Cascade control recommendation tips


Cascade control is most effective when the lower loop is 4 times or more faster than the upper loop (cascade rule). If the lower loop is too slow, the upper PID must be prevented from changing the lower PID setpoint faster than the lower loop can respond to prevent a burst of oscillations for large and fast disturbances or setpoint changes. 

Cascade control can greatly reduce the effect of disturbances entering the lower loop. If the cascade rule is not violated, disturbances are also reduced that enter the upper loop. The ultimate period of the cascade loop is less than the ultimate period of a single loop with the same dynamics. 

Lower loops can also isolate nonlinearities from the upper loop and better enforce limits on lower process variables (e.g. coolant temperature). Lower flow loops enable flow feedforward and process simulation, metrics, and analysis. 


  1. Use valve positioners (digital valve controllers) on all control valves.
  2. If the valve stroking time (time for 100% stroke) is significantly greater than the reset time of the PID manipulating the valve, add volume booster(s) on valve positioner output(s). Open the booster bypass enough when stroke testing the valve to prevent high frequency cycling of valve position (e.g. 1 cps).
  3. Tune valve positioner for fast response avoiding the use of integral action.
  4. Use lower flow loops wherever possible to compensate for nonlinear installed flow characteristics and to provide the measurements needed for mass, mole and energy balances and cost analysis.
  5. Use jacket and coil temperature loops for bioreactor, crystallizer, and chemical reactor temperature control.
  6. Tune the lowest loop first, then the next lowest loop, and ending up tuning the upper most loop last. For example, tune the valve positioners first, then the flow loop, then the jacket temperature loop, and finally the reactor temperature loop.  
  7. Tune lower loops to be as fast as possible.
  8. Do not use setpoint filters on lower loops. Use lambda tuning for coordination of flow loops (e.g. maintaining stoichiometry for inline blending and reaction control). Set the lambda (closed loop time constant) of each loop to match the lambda of the slowest loop.
  9. If the lower loop cannot be made 4 times faster than the upper loop, tune the upper loop slower by making the upper loop lambda 4 times larger than the lower loop lambda. Consider abandoning cascade control using either the lower or upper loop process variable for single loop control depending upon whether the disturbance size and speed is more problematic in the lower or upper process. The temperature control of some bioreactors is best done by simple jacket temperature control because the jacket volume is comparable to the process volume and process disturbances from cell growth are incredibly small and slow.
  10. Use external reset feedback of lower loop PV to upper loop PID so the upper loop PID output does not try to change faster than the lower loop can respond.
  11. If flow measurement rangeability is insufficient, there may need to be a switch to direct throttling of the control valve, a common practice in boiler drum level control at low steam rates and start up. A better solution is a computed flow from the installed valve characteristic and a bumpless transition to an inferential flow measurement maintaining cascade control.
Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.


  • <p>It would be very useful if the readers could be better informed than the recommendation listed in the last Control Update under the title: “Cascade Control Recommendation Tips...</p> <p>...8.Do not use setpoint filters on lower loops. Use lambda tuning for coordination of flow loops (e.g. maintaining stoichiometry for inline blending and reaction control). Set the lambda (closed loop time constant) of each loop to match the lambda of the slowest loop.”</p> <p>I find very saddening that even after someone of the stature of Greg Shinskey gets his paper “A case against Lambda tuning” published in your May 2012 magazine, and that over the years several people from industry and academy, including one of the inventors of the IMC tuning, have acknowledge the inherent deficiency of the method to cope with the most frequent dynamic behavior found in process control, and yet your magazine keeps promoting it through this type of recommendations.</p>


