"Ask the Experts" is moderated by Béla Lipták, process control consultant and editor of the Instrument Engineer's Handbook (IEH). The 4th edition of Volume 3, Process Software and Networks, is in progress. If you are qualified to contribute to this volume or if you are qualified to answer questions in this column or want to ask a question, write to firstname.lastname@example.org.
Q: I've read the discussion about the tempered water system and heating to an exothermic reactor with interest because we too have systems like that—in our case, batch reactors. We charge the reactor and then heat the contents. We want to reach the desired temperature as quickly as possible, and we want stable temperature control as we begin charging reactants and start an exothermic reaction.
We tried moving the split range away from 50% to compensate for different process gains between heating and cooling, but we had an additional problem. Our boiler response is slow, so the steam valve should not open or close too fast. Because we still want fast control during the exothermic reaction, when mostly cooling is required, we had to limit the rate of change of steam valve opening (%/minute). Therefore, we have not only different gains during heating and cooling, but the response during heating is also limited.
I have thought of two ways to handle this problem.
The first, as was confirmed in the trial, was to switch tuning parameters as we move from heating to cooling, based on controller output. This way we can limit the speed of steam valve movement to 10%/minute, resulting in a very slow response. I think the tuning of the PID in this service should be such that the output rate of change limit is not touched. Would a first-order exponential filter instead of the linear output rate of the limiter be better?
Another way to slow the change in steam flow during heating would be to slow the setpoint (SP) ramping, so it would not be too restrictive during the exothermic reaction. The rate of moving the SP could be set differently for the two phases. Here, would a first-order exponential SP rate of change limit not be better than the linear one? Wouldn't changing the rate of SP change cause upsets at the boilers?
Right now, in order to avoid overshooting at the end of the heat-up phase, I limit the rate of opening of the steam valve. I don't allow the percentage opening to exceed the controller error in percent times a gain factor. This way, when we reach the target temperature, the steam valve is closed and the temperature will not overshoot unless the steam valve is leaking.
Do you see any other options ?
A: Insert a rate limiter between the secondary controller output and the steam valve. Then connect the limiter output to the external reset feedback output of the controller, if available. In this way, the controller will stay tuned properly regardless of adjustments to the rate limiter. However, if the valves are split-ranged from a single controller output, the rate-limiter will have to be bypassed while the coolant valve is operating.
A: In a well-designed plant, the availability of utilities should not limit the control of the process.
As to the best control option, I agree with Greg Shinskey (above). Figure 1 shows how you might operate the reactor, while limiting the rate at which the demand for steam can be changed during heating. This involves the addition of two functions (shown in red). In black, I show the basic split-range reactor temperature control system used in installations where there is no limitation on the rate at which utility demand can change.
One software function that I added is the rate limiter (RL), which has an adjustable setpoint (SP). This way, the limit on the rate at which steam demand can change can be automatically changed. So, when the slow heat-up phase is over and the reactor is switched to the reaction phase, the response of the temperature control loop is slow if the changes required in steam rate is larger, and fast if it is smaller.
The output of RL throttles the steam valve and likewise provides the external reset (ER) to the cascade master temperature controller (TRC with PID control modes) only during heating. The purpose of the switching function (SS) is to switch the ER to be received from the measurement of the slave temperature controller (TRC) when the exothermic reaction is started (the mostly cooling phase). In this way, during heat-up, the ER signal is received from the RL during cooling from jacket temperature (TT), and the setpoint of the RL can be automatically changed to vary the speed of response as a function of the sizes of upsets.
Naturally, before doing all that, first I would visit the utility building and would try to speed up the boiler(s).
We always suggest using proper design and appropriate control strategies; we try to use ramps, limiters and other handcuffs for security.
Q: Is there a difference between the definitions of smart actuators and smart positioners? If there is, what are the main features of each considering their capability, limitation, advantage and future trends?
A: An actuator is any device capable of changing the opening of a valve or modifying the position, speed or any other operating condition of pumps, compressors, etc. They can be as simple as a coil operating an on-off solenoid or more sophisticated pneumatic or electric throttling devices, including variable-speed drives (VSD). VSDs are superior to valves because of their speed and energy savings, and they have no hysteresis.
The actuator by itself is not necessarily provided with feedback; it does not "know" if the desired position was in fact achieved. The term "smart"does not mean much. It usually means fieldbus connectivity, but it can still be just a sales gimmick, depending on the information provided. It can have real value in the areas of self-diagnostics, historical data collection, visual displays, maintenance scheduling, process property measurements, etc.
The positioner is a position control loop consisting of a position sensor and a controller. Its job is to eliminate the difference between the measurement and setpoint of this position controller. In addition, you can think of a positioner as a cascade slave controller in which the cascade master is a temperature, pressure, level, flow or any other variable.
The positioner can also be used to change the characteristics of the control valve artificially if the wrong valve was installed, or to change the dynamics of the loop if needed. Finally, the positioner can be part of and supplied with the actuator. As to the term "smart positioner," it can either imply just a sales gimmick, or provide valuable extra features, similar to those in "smart actuators." To determine what the term really means requires careful analysis of the bids summitted.