Emerson Exchange 365

The major types of process responses and the different worlds of process applications are presented. Additionally, the oversight of not including and understanding the contribution of automation dynamics is addressed. The discussion of the response of the PID is started. The following key points support and augment the PID Options and Solutions - Part 1 recording for the ISA Mentor Program. 1-1: The most common type of process by virtue of the very common flow loop has a self-regulating response where the process will reach a steady state with the PID in manual for a given disturbance or change in PID output due to negative feedback within the process. 1-2: The response of the more important continuous processes for temperature, composition and pH control more directly related to product quality and process efficiency and capacity usually have an open loop time constant that is more than 4 times the total loop deadtime. In these cases, the PID controller, in its typical response time frame of 4 dead times, sees a ramping process. So far as the PID is concerned, the process appears to be integrating. It is extremely used to treat these processes as near-integrating and analyze them and tune them as integrating processes. You can easily convert between the integrating process open loop gain and the self-regulating process open loop gain and time constant. 1-3: Batch processes for temperature, composition and pH control typically have a true integrating response (no steady state) that is caused by zero negative feedback within the process in the normal operating range.  Level and gas pressure systems have a true integrating response. 1-4: The temperature control of highly exothermic chemical reactors, the cell concentration of bioreactors in the exponential growth phase, and the speed control of some axial compressors in surge have a runaway response characterized by acceleration due to positive feedback within the process. 1-5: Processes with slow or no negative feedback or that have positive feedback rely much more on the PID controller to provide the missing negative feedback that is best done by PID gain action. PID integral action is not as effective and can be detrimental if it is greater than the PID proportional action. This will be seen in in subsequent recordings as the slow large oscillations caused by a PID gain too small relative to the reset time.  Process control is achieved by negative feedback. The less negative feedback in the process, the more you need to rely upon the negative feedback provided by the proportional mode. 1-6: To compensate for a disturbance or to achieve a new setpoint for near and true integrating and runaway processes, the PID output must be driven past its Final Resting Value (FRV). 1-7: The most common metric for loop performance cited in the literature is an integrated absolute error (IAE).  This IAE is the area between setpoint (SP) and process variable (PV) on a trend recording. It can be representative of the total amount of material that is off-spec. 1-8: Of more immediate value is the peak error that is the maximum excursion of the PV from the SP. A large peak error can trigger side reactions, Safety Instrumented System (SIS) activation, cell death, and relief devices to blow. 1-9: PID PV overshoot for a setpoint change can have many of the undesirable effects of a peak error. 1-10:  PID output overshoot of the FRV is problematic for many self-regulating hydrocarbon processes with extensive heat integration and recycle. 1-11: Minimizing time to reach setpoint (rise time) is important to minimize the startup and transition time of continuous processes and the cycle time of many batch processes. There is typically a tradeoff between minimizing rise time and PV overshoot. For bioreactors, the cells are so sensitive and the batch cycle times are so slow that minimizing the overshoot for temperature and pH shifts is much more important, particularly for mammalian cell cultures. 1-12: Pulp, paper, food and polymer inline, extrusion and sheet processes may be dead time dominant if there is no heat transfer involved and valve or sensor lags are less than transportation delays. 1-13: While the process control literature typically uses the terms “process gain”, “process dead time” and “process time constant”, in most industrial processes the contribution of the automation system dynamics needs to considered particularly for processes whose dead time or time constant is less than 10 minutes or when control valves or sensors have a poor dynamic response or when at-line analyzers are used.  The effect of control valve or Variable Frequency Drive (VFD) gain, transmitter calibration and PID scale must always be considered in that the open loop gain is a product of the valve or VFD gain, process gain and measurement gain where engineering units cancel out giving an open loop gain self-regulating process that is dimensionless and an open loop gain integrating process gain of 1/sec. 1-14: In a first order plus dead time approximation for self-regulating processes, all of the time constants less than the largest time constant need to be taken as an equivalent dead time by multiplying the small time constant by a factor ranging from 0.28 to 0.88 as the ratio of the small to large time constant decreases. Since there is a tendency to underestimate dead time, I usually just sum up all the very small time constants as additional dead time. The sources of the largest contribution to the total dead time should be investigated for methods to reduce them. The good news is that as automation engineers, we may be able to significantly reduce the total loop dead time by a design and installation that takes into account the detrimental effect of dead time. If there was no dead time I would be out of a job, the PID gain could be set as high as wanted and perfect control would be possible for noise free measurements.  I will provide additional synergistic Key Points in future Control Talk Blogs. Until then check the photocopy machine for your copies of the latest memos on a paperless office.

The post, PID Options and Solutions - Part 1 , first appeared on the ControlGlobal.com Control Talk blog . The major types of process responses and the different worlds of process applications are presented. Additionally, the oversight of not including and understanding the contribution of automation dynamics is addressed. The discussion of the response of the PID is started. The following key points support and augment the PID Options and Solutions - Part 1 recording for the ISA Mentor Program. 1-1: The most common type of process by virtue of the very common flow loop has a self-regulating response where the process will reach a steady state with the PID in manual for a given disturbance or change in PID output due to negative feedback within the process. 1-2: The response of the more important continuous processes for temperature, composition and pH control more directly related to product quality and process efficiency and capacity usually have an open loop time constant that is more than 4 times the total loop deadtime. In these cases, the PID controller, in its typical response time frame of 4 dead times, sees a ramping process. So far as the PID is concerned, the process appears to be integrating. It is extremely used to treat these processes as near-integrating and analyze them and tune them as integrating processes. You can easily convert between the integrating process open loop gain and the self-regulating process open loop gain and time constant. 1-3: Batch processes for temperature, composition and pH control typically have a true integrating response (no steady state) that is caused by zero negative feedback within the process in the normal operating range. Level and gas pressure systems have a true integrating response. 1-4: The temperature control of highly exothermic chemical reactors, the cell concentration of bioreactors in the exponential growth phase, and the speed control of some axial compressors in surge have a runaway response characterized by acceleration due to positive feedback within the process. 1-5: Processes with slow or no negative feedback or that have positive feedback rely much more on the PID controller to provide the missing negative feedback that is best done by PID gain action. PID integral action is not as effective and can be detrimental if it is greater than the PID proportional action. This will be seen in in subsequent recordings as the slow large oscillations caused by a PID gain too small relative to the reset time. Process control is achieved by negative feedback. The less negative feedback in the process, the more you need to rely upon the negative feedback provided by the proportional mode. 1-6: To compensate for a disturbance or to achieve a new setpoint for near and true integrating and runaway processes, the PID output must be driven past its Final Resting Value (FRV). 1-7: The most common metric for loop performance cited in the literature is an integrated absolute error (IAE). This IAE is the area between setpoint (SP) and process variable (PV) on a trend recording. It can be representative of the total amount of material that is off-spec. 1-8: Of more immediate value is the peak error that is the maximum excursion of the PV from the SP. A large peak error can trigger side reactions, Safety Instrumented System (SIS) activation, cell death, and relief devices to blow. 1-9: PID PV overshoot for a setpoint change can have many of the undesirable effects of a peak error. 1-10: PID output overshoot of the FRV is problematic for many self-regulating hydrocarbon processes with extensive heat integration and recycle. 1-11: Minimizing time to reach setpoint (rise time) is important to minimize the startup and transition time of continuous processes and the cycle time of many batch processes. There is typically a tradeoff between minimizing rise time and PV overshoot. For bioreactors, the cells are so sensitive and the batch cycle times are so slow that minimizing the overshoot for temperature and pH shifts is much more important, particularly for mammalian cell cultures. 1-12: Pulp, paper, food and polymer inline, extrusion and sheet processes may be dead time dominant if there is no heat transfer involved and valve or sensor lags are less than transportation delays. 1-13: While the process control literature typically uses the terms “process gain”, “process dead time” and “process time constant”, in most industrial processes the contribution of the automation system dynamics needs to considered particularly for processes whose dead time or time constant is less than 10 minutes or when control valves or sensors have a poor dynamic response or when at-line analyzers are used. The effect of control valve or Variable Frequency Drive (VFD) gain, transmitter calibration and PID scale must always be considered in that the open loop gain is a product of the valve or VFD gain, process gain and measurement gain where engineering units cancel out giving an open loop gain self-regulating process that is dimensionless and an open loop gain integrating process gain of 1/sec. 1-14: In a first order plus dead time approximation for self-regulating processes, all of the time constants less than the largest time constant need to be taken as an equivalent dead time by multiplying the small time constant by a factor ranging from 0.28 to 0.88 as the ratio of the small to large time constant decreases. Since there is a tendency to underestimate dead time, I usually just sum up all the very small time constants as additional dead time. The sources of the largest contribution to the total dead time should be investigated for methods to reduce them. The good news is that as automation engineers, we may be able to significantly reduce the total loop dead time by a design and installation that takes into account the detrimental effect of dead time. If there was no dead time I would be out of a job, the PID gain could be set as high as wanted and perfect control would be possible for noise free measurements. I will provide additional synergistic Key Points in future Control Talk Blogs. Until then check the photocopy machine for your copies of the latest memos on a paperless office.

The post The Good, Bad and Ugly of Thermal Mass Flowmeters first appeared on the ISA Interchange blog site. The following technical discussion is part of an occasional series showcasing the ISA Mentor Program , authored by Greg McMillan , industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient and retired Senior Fellow from Solutia Inc (now Eastman Chemicals). Greg will be posting occasional questions and responses from the ISA Mentor Program , with contributions from program participants. Thermal mass flowmeters can be a relatively inexpensive flowmeter that can handle extremely small flows (e.g., inline 1/16 inch meters) and large flows (e.g., multipoint insertion type in 60 inch ducts). Thermal flowmeters introduce heat into the flow stream and measure how much heat is absorbed using one or more temperature sensors. However, the meter requires a fixed specific heat capacity, no heat loss or gain from ambient conditions, a fixed composition, predictable heat distribution, no change in phase and no change in heat transfer coefficient unrelated to velocity (e.g., surfaces must be clean and dry). Thermal mass flowmeters are most frequently used for gas flow since heat absorption in liquids and solids can be problematic. The greatest success is seen with properly installed inline meters measuring single component gas flows in a very controlled environment. The flow measurement uses two temperature sensors, the upstream one being heated by an electrical current. Flow is inferred either from the temperature rise for a constant current or from the amount of current needed to maintain a constant temperature differential. Total shipments of thermal mass flowmeters have been increasing by about $5 million per year to become about 2 percent of the total worldwide market for all types of flowmeters that was about $5 billion in 2009. Questions from ISA Mentor Program Participant Adrian Taylor We have some installed on nitrogen lines to reactor inserts, these give poor agreement and we have a current project to change these out for Coriolis meters (although I suspect we are suffering the effects insufficient upstream piping diameters on these particular meters). We also have some thermal mass flow switches on suction lines from some sample cabinets up to our hygiene vacuum system. The duty is air most of the time from a vent to the sample cabinet but following a sample there are also HCl vapors. The switch is used in an interlock to prevent the cabinet door being opened with insufficient vacuum for fumes from the sample. At the time of installation the switches worked perfectly but we’ve found over time there is a shift in the flow/voltage curve which has to be adjusted for. The elements are Hastelloy, which I think should be fine for the occasional ambient temperature HCl vapors. All I can think is the shift is due to either some kind of coating forming, or due to droplets forming on the element from the occasional vapors. Finally as part of a turnkey burner replacement project we have just installed an array of thermal mass elements in a square combustion air duct on our site. This is intended to work in a similar way to a pitot array but for average mass flow rather than volume. The system has only recently been commissioned so I can’t comment on performance yet. The ISA Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career. Click this link to learn more about how you can join the ISA Mentor Program. Answers from ISA Mentor Program Participant Hunter Vegas Rule #1 (and it is a huge one) – If you are trying to use a thermal flow meter on air/gas/vapors never install it in a service where it can ever see a gas/vapor approaching dew point.  Even a tiny bit of liquid will create erratic flow readings. Rule #2 – Most thermal flow meters do not promise high accuracy particularly if they are the insertion type. If you need highly accurate flow readings you should probably investigate a different meter type. However they do work well as a flow switch or a go/no go type of flow application like purges and the like. I’m a little surprised you had trouble on nitrogen. Normally dry nitrogen is a good application, but you do need to have a decent meter run to get accurate results. Can’t really speak to the sample cabinet meters, but if you are getting vapors that condense the reading will be off.  However it should recover if you are just flowing air unless you are getting some kind of a coating as you say. In that case it would definitely impact the meter. Realize too that different gasses will require different calibrations so if you have a mix of gases your accuracy will be impacted. The array meter will only work if you are downstream of the economizers and have hot air coming through that is well above the dew point.  If the meters are in the suction duct where they’ll see outside air at ambient temperature they’ll work great until it is foggy or rainy and then the readings will likely turn unstable. Greg McMillan’s Comments I have seen thermal mass flow meters extensively use in laboratories and pilot plants to measure air, oxygen and carbon dioxide flows for bioreactors. As you can imagine, the gas and ambient conditions are exceptionally controlled, which may explain their success. For more on the physical principles and practical considerations as to selection and installation of thermal mass flowmeters and all other types of measurements see my 2010 ISA book Essentials of Modern Measurements and Final Elements in the Process Industry: A Guide to Design, Configuration, Installation, and Maintenance. See the ISA book  101 Tips for a Successful Automation Career  that grew out of this Mentor Program to gain concise and practical advice. See the  InTech  magazine  January/February 2013 feature article “ Enabling new automation engineers ” for candid comments from some of the original program participants. See the May 2015  Control Talk  column “ How to effectively get engineering knowledge ” with the Mentor Program protégée Keneisha Williams on the challenges faced by young engineers today. Discussion and answers are provided by Greg McMillan, Hunter Vegas (co-founder of the ISA Mentor Program and project engineering manager at Wunderlich-Malec), Brian Hrankowsky (consultant engineer at a major pharmaceutical company), Michel Ruel (executive director, engineering practice at BBA Inc.), Leah Ruder (process systems automation group manager at the Midwest Engineering Center of Emerson Process Management) and Nick Sands (ISA Fellow and Manufacturing Technology Fellow at DuPont).

