Store 
User Stories
BlueScope Steel Eliminates Environmental Impact of Discharged Wastewater with pH Control System
Download this user story 209k
|
|
|
BlueScope Steel's Port Kempla steelworks produces more than five million tons of finished and semifinished flat steel each year. The production of each ton of steel generates about 2,700 liters of wastewater, posing a formidable disposal challenge.
To comply with Australia’s National Pollutant Inventory regulations and safeguard the environment, the BlueScope steel plants collect the wastewater in a mixing tank and treat it with pH-neutralizing agents before pumping it out of the plant. BlueScope Steel's existing treatment system was providing inconsistent pH control. To ensure that discharged wastewater remained environmentally safe, BlueScope Steel designed a new control system. Using MathWorks products and Model-Based Design, they completed the project in just three months.
"We were unable to solve this problem for two years, as we didn't have modeling capability or knowledge," says BlueScope Steel automation engineer Dr. Amid Bakhtazad. "MathWorks products enabled us to quickly implement the system--with very successful results for BlueScope Steel and the environment."
Challenge
Wastewater from four separate plant streams at BlueScope Steel is collected and analyzed in a storage tank. The plant controller adjusts the pH level by cycling valves to deliver lime, acid, ferric sulphate, or other agents into the neutralizing system. BlueScope Steel’s previous control system consisted of a cascade loop and feedforward compensation. It frequently malfunctioned, leading to blockages in the dosing valves and undesirable pH levels for discharged streams—endangering aquatic life and contaminating drinking water.
To prevent further failures, BlueScope Steel needed to design a new controller as quickly as possible. They required software that would enable them to evaluate various control strategies and test a range of parameter settings. The software would need to support modeling, testing, and system validation.
"Using Model-Based Design with MATLAB and Simulink enabled us to quickly try out multiple control strategies and validate the system before putting it into operation. This approach saved us time and ensured a better product."Dr. Amid Bakhtazad,
BlueScope Steel
Solution
The BlueScope Steel team began by modeling the treatment process in MATLAB and the System Identification Toolbox, using data collected over a wide range of plant operating conditions. This approach enabled them to examine several potential model structures. After validating each model against actual process data, they selected a 3-input by-1-output linear model. They used this model and the Control System Toolbox to develop a controller, and then to develop optimization routines to find the best parameter settings for the proposed control structure.
Working in Simulink, they simulated the treatment system using the previously identified model and a variable-gain PID-based controller. They studied the stability of each control system and evaluated several feedback control structures, arriving at a cascade feedback structure that eliminated the need for the feedforward loop.
With Simulink Response Optimization, they further improved the control system by tuning the controller parameters to best match the simulation performance and the observed plant behavior.
The process control engineers implemented the new control strategy on the wastewater treatment plant’s Yokogawa distributed control system. They tested the system rigorously by adding acid to the inlet streams for 30 minutes. The controller operated the lime flow valve to maintain the proper pH of the wastewater. When the acid addition ceased, the controller throttled the lime flow valve to close, confirming the system’s ability to withstand disturbances.
Results
- Environmental impact eliminated. The previous controller could maintain acceptable pH levels only 84% of the time. The new controller maintains these levels 100% of the time.
- Controller performance improved. "We monitored the system for a week," says Bakhtazad. "The collected data showed improved pH control. The pH level was not only better controlled, but it fluctuated evenly around the set point, which ensured better neutralization."
- Development streamlined. By using MathWorks tools for modeling, control algorithm design, simulation, and optimization, BlueScope Steel completed the project in just three months— avoiding the extended effort and cost of trial-and-error approaches.
Learn more about BlueScope Steel
