Control system integrator helps stainless steel manufacturer save energy, reduce reprocessing
By Chetan Chothani
J&L Specialty Steel operates a continuous direct roll, anneal, and pickle (DRAP) line at its Midland, Penn., plant. The unit is designed to process 300- and 400-series stainless steel. With a capacity of 60 tons per hour, the line can handle coils from 0.034- to 0.312-inch gauge, and 26 to 62 inches wide. Line speed varies from 20 to 150 feet per minute. The maximum cold reduction is 62%.
J&L retained Adaptive Resources to design model predictive controls (MPC) to manage the annealing process. When going from one type of coil to another, the strip annealing temperature at the furnace exit must remain relatively constant. The excess oxygen in each zone must stay within a narrow band to eliminate one source of oxides forming on the strip.
Annealing: high temperatures, rapid cooling
Annealing involves heating metals to high temperatures and then rapidly cooling them. The process removes stresses; induces softness; alters ductility; imparts toughness, electrical, magnetic, or other physical properties; refines the crystalline structure; removes gases; or produces a definite micro-structure. In annealing, the temperature of the operation and the rate of cooling depend upon the material being heat-treated and the treatment's purpose.
The furnace portion of J&L's line consists of two unfired preheat zones and five top- and bottom-fired heating zones. Temperature control in each zone relies on roof temperature measurement as the process variable, with the output setting the top and bottom firing rates of the fuel/air controllers. Each zone temperature set point is generated by the host computer, which contains the furnace model.
Dynamic variables prove challenging
The dynamic variables of the process are:
- Transition from one series of stainless to another
- Change in the gauge reduction being applied
- Weld movement through the furnace
- The strip's cross-section
- Emissivity of the strip
- Line speed change
The line's conventional PID controls could handle the mission-critical Zone 5 (furnace exit) temperature under steady-state conditions. But variances in process dynamics caused upsets from which the PID controls could not recover fast enough. This resulted in processing problems and reprocessing of certain coils.
A particular problem occurs when a coil is cold-rolled; a length of metal on each side of a butt weld cannot be rolled. When this section of strip moves through a zone, the load could go up or down as much as 62% (the maximum gauge reduction applied) because the weld section is virtually the same length as the length of a zone. (Since the original installation, the plant started rolling the welds; however, the other challenges remain, and the MPC control strategy continues to be necessary.) A section of strip moves through a zone between 12 and 90 seconds, depending upon line speed.
Models suggest using independent predictive control blocks
Offline models were created automatically from historical data using Adaptive Resources QuickStudy APC software. During this process, the system integrator learned that replacing the five zone temperature controllers with five independent predictive control blocks within the APC could predict each zone temperature closely. The inputs to each block were zone temperature for process variable, with line speed, cross-section, and strip weld events as disturbance variables. The outputs went directly to the fuel/air controls of the respective zones.
With this established, the modeling software was put online and fine-tuned in the adaptive mode. As a result, zone temperatures are now controlled within 5°F, and coil transitions do not cause firing control problems. This provides better fuel/air control with less excess oxygen.Speed ramp control was implemented to prevent strip overheating when ramping up, especially during short line stops or slowdowns. Ramp-ups occur during startup and when changing to thinner gauges. This control's objective was to enable faster ramp-ups with tighter temperature control, increasing energy efficiency and throughput.
An existing industrial PC running a Windows XP operating system is used to implement the MPC application. This PC is networked over Ethernet to PLCs, which controls the DRAP line. The APC software communicates with the PLC via OPC. An operator display on the existing workstations provides access to the APC. Logic in the furnace PLC supplies the necessary interlocks and "bumpless" transfers between the APC and the conventional PID controls.
The APC furnace control enabled smooth transitioning from coil to coil. Additionally, the APC furnace control reduced temperature variability, which resulted in:
- Improved quality
- Reduced reprocessing of coils
- Increased throughput at lower operating costs
The APC speed ramp control enabled faster speed ramps and tighter strip temperature control, increasing energy efficiency and throughput.
ABOUT THE AUTHOR
Chetan Chothani (email@example.com) is president of Adaptive Resources, Carnegie, Penn., a member of the Control System Integrators Association (CSIA). You may contact him at +1 (412) 431-4662. To learn more about CSIA, visit the association's new website, www.controlsys.org.