The post, Secrets to Good Vessel Temperature and pH Control , first appeared on ControlGlobal.com's Control Talk blog. The key aspects of good vessel temperature and pH control are often not recognized. Here we provide some simple guidance and overview of the benefits of a particular strategy and implementation guidelines that should be commonly used. A fundamental understanding is provided that may also be beneficial for other primary control loops. In case you are busy let’s cut to the chase. Let’s start with pH. The cascade control of vessel pH to inline pH provides a simple well known open loop gain for the vessel pH loop (primary loop). The effect of the nonlinear pH titration curve is isolated from the vessel pH loop and dealt with by the fast inline pH loop (secondary loop). The open loop gain (also commonly known as the process gain) for the vessel loop is simply dependent upon the scales of the vessel and inline pH loops. If the pH loops have the same scales, the open loop gain is simply one. The well-known open loop gain makes the tuning of the much slower vessel pH loop much easier. The biggest benefit is that you can much more easily avoid having a vessel PID gain that is too high or too low. The less recognized problem of too low a vessel PID gain stems from the vessel loop response being near integrating and true integrating for continuous and batch operations, respectively. A near or true integrating response will cause incredibly slow oscillations (e.g., period that is 40 times dead time) if the PID gain is too low. In some ways violation of the low gain limit is worse that violation of the high gain limit because the oscillations are much larger besides being much slower, which means they are not as effectively smoothed out by downstream volumes. They are also continuative. The minimum PID gain is inversely proportional to the product of the open loop integrating process gain and PID reset time. You don’t want this to be dependent upon the slope of the titration curve that varies enormously with pH and feed concentrations or how well you tuned the reset time. Most reset times are set way too small since people tend to use too much integral action and not enough proportional action on vessels or columns (any volume with mixing either due to turbulence or agitation). Often the reset time needs to be increased by two orders of magnitude. By simplifying the open loop gain to a number that is constant and easy to calculate, the PID gain can be more readily set to be above the minimum besides below the maximum. The reset time is simply a factor of the vessel dead time. For an arrest time (lambda) of two deadtimes in lambda tuning for integrating processes, this corresponds to a reset time of 5 deadtimes. Note that smaller arrest times can be used by making the process gain linear and relatively constant by the recommended cascade control. For a well-mixed vessel, the process time constant is approximately the residence time that is simply the vessel liquid mass divided by the process input mass flow. For inline pH loops on the influent, this is the influent feed flow. For inline pH loops on a recirculation line, this input flow is basically the recirculation flow since the recirculation flow is typically much greater than the feed flow. The near-integrating process gain is the open loop gain (dependent upon primary and secondary scales) divided by the process time constant (dependent upon residence time). If the pH loops have the same scale, the integrating process gain is simply the process input mass flow divided by the liquid mass in the vessel. This calculation can be done on a volumetric basis. You just need to have consistent flow and liquid inventory units. The cascade control of vessel temperature to inline feed temperature or recirculation temperature, the computations are similarly simple. If the secondary loop is jacket inlet temperature with constant jacket recirculation flow, the open loop gain is again simply dependent upon primary and secondary loop temperature scales. The residence time is the liquid mass divided by the process input mass flow. Note that the jacket input temperature loop is an inline loop where the jacket temperature is a blend of a manipulated makeup coolant flow with a constant jacket recirculation flow where increases in the manipulated jacket coolant makeup flow result in corresponding changes in coolant return flow by jacket outlet pressure control. For jackets that require some heating, the inline temperature controller manipulates the steam to an inline injector and the return flow is a hotter return water flow. The secondary pH and temperature loops are left to deal with the nonlinearities of the process. They should be tuned for an aggressive setpoint response minimizing rise time even at the expense of an increase in overshoot. Setpoint feedforward can help where 50% of the inline loop setpoint change is translated to the corresponding change in inline PID output to achieve this setpoint change. The setpoint feedforward is simple added to the inline PID output via the PID feedforward summer. For both loops, the process gain is inversely proportional to the inline flow which is kept constant for vessel and jacket recirculation flows. For these loops, a linear installed flow characteristic is best. For inline pH and temperature loops on vessel feed flows, an equal percentage flow characteristic helps compensate for the process gain being inversely proportional to feed flow by a valve gain being proportional to flow. Note to insure the installed flow characteristic is close to the inherent flow characteristic and not distorted by changes in available pressure drop, the ratio of valve drop to system pressure drop should be greater than 0.25. Providing enough pressure drop to the control valve also prevents a severe loss in valve rangeability, an aspect of the practical reality due to deterioration from friction and slope of the valve installed flow characteristic near the closed position not discussed by nearly anyone but me. Statements in the literature and catalogs about valve rangeability are generally flat out wrong. The secondary pH and temperature loop outputs go directly to the control valve. The manipulation of a coolant or reagent flow loop setpoint is not advisable because most of the flow measurements used have insufficient rangeability. In general, the biggest problem with cascade control systems that manipulate a flow setpoint rather than a valve positioner or variable frequency drive speed, is the erratic flow measurement signals at low flows. The signals get noisy and some are simply set to drop out. You can add logic to substitute an inferential flow measurement based on valve position using its installed flow characteristic but this leaves the loop vulnerable to knowledge of the installed flow characteristic and precision of the valve. If you need to use a flow loop, the best solutions are a magnetic flowmeter or even better, a Coriolis mass flowmeter, with a rangeability of 50:1 and 200:1, respectively. Differential head meters with dual range differential pressure (d/p) transmitters and vortex meters have a best case rangeability of 15:1 that is difficult to achieve in practice. A more typical rangeability for dual d/p head and vortex meters is about half the best case (e.g., 8:1). The valve must be a true throttling valve, not an on-off valve posing as a throttling valve. A diaphragm actuator capable of providing 150% of thrust requirement should be used with a goal of 0.2% resolution, 0.3% deadband and 86% response time (T86) of less than 4 seconds at a starting position of 10% besides 50%. A smart digital positioner tuned with aggressive gain and rate action is needed. The T86 performance and actuator thrust objectives are relaxed here from 2 sec and 200%, respectively that were stated in the March 2016 article “ How to specify valves and positioners that do not compromise control ” and the associated white paper “Valve Response – Truth or Consequences”. Don’t be surprised if the valve supplier will only provide the resolution, deadband and T86 for a starting position of 50% because of the deterioration of these metrics due to seating and seal friction near the closed position. The inline pH loop must still deal with the titration curve. Signal characterization can help to translate the controlled variable in pH that is the Y axis to a % reagent demand that is the X axis of the titration curve. The linearization will not be perfect. A simple standard signal linearization block where you enter 20 or so X,Y pairs for a piecewise linear fit is best. Since you are doing the opposite of the process for linearization, the X value is the ordinate (pH) and the Y value is the abscissa (reagent demand) of the titration curve. If just part of the nonlinearity is addressed, you are way ahead in the game in terms of tuning and performance. Plus, the characterization reduces the noise seen by the controller for setpoints on the steeper portion of the titration curve. Even without such linearization, the inline pH loop can usually correct for the disturbances within a minute using mostly integral action. For more on pH and temperature control, see the ISA books “ Advanced pH Measurement and Control 3 rd Edition " and “ Advanced Temperature Measurement and Control 2 nd Edition ". To summarize: 1. Use cascade control of vessel pH and vessel temperature to inline pH and inline temperature control to provide a relatively constant and easily estimated open loop (process) gain that can readily prevent the common occurrence of violating the low PID gain limit for near integrating and true integrating processes. Also more aggressive tuning can be used by a more linear and constant process gain seen by vessel loop. 2. The inline pH and inline temperature controller outputs should directly manipulate precise and fast throttling valves with the right installed flow characteristic. 3. The inline pH and inline temperature controller should be tuned for an aggressive setpoint response. 4. Signal characterization for the inline pH loop PV using a piecewise linear fit should be used to help deal with the nonlinearity of the titration curve when the normal range of inline pH setpoints are on the steeper portion of the curve.

The post PID Tuning for Near and True Integrating Processes first appeared on the ISA Interchange blog site. The following technical discussion is part of an occasional series showcasing the ISA Mentor Program , authored by Greg McMillan , industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient and retired Senior Fellow from Solutia Inc (now Eastman Chemicals). Greg will be posting occasional questions and responses from the ISA Mentor Program , with contributions from program participants. The most important loops on vessels and columns typically have a near or true integrating response. A self-regulating process is classified as near integrating if the process time constant is larger than four times the process dead time. The composition, pH, and temperature response of continuous columns tend to have a time constant to dead time ratio of about 6:1 for changes in the liquid material balance. The composition, pH, and temperature response of continuous well mixed vessels tend to have a time constant to dead time ratio of 50:1 or larger. Level and gas pressure have a true integrating response since changes in level or pressure have a negligible effect on the discharge flow notable exceptions occurring for gravity flow for liquid level control and relatively high but non critical pressure drops for gas pressure control. The composition, pH, and temperature of batch columns and batch vessels have a true integrating response in the normal operating range. A notable exception is the composition control of a batch reaction when there is no deficiency of any reactant concentration. Here the response is near-integrating with a very large time constant to dead time ratio making differentiation between true and near integrating inconsequential. Reactors with a potentially runaway response are treated as true integrators with the intent being that control action is sufficient to prevent the loop from seeing the acceleration from a runaway response. The designation of having an integrating response is critical in terms of tuning and recognizing there is a window of allowable controller gains, where too low of a PID gain as well too high of a PID gain will cause excessive oscillations. For a PID gain that is too low, the oscillations tend to be much larger and 10 times slower (e.g., period is 40 dead times for low PID gain and four dead times for high PID gain). For a PID gain greater than the ultimate gain, the oscillations can grow and the loop becomes unstable. For a PID gain that is too low, the oscillations will always decay but the decay rate becomes incredibly slow as the PID gain is decreased. For a runaway reaction, too low of a PID gain is disastrous in that the process can runaway reaching a point of no return. The Lambda tuning rules switch from a Lambda being the closed loop time constant for a setpoint change for self-regulating processes to Lambda being an arrest time for a load disturbance with the objective of stopping the ramping effect of integrating processes and potential acceleration of runaway processes. Questions from ISA Mentor Program Participant Hector Torres How do you calculate Lambda for near-integrating processes? I understand we should follow the Integrating Process rules but I am not clear as of how to determine the desired arrest time. You mention that for maximum unmeasured disturbance rejection a Lambda equal to the dead time is used. Also it is stated that a Lambda equal three dead times minimizes consequences of nonlinearities, inverse response and resonance. Why should we consider Lambda of one or three times dead time in these rules? To be identified as a near integrating process the time constant should be four times greater than the dead time. Why should we make Lambda a factor of dead time here? I remember it was mentioned that Lambda should be set at three or four times the largest of the dead time or the time constant. Would this apply here? Am I mixing in my mind the rules for self-regulating and integrating processes? I understand that integrating processes can have an inverse response that is problematic. What could be an example of inverse response? What do you mean by this? The ISA Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career. Click this link to learn more about how you can join the ISA Mentor Program. Greg McMillan’s Answers 1) In integrating tuning rules, Lambda is the arrest time, which means for a step disturbance or step change in PID output, how long does it take for the PV to halt its excursion and start its return to setpoint. If you multiply the integrating process gain (%PV/sec/%CO) by a the change in controller output (CO%) required to get a ramp rate (%PV/sec) and then Lambda, you have the peak error (maximum excursion in %PV). For level and pressure it is easier to visualize in that the maximum PV excursion (peak error) for a maximum expected change in controller output added to the setpoint must not hit an alarm or trip point. The integrated error (% sec area between the PV and SP on trend chart) is the peak error (%) multiplied by Lambda (sec). Thus, the peak error is proportional to Lambda and integrated error is proportional to Lambda squared where Lambda is the arrest time set relative to the dead time. 2) The ability of a loop to handle changes in gain and dynamics is expressed by the gain margin and phase margin, which are both a function of Lambda relative to dead time. The Lambda tuning rules reduce to tuning rules commonly used for the last six decades if you realize Lambda should always be thought of and set relative to dead time and not a time constant or an integrating process gain (as mistakenly shown in various publications). Also for large time constants you have a near-integrating process and must switch to integrating tuning rules. In the 5th edition of the Process/Industrial Instruments and Controls Handbook (1999 edition for which I became chief editor), Bialkowski on pages 10.52 and 10.53 shows how the gain margin and phase margin are a function of Lambda varying from one to five dead times. Elsewhere he talks about the concept of near-integrating processes. I think the rule sometimes states of choosing the largest of three times the dead time or three times the time constant are not in tune with advancement in understanding of Lambda always being thought of as a value relative to dead time. While Bialkowski did not say when to switch integrating tuning rules, I estimated that the switch point of when the time constant to dead time ratio was greater than four would result in tuning rules similar to what has been practices for the last six decades. This rule plus realizing that Shinskey essentially was using a Lambda of 0.6 x dead time to give the impressive maximum disturbance rejection results he has in his articles and books. His response is oscillatory. The acknowledged practical limit for a smooth response even if you exactly know the process dynamics and they never change is a Lambda equal to the dead time. Thinking of Lambda as being three dead times is a good rule for both self-regulating and integrating tuning rules. For self-regulating processes with a time constant to dead time ratio between one and two , there might be some advantage of using three times constants instead of three dead times for PI control but I think the advantage is minimal and is negligible compared to other issues. If you have a nonlinear valve or process, you may need to increase Lambda to be five or six dead times unless you do signal characterization, gain scheduling or adaptive control. To summarize, for all Lambda tuning rules (self-regulating and integrating), the most aggressive tuning if the process dynamics are fixed and exactly identified (rare case), is a Lambda of 0.6 dead times. This is case is normally only used to show how well Lambda tuning can do compared to other tuning methods (showcase test for gamesmanship). Normally, to deal with unknowns and nonlinearities a Lambda of three dead times is used but may be increased for more uncertain applications and greater changes in dynamics whether due to the process or valve. For bioreactors where the disturbances are extremely slow, rise time is inconsequential for setpoint changes, and process gains can change dramatically from the pre-exponential to the exponential growth phases, Lambda may be as large as 10 dead times. To better understand different process responses and tuning objectives, watch the three-part ISA Mentor Program webinar on PID options and solutions: PID Options and Solutions Part 1 PID Options and Solutions Part 2 PID Options and Solutions Part 3 3) Inverse response is where the initial response of the PV is in the opposite direction of the final response.  If a feedforward correction arrives too soon it can cause inverse response. For feedback control, inverse response originates typically occurs when the feed stream throttled that has a temperature less than the operating temperature of the equipment. The classic case of inverse response is boiler drum level. An increase in feed water flow being colder than the boiling water in down comers will cause bubbles to collapse which will cause fluid to go down from the drum into the down comers causing shrink (decrease in drum level). Eventually the increase in feed water is heated enough in the down comers to increase the drum inventory (increase drum level). For a decrease in feed water flow, the bubbles in the down comers increase in number and size pushing fluid up into the drum causing swell (increase in drum level). This shrink and swell is quite common and can be reduced by feed water preheaters. For furnace and reactor temperature an increase in air flow or reactant flow that is colder than furnace or reactor temperature will cause the equipment temperature to decrease until the firing rate and reaction rate generates enough heat to increase the equipment temperature. This is the main reason plus the time constant of the concentration response why reactor temperature should not be controlled by manipulating reactant flow. This is true for liquids and polymers. For gases in fluidized bed reactors, the reaction rate and concentration time constant are so fast, inverse response is imperceptible. See the ISA book 101 Tips for a Successful Automation Career that grew out of this Mentor Program to gain concise and practical advice. See the InTech magazine January/February 2013 feature article “ Enabling new automation engineers ” for the candid comments of some of the original program participants. See the May 2015 Control Talk column “ How to effectively get engineering knowledge ” with the 2014 addition to Mentor Program Keneisha Williams on the challenges faced by young engineers today. Providing discussion and answers besides Greg McMillan and co-founder of the program Hunter Vegas (project engineering manager at Wunderlich-Malec) are resources Brain Hrankowsky (consultant engineer at a major pharmaceutical company), Michel Ruel (executive director, engineering practice at BBA Inc.), Leah Ruder (process systems automation group manager at the Midwest Engineering Center of Emerson Process Management) and Nick Sands (ISA Fellow and Manufacturing Technology Fellow at DuPont).

