BLUESCOPE Steel has installed a new wastewater pH control system at its Port Kembla Steelworks in New South Wales.
The plant is capable of producing five million tonnes per year of finished and semi-finished flat steel products for Australian and international customers.
As with any large-scale industrial facility, waste products need to be treated correctly to meet regulatory and ecological commitments.
The new control system, developed using Matlab and Simulink software, ensures that the wastewater discharged at the steelworks remains environmentally safe.
One of the wastewater treatment facilities at the site collects waste from four separate plant line streams. The waste is stored in a mixing tank, where it is analysed and treated to ensure it can be safely pumped to the environment.
Because the stream of wastewater is continuous, treatment is a highly automated process.
Because the former control process did not deliver the level of consistency required by the company, a new system was developed using Matlab mathematical software from The MathWorks.
According to the automation engineer at BlueScope Steel, Dr Amid Bakhtazad, the desired pH level is reached by neutralising plant waste through a series of valves that open and close, delivering neutralising agents such as lime, acid or ferric sulphate into the tank.
“The previous system did not always deliver acceptable accuracy,” said Bakhtazad. “Also, malfunctions in the control process could sometimes lead to blockages in the dosing valves and eventually undesired pH for the discharged streams.”
The new control system, developed using a classic model-based design, ensures that discharged wastewater remains constantly within the desired environmental pH range.
The process was first modelled in Matlab using collected data, a control system was created using the control system toolbox, and then an optimisation routine was used to find the best parameter settings for the proposed control structure.
This algorithm allowed Bakhtazad to satisfy some of the process and parameter constraints that existed in the real-world process.
Using The MathWorks’ Simulink program, the control techniques were then examined in a simulation environment.
This model was validated against plant data in different situations and scenarios, and tested before implementation.
The final system structure selected provided the best stability and consistency.
This approach ensured an accurate system was built and tested before being put into operation.
The system took only three months to develop.
“We monitored the new control system for a week,” said Bakhtazad. “The data collected showed an improvement in the performance of pH control. The pH level was not only being controlled better but fluctuated evenly around the set point, which ensured a better neutralisation and reduced the impact on the environment.
“We also experienced fewer line blockages in the valves, previously caused by malfunctions in the control system,” said Bakhtazad. “The new system has been implemented with very successful results for BlueScope and the environment.”