Application of fuzzy logic to control acetylene hydrogenation reactor of olefin unit

Dissertation for receiving the M.Sc degree

Electrical Engineering - Control

Abstract

Given the high consumption of petrochemical products in the world today and the high potential of Iran for the development and supply of petrochemical feed, petrochemical complexes are of special importance. Therefore, one of the important reactions called acetylene hydrogenation has been investigated and controlled in one of the parent petrochemical units, i.e., the olefin unit. In this research, after reviewing the scientific reference books, scientific articles and treatises available in this field, a nonlinear acetylene hydrogenation reactor system that is being used in the petrochemical industry in the south of the country was selected and simulated using the mass and energy balance equations by the MATLAB specialized program. Then, by applying changes of a step value in reactor temperature equations, it is tried to be approximated and modeled with a suitable transformation function, with a very low error. Also, by using the step response of the approximated system, classical controllers and then fuzzy controller should be designed first, and then by showing that none of the controllers are suitable for controlling the reactor outlet temperature in the acetylene hydrogenation process, the PI_Smith fuzzy controller design is proposed with the conditions mentioned in the fourth chapter. Finally, by comparing the results of the design of different controllers for the mentioned system, it is concluded that the PI_Smith fuzzy controller can be a more suitable controller for systems with delays. rtl;">fuzzy controller design, classic controller design, acetylene hydrogenation reactor modeling

Preface

In the chemical industry, in order to reduce operational costs and develop new markets, there is always a special attention to improve product quality and reduce production waste. To achieve these goals, it is very necessary to use control systems. There are always many non-linear processes in the industry that cannot be controlled by classical methods, also if there are restrictions on the inputs and outputs of a process, the use of classical controllers will become much more difficult [1]. In the meantime, the fuzzy controller that has a non-linear behavior is a developing and very useful control method for non-linear processes. In addition to having a very high efficiency, it shows much more suitable behaviors in dealing with the constraints on the controlling variables, the controlled and other characteristics of the problem, and it is also suitable for systems with delays or in other words with slow dynamics.

1-2- Statement of the problem

Ethylene production [1] is one of the main indicators of growth in the petrochemical industry of any country, which is mainly used in polymer industries, especially polyethylene production. Olefin units in the petrochemical industry, where ethylene is produced, are among the mother phases. The conditions of the ethylene gas preparation process require that during the reaction, the process be free of any impurities, especially acetylene gas [2] with the symbol, ethane gas [3] with the symbol and carbon monoxide gas [4] with the symbol. For example, a small amount of acetylene impurity causes the reaction catalyst to be deactivated and thus the ethylene production process stops. Therefore, these impurities must be separated with a specific and reliable method. Acetylene is separated from ethylene by a process called hydrogenation[5] in a hydrogenation reactor. Since this process is very sensitive to temperature changes, therefore, the temperature of the hydrogenation reactor input materials should be controlled in an optimal way in order to increase the output pure ethylene gas production efficiency. Among the common problems of this reactor are its very long time delay (on the clock scale) and its nonlinear behavior.Therefore, it is expected that by considering the conditions of the acetylene hydrogenation process, by designing a suitable controller to control the temperature of the acetylene hydrogenation reactor input materials, a quality reaction will be created along with a specific production rate of the desired product of ethylene gas under optimal economic conditions [2]. rtl;">          Considering the importance of ethylene, we know that it is used as a feed for other petrochemical units, so by increasing the production efficiency of ethylene gas output from the hydrogenation reactor, it increases the quality of materials used in petrochemical industries. Therefore, by trying to increase the efficiency of ethylene production by designing suitable controllers on the acetylene hydrogenation reactor, we can overcome this goal. Also, among other things of the necessity of this research, it can be mentioned to take a positive step towards gaining the control of the important industrial processes of the country by Iranian experts using modern control sciences, such as fuzzy logic. accepted can be summarized as follows.

As a scientific goal, it can be pointed to raise the level of knowledge of researchers and experts in the field of designing and controlling a non-linear system with delay, important and widely used in the country's industries based on fuzzy logic.

In terms of the practical goal, it is possible to improve and upgrade conventional classical control methods used in important industrial hydrogenation processes in Petrochemicals of the country pointed out.

Proposing a new control method, in order to reduce ethylene losses and increase the efficiency of hydrogenation reactor products, as well as considering the actual conditions in the hydrogenation reactor, in order to be closer to the actual working conditions in the country's petrochemical units.

Designing a modern controller, which leads to reducing the response time of industrial systems Especially the delay of the acetylene hydrogenation reactor and the increase in the speed of action in removing the disturbances entering the system, without having any losses. gives, has complexity in the reactor model. This has made it impossible to identify the model of the reactor by collecting input and output data in the usual way. Therefore, as the full description of the hydrogenation reactor modeling will be discussed in the third chapter, the reactor model should be approximated by solving the mass and energy balance equations. This approximation in the model makes it impossible to always get an accurate model of the reactor for controller design. This issue in its kind is considered as a big limitation in reaching the dynamic model of the reactor and controller design. Also, among other limitations of conducting this research, we can mention the lack of access to detailed information about chemical reactors in the industry due to information protection. of the important reactions in the hydrogenation of acetylene in the main of one of petrochemical's olefin unit, too.

In this research, after review of the scientific reference books, research papers and dissertations in the field, in the hydrogenation of acetylene nonlinear reactor systems that are used in the petrochemical industry in the south is selected by using the Mass and energy balance equations Community Planning is the MATLAB simulation. Then it is tried using a small step change in reactor temperature equation, with an appropriate conversion function, approximated by modeling error is very low. The step response approximated system, the first classic controller and fuzzy controller are designed.