The post, Control Strategies to Improve Process Capacity and Efficiency - Part 3 first appeared on ControlGlobal.com's Control Talk blog . Here we provide the key points for the final Part 3 of the recordings of my recent AIChE course sponsored by MYNAH Technologies. The focus here is on improving the performance of reactors, fermenters, crystallizers, evaporators, distillation columns, dryers and neutralizers. The improvements typically can be done by the selection of PID options, pairing of variables, and the addition of PID controllers and simple standard function blocks. The configuration improvements can often be done and tested in a few days. My recent AIChE St. Louis Section short course sponsored by MYNAH Technologies “Control Strategies to Improve Process Capacity and Efficiency” is viewable in three parts, each video about one hour in length on www.MYNAH.com by selecting “Videos” at the bottom of the webpage or clicking on the following direct links to each video: Part 1: Process Dynamics Review, General PID Solutions, Feedforward and Ratio Control, Steam Systems Part 2: Compressors, Jackets & Heat Exchangers, Valve Position Control, Continuous and Batch Control Part 3: Reactors, Fermenters, Crystallizers, Evaporators, Distillation Columns, Dryers, and Neutralizers Part 3 Key Points 3-1: Residence time control of a continuous reactor simply consists of the level setpoint being adjusted as the production rate and total feed rate change to keep constant the level mass divided by total flow that is the residence time as seen in slide 30. Keeping this residence time constant and large enough to give the desired conversion helps provide a more consistent product concentration. For batch reactors, this time is set by batch cycle time. Often, batches are held longer than necessary not taking advantage of research reports, prediction of end points and inferential measurements of conversion (slide 22). 3-2: The reactant ratio can be trimmed by the online measurement of the key excess reactant concentration in a recirculation stream that is needed to make sure there is just enough excess reactant and the right relative concentrations per the reaction stoichiometry. 3-3: The use of equal setpoint filters on the reactant feed setpoints (simultaneously adjusted for production rate or reactant concentration control) can provide synchronized changes in reactant feed rates preventing temporary unbalances. The use of a setpoint filter rather than via a lambda of the slowest loop for all loops enables each loop to be tuned with individual lambdas that provide the best disturbance rejection. 3-4: The use of a Valve Position Controller (VPC) with external reset feedback and fast setpoint up and slow setpoint down rate limits on the feed setpoint to be maximized (pushing a coolant valve to its maximum throttle position) can provide a smooth optimization, quick getaway for abnormal conditions and minimization of interaction. A feedforward signal representative of upsets can also be added to the VPC and temperature controller output to help it deal with abnormal conditions. 3-5: The use of external reset feedback with fast readback of the reactor coolant valve position by the secondary temperature controller along with aggressive tuning can minimize the oscillations from stiction and backlash in the coolant valve that would be propagated as feed changes by the VPC. See the March 2016 Control March 2016 Control magazine article “ How to specify valves and positioners that don’t compromise control ” and the associated white paper “Valve Response – Truth or Consequences”. 3-6: If excess reactants in the product stream are recovered and recycled as reactor feeds, a snowballing effect will occur if there is no fixed flow control in the path of the recovered to recycled reactant. For this reason, the reactor level controller does not manipulate the product discharge flow but manipulates the makeup leader reactant feed flow as seen on slide 31. The product stream discharge flow controller prevents snowballing. This flow sets production rate and can be gradually maximized. The follower makeup reactant feed flow is set via the ratio station. The makeup reactant flow setpoints are corrected by the amount of the respective recycled reactant flow to give a net makeup requirement. If the makeup and recycled reactant concentrations are different, the makeup flows need to be corrected accordingly. Hi accuracy density measurements by Coriolis mass flow meters can provide a fast inline measurement of reactant concentration for each feed. 3-7: For gas and liquid reactants with a liquid product (no gas product), a pressure controller manipulating the gas reactant feed flow will inherently add just enough gas reactant to maintain the reaction as shown in slide 32. The gas purge flow must be optimized to minimize the amount of gas reactant purged and minimize the buildup of inerts in the gas phase that would reduce the driving force for reactant mass transfer for reaction in the liquid phase. The liquid product stream is manipulated by the reactor level controller to provide residence time control. The liquid reactant feed flow sets production rate and can be maximized by a VPC. 3-8: For gas and liquid reactants with a gas product (no liquid product), a level controller manipulating the liquid reactant feed flow will inherently add just enough liquid reactant to maintain the reaction as shown in slide 33. The liquid purge flow must be optimized to minimize the amount of liquid reactant purged and minimize the buildup of inerts in the liquid phase that would reduce the driving force for reactant mass transfer for reaction in the gas phase. The gas product stream is manipulated by the reactor pressure controller. The gas reactant feed flow sets production rate and can be maximized by a VPC. 3-9: For liquid reactants and a primary liquid product and secondary gas co-product you have a condenser recycling as reflux the liquid product that appears in the gas phase as shown in slide 34. Here the control scheme is similar to what was shown in slide 32 but with more opportunities for VPC to push condenser and vent pressure valves besides the jacket coolant valve to their maximum throttle position to maximize production rate. The lowest of the VPC outputs is selected to prevent the furthest open valve from going too far open. This strategy also works for batch operations where there is the additional challenge of tight condenser temperature control for big differences in vapor load and loop dynamics as a result of different states in the batch operation. Adaptive control is often needed to provide aggressive control to catch up with an ever shifting load. Due to the heat transfer lag, the response is most likely near integrating. If precise valves and adaptive control can compensate for changes in dynamics including those occurring due to valve nonlinearity, the lambda arrest time should be reduced to be about two deadtimes to help catch up with the shifting load. 3-10: For a fluidized bed reactor with consequently gas reactants and a gas product, a temperature controller using the highest average bed temperature manipulating the leader gas reactant feed will inherently maximize production rate to limit of the cooling capability of the boiler feed water (BFW) coils being used to generate steam as seen on slide 35. Production rate is set by taking BFW coils in and out of service. A feedforward based on changes of coils in service can help with production rate changes. The averaging of temperature sensors in a cross section helps reduce noise from channeling and stratification. The use of the maximum average helps deal with hot spots that can cause byproduct formation. Since the residence time is so small and the reaction so fast, there is no inverse response from manipulating reactant feeds like there would be with liquid reactors making this a very viable strategy for temperature control. The thermowell design must minimize the thermal lags and the valve responses must be fast and precise for tightest control and best maximization of production rate. 3-11: Valve position control can optimize compressor or pump pressure, boiler pressure, chiller temperature, waste reagent or fuel use, reactor and column production rate and ratio control (feedforward accuracy) as summarized in the table on slide 36. 3-12: The key VPC features for best optimization, best disturbance rejection and least interaction is directional move suppression (up and down setpoint rate limits), external reset feedback (the essential feature for all of these opportunities), adaptive tuning, feedforward, and an enhanced PID as summarized in the table on slide 37. 3-13: The use of an enhanced PID with an online corn fermentable starch analyzer can provide an easily setup and tuned production rate controller for the front end of an ethanol plant that inherently maximizes ethanol yield per slide 38. The production rate controller provides an easy way of matching front end production to the capability of the back end. The PID gain can be simply set to be as large as the inverse of the open loop gain and the reset time as small as a few seconds. 3-14: Dissolved Oxygen (DO), pH and temperature of a bioreactor can be tightly controlled by the use of split ranged control with directional move suppression (made possible by external reset feedback) to prevent unnecessary crossings of the split range point in slide 39. 3-15: For mammalian bioreactors, there is an optimum ratio of the feed rates of sugar (glucose) and amino acid (glutamine). This ratio can be adjusted as shown in slide 40 by online measurements of these substrate concentrations provided by an at-line Nova Bioprofile Flex Analyzer that offers precise fast measurements of Glucose, Lactate, Glutamine, Glutamate, Ammonium, Sodium, Potassium, Osmolality, Cell Diameter, Cell Density, and Cell Viability. Since the analysis takes just a few minutes, an analyzer can be multiplexed to serve many bioreactors. The price is reasonable enough to even enable redundancy. 3-16: Bioreactor profile optimization can be added by computing the rate of change via the simple setup of the deadtime blocks as noted in Part 2 with a deadtime chosen to optimize the signal to noise ratio. The deadtime block inputs are viable cell concentration and product concentration to provide inferential measurements of cell growth rate and product formation rate, respectively. These rates that are the slopes of the batch profile of viable cell and product concentration that can be optimized as a function of batch time and used to predict batch end points. 3-17: A virtual plant can be setup and adapted rather easily as shown in slide 42. The virtual plant has the same setpoints and controllers as the actual plant. Given the controllers achieve good control keeping the process variables at their respective setpoints, differences between the virtual and actual plants show up as differences in the manipulated variables (e.g., manipulated process input flows). A model predictive control can be setup and models identified by just running the virtual plant. The adaptation can be used without changing anything in the actual plant (completely nonintrusive). When the adapted virtual plant has demonstrated its fidelity by matching manipulated flows, inferential measurements from the virtual plant can be used to provide optimization of the actual plant. 3-18: The virtual plant synergy enables the more effective use of modeling tools and advanced control tools as depicted in slide 43 leading to greater process knowledge, the key to PCI and increased process capacity and efficiency. 3-19: The virtual plant continuity offers the opportunity of a working installation for exploration, discovery, prototyping, deployment and training enabling continuous improvement as shown on slide 44. Hundreds of pages even if they could be written cannot describe the knowledge embedded and gained in a virtual plant. The August and September Control Talk Columns will give the views of DuPont and ProSys on this incredible often overlooked opportunity that becomes obvious when models become a working prototype. 3-20: There are many automation opportunities for batch operations as listed on slides 45 and 46 that may not have been realized. Many just involve configuration changes. All can be exploited with the greatest confidence and the least disruption by means of a virtual plant. 3-21: Crystallizer operation can be greatly improved by the use of a crystal size and concentration meter in a recirculation line as shown on slide 47 to optimize the setpoint of the crystallizer temperature controller. The secondary coil inlet temperature controller setpoint manipulated by this controller has a low setpoint limit to prevent cold spots. For batch operations and the startup of continuous operations, there is an optimum cooling curve that is the mirror image of the normal cooling rate curve. The coil inlet temperature must be higher in the beginning and then optimally decreased. The cooling rate is initially low to prevent the formation of small crystals (fines) that coat heat transfer surfaces and create product quality and recovery problems. 3-22: Double effect evaporators use an online measurement of product solids concentration often afforded by an accurate density measurement of a Coriolis Meter to correct the product discharge flow as shown in slide 48. The level controller on the second effect manipulates its feed flow accordingly that is then followed by a level controller on the first effect that manipulates its feed flow. A ratio controller maintains the necessary steam flow to feed ratio. A correction of the ratio can be greatly beneficial based on feed solids concentration measurement afforded by a Coriolis meter on the feed. 3-23: The best distillation column temperature point for control is the one that shows the largest and most symmetrical change in temperature for a change in the manipulated ratio (e.g., distillate to feed ratio). On slide 49 this corresponds to stage 5. This control point can provide tighter concentration control beyond considerations of linearity. The sensitivity to concentration changes of a distillate or bottoms product can be checked as shown on slide 50. The benefit in terms of disturbance rejection (e.g., feed increase or decrease) from the best control point is greater than might be expected as seen on slide 51. All of these tests can be done by high fidelity simulations. 3-24: For temperature control of either a single distillate or bottoms product, the ratio manipulated by the temperature controller is chosen for tightest concentration control provided the level controllers can do their job as shown on slides 52-55. For top temperature control, the preferred setup is the manipulation of the distillate to feed ratio. Tight distillate receiver level control is then used to translate manipulated distillate changes into reflux flow changes. This setup also provides some internal reflux control by inherently correcting reflux for changes in distillate temperature due to changes in operating or ambient conditions. For bottom temperature control, the preferred setup is the manipulation of steam to feed ratio to avoid the inverse response from manipulation of steam for sump level control. In all cases, feed rate is used to provide feedforward control by means of ratio stations. 3-25: For dryer control, the outlet temperature can be optimized by an inferential measurement of outlet moisture computed based on the temperature rise in the falling drying rate zone. After startup or batch operation when the product’s interior moisture has been depleted, dry spots start to appear on the particle surface. The driving force for evaporation based on decreases in the area of surface moisture starts to drop causing a rise in outlet temperature indicative of the falling drying rate zone. Since the inferential measurement uses outlet temperature, it must be filtered with a time constant greater than the reset time of the temperature controller to prevent positive feedback. The inferential moisture calculation can be corrected by product moisture measurement. Most dryers are significantly over drying the product. Optimizing the product moisture increases efficiency by reducing energy use and increases capacity by selling moisture as product. 3-26: For pH control, it all comes down to the slope of the titration curve, the degree of mixing, the precision of reagent addition and deadtime. Slide 57 shows how there are no straight lines on a titration curve, how pH can look like a runaway process and what is needed to get a good titration curve. Slide 58 shows how moving the setpoint to a flatter portion of the titration curve can not only save reagent but also reduce the pH oscillations seen from imperfect mixing and valve response. Slide 59 shows how two electrodes create more questions than answers and notes how critical middle signal selection of three electrodes is in terms of minimizing maintenance besides enabling tighter more reliable pH control. Slide 60 shows many of the mistakes made in pH system design and slide 61 shows how an inline system can be used on a vessel to provide fast pH control with oscillations smoothed out by the vessel volume. The tuning of this inline loop is easier due to the small deadtime. A signal characterizer for setpoints on the steep part of the titration curve can help enormously in reducing oscillations on the controller faceplate making the operators and process engineers much happier. Not shown is that the addition of a pH loop on the vessel can provide correction of the inline setpoint. By having this vessel pH loop manipulate the secondary inline pH loop rather than the reagent directly, the process gain seen by the vessel pH loop is simply just one isolating the nonlinearity of the titration curve from the vessel pH loop. Also, the danger of a controller gain being too low as well as too high is greatly reduced. This realization of the window of allowable gains for near integrating and integrating processes complicating pH control has not been discussed despite being a very real and pervasive problem when the vessel pH controller directly manipulates reagent flow and consequently is subject to the nonlinearity of the titration curve. I will provide synergistic Key Points of my ISA Mentor Program WebEx Recordings in future Control Talk Blogs. Until then enjoy the dynamic world of process control using a virtual plant as your working prototype for education, exploration, discovery, prototyping, testing, deploying and training.