  • <p>"Let me first say I think Greg Shinskey has provided an understanding of the effect of process and equipment design on dynamics and PID control in his books that are go far beyond what anyone has done. I doubt there will ever be another individual that will do as much as what he did to provide a deep knowledge based on chemical and mechanical engineering principles. I recommend practitioners read all of his books. Unfortunately publishers do not realize the significance of the unparalleled knowledge in his books and have let most of them go out of print."</p> <p>"I shared Shinskey's concern about lambda tuning for over a decade. The major concern centered on a reset time that was much larger than what is needed to return to setpoint after a load disturbance to self-regulating processes with a large time constant to dead time ratio. Many of these loops are very important in that they involve concentration, pH, and temperature control of large continuous liquid unit operations such as distillation, neutralization, and reaction. A second concern was the lack of derivative action for these processes."</p> <p>"I realized about 2 years ago that if I switch from lambda tuning rules for self-regulating processes to lambda tuning rules for integrating processes with derivative action when the time constant to dead time ratio became greater than 3 and used a lambda equal to half the dead time, the problem that Shinskey raised about the recovery from load disturbances largely goes away. In fact the tuning settings converge to those of Ziegler Nichols reaction curve method and many other tuning methods based on minimizing the peak and integrated error for disturbances at the process input (load disturbances). In the lambda integrating process tuning rules the lambda becomes the arrest time for the load response which is more appropriate for these processes. In the lambda tuning rules for self-regulating processes, lambda was the closed loop time constant for a setpoint response. The rate time is set equal to the secondary time constant in a second order plus dead time approximation. If the secondary time constant is not identified, I use 1/2 of the dead time as an inferred secondary time constant."</p> <p>"To provide more robustness and less tendency for the loop to oscillate I have since used the rule of thumb to switch from lambda tuning rules for self-regulating processes to lambda tuning rules for integrating processes when the time constant to dead time ratio is greater than 4, and use a lambda equal to the dead time. To deal with unknown nonlinearities that cannot be corrected by adaptive tuning and for interaction that cannot be corrected by decoupling and to reduce resonance from disturbance oscillations with a period close to the ultimate period of the loop, I have increased lambda to be 3 dead times. You can also readily compute an arrest time that maximizes the absorption of variability for surge tank level control. Much of excess process variability in processes has been traced to surge tank level controller oscillations and excessive integral action (too small of a reset time). The lambda tuning rules automatically enforce the product of the PID gain and reset time to be greater than 2 times the inverse of the integrating process gain to prevent slow rolling oscillations. Most integrating processes do not use the maximum PID gain that is possible and end up with a reset time that is several orders of magnitude too small. Users do not realize that too small of a gain causes slow oscillations (10 times the ultimate period) from violation of the inequality besides the fast oscillations from too high of a PID gain learned in courses and seen throughout the control literature. This window of allowable PID gains where too small besides too large of a PID gain causes oscillations occurs for near-integrating besides true integrating processes."</p> <p>The approximation of a self-regulating process as a near-integrating process is consistent with what the PID sees in the time frame of an aggressive load response. The PID makes most of its correction in the first two dead times. During this time the process appears to be ramping like an integrator. The literature often only shows the lambda tuning rules for self-regulating processes and uses a lambda factor instead of lambda leading to considerable miss-understanding. I think I am the only one who is advocating the switch to integrating tuning rules when the time constant to dead time ratio is greater than 3 or 4 with a lambda that ranges between 1/2 to one dead time depending upon the degree of robustness needed."</p> <p>"The other part of the recommendation that setpoint filters not be used on secondary control loops is consistent with Shinskey's recommendations. The idea here is that the secondary loop should provide a more immediate reaction to demands of the primary that is gained by mostly proportional mode action. For the primary loop, a setpoint lead-lag would be used to help a PID with load response tuning have minimal overshoot in its setpoint response. The lag is set equal to the reset time of the primary PID and the lead equal to 25% of this lag time. I have since realized that a 2 degrees of freedom structure can give similar results to a setpoint lead-lag with beta=0.5 and gamma=0.25, which is useful because some PID blocks offer this structure whereas a setpoint lead-lag would require a configuration change."</p> <p>"Lambda tuning allows setting lambda for a much larger spectrum of objectives than minimizing the peak and integrated error in a load response. For the case cited in the recommendation, the use of the same lambda for the flow loops reduces temporary unbalances in the stoichiometry of the blend from a primary loop for analyzer feedback control and for production rate changes or product grade or type transitions. In an inline blend system, you have essentially plug flow (no back mixing or axial mixing) and hence no primary time constant for filtering out short term transients. The residence time becomes mostly a transport delay. If there is no back mixed volumes downstream these short term fluctuations appear in the product. The worst case is where the product is a plastic or paper sheet. Here all the short term variability upstream is captured in the sheet causing sheet quality problems. Lambda tuning originated in the pulp and paper industry because their upstream unit operations where mostly inline and the product was a sheet traveling at 60 mph. A sheet break dumps a heck of lot of sheet on the floor. In the last 10 years lambda tuning has been used more and more in chemical processes, particularly in batch unit operations (known to be integrating processes) that clearly demanded the use of lambda integrating process tuning rules."</p>


  • <p>I forgot to mention that the equations in many of the older publications do not show the lambda tuning method advocated today. Lambda factors are not used. Lambda is entered directly as the arrest time for near and true integrating processes and as a closed loop time constant for self-regulating processes. As documented in 1999 lambda is thought of as a value that is a multiple of the dead time for gain margin analysis. Furthermore, a low limit to the reset time of 4 times the dead time for near and true integrating processes is used. The numerator for the PID gain calculation still uses 2 times lambda plus the dead time. Runaway processes are treated as integrating processes and a low limit to the rate time of 1/2 the loop dead time is used to help prevent an accelerating response (e.g. heat of exothermic reaction exceeds cooling rate).</p>


RSS feed for comments on this page | RSS feed for all comments