Scribd
Upload
254 views

Modeling Identification and Control Od Cement Kiln

This document discusses developing a model-based predictive control strategy for cement kilns. Soft sensors will be developed for two variables not measured in real time: the amount of free lime in the clinker and the burning zone temperature. Developing effective control systems for cement kilns is challenging due to the complex heat exchanges and chemical reactions. A model will be developed and validated, then used to develop the soft sensors and investigate interactions in the multi-loop control structure. The goal is to maximize productivity while maintaining acceptable clinker quality.

Uploaded by

miltoncoba
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
254 views

Modeling Identification and Control Od Cement Kiln

This document discusses developing a model-based predictive control strategy for cement kilns. Soft sensors will be developed for two variables not measured in real time: the amount of free lime in the clinker and the burning zone temperature. Developing effective control systems for cement kilns is challenging due to the complex heat exchanges and chemical reactions. A model will be developed and validated, then used to develop the soft sensors and investigate interactions in the multi-loop control structure. The goal is to maximize productivity while maintaining acceptable clinker quality.

Uploaded by

miltoncoba
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

Modeling, Identification & Control of Cement Kiln

CAPEC
Bao Lin Computer Aided Process Engineering Center, Department of Chemical Engineering, Technical University of Denmark, DK-2800 Lyngby, Denmark

1. Purpose
The objective of this project is to develop a model-based predictive control (MPC) strategy for cement kilns. For this purpose, two soft sensors are to be developed for variables which are not measured in real time: the amount of free lime in the clinker and burning zone temperature.

4. Reactions Network in Cement Kiln

2. Motivation
Cement processes are extremely energy-consuming. The rotary kiln represents the largest portion of energy consumption and CO2 emission. In addition, rotary kilns release other emissions of environmental concerns, such as CO, SO2, NOx in gas phase and dust particles. In most cases, cement plants operate with an energy efficiency of less than 40%, which indicates the large opportunities for process improvement. The heat exchanges coupled with a series of chemical reactions taking place simultaneously, pyroprocessing in rotary kilns constitutes a very complex dynamic process. The design of effective control systems for cement kilns therefore constitutes a challenging task. Due to the poor performance of PID control, manual operation is most common for cement kilns. Rule-based methods, such as expert systems, have been designed and implemented for enhanced performance.1 If the essential process dynamics can be modelled accurately, model-based control systems will be more effective, since an expert system requires much more information to achieve similar performance. The objective of this project is to develop model-based predictive control (MPC) strategy for cement kilns.

5. Key Issues 3. Process Description


In a typical dry process kiln system,the feed materials are preheated by hot gas from the rotary kiln. A secondary burner in the precalciner is included to improve energy efficiency. The mixture of preheated and precalcined materials enters the rotary kiln, which is a counter-flow reactor. Clinker quality is indicated by the amount of free lime. The time delay from lab analysis is often an hour. It is desirable to develop a soft sensor that is able to accurately predict the content of free lime in real time to be used for effective quality control. Burning zone temperature is widely accepted as an indication of kiln operation and clinker quality. The temperature of gas phase is over 1500C, and difficult to measure reliably. A soft sensor for burning zone temperature is to be developed. Interactions between the loops in the basic multi-loop control 4. therefore, a multiple variable MPC structure are expected, Work Plan strategy will be developed. The performance index is defined as maximizing productivity with acceptable clinker quality. Approaches for optimizing performance will be investigated.

6. Project Plan
A model for control will be developed and validated using a detailed first principle model in the FLS Automation CEMulator. A soft sensor for free lime will be derived based on industrial process data. Based on the validated model, a soft sensor for burning zone temperature will be developed. The model will be employed for the investigation of interactions within multi-loop control structure. Together with a nonlinear analysis, an MPC control system will be designed and investigated.
References:
1. FLSmidth Automation, " ECS/FuzzyExpert - Cement Kiln Control", 2003 2. G. Locher, Mathematical models for the cement burning process, ZKG Int, 2002

Fuel together with air enter from the opposite end. The solid feed is heated to an extremely high temperature in the burning zone such that raw materials react and form the nodular clinker. Clinker exits the kiln at about 1200C, and is cooled down to less than 200C by cross flowing air. Part of the heated air from cooler enters the kiln, another portion flows to the precalciner, the rest is released as exhaust air.

You might also like