The post, Control Strategies to Improve Process Capacity and Efficiency - Part 2 first appeared on ControlGlobal.com's Control Talk blog. Here we provide the key points for Part 2 of my recent AIChE St. Louis Section course sponsored by MYNAH Technologies. The focus here is on improving the performance of compressors, jackets, heat exchangers and batch operations and the key role of valve position control (VPC). The improvements typically can be done by the selection of PID options, pairing of variables, and the addition of PID controllers and simple standard function blocks. The configuration improvements can often be done and tested in a few days. My recent AIChE St. Louis Section short course sponsored by MYNAH Technologies “Control Strategies to Improve Process Capacity and Efficiency” is viewable in three parts, each video about one hour in length on www.MYNAH.com by selecting “Videos” at the bottom of the webpage or clicking on the following direct links to each video: Part 1: Process Dynamics Review, General PID Solutions, Feedforward and Ratio Control, Steam Systems Part 2: Compressors, Jackets & Heat Exchangers, Valve Position Control, Continuous and Batch Control Part 3: Reactors, Fermenters, Crystallizers, Evaporators, Distillation Columns, Dryers, and Neutralizers I am exceptionally proud of this course. It is an updated concise view of what I have learned to be truly important for achieving significant benefits from process control. Last month we covered Part 1. Here are my Part 2 Key Points that provide a synergistic understanding with the viewing of the videos: Part 2 Key Points 2-1: The best compressor setpoint for compressor suction flow to prevent surge and maximize efficiency is a setpoint curve on the plot (compressor map) of compressor Pressure Rise versus Suction Flow that is parallel to the surge curve. 2-2: The offset must provide enough time for the surge control system to react and prevent crossing of the surge curve and ideally be near the center of the ellipses for compressor efficiency. 2-3: The surge curve coincides with the points of zero slope on the characteristic curves for different speeds or inlet vane positions signifying the loss of negative feedback in the process given by a negative slope of characteristic curves (processes without negative feedback are inherently unstable). 2-4: To the left of the surge point, the slope of the characteristic curve becomes positive creating positive feedback causing the operating point to jump to a point on the characteristic curve that has negative slope and negative feedback. Unfortunately, this is typically a negative suction flow. The operating point then walks along the negative slope characteristic until it reaches the positive slope characteristic at which point it jumps to a positive flow at a previous operating point in its approach to the surge curve. The operating point then moves toward the surge point repeating the cycle. 2-5: The precipitous drop in flow can occur in 0.03 seconds and the surge cycle period tends to be 1 to 2 seconds (faster for the experimental setup in presentation). This incredibly fast jump to negative flows and back to positive flows is extremely disruptive to downstream users and potentially damaging to the compressor. Surge causes high radial vibration and high dynamic axial thrust forces eventually damaging the journal bearing and thrust bearing. There is can be a consequential loss in compressor efficiency at all operating conditions from successive surge cycles. For axial compressor the damage can be particularly severe. A large high speed high pressure compressor I had to protect could destroy its million dollar rotor in a matter of seconds. 2-6: The suction flow controller whose setpoint is a curve parallel to the surge curve must come off of its output limit in time to prevent reaching the surge curve. This typically must occur before the operating point crosses the surge setpoint (suction flow drops below its setpoint) unless the setpoint is far to the right of surge curve. Since reset action tends to try to keep the surge valves closed until crossing setpoint, intelligent tuning must provide more proportional than integral action and logic to specify when the surge valve starts to open is often needed. 2-7: A feedforward of user feed flow can help preemptively position the surge valves (not shown on slide). Since the surge valves should have linear installed characteristics, this feedforward can be simply added to the suction flow controller if the feedforward signal is scaled in terms of surge valve capacity similar to what was stated for the steam header letdown valves. A lead-lag may be necessary as dynamic compensation to make sure the feedforward correction to the compressor suction flow does not arrive too soon or too late. 2-8: External reset feedback and analog output block setpoint rate limits can provide directional move suppression to give fast opening and slow closing surge valves without retuning the surge controller. 2-9: An open loop backup is needed to help prevent the start of surge and the recovery from surge. The surge phenomena and cycles are too fast and confusing to rely on feedback control. The open loop backup computes the rate of change of compressor suction flow to predict a crossing of the surge curve at least one deadtime into the future. A variable deadtime block is used to provide the desired signal to noise ratio in the prediction. See the 6/28/2012 Control Talk Blog “ Future PV Values are the Future ” 2-10: The controller execution time must be 0.1 seconds or faster, the transmitter damping time constant must not exceed 0.2 seconds and valve 86% response time must be 2 seconds or faster. 2-11: Fast valve response time requires backlash and resolution limits of 0.2% or better, a sensitive positioner and actuator with volume boosters on the positioner output(s) to the actuator with booster bypass valve(s) opened just enough to prevent oscillations. For more on how to achieve good valve response see the March 2016 Control magazine article “ How to specify valves and positioners that don’t compromise control ” and the associated white paper “Valve Response – Truth or Consequences”. 2-12: A valve position control (VPC) can smoothly optimize the compressor setpoint by pushing the largest opening of user feed valves to their maximum throttle position. External reset feedback and setpoint rate limits can provide directional move suppression to provide a gradual optimization with a quick getaway for abnormal conditions. 2-13: See the May 2016 Control Talk Column “ Finer points of compressor control ” for more details on the many challenges of compressor control. 2-14: A deadtime block can account for the transportation delay to synchronize jacket or coil inlet temperature with the outlet temperatures to compute the rate of cooling or heating to provide an inferential measurement of reaction and crystallization for batch and continuous operations. 2-15: Constant jacket or coil recirculation flow eliminates increases in process gain and dead time at low cooling or heating rates providing better tuning and preventing a burst of oscillations at low production rates. 2-16: Jacket or coil inlet temperature control offers faster correction of jacket utility temperature disturbances by seeing these disturbances before they enter jacket or coil. Limits can be placed on inlet temperature setpoint to prevent cold and hot spots. 2-17: Jacket or coil outlet temperature control reduces process noise by allowing time for more mixing and changes in phase (bubbles to condense and droplets to vaporize). Outlet temperature control can help see the start of a runaway condition sooner. 2-18: Manipulation of heat exchanger bypass flow for jacket or coil temperature control provides a faster secondary loop in the cascade control of vessel temperature to jacket or coil temperature. A VPC can increase the heat transfer coefficient and extend the rangeability of jacket or coil temperature control by minimizing the exchanger bypass valve position. 2-19: Steam injection heaters instead of split ranged operation of steam and coolant valves can eliminate the split range discontinuities and changes in phase providing much faster and smoother jacket or coil temperature control. 2-20: For heat exchangers in recirculation lines, the manipulation of the heat exchanger bypass flow for recirculation temperature control provides much faster initial response and separation of the self-regulating response at the exchanger and the integrating response from the vessel. A VPC can increase the heat transfer coefficient and extend the rangeability of recirculation temperature control by minimizing the exchanger bypass valve position. 2-21: The feed rate of continuous and fed-batch reactors can be maximized by a VPC. 2-22: External reset feedback with fast down and slow up setpoint limits on reactor feed can provide a directional move suppression enabling a gradual smooth optimization and a fast correction for abnormal conditions. 2-23: An enhanced PID for VPC developed for wireless and analyzers can ignore limit cycles in valve. 2-24: Coordinated ratio control of continuous and fed batch reactants by the use of setpoints with the same setpoint filter minimizes upsets to stoichiometry from changes in production rate created by VPC. 2-25: The distinguishing characteristic of a batch operation is a closed liquid discharge valve during processing causing an integrating response in composition, pH and temperature. 2-26: Fed-batch reactors have simultaneously PID controlled reactant flows and are for this reason sometimes called semi-continuous. However, the closure of the liquid discharge valve causes the same integrating response as traditional batch where reactants are sequenced. 2-27: Level and residence time control are not applicable to batch reactors. 2-28: Continuous reactor composition control techniques are applicable to fed-batch if the controlled variable is converted to a batch profile slope, which creates a steady state not seen before translation. 2-29: Traditional batch and fed-batch reactors can be held until a reactant is depleted whereas some of the reactant is in the discharge flow of continuous reactors in order for the reaction to continue. 2-30: The use of a deadtime block to create a inferential measurement of conversation rate from batch slope is useful for not only PID and Model Predictive Control of conversion rate but also for prediction of batch end points enabling detection of when a batch is complete and making economic decisions for the tradeoff between increasing yield and reducing batch cycle time. I will provide Key Points of Part 3 concluding this series next month. Until then live large by process control improvement to advance you, your company and our profession.

