Optimizing Soy Processing with Model Predictive Control
Every efficiency counts in industrial processing plants. For one leading soy processing facility, managing energy consumption and improving product quality were key priorities. This client turned to Interstates for a cutting-edge solution—Model Predictive Control (MPC)—to optimize their Disolventizer Toaster (DT) operations. The results? Reduced steam utility usage, enhanced temperature stability, and cost savings that exceeded their expectations in record time.
What is Model Predictive Control?
You can think of MPC as a brain that anticipates and automatically adjusts for the future. Unlike traditional control systems, which react to single variables in real time, MPC uses mathematical models to evaluate multiple variables simultaneously. By analyzing current and past data, MPC predicts how a system will behave and makes proactive adjustments to optimize performance.
In industrial settings, this predictive power allows owners and operators to achieve unparalleled precision, particularly for their most complex processes. For our soy processing client, MPC became the key to tackling inefficiencies in steam usage and temperature management. With this solution, they were able to reduce variability while pushing process limits and maintaining quality.
Overcoming Challenges with MPC
The soy processing client faced some technical challenges related to production, including needing to reduce the plant’s steam consumption to manage high energy costs. The client also wanted to improve temperature stability at the facility because inconsistent temperatures were affecting their product quality. In addition, a higher-than-necessary temperature set point for end-product quality was leading to higher energy usage and costs. To address these pain points, Interstates’ analytics team deployed Rockwell Pavilion8 software to create an MPC solution that met the client’s needs. Utilizing MPC means that machines don’t have to wait until there are problems to make a change; instead, MPC constantly makes tiny, proactive corrections.
On this project, our MPC approach incorporated both manipulated and control variables. Manipulated variables are elements that the MPC can control, such as flow control valves and extruder speeds. The control variables are the outcomes or results that should be optimized. On this project, this included dome temperature, exhaust pressure, and gearbox amperage.
These key variables were tightly integrated into the MPC system, allowing continuous, proactive adjustments that ensured optimal performance.
The Project Process
Interstates followed a structured and collaborative approach to ensure the success of the MPC deployment at this site:
- Project Kickoff: The team began by meeting with the client to draft design specifications, analyze process variables, and define clear success metrics.
- Technology Deployment: This phase involved reviewing control system health, performing data readiness checks, installing software, and connecting it to existing systems. This information was compiled into a feasibility study that outlines an estimated return on investment for implementing the MPC and estimates a variance reduction over the targeted process.
- MPC Enablement: Custom logic was developed and tested to integrate manipulated and control variables. Calculations were refined to ensure the MPC application would perform as expected.
- Step Testing: Step test experiments were designed and executed to fine-tune the system. These tests were critical to understanding the system’s dynamic responses and achieving optimal settings.
- Commissioning: The final phase included steady-state assessments, setting up KPIs, operator training, and project closeout.
By taking this methodical approach, we completed the project in just three months—less than a third of the estimated timeline. The client achieved ROI in just nine months.
Results and ROI
The results of our MPC deployment at this soy processing facility were noteworthy:
- Steam Savings: Steam consumption was reduced by one pound per bushel, translating to $162,000 in annual savings.
- Temperature Stability: The DT high-temperature set point decreased by 6.8% (from 161°F to 151°F), improving consistency and product quality.
These improvements allowed the client to achieve a faster return on investment while gaining a more stable and efficient process. Beyond the numbers, the enhanced quality and consistency of their product have positioned the client to better meet market demands.
Looking Ahead: Expanding the Benefits of MPC
The success of this project has sparked interest in leveraging MPC across other areas of the client’s operations, including scaling this solution to sister sites. By partnering with Interstates, the client optimized their DT process and demonstrated the potential of advanced industrial analytics to drive significant cost savings and operational improvements.
Interstates’ expertise in implementing Rockwell Pavilion8 MPC lets us empower our clients with innovative solutions that strengthen their competitive advantage. Whether it’s improving energy efficiency or enhancing product quality, MPC offers a powerful way to unlock new levels of performance.
At Interstates, we’re committed to helping clients stay ahead in an ever-changing industry. Want to learn how our analytics solutions can benefit your facility? Contact us today.