In the operation of distillation columns, controlling the bottoms level is crucial for maintaining product quality and overall process stability. Typically, the bottoms withdrawal flow is used as the primary means of regulation. However, there are complex process control challenges, particularly when dealing with special operating conditions. These include instances where the bottoms withdrawal exhibits intermittent operation, microflow conditions, or when it must be fixed due to process constraints. In these cases, the traditional level-to-withdrawal control loop fails to maintain stability. To address these issues, alternative strategies such as adjusting the reboiler steam flow, controlling side-stream withdrawal, or intervening in the reflux system are required.
In recent advanced process control (APC) projects, two distillation systems have shown significant insights into control strategy optimization. These columns do not directly send the bottoms withdrawal to external systems but instead return it to an upstream stripper tower. The core idea behind this design is that the bottoms of the distillation column contain recoverable product components. By returning them to the upstream column, both product purity can be enhanced, and valuable components can be recovered. From a system architecture perspective, the upstream tower essentially assumes the role of the “lower column” in a two-column distillation system, while the distillation column itself serves as the “upper column.” This dual-column setup significantly improves system efficiency.
Case Study 1: Managing Fluctuations in Recycled Material
In one of the facilities, the original control strategy involved controlling the reboiler steam flow to regulate the bottoms temperature (TC) and controlling the bottoms withdrawal flow to regulate the level (LC). However, during distillation column fluctuations, the level control loop caused upstream tower oscillations, which in turn triggered system-wide instability. The operations team initially attributed this to environmental temperature changes, but the implementation of APC revealed that despite optimizing the bottoms temperature with advanced controls, level control-induced fluctuations persisted.
The operations team had experience in manually stabilizing the bottoms withdrawal, but this approach lacked automated control support, making level fluctuations the primary bottleneck to system stability, ultimately resulting in low APC system usage.
To address this challenge, we restructured the control architecture: the bottoms withdrawal was converted to flow control (FC), while the level was regulated by adjusting the reboiler steam flow. This solution was designed based on the following principles:
- Excess bottoms withdrawal: When the bottoms contains sufficient light components and has been fully stripped, there is no need for stringent control of the reboiler temperature.
- Insufficient withdrawal: If product quality is compromised due to insufficient withdrawal, increasing the withdrawal flow will rectify the issue.
This redesign aligned perfectly with the operational logic:
- Once the withdrawal flow stabilized, fluctuations in the upstream tower’s recycled material were reduced by 42%.
- The disturbance propagation path in the four-column distillation system was effectively blocked.
- Without modifying the underlying control logic, system stability was achieved through strategic optimization.
Case Study 2: Level Control Under Fixed Withdrawal Conditions
Another facility faced a process constraint requiring the bottoms withdrawal flow to remain constant. The original APC system was in conflict with the operational needs, resulting in the system being largely inactive. Using a similar design approach as in Case Study 1, we fixed the bottoms withdrawal flow and controlled the level through reboiler steam flow adjustments.
This approach resulted in a significant reduction in manual interventions, with the system usage rate consistently maintained at 100%. The control logic was in harmony with the operators’ experiential models, ensuring smooth operations.
Core Insights from APC Implementation
Human Factors Engineering Priority: Control strategies must align with operators’ cognitive models. Focusing solely on algorithm complexity without considering user experience will likely result in system abandonment. In our case, ensuring compatibility with operators’ mental models was key to successful APC implementation.
Value-Creation Orientation: Advanced control should focus on creating value through the “automation improvement – reduction in manual interventions” feedback loop. If additional PID tuning or redesigns are required post-implementation, the system design has a fundamental flaw.
Disturbance Source Identification: In multi-column systems, recycled material fluctuations are often a hidden disturbance source. To achieve stable operation, it’s essential to restructure the control architecture to isolate such disturbances.
Conclusion
The challenges faced in implementing APC effectively highlight a gap between “technological perspectives” and “operational perspectives.” In many industries, APC systems fail because there is a mismatch between the algorithmic expertise of control engineers and the domain knowledge of operations engineers. The key to successful APC implementation is integrating the strengths of both disciplines. By combining advanced control algorithms with a deep understanding of operational processes, a truly effective and sustainable control system can be created.