The post, Control Strategies to Improve Process Capacity and Efficiency - Part 1 , first appeared on ControlGlobal.com's Control Talk blog. Many simple additions to PID control loops are presented here that can increase the plant production rate, yield, product quality, and on-stream time. The improvements typically can be done by the selection of PID options, pairing of variables, and the addition of PID controllers and simple standard function blocks. The configuration improvements can often be done and tested in a few days. My recent AIChE St. Louis Section short course sponsored by MYNAH Technologies “Control Strategies to Improve Process Capacity and Efficiency” is viewable in three parts, each video about one hour in length on www.MYNAH.com by selecting “Videos” at the bottom of the webpage or clicking on the following direct links to each video: Part 1: Process Dynamics Review, General PID Solutions, Feedforward and Ratio Control, Steam Systems Part 2: Compressors, Jackets & Heat Exchangers, Valve Position Control, Continuous and Batch Control Part 3: Reactors, Fermenters, Crystallizers, Evaporators, Distillation Columns, Dryers, and Neutralizers I am exceptionally proud of this course. It is an updated concise view of what I have learned to be truly important for achieving significant benefits from process control. Here are my Part 1 Key Points that provide a synergistic understanding with the viewing of the video: 1-1: A process can be made to move smoothly in concert to different production rates or deal with disturbances by keeping a series of mass flow rates and/or energy transfer rates in the right ratio. 1-2: The ratios can be seen directly or simply computed from the process operating conditions given on Process Flow Diagram (PFD). 1-3: Most feedforward control systems are addressing this common need to keep flow rates and energy transfer rates in the right ratio. 1-4: Nearly all feedforward systems need feedback correction as shown in the course by a primary PID controller measuring process composition, level, pH, pressure, or temperature. 1-5: When the primary PID controller output directly manipulates a valve signal or power input, the implementations can be simply done via a feedforward summer embedded in the PID controller. 1-6: When the PID controller output is cascaded to a secondary PID flow or speed controller setpoint, the implementation is best done via a Ratio and a Bias block. 1-7: The operator must be able to not only manually set the flow or speed manipulated by the primary PID but also locally set a desired ratio for startup and abnormal operation. 1-8: The actual ratio per measured flow rates or energy transfer rates, the normal ratio set locally by operator, and the corrected ratio set by the primary feedback control must be displayed to operator. 1-9: The feedforward input is designated as the “leader” and the feedforward output as the “follower”. 1-10: Each feedforward input (leader flow or speed) is multiplied by a ratio factor (feedforward gain or ratio setpoint). A bias is then added or subtracted before becoming the feedforward output (follower flow or speed). 1-11: For vessels and columns, the primary PID controller corrects the ratio factor. 1-12: For inline systems and sheet lines, the primary PID controller corrects the bias. 1-13: A ratio or bias not corrected by the primary PID controller can be slowly adapted via the output of a simple integral-only controller that seeks to minimize the feedback correction needed from the primary controller. 1-14: The result of the feedforward output (follower) must arrive at the same point in the process at the same time as the disturbance (leader) as visualized on the block diagram of dynamics for the loop. 1-15: The result of the feedforward output must be equal to and opposite in sign to the disturbance. 1-16: Filtering of the feedforward signal should be just enough to prevent follower valve reaction to feedforward noise. 1-17: A result of the feedforward output that arrives too soon or is too large will cause a response in the opposite direction (inverse response) that is terribly confusing to the primary PID controller. 1-18. Feedforward gain and ratio factor settings are conservatively set to prevent overcorrection. 1-19: Dynamic compensation by an insertion of a deadtime block and lead-lag block on the feedforward signal is needed to achieve the correct timing to insure the arrival of the result is not too early or late and is not too fast or slow. 1-20: The feedforward deadtime is set equal the disturbance (load) path delay minus the feedforward path delay to make sure the result of the feedforward does not start too soon. If the feedforward path delay is larger than the disturbance (load) path delay the ratio factor must be decreased. If the feedforward path delay is larger by more than the total loop deadtime, feedforward be more detrimental than beneficial. 1-21: The feedforward lead time is set equal to the largest lag time (time constant) in the path of the result of feedforward to the same point in the process as the disturbance (load). 1-22: The feedforward lag time is set equal to the largest lag time (time constant) in the path of the disturbance (load) to same point in the process as the result of feedforward. 1-23: For a similar large lag time in both the disturbance (load) and feedforward paths, the lead or lag time can be simply increased to help the feedforward to provide a faster or slower correction, respectively. 1-24: For steam header systems, the feedforward input for each header pressure control output is the summation of the letdown flow to the lower header (acting as a half decoupler besides mitigating letdown flow disturbances) plus the flows of steam users and generators on the respective header. User steam flows have a plus sign and generator steam flows have a minus sign in the feedforward summation. 1-25: The feedforward decoupler does not need dynamic compensation if the letdown valves from upper header and to lower header are in the same relative location. 1-26: The header user and generator feedforward signals need dynamic compensation based on the deadtime and secondary time constant of the integrating pressure response where the input is the user or generator flow measurement and the output is the pressure measurement used for header control. 1-27: For user, generator and letdown steam flows in the same mass flow units the theoretical feedforward gain is 1.0 with the feedforward scale set equal to the linear letdown valve flow capacity. 1-28: A feedforward signal of steam demand when a Cogen high flow override controller output is selected to override a boiler master controller is added to the respective header’s pressure controller output to provide a faster correction of letdown flow to prevent a Cogen steam generation that exceeds permit flow. I will provide additional synergistic Key Points in future Control Talk Blogs. Until then don’t procrastinate on optimization even though it optimizes living in the past.

This post, Sizing up Valve Sizing Opportunities , first appeared on ControlGlobal.com's Control Talk blog. Specifying the best valve size can reduce loop variability and enable tighter control. A side benefit can be easier PID tuning and a smaller valve. By understanding key principles, you can become a valuable resource for getting the best automation component that directly affects the process, the control valve. I became sensitized to control valve sizing early in my career when I found most of the pH reagent valves and many of the Fed-batch reactant valves were riding the seat. This is a bumpy ride. The friction and consequently the limit cycle from stick-slip are greatest near the seat due to the rubbing of the plug against the trim seat for sliding stem (globe) valves. For many rotary valves the cycling is worse. The friction of the ball or disk rubbing against the seal and resulting friction is greater and persists for larger valve positions, causing a greater and more prevalent limit cycle. We are pretty diligent about making sure the valve can supply the maximum flow. In fact, we can become so diligent we choose a valve size much greater than needed thinking bigger is better in case we ever need more. What we often do not realize is that the process engineer has already built in a factor to make sure there is more than enough flow in the given maximum (e.g., 25% more than needed). Since valve size and valve leakage are prominent requirements on the specification sheet if the materials of construction requirements are clear, we are setup for a bad scenario of buying a larger valve with higher friction. The valve supplier is happy to sell a larger valve and the piping designer is happier that not much or any of a pipe reducer is needed for valve installation. The process is not happy. The operators are not happy looking at trend charts unless the trend chart time and process variable scales are so large the limit cycle looks like noise. Eventually everyone will be unhappy. The limit cycle amplitude is large because of greater friction near the seat and the higher valve gain. The amplitude in flow units is the percent resolution (e.g., % stick-slip) multiplied by the valve gain (e.g., delta pph per delta % signal). You get a double whammy from a larger resolution limit and a larger valve gain. If you further decide to reduce the pressure drop allocated to the valve as a fraction of total system pressure drop to less than 0.25, a linear characteristic becomes quick opening greatly increasing the valve gain near the closed position. For a fraction much less than 0.25 and an equal percentage trim you may be literally and figuratively bottoming out for the given R factor that sets the rangeability for the inherent flow characteristic (e.g., R=50). What can you do to lead the way and become the “Go To” resource for intelligent valve sizing? You need to compute the installed flow characteristic for various valve and trim sizes as discussed in the Jan 2016 Control Talk Column “ Why and how to establish installed valve flow characteristics ” You should take advantage of supplier software and your company’s mechanical engineer’s knowledge of the piping system design and details. You must choose the right inherent flow characteristic. If the pressure drop available to the control valve is relatively constant, then linear trim is best because the installed flow characteristic is then the inherent flow characteristic. The valve pressure drop can be relatively constant due to a variety of reasons most notably pressure control loops or changes in pressure in the rest of the piping system being negligible (fictional losses in system piping negligible). For more on this see the 5/06/2015 Control Talk Blog “ Best Control Valve Flow Characteristic Tips ” On the installed flow characteristic you need to make sure the valve gain in percent (% flow per % signal) from minimum to maximum flow does not change by more than a factor of 4 (e.g., 0.5 to 2.0) with the minimum gain greater than 0.25 and the maximum gain less than 4. For sliding stem valves, this valve gain requirement corresponds to minimum and maximum valve positions of 10% and 90%. For many rotary valves, this requirement corresponds to minimum and maximum disk or ball rotations of 20 degrees and 50 degrees. Furthermore, the limit cycle amplitude being the resolution in percent multiplied by the valve gain in flow units (e.g., pph per %) and by the process gain in engineering units (e.g., pH per pph) must be less than the allowable process variability (e.g., pH). The amplitude and conditions for a limit cycle from backlash is a bit more complicated but still computable. For sliding stem valves, you have more flexibility in that you may be able to change out trim sizes as the process requirements change. Plus, sliding stem valves generally have a much better resolution if you have a sensitive diaphragm actuator with plenty of thrust or torque and a smart positioner. For pH reagent, Fed-batch reactant and fermenter or bioreactor air flow, the rangeability requirement can be extraordinary. If you cannot achieve the rangeability with one valve, you may need to add a small trim valve in parallel with the big valve. This is preferable to split ranged operation because it eliminates the split range discontinuity and adds the ability to make more precise flow adjustments all the time by keeping the small trim valve continually available for direct throttling by the flow or process controller. The large valve is throttled by a valve position controller (VPC) whose controlled variable is current trim valve position and whose setpoint is optimum trim valve position (e.g., 50%). The VPC tuning has traditionally been integral-only with integral action 10 times slower than the flow or process controller integral action to reduce interaction between the loops. Sometimes the VPC action is reduced or eliminated for small deviations of the small valve position from its optimum position (e.g., no VPC action for small valve positions between 40% and 60%). There are now more effective ways of doing VPC by means of external reset feedback, directional move suppression, and an enhanced PID as described in the June 2015 Control feature article “ Don’t Overlook the Virtues of PID when Optimizing Processes ” If you take advantage of this knowledge, your supplier will be happy because you are happy. The supplier may also be able to sell you that small trim valve besides the regular valve. For isolation, the supplier can also sell you an on-off valve with low leakage. All of this and more is revealed in the Control May 2016 feature article “ How to specify valves and positioners that don’t compromise control ” If you specify the best valve and best size, you might just be famous by Friday. “Thank goodness for Friday.” For a concise presentation of the concepts and details on the effects of control valves on PID control see my ISA book Good Tuning: A Pocket Guide, 4 th Edition . If you are up for a more comprehensive view, see my Momentum Press book Tuning and Control Loop Performance - 4 th Edition .

