PID Control TUNING
Process industries rely upon PID controllers to maintain production rates and production quality. Few engineers and technicians can tune PID controllers. Lack of process knowledge and lack of understanding control priorities causes much of this inability to successfully tune a process.
Over 30% of the PID controllers are in manual for most processing plants. This creates a burden upon operations:
The Manual Control Problem
This adds Extra responsibilities on the operators when operating controls in manual,
which leads to:
- Off-spec product
- Over-specification product, i.e. product give away
- Inability of operators to handle process upsets
- High frequency of plant trips and excessive production downtime
- In Short, loss of revenue
The principal reasons for controls in manual are:
- Poor tuning of controls
- Lack of operator understanding or operator trust in the control scheme
- Poor training of the operator
- Lack of on line graphics to explain the controls
- Lack of operator involvement with the control design
- Difficulty in bumplessly initiating automatic control
- Improper handling of control constraints
- Poor design of the loop such that it does not function properly
Tuning Goals
TUNING GOALS
The tuning goal for most engineers and operators is: Control the process such that all controllers always maintain a minimum error between setpoint (SP) and the measured variable (PV).
Unfortunately, that goal is only feasible as an average, not instantaneously. In any process system, inputs will change, such as the desired production rate, the ambient temperature, fuel quality, etc. These changes will upset the process. Tuning everything tightly results in instability.
A major control misconception is that all variables must be tightly controlled. Temperatures which control product quality (most temperatures) should be tightly controlled. Most pressures should be tightly controlled.
Most levels do not impact product quality and can be loosely controlled. Loose level control dampens out process changes and allows the process to glide into new operating conditions, instead of bouncing to new conditions.
Another misconception in tuning is that response times and delay times (process dynamics) are consistent. Process dynamics change with production rates, quench flows, recycle rates, catalyst activity, and many other process conditions. When tuning, the goal is often to tune such that for a large range of dynamics the process stays stable and does not oscillate. Aggressively tuned controls can lead to instability when dynamics change.
Another misconception is that mechanical issues can be overcome by tuning; they cannot.
The true goal of the control engineer is to tune the process such that the plant minimizes deviation from targets (Setpoints) for variables which effect product quality or process safety, while maintaining non-quality effecting variables (think levels) within ranges; all while minimizing process oscillations.
Why are PID controllers left in manual?
The principal reasons for controls in manual are:
- Sensor input problems and valve problems
- Poor tuning of controls
- Lack of operator understanding or operator trust in the control scheme
- Poor training of the operator
- Lack of on line graphics to explain the controls
- Lack of operator involvement with the control design
- Difficulty in bumplessly initiating automatic control
- Improper handling of control constraints
- Poor design of the loop such that it does not function properly
Collaborative controls PID Tuning Assistance
Our senior engineers have been troubleshooting and tuning processes for nearly 50 years; this allows Collaborative Controls to quickly:
- Review the structure and timing of control strategies to find errors in the design
- Review the process flow to define proper tuning goals for each loop.
- Evaluate large groups of PID controllers to determine if the tuning falls within process norms for the control service.
- Recommend initial changes after observation of trends
- Evaluate loops which operators identify as “problematic” for design or structural issues.
- Tune critical loops or loops with complex dynamics (primarily temperatures, and some pressures and levels) by step tested for frequency domain identification, and tuning using RLSTuner
- For Complex Loops: operator discussions and configuration evaluation are performed to recommend improved configuration, detailed graphics, or revised schemes
- Train operators in control schemes
Much of the evaluation and observations can be performed remotely via online meetings or in site visits
Work is preferred to be performed along with a client engineer, such that the client is trained in control evaluation