proposed method is its non-dependence on experimental data.

Formulation of an optimal academic exam

espanolEl objetivo de este trabajo es formular un examen academico optimo para una materia dada. Para ello, primero, se modela la probabilidad de que un estudiante apruebe el examen en funcion del numero de unidades que estudia y de las que el profesor evalua. Ese modelo de simulacion es desarrollado realizando un analisis probabilistico. Un examen optimo es luego definido como aquel que asigna la nota que el estudiante merece. Por lo tanto, en un examen optimo, aprueban quienes merecen aprobar, y desaprueban quienes no merecen aprobar. Ademas, el examen debe respetar las limitaciones de tiempo y esfuerzo que el profesor impone. En base a esta definicion y usando el modelo de simulacion, se formula un modelo de optimizacion del tipo INLP. Este modelo de optimizacion determina el numero de unidades que el profesor debe evaluar para maximizar la probabilidad de conseguir un examen optimo. EnglishThe aim of this paper is to formulate an optimal academic exam for a given subject. To do this, the probability is first modelled of a student passing the exam according to the number of units he studies and the professor evaluates. That simulation model is developed by performing a probabilistic analysis. An optimal exam is then defined as the one that awards the grade that the student deserves. Therefore, in an optimal exam, approve those who deserve to approve, and disapprove those that do not deserve to approve. Besides, this exam must respect the limitations of time and effort that the professor imposes. Based on this definition and using the simulation model, an INLP type optimization model is formulated. This optimization model determines the number of units the professor must evaluate to maximize the probability of getting an optimal exam.