The post, How to be a World Traveler in Process Control first appeared on ControlGlobal.com's Control Talk blog. There are many different worlds of process control. Here you can learn to understand and appreciate the differences enabling you to extend your expertise in process control. You can benefit from the experiences in other worlds, expanding your capability and marketability helping to make the profession stronger by building on the talents and knowledge of exceptional experts, such as James Beall, Mark Coughran, Lou Heavner, Sigifredo Nino, Michel Ruel, Greg Shinskey, Jacques Smuts and Terry Tolliver. The different worlds are the result of the extreme diversity of products in the process industry with different dynamics and objectives. I came out of the world of continuous chemical processes, the principal and principle world of Greg Shinskey. His books Process Control Systems, Distillation Control, pH and pION Control in Process and Waste Streams, Feedback Controllers for the Process Industries, Controlling Multivariable Processes, and Energy Conservation through Control each have orders of magnitude more practical knowledge on process dynamics and relationships and how they affect control than any book I have written or seen. My associate Terry Tolliver, the world’s foremost expert in distillation control feels the same way. Our deep knowledge of first principles all originates from Shinskey. I was so honored to be inducted into the Control Hall of Fame the same year as Shinskey. As my career progressed I expanded my world of application into batch control, fermenter and bioreactor control, plug flow reactor control, extruder control and plastic sheet control. However, the worlds of gas and oil, pulp and paper, and food and beverage processes were and still are foreign to me. Bill Bialkowski principally specializing in pulp and paper applications developed lambda tuning. In the same time frame academics developed and popularized Internal Model Control (IMC). The tuning parameter for both was the closed loop time constant (time to reach 63% of a setpoint change) for self-regulating processes (called lambda for lambda tuning and gamma for IMC tuning). Both methodologies concentrated on setpoint response and considered disturbances to be on the process output, which then had similar tuning requirements for load response as setpoint response. In one case there was a step change in the process variable (PV) and in the other case a step change in the setpoint (PV). For a structure of PID on Error, the tuning was essentially the same. Shinskey’s complaint was that the emphasis should be on load disturbances at the process input that have to pass through the process time constants and consequently on being able to deal with processes with large time constants. If you used lambda or IMC tuning for such processes, the reset time could be an order of magnitude or more larger than what would provide the minimum integrated absolute error (IAE), the common criterion in the literature for load disturbances. Furthermore, tuning tests should be made by emulating a load disturbance by momentarily putting the PID in manual and making a step change in PID output. Shinskey pointed out the setpoint response would also be good for good load response tuning by simply adding a setpoint lead-lag. Alternatively, for today’s modern controller with many choices of PID structure, you can get the same results as a lag time on the setpoint equal to the reset time by the use of a “PD on PV, I on error” structure as noted in the ISA Mentor Program 1/26/2106 post “ Equivalent Methods to Eliminate proportional Step and Derivative Kick ”. Additionally, the use of a “Two Degrees of Freedom” PID structure can give similar performance to a setpoint lead-lag. I found that if you use lambda rather than lambda factor setting it equal to about ¾ the dead time and switching to lambda integrating process tuning rules when the process time constant becomes greater than 4 times the dead time, you end up with good load response tuning. Whether this is robust enough considering nonlinearities and unknowns and whether you want to move the PID output so aggressively is another story. Additionally if you simply add the requirement that rate time is not only the secondary time constant but also ½ the dead time, you end up with tuning similar to what you would get for minimization of IAE. Also, the focus is switched from a setpoint response metric (lambda being a closed loop time constant) to a load response metric (lambda being an arrest time - time for the PV to reach the maximum of its excursion to start its return to setpoint) in lambda integrating process tuning rules. If you are concerned about abrupt changes in the PID output from affecting other loops, you can put a setpoint rate limit on the analog output or secondary loop setpoint and turn on external reset feedback to provide directional move suppression. No detuning is needed to provide a smoother output change. Furthermore, for small process time constants, which is the case in pulp inline and paper sheet line loops, whether the disturbance is on the input or output of the process doesn’t make much difference. Finally, for processes where the time constant is less than the dead time, putting a limit on lambda and thus reset time being about ½ the dead time, prevents the transition to an integral-only type of tuning. For more on how the size of the process time constant has caused such different perspectives, see my 10/24/2013 Control Talk Blog “ The Primary Source of Disagreement in Process Control Tips ”. What I am offering is that you can be a fan of both Shinskey fan and Bialkowski taking advantage of the knowledge they respectively offer on the value of knowing chemical engineering principles and different tuning objectives if you just know how to modify tuning rules and use key PID options. For more on this perspective see my Control 10/16/2014 white paper “ So Many Tuning Rules, so Little Time”. This is probably much more background you have time for, so let’s cut to the chase. Frist we need to explain that the term is Final Resting Value (FRV) is what the PID output ends up after a load disturbance or setpoint change. The generalization that FRV overshoot should always be eliminated by some experts is a prime example of not understanding other worlds. Overshoot of FRV is necessary for a setpoint and load response for integrating and runaway processes. In fact the elimination of FRV overshoot leads to level, pressure and temperature trips in vessels and columns and thus a safety issue. Here are some of the many loop objectives : Minimum PV peak error in load response to prevent: – Compressor surge, SIS activation, relief, undesirable reactions, poor cell health Minimum PV integrated error in load or setpoint response to minimize: – total amount of off-spec product to enable closer operation to optimum setpoint Minimum PV overshoot of SP in setpoint response to prevent: – Compressor surge, SIS activation, relief, undesirable reactions, poor cell health Minimum Out overshoot of FRV in setpoint response to prevent: – Interaction with heat integration and recycle loops in hydrocarbon gas unit ops Minimum PV time to reach SP in setpoint response to minimize: – Batch cycle time, startup time, transition time to new products and operating rates Minimum split range point crossings to prevent: – Wasted energy-reactants-reagents, poor cell health (high osmotic pressure) Maximum absorption of variability in level control to prevent: – Passing of changes in input flows to output flows upsetting downstream unit ops Here are some of the different worlds : Hydrocarbon processes and other gas unit operations with plug flow, heat integration & recycle streams (e.g. crackers, furnaces, reformers) – Fast self-regulating responses, interactions and complex secondary responses with sensitivity to SP and FRV overshoot, split range crossings and utility interactions. Chemical batch and continuous processes with vessels and columns – Important loops tend to have slow near or true integrating and runaway responses with minimizing peak and integrated errors and rise time as key objectives. Utility systems (e.g., boilers, steam headers, chillers, compressors) – Important loops tend to have fast near or true integrating responses with minimizing peak and integrated errors and interactions as key objectives. Pulp, paper, food and polymer inline, extrusion and sheet processes – Fast self-regulating responses and interactions with propagation of variability into product (little to no attenuation of oscillations by back mixed volumes) with extreme sensitive to variability and resonance. Loops (particularly sheets) can be dead time dominant due to transportation delays unless there are heat transfer lags. Biological vessels (e.g., fermenters and bioreactors) – Most important loops tend have slow near or true integrating responses with extreme sensitivity to SP and FRV overshoot, split range crossings and utility interactions. Load disturbances originating from cells are incredibly slow making load response a nonissue. Maybe I will write an article and addendum like I did for “ Valve Response - Truth or Consequences ” on the different worlds. At least for now this blog is a starting point. I hope I have given you a way of seeing what world you are in and expanding your horizon and understanding of how it can all come together. A more unified profession is a stronger profession. What world do you live in and where would you like to go? For a concise presentation of the concepts and details on PID control see my ISA book Good Tuning: A Pocket Guide, 4 th Edition . If you are up for a more comprehensive view, see my Momentum Press book Tuning and Control Loop Performance - 4 th Edition .

The post, The Most Disturbing Disturbances are Self-Inflicted , first appeared on ControlGlobal.com's Control Talk blog. The most disturbing disturbances, the ones that are frequent, fast and furious (FFF), are self-inflicted either in the design, installation, maintenance or operation of the process control system. The good news is that through better application of advances in technologies and the better education of everyone responsible for the implementation, maintenance, operation and technical support of the plant's systems, we can eliminate most of these disturbances. This is an opportunity for the synergy between modeling and control to identify, prototype, test, install and continuously improve solutions not forgetting that training plant people is the key for long term effectiveness. We can revitalize the automation profession by finding and implementing process control improvements that goes beyond a simple migration project. Keep in mind as we go through the sources of these most disturbing disturbances that this dialog is not meant as criticism. It is intended to open minds and provide insights as to opportunities. Disturbances most often originate from the Advanced Process Control (APC), Basic Process Control Systems (BPCS), Sequential Operations, Safety Instrumented Systems (SIS), maintenance and operators. The disturbances from outside influences (e.g., raw materials and weather) and internally (fouling of surfaces and deactivation of catalysts) tend to be slower, smaller and less often. The disturbances originating from people, valves, and PID controllers tend to be FFF. Cyclic disturbances are the epitome of frequent disturbances. Cyclic disturbances pose additional problems in terms of the disturbance frequency being much faster than a loop’s natural frequency becoming effectively noise or even worse, the disturbance frequency being around the loop frequency causing resonance. Cyclic disturbances tend to originate from PID controllers due to poor tuning or poor choice of structure, throttle valves due to backlash, stiction, and positive feedback from recycle steams, on-off valves due to sequential operations, regeneration of catalysts, defrosting of crystal coated surfaces and cleaning of fouled surfaces. Cycles can appear that coincide with shift changes due to operators moving the process to their preferential operating points. SIS actions, shutdown and startup are extremely disruptive and can appear to be cyclic if there are many parallel unit operations that are going up and down. If you go a little deeper, you realize nearly all the disturbances begin with valve movement and speed changes since these are ultimately what are manipulated to affect the process. If the valve positions and speeds were not changed, the process would not be at the right operating conditions and possibly unsafe but there would be few cyclic disturbances. We should not lose sight of this fundamental when trying to track down and eliminate a disturbance. We should always seek to find out what valve moved first and what happens if the valve does not move (loop momentarily in manual). Finally, we should ask does that valve need to move and can the movement be slower? For a short term disturbance whose amplitude is not going to cause a trip or damage equipment, the best correction may be no correction. Any feedback correction may be late creating a second disturbance. We will focus in this blog on the BPCS and Sequential Operations since they are the primary source of the above cyclic disturbances. Well-designed Model Predictive Control (MPC) takes into account interactions and dynamics and tends to move the process slowly through proper adjustment of penalty on move (move suppression) and penalty on error. However, if the MPC in the process of pushing the process closer to constraints causes an activation of the SIS, the consequences are severe. MPC by nature is pushing the limits so there needs to be caution exercised. Valve Position Control (VPC) also pushes the limits with the added risk of a VPC not providing decoupling and the PID being tuned improperly. We will address minimizing the disruption by a VPC as part of the discussion of PID control. The overuse of integral action is the primary culprit since it is has no sense of direction and is never satisfied continually moving the process. Integral action is often used in lieu of proportional action. Before we even go any further, let’s make sure that any concern about excessively abrupt changes by proportional action for setpoint changes is alleviated by the use of PID options such as structure or a setpoint filter as noted in the ISA Mentor Program 1/26/2106 post “ Equivalent Methods to Eliminate proportional Step and Derivative Kick ”. For all types of changes, external reset feedback along with up or down rate limits on setpoints can provide directional move suppression, even more flexible than what we get from an MPC. This move suppression prevents upsets to other loops from more proportional action, whether a speed or valve is being manipulated. Simply turning on external reset feedback suspends integral action that would cause the manipulated loop, valve or speed setpoint to change faster than the loop, compressor, fan, pump or valve can respond. This eliminates the confusing burst of oscillations for large disturbances or setpoint changes that occurs from violation of the cascade rule. External reset feedback eliminates the limit cycle from valves that have excessive backlash, a slow response of a large actuator or an insensitive poor positioner design. External reset feedback also enables a VPC to more effectively do its job to give a gradual optimization but a fast recovery for upsets. External reset feedback is also the key feature of an enhanced PID that suppresses oscillations from the excessive dead time of analyzers as discussed in the 7/6/2015 Control Talk Blog “ Batch and Continuous Control with At-Line and Offline Analyzers Tips ”. I could go on and on about all the benefits I have found. Simply turning on external reset feedback and making sure the feedback signal is fast and representative of the response of what is manipulated can prevent oscillations. The PID gain can be increased eliminating the slow oscillations from too low of a PID gain in counterintuitive situations where oscillations get worse as we decrease the PID gain. We are taught that a high PID gain causes oscillations not realizing there many important situations where too low of a PID gain causes even more disruptive oscillations because the amplitude is larger and the cycling is perpetual. We have the counterintuitive situation where for composition, gas pressure, level, pH and temperature control loops on vessels and columns, we have oscillations that get worse as we decrease the PID gain. The process response in these loops is near-integrating, true integrating or runaway. Another common non-intuitive situation is where the limit cycle from backlash and stiction in these processes is reduced by increasing the PID gain. In self-regulating processes, external reset can stop an oscillation from stiction. The offset can be corrected by a setpoint change from an upper loop or the operator. The oscillations from insensitive positioners (e.g., most single stage and spool positioners from the last century) and from the slow response of large actuators (aggravated by low bleed positioners) can be eliminated in most cases by increasing the PID gain after turning on external reset feedback. Manual actions are abrupt, subjective, non-repeatable and usually late and are guesses at best. All manual actions by operators including those during startup, transitions, maintenance activities and abnormal operations should be automated preferably by the use of PID control. Feedforward control, principally ratio control, should be used to keep the process unit operations working together in unison. The scheduled addition of batch feed, reagent, substrate, air and utility flows, should be made more continuous by ratio control and by fed batch control of composition, dissolved oxygen, pH or temperature. The effect of batch operations on downstream equipment should be moderated by intervening tank volumes whose level control is tuned for maximum absorption of flow variability being careful the counterintuitive situation is not created by too small of a PID gain and reset time. The product of the PID gain and reset time should be larger than twice the inverse of integrating process gain. This can be assured by lambda tuning with the arrest time set to prevent the maximum rate of change of level from activating a low or high level alarm. Turning batch, manual, discrete, and sequential operations over to a PID loop intelligent options and tuning with move suppression provides better control and smoother more repeatable manipulation of flows leading to less FFF and better recognition and attainment of process control improvement. I have run out time here and I suspect we are bordering on overload , so let’s simply say that mitigating and even eliminating the FFF disturbances comes down to slowing down the movement of on-off valves for sequences and of throttle valves and speed setpoints in loops by the use of move suppression and external reset feedback. The stroking of on-off valves can be slowed down by needle valves in the actuator air lines but a more effective solution is a smart positioner to provide precise intelligent adjustment and monitoring of the stroking time. The installed flow characteristic curve of these valves can be used to set the valve signal rate of change to slow down and speed up on the steep and flat part of the curve, respectively to provide a more constant less disruptive rate of change of flow. For a concise presentation of the concepts and details on PID control see my ISA book Good Tuning: A Pocket Guide, 4th Edition. If you are up for a more comprehensive view, see my Momentum Press book Tuning and Control Loop Performance - 4th Edition.

