Computer Optimization Control of Reheating Furnace in Seamless Steel Tube Hot Rolling Production Line

Literature Overview

This paper by Jiang Zeyi, Zhang Xinxin, Chen Xiaodan, and Zhang Haiyan (2003), published in Industrial Heating, presents the development and implementation of a computer optimization control system for the ring reheating furnace and the billet reheating furnace on a seamless steel tube hot rolling production line. The work was conducted jointly by University of Science and Technology Beijing and Baoshan Iron and Steel Group Co., Ltd. The authors established mathematical models for both furnaces and implemented a computer-based temperature tracking and optimization control system that achieved satisfactory control performance in actual production operations.

Process Context and Control Challenges

The seamless steel tube hot rolling process involves a sequence of thermal and mechanical operations, with reheating being the critical first step that determines the quality of the subsequent piercing, rolling, and finishing operations. The ring reheating furnace is used for the initial heating of steel billets, while the billet reheating furnace (or rough tube reheating furnace) provides the secondary heating for rough tubes before the final rolling pass. The control of these furnaces presents several significant challenges:

Thermal inertia: The massive refractory lining of the furnace creates significant thermal lag, making real-time temperature control inherently difficult.
Variable charge: The dimensions, alloy composition, and initial temperature of incoming billets vary, requiring adaptive control strategies.
Multi-zone temperature profile: The furnace must maintain a specific temperature distribution along its length to ensure uniform heating without excessive scale formation or grain growth.
Energy efficiency: Fuel consumption is a major operational cost, and over-heating or under-heating both lead to quality defects and energy waste.

Mathematical Model Development

The core contribution of this paper is the establishment of mathematical models for both furnaces. These models typically incorporate the following physical phenomena:

Model Component	Description	Governing Equations
Heat conduction	Heat transfer through billet cross-section	Fourier heat equation with time-dependent boundary conditions
Heat convection	Heat exchange between furnace gas and billet surface	Newton's law of cooling with radiation component
Radiation	Thermal radiation from furnace walls and burners	Stefan-Boltzmann law with view factors
Fuel combustion	Heat release from fuel-air mixture	Stoichiometric balance and enthalpy calculations
Gas flow	Furnace atmosphere circulation	Continuity and momentum equations

The mathematical models are typically reduced to a set of algebraic or differential equations that can be solved iteratively within the control loop. The key challenge is balancing model accuracy with computational speed, as the control system must provide setpoint adjustments within seconds to maintain stable furnace operation.

Computer Control Architecture

The computer optimization control system described in the paper likely follows a hierarchical architecture:

Process monitoring layer: Temperature sensors (thermocouples, infrared pyrometers) distributed along the furnace length provide real-time temperature measurements.
Model computation layer: The mathematical models are executed to predict the billet temperature evolution and compute optimal fuel flow rates and burner settings.
Control execution layer: The computed setpoints are transmitted to the field control system, which adjusts fuel valves, air dampers, and burner positions.
Feedback and correction layer: Measured temperatures are compared with predicted values, and the model parameters are updated to compensate for model-plant mismatch.

The temperature tracking capability mentioned in the paper refers to the system's ability to monitor and control the temperature of individual billets as they move through the furnace, which is particularly important for seamless tube production where the chemical composition and dimensions of each billet may differ.

Engineering Practice and Control Performance

The implementation at Baoshan Iron and Steel demonstrates that computer optimization control can significantly improve furnace operation. Typical performance improvements include:

Temperature uniformity: Reduction of billet temperature variation from ±30°C to ±10°C or better.
Energy savings: Reduction of fuel consumption by 5-15% through elimination of over-heating.
Scale reduction: Lower scale thickness due to optimized heating rates, reducing subsequent descaling costs.
Throughput improvement: Shorter heating times through optimized temperature profiles, increasing furnace productivity.

Key Technical Considerations

Several technical aspects deserve emphasis for engineers considering similar implementations. First, the accuracy of the mathematical model is paramount. Model parameters such as heat transfer coefficients, furnace wall emissivity, and gas flow distribution must be calibrated against measured data. Second, the control system must handle disturbances such as variations in billet charge, fuel quality fluctuations, and ambient temperature changes. Third, the system must include safety interlocks to prevent dangerous conditions such as furnace over-temperature, fuel gas leaks, or inadequate combustion air supply.

The paper, while somewhat brief in its technical details, represents an important early contribution to the application of computer optimization control in steel tube production. The methodology described is consistent with modern process control principles, and the fundamental approach remains applicable to contemporary furnace control systems, which now benefit from advanced sensor technology, faster computing platforms, and more sophisticated optimization algorithms.

Study Insights

The integration of mathematical modeling and computer control in industrial heating processes exemplifies the power of model-based control in process industries. The seamless steel tube production line serves as an excellent case study because it involves complex thermal processes with significant quality and energy implications. The work by Jiang and colleagues demonstrates that even relatively simple mathematical models, when properly implemented in a computer control system, can yield substantial operational improvements. This principle continues to underpin modern industrial control systems, where the emphasis has shifted from model development to model maintenance and adaptation in response to changing operating conditions.