The post What Are the Equivalent Methods to Eliminate Proportional Step and Derivative Kick? first appeared on the ISA Interchange blog site. The following technical discussion is part of an occasional series showcasing the ISA Mentor Program , authored by Greg McMillan , industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient and retired Senior Fellow from Solutia Inc (now Eastman Chemicals). Greg will be posting questions and responses from the ISA Mentor Program , with contributions from program participants. In the previous post on bumpless transfer , it was mentioned that there are ways to optimize the immediate change you get in the PID output for a setpoint change. For a structure of PID on error, there is a step from the proportional mode and a kick from the derivative mode in the PID output. This fast and immediate change in PID output makes the setpoint response faster but operators and other loops may get upset by the large abrupt change in PID output. A gradual and smooth response of the PID output for a setpoint change may increase the time to reach setpoint by a matter of seconds for pressure loops and minutes for temperature loops but this may be inconsequential. Also, the prevention of setpoint overshoot is more easily achievable when there is no proportional step. For bioreactor temperature and pH control, overshoot must be totally eliminated and increases in setpoint response time are insignificant because the batch cycle times are days to weeks. Additionally, for many gas unit operations in hydrocarbon processes, abrupt changes and overshoot of final resting value by the PID output must be avoided at all costs due to severe interactions from heat integration and recycle streams. Brian Hrankowsky, an ISA Mentor Program resource, has found several equivalent methods that give the user a lot of options to totally eliminate the proportional step and derivative kick for a setpoint change without affecting load response. His use of Laplace transforms not only provides conclusive proof but shows that this knowledge gained in university control theory courses has considerable practical value. Laplace transforms are effectively used by Brian here to provide a unified view and a valuable perspective showing that the user has a lot of ways of achieving this objective. The ISA Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career. Click this link to learn more about how you can join the ISA Mentor Program. Brian Hrankowsky’s Derivation of Equivalent Methods Vendors provide PID features in various combinations that often are intended to address the same control problem or objective. It is not always clear how to translate or compare these features from one vendor PID algorithm to another. The purpose of these derivations is to illustrate that several methods for eliminating or controlling the proportional step and derivative kick in a PID output resulting from a change of setpoint are equivalent by seeing they end up with the same Laplace transform equations. All algorithms below are based on the ISA standard form. To identify control actions taken on process value vs. error, a hyphen is used to separate the two. Actions to the left of the hyphen are taken on error. Actions to the right are taken on PV. For example, a PID controller with a structure of proportional and integral on error and derivative on PV is shown as “PI-D” and a structure with integral on error with proportional and derivative on PV is shown as “I-PD”. Setting a Setpoint Filter Equal to the Reset Time Setting the PID Algorithm to Use a Proportional and Derivative on PV Structure Setting the PID Algorithm to Two Degrees of Freedom (2DOF) Structure with Beta and Gamma Set to Zero After setting beta and gamma to zero, the Laplace transform is the same as the one for “I-PD” structure previously shown so there is no need to repeat the remaining equations. Setting the Setpoint Lead-Lag (SPLLAG) Factor to Zero Some controllers use a setpoint lead and lag where the lag is automatically set to the RESET time and the lead is a factor of the reset time between zero and one inclusive. Below, I’ve used beta to represent the factor. The resulting equation is the same as for the 2DOF with beta and gamma set equal to zero. We see that there are four equivalent methods of eliminating the proportional step and derivative kick from a setpoint change. Since not all PID controllers have a built in option for adding a lead-lag or filter to the setpoint or a structure with Integral action on error and proportional and derivative action on the process variable or two degrees freedom, this recognition of equivalent functionality has significant practical value as it allows the control engineer to achieve the desired result with whichever of the options are available for the particular control system. The user can readily eliminate abrupt changes in the PID output and prevent overshoot of the setpoint by the PV and overshoot of the final resting value by the PID output when required. Any change that is to be made to a control system must be thoroughly functionally tested by realistic simulations of the process’s dynamic response. The ability of the control system improvement to deal with abnormal besides normal operating conditions must be verified. The commissioning and performance of improvements should be closely monitored to ensure they meet plant requirements. See the ISA book 101 Tips for a Successful Automation Career that grew out of this Mentor Program to gain concise and practical advice. See the InTech magazine January/February 2013 feature article “ Enabling new automation engineers ” for the candid comments of some of the original program participants. See the May 2015 Control Talk column “ How to effectively get engineering knowledge ” with the 2014 addition to Mentor Program Keneisha Williams on the challenges faced by young engineers today. Providing discussion and answers besides Greg McMillan and co-founder of the program Hunter Vegas (project engineering manager at Wunderlich- Malec) are resources Brain Hrankowsky (consultant engineer at a major pharmaceutical company), Michel Ruel (executive director, engineering practice at BBA Inc.), Leah Ruder (process systems automation group manager at the Midwest Engineering Center of Emerson Process Management) and Nick Sands (ISA Fellow and Manufacturing Technology Fellow at DuPont).

The post, Effect of Measurement Span on Loop Performance first appeared on the ControlGlobal.com Control Talk blog. The advent of smart transmitters has reduced the effect of large measurement spans on accuracy but most measurement accuracies are still a function of per cent of span, albeit possibly detailed by more sophisticated equations than simply the error being a percent of span. Bigger considerations these days concern the effect of measurement span on controller tuning and the effect of flow meter size on measurement rangeability. It has been a common practice to narrow the span of temperature transmitters to improve the accuracy of this measurement that in many loops is critical since temperature is often the key to product quality by determining the composition (e.g., distillation tray temperature), the formation of products (e.g., bioreactor and chemical reactor and crystallizer temperature) and the separation of phases (e.g., evaporator and dryer temperature). The advent of smart transmitters reduces the error from span but the practice of narrowing the span is still worthwhile especially in avoiding unnecessarily high scale ranges. The benefit of a separate transmitter to handle lower operating ranges on startup can be valuable. The second transmitter also offers reliability and better online recognition of performance. For pH measurement, triple electrodes and middle signal selection are essential to improve reliability and reduce the effect of a coated electrode or measurement noise that ultimately determines repeatability. Most flow meters have a low flow limit to their rangeability that is based on fluid velocity in the meter. To get the meter rangeability shown in the catalog necessitates your max flow requirement exactly matching the meter size capacity. This is a rare occurrence. You are better off choosing a smaller size meter than line size that enables a lower flow at the low velocity limit. In some cases, a second meter or second transmitter with a lower range is used. For flow feedforward, which is really flow ratio control, flow meter rangeability is a limiting factor especially for startup when often ratio control is more important. For example, distillation columns run solely on ratio control until the column reaches operating conditions where temperature is an inference of composition and can be used to correct the ratio. A feedforward summer is particularly appropriate for volumes with mixing from agitation or recirculation or boiling like columns and is less sensitive to feedforward scaling and flow measurement problems especially at low production rates as noted in the 5/30/2015 Control Talk Blog “ Essential Feedforward Control and Operator Interface Tips ”. Most of the literature talks about a process gain and its effect on tuning. This leads one to overlook the effect of instrumentation on tuning. More appropriate terms are an open loop self-regulating process gain and an open loop integrating process gain. These open loop gains are the product of a manipulated variable gain (e.g., valve gain), a process gain (including a hidden factor for composition, pH and temperature per last month’s blog), and a measurement gain. The measurement gain is the simplest and most linear of all gain calculations (assuming linearization has occurred for differential head meters and temperature sensors) being simply 100% divided by the measurement span in the same engineering units as the process gain. The proper use of units gives an open loop self-regulating process gain that is dimensionless and an open loop integrating process gain with units of 1/sec. The end result is that if you reduce the measurement span by one half, you have doubled the open loop gain. If your PID gain was much lower than possible, often the case for level and composition control, the performance of the loop will be better in terms of a smaller peak error and integrated error for load disturbances. If the PID gain was close to the maximum, the PID gain should be halved for the halved measurement span. The problem here of more oscillation may only show up at low production rates for composition, pH and temperature control when the process gain is higher due to the hidden factor. The final message is to realize the effect of measurement span on loop performance in terms of truly knowing what the process is doing (measurements being the window into the process) and enabling the best effect on the process in terms of controller action (feedforward and feedback control being the way of correcting the process). Please don’t skimp on technology and number of transmitters needed to achieve the 5 Rs (rangeability, resolution, repeatability, response time and reliability) detailed in the 9/9/2015 Control Talk Blog “ What is Truly Important for Measurements and Valves ”. We often don’t give our profession and ourselves enough credit to know and make a case for more and better instrumentation. Most of my mistakes as a user early in my career were the result of a misguided attempt to reduce project costs. This is not me putting on a sales hat. I never have been or can be a sales person. This is my honest plea to not sell yourself or our profession short by focusing on cost to the exclusion of performance. Taking the advantage of new technology is the key to advancement. Better loop performance can provide benefits in many unexpected ways. Just have the confidence to do it. Migration projects that simply translate the configuration and do not upgrade the instrumentation and improve the use of PID features and tuning are doing the plant and our profession a great disservice. For more details on the use of terms such as process gain that are often misunderstood see the 8/24/2015 Control Talk Blog “ Understanding Terminology to Advance Yourself and the Profession ”. For a concise presentation of the concepts and details on the effects of measurements on PID control see my ISA book Good Tuning: A Pocket Guide, 4th Edition . If you are up for a more comprehensive view, see my Momentum Press book Tuning and Control Loop Performance - 4th Edition .

The post What Is the Definition of PID Bumpless Transfer? first appeared on the ISA Interchange blog site. The following technical discussion is part of an occasional series showcasing the ISA Mentor Program , authored by Greg McMillan , industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient and retired Senior Fellow from Solutia Inc (now Eastman Chemicals). Greg will be posting questions and responses from the ISA Mentor Program, with contributions from program participants. The PID controller is an essential component of nearly all control loops in the process industry. We take for granted that a change in the PID controller mode or setpoint will not disturb the process. More needs to be known about bumpless transfer and how a new operating point can be achieved quickly and smoothly after the PID is switched to the automatic mode.The following question by Brian Hrankowsky, a resource in the ISA Mentor Program, helps us seek a good definition of PID bumpless transfer, how it is achieved in the PID and how options such as setpoint tracking, PID structure, setpoint dynamics and external reset feedback come into play. Brian Hrankowsky’s Question In our plants, I have seen two definitions for PID bumpless transfer: Ninety-nine percent use bumpless transfer to mean that on transition from auto to manual, the control output does not change and on a transition from manual to auto, the SP matches the PV and no control action is taken from the current controller output until the SP or PV changes (output does not jump to previous or other value; when an error occurs, output is driven by PID action from the current output value). One percent use bumpless transfer to mean that on transition from auto to manual, the control output does not change and on a transition from manual to auto, control action is taken from the current controller output in response to the SP not being not equal to the PV (output does not jump to previous or other value; since an error already exists, output is driven by integral action from the current output value). Note that the only difference in the two definitions is the removal of “ the SP matches the PV and “. In reading on line, it seems that both ways of using the term are in use. The majority seems to be definition No. 1. However, I can imagine that configuring some systems to work like definition No. 2 suggests would be beneficial. In fact, I think the way some skid mount OEM equipment works is really like definition No. 2: The controller setpoint is fixed by the selected “format” or “recipe” and starting/stopping the machine just changes the controller mode. But generally in our DCS implementations, we configure the loop to behave as in definition No. 1. The ISA Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career. Click this link to learn more about how you can join the ISA Mentor Program. What is the official (ISA?) definition? Greg McMillan’s Answer Good question. The transfer from manual or remote output to auto should be bumpless regardless of the current setpoint. This gives the most options. This is commonly achieved by biasing the contribution from the integral mode to counter act any contribution from the proportional mode. The ISA Dictionary 4th Edition says as much and even has a little equation that shows the bias (b) is set equal to the manual output (m o ) minus the product of the gain and error (K c e) to provide bumpless transfer. The simple equation does not show the control action sign or the effect of a beta setpoint weight factor. In real life it gets more complicated but the goal should be the same. b = m o – K c e You may be able to simply check a box to enable a SP track PV option in a modern DCS. In chemical processes where we have seen benefit from the proportional step in the PID output on a setpoint change to get to the new setpoint sooner, we use a structure of PI on error and D on PV, enable the SP track PV option, and then have the batch or startup sequence automatically change the primary PID setpoint at the right time. A setpoint lead-lag or two degrees of freedom structure can minimize setpoint overshoot by optimizing the contribution from the proportional and derivative modes to the PID output. This of course only works well if other systems are not upset by a sudden large change in the PID output. Tight pressure control of the source of the manipulated flow can reduce the upset to other users of the source (e.g., utility). However, operators may get concerned even if there is no detrimental effect. To prevent this, you can use setpoint rate limits on the analog output block or secondary PID setpoint and enable external reset feedback (dynamic reset limit) in the primary PID to slow down the change the in manipulated flow (now famously known as directional move suppression) and the consequential upset to others (including operators) without having to retune the primary PID. In other words, “Bumpless Transfer” is a basic MUST for all loops regardless of setpoint value, and the “SP Track PV” option is chosen based on process and operator requirements. To achieve a new setpoint faster after a PID mode makes the transition from manual or remote output to auto, this track option is used and the setpoint is then changed to the desired operating point after the PID goes to auto to give a quick change in the contribution from the proportional and derivative modes as determined by the PID structure or setpoint dynamics. External reset feedback and setpoint rate limits of the manipulated flow can slow down the actual change in the PID output to reduce the upset to other loops and operators. Any change that is to be made to a control system must be thoroughly functionally tested by realistic simulations of the process’s dynamic response. The ability of the control system improvement to deal with abnormal besides normal operating conditions must be verified. The commissioning and performance of improvements should be closely monitored to ensure they meet plant requirements. See the ISA book 101 Tips for a Successful Automation Career that grew out of this Mentor Program to gain concise and practical advice. See the InTech magazine January/February 2013 feature article “ Enabling new automation engineers ” for the candid comments of some of the original program participants. See the May 2015 Control Talk column “ How to effectively get engineering knowledge ” with the 2014 addition to Mentor Program Keneisha Williams on the challenges faced by young engineers today. Providing discussion and answers besides Greg McMillan and co-founder of the program Hunter Vegas (project engineering manager at Wunderlich- Malec) are resources Brain Hrankowsky (consultant engineer at a major pharmaceutical company), Michel Ruel (executive director, engineering practice at BBA Inc.), Leah Ruder (process systems automation group manager at the Midwest Engineering Center of Emerson Process Management) and Nick Sands (ISA Fellow and Manufacturing Technology Fellow at DuPont).

Composition, pH and temperature loops that largely determine product quality have a hidden factor that affects the loop linearity and particularly the ability to perform well at low production rates. Here we detail the factor, the consequences and the solutions for continuous unit operations and fed-batch reactions. More...

The post How Do You Measure Control System Performance? first appeared on the ISA Interchange blog site. The following technical discussion is part of an occasional series showcasing the ISA Mentor Program , authored by Greg McMillan , industry consultant, author of numerous process control books, 2010 ISA Life Achievement Award recipient and retired Senior Fellow from Solutia Inc (now Eastman Chemicals). Greg will be posting questions and responses from the ISA Mentor Program, with contributions from program participants. For the automation profession to gain the recognition it deserves, we need to show the impact of our achievements on improving plant performance. This question is from Sridhar Dasani, a recent addition to the ISA Mentor Program, with my answer and additional input from co-founder of the program Hunter Vegas (project engineering manager at Wunderlich-Malec) and resource Brian Hrankowsky (consultant engineer at a major pharmaceutical company). The question by Sridhar Dasani helps us seek a suite of practical metrics in terms of key performance indicators (KPIs) that can have an impact on plant performance. At the end is a link to a Control Talk column that offers metrics more directly related to plant efficiency and plant capacity. Sridhar Dasani’s Question As a new engineer supporting plant operations, I regularly attend monthly meetings where the operations team presents its key performance indicators in terms of the production rates, savings, and uptime, and maintenance personnel present their SAP-scheduled maintenance numbers. I always wonder how an automation engineer supporting DCS system administration or process control can present his/her performance? Based on your experience, would you be able to suggest what KPIs can be presented during these meetings, anything from system health to loop performance to start with? The ISA Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career.  Click this link to learn more about how you can join the ISA Mentor Program. Hunter Vegas’s Answer If the plant is continuous you can look at: (1) Percent of time the control loops are in their highest mode (e.g., auto, cascade, remote cascade). (2) Alarm rates (per week, per month). Obviously this is somewhat dependent on the plant operations but if you are doing an alarm rationalization project it gives you a feel for how well you are doing. (3) If you have made a control system change to the process you can document before/after performance (within range of setpoint, production rate, quality, etc.). (4) Non-routine instrument work orders generated per month gives you a clue on instrumentation reliability. Those are some that immediately come to mind. No doubt Greg will have a longer list. Greg McMillan’s Answer (5) Percent of time and number of times PID output hits output limit. (6) Standard deviation of the process variable. (7) Number of oscillations where the process variable has crossed the setpoint by more than a deadband or noise band (may or may not want to detect noise) forgetting results older than an analysis period. (8) Peak error for a constant setpoint forgetting results older than an analysis period. (9) Integrated absolute error (IAE) forgetting results older than an analysis period. (10) Rise time: time to reach setpoint (within deadband around setpoint) for setpoint change. (11) Overshoot and undershoot for setpoint change. (12) Settling time: time to settle out at setpoint (stay for more than four dead times within deadband around setpoint) after setpoint change. Even better as noted by Hunter is if you could tie what you do to operations KPIs on process capacity (production rate and onstream time) and efficiency (energy and raw material use and waste). This could be done by noting if the KPIs are better because of changes you made or from better loop metrics you achieved. For a view on how to use a virtual plant to develop and demonstrate online KPIs, see the April 2015 Control Talk column “ Getting innovation back into process control. ” Brian Hrankowsky’s Answer (13) How many operator actions are logged in the audit trail/event history? (14) How long are the “waits” in the process? (How long is the heat up or cool down before the next value added step can take place and what is the duration of prompts lingering for an operator?) (15) How many batch “hold” events are there and what has been the duration? (16) How often are interlocks tripping? (17) How often are estop or shutdown events occurring? (18) How much paper work does it take to run the process? Any change that is to be made to a control system must be thoroughly functionally tested by realistic simulations of the process’s dynamic response. The ability of the control system improvement to deal with abnormal besides normal operating conditions must be verified. The commissioning and performance of improvements should be closely monitored to ensure they meet plant requirements. See the ISA book 101 Tips for a Successful Automation Career that grew out of this Mentor Program to gain concise and practical advice. See the InTech magazine January/February 2013 feature article “ Enabling new automation engineers ” for the candid comments of some of the original program participants. See the May 2015 Control Talk column “ How to effectively get engineering knowledge ” with Keneisha Williams, who joined the ISA Mentor Program in 2014, on the challenges faced by young engineers today. Providing discussion and answers besides me and co-founder of the program Hunter Vegas (project engineering manager at Wunderlich- Malec) are resources Brian Hrankowsky (consultant engineer at a major pharmaceutical company), Michel Ruel (executive director, engineering practice at BBA Inc.), Leah Ruder (process systems automation group manager at the Midwest Engineering Center of Emerson Process Management) and Nick Sands (ISA Fellow and Manufacturing Technology Fellow at DuPont).

The ISA Mentor Program, started by Greg McMillan and Hunter Vegas in 2011, is extending its invitation to prospective protégés/protégées and resources to include any engineer or technician currently employed in a plant or at an engineering center or supplier of services that is applying or maintaining process control systems. More...

The post How Do You Convert Tuning Settings of an Independent PID? first appeared on the ISA Interchange blog site. The following technical discussion is  part of an occasional series showcasing the ISA Mentor Program, authored by Greg McMillan , industry consultant, author of numerous process control books and a retired Senior Fellow from Monsanto. The PID controller has been used in more than 99 percent of the control loops in the process industry for the last 80-plus years. The loop performs only as well as the controller is tuned. Considerable confusion exists as to how the PID Form affects tuning settings. The lack of understanding of Form and units of settings can lead to disastrous results. Users may not realize that the algorithm shown in text books often implies an Independent Form that is also known as the Parallel Form and consequential engineering units that are seldom used in the process industry. Some of these algorithms even imply the PID computation is done in engineering units rather than in percent of input and output scales that is almost universally done in industry. The lack of recognition of the Form and units employed by a PID algorithm can cause settings to be off by orders of magnitude. ISA’s Mentor Program enables young professionals to access the wisdom and expertise of seasoned ISA members, and offers veteran ISA professionals the chance to share their wisdom and make a difference in someone’s career. Click this link to learn more about the ISA Mentor Program Here we look at the Independent (Parallel) Form and how to convert settings to the ISA Standard Form that is also known as the Ideal Form. The answer also discusses the Series Form predominantly used in analog and early DCS controllers that is also known as the Interacting Form due to interaction of the derivative mode with the other modes settings in the time domain. Note that the tuning settings of the ISA Standard Form are non-interacting in the time domain but are interacting in the frequency domain. The inconsistent names for the Forms increase confusion. I will be posting an occasional series of technical discussions from the ISA Mentor Program, with contributions from program participants. This question is from Héctor Torres, an original member of the ISA Mentor Program, with my answer and additional input from program participants Hunter Vegas (project engineering manager at Wunderlich-Malec) and Michel Ruel (executive director, engineering practice at BBA Inc.) Héctor Torres’s Question I am moving some PID’s currently residing in a PLC to a DCS. I would like taking advantage of the tuning already existing in the PLC to minimize commissioning time at startup. This is the information I got about this control loops in the PLC: Loop 1: PID Type: Independent Control Action: E=PV-SP Settings: Proportional:     0.9 Integral:             0.75 repeats /min Derivative:         0 min Loop 2: PID Type: Independent Control Action: E=SP-PV Settings: Proportional:     1.5 Integral:             0.50 repeats/min Derivative:         0 min How can I translate these tuning parameters to a Form commonly used in a today’s DCS. I would like to use the Series Form of the PID. We are migrating from an older DCS that used the Series Form. The new DCS offers the ISA Standard Form and the Series Form. Greg McMillan’s Answer My understanding is that the independent PID is what is more commonly called the parallel PID because the contribution of the modes is independent in the time domain because the PID gain is not applied to the contribution from the integral and derivative modes (see Figure K-3 in Appendix K of Tuning and Control Loop Performance – 4 th Edition). Fortunately, your rate time is zero so we don’t need to be concerned about whether you are using the Series or ISA Standard Form. The gain for the independent PID is the same as the ISA Standard PID and for the case of zero rate time is also the same as the Series Form. One thing to be careful about is to make sure the PLC algorithm is working with signals in percent rather than engineering units. Nearly all DCS and most PLC PID algorithms work with signals in percent. The user must thoroughly test and confirm all changes are correct by a real-time simulation before using them in an actual plant. Assuming the proportional settings are a dimensionless gain and not Proportional Band, the PID algorithm works in percent signals and the Independent PID is what we call the Parallel Form, the equivalent Series Form settings are: Loop 1: Gain = 0.9 Reset = 0.9 * 60/0.75 = 0.9 * 80 = 72 sec Rate = 0 sec Loop 2: Gain = 1.5 Reset = 1.5 * 60/0.5 = 1.5 * 120 = 180 sec Rate = 0 sec It would be good if Michel and/or Hunter would check these results in case I am missing something or am in error in some way. Hunter Vegas’s Answer I didn’t work through Greg’s numbers but generally when I am faced with something like this I have been able to find “conversions” from one system to another. You didn’t specifically mention the PLC you were using but I have found tuning software packages can convert and even a scan on the internet can turn up some decent conversion equations. Do note that some PLCs have issues where scan time can alter the execution time of the PID block and effectively change your tuning constants accordingly. Michel Ruel’s Answer Be careful with terminology… most suppliers use independent for parallel and dependent for series but some suppliers use independent for ideal form and dependent for series. You need to look at the structure and read the description. Greg McMillan’s Follow Up If you scroll down on the webpage Comparison of PID Algorithms to the Parallel PID noted as the PID algorithm with “independence” you see the equations to convert from the Parallel Form to the Ideal Form (ISA Standard Form) are the same as the ones I used below and in my book but with different nomenclature. Some parallel algorithms have integral gain instead of an integral time (seconds) or the inverse of integral time (repeats/second) and a derivative gain rather than a derivative time. The general methodology is: Use Chapter 1 Equation 1.2b to convert the integral mode reset setting from repeats per minute to seconds (seconds per repeat). Use Appendix K Equations K.7, K.8, and K.9 to convert settings from the Parallel PID (Independent PID) to the ISA Standard Form (Ideal PID). If the rate time is not zero and the integral time is much greater than four times the rate time, use Equations K.4, K.5, and K.6 to convert from the ISA Standard Form to the Series Form. Note that if the rate time is zero, Equations K.4 and K.5 show the Series Form settings are the same as the ISA Standard Form settings since the division by 2 is cancelled by the multiplication by 2. To convert the tuning settings from older DCS Series Form to the newer DCS default ISA Standard Form: Note that if the rate time is zero, the ISA Standard Form and Series Form settings are identical because the correction factor applied becomes 1. When using the ISA Standard Form, if the rate time is greater than one-fourth the reset time the response can become oscillatory. If the rate time exceeds the reset time, the response can become unstable from a reversal of action from these modes. The Series Form inherently prevents this instability by increasing the effective reset time as the rate time is increased. Nomenclature: See the ISA book 101 Tips for a Successful Automation Career that grew out of this Mentor Program to gain concise and practical advice. See the InTech January/February 2013 feature article “ Enabling new automation engineers ” for the candid comments of some of the original program participants.

We have probably heard of the benefits of a signal character in compensating for the gain nonlinearity of a control valve flow characteristic and the pH measurement. What we often don’t recognize is how the use of signal characterizers enables a more accurate loop dynamics identification and adaptation, restores a process time constant and reduces deadtime and limit cycle amplitude. Here we look at how these benefits arise and the “ins and outs” for a successful signal characterizer implementation. Read Greg's complete post, Unexpected Benefits of Signal Characterizers , on his ControlGlobal.com Control Talk blog.

The post How to Improve Automation by Understanding Integrating and Runaway Process Dynamics first appeared on the ISA Interchange blog site. The following insights are part of an occasional series authored by Greg McMillan , industry consultant, author of numerous process control books and a retired Senior Fellow from Monsanto. This is the second part of a two-part series. Many important process control loops on vessels and column do not have a steady state in the time frame of the PID response and are classified as near-integrating as discussed in the last post. Integrating and runaway processes have no steady state at all in the operating range of the process. If there is insufficient process control action, these processes can ramp and even accelerate to the violation of an equipment limit or a trip of a Safety Instrumented System (SIS). Integrating processes are defined by a total loop dead time and an integrating process gain (see the figure below). A secondary time constant may be used to describe the initial lag (bend) in the response. The integrating process gain is the ramp rate in % of PID input per second divided by the % change in PID output giving a gain of reciprocal time units (1/sec). Level is the most common loop with an integrating response. Integrating Process Response Gas pressure has a near integrating or true integrating response depending on the size of the pressure drop across the manipulated valve relative to the changes in pressure during an open loop test. Batch composition, pH, and temperature loops have essentially an integrating response unless altered by a reaction. An integrating process will exhibit self-regulating closed loop response for a proportional only controller. The distance of the new from the initial operating point decreases as the PID gain increases. Integrating and “near integrating” processes require aggressive proportional action. The steady state gain divided by the open loop time constant of a near integrating process is effectively an open loop integrating process gain. The maximum PID gain is inversely proportional to the process time constant or integrating process gain. Most integrating processes are so slow (integrating gain so small) and the dead time is so relatively small that the maximum PID gain so large that the primary limit to how high you set this gain is user knowledge and noise. The degree of internal negative feedback within the process determines whether the process will quickly settle, ramp, or diverge when the controller is in manual and the PID output is changed. For integrating and runaway processes there is no internal negative feedback and hence no steady state. When internal process negative feedback is absent or even worse is replaced with positive feedback in runaway processes, the amount of negative feedback from the controller must increase principally by an increase in proportional action (controller gain). The best controller tuning settings can be determined by identifying key dynamic variables that characterize the process response. See the recently extensively updated Good Tuning: A Pocket Guide 4 th Edition for concise guidance on this topic and everything else you need to know to get the most out of your PID controller. Runaway Process Response Runaway processes are defined by a total loop dead time, a positive feedback open loop time constant, and an open loop runaway process gain (see the figure). The positive feedback time constant causes the continual acceleration. A secondary time constant may be used to describe the initial lag (bend) in the response. The runaway process gain is best estimated via the ordinary differential equations for the reactor energy balance per Appendix F in the ISA book Advances in Reactor Measurement and Control . Open loop tests are rarely done in runaway processes because of the safety concerns from the acceleration. Manual control of a runaway process is extremely difficult. Most true runaway processes are always operated in automatic or a higher mode. Tuning tests are done with the PID loops in auto. The large controller gains and integral times used for these loops provide a step change in controller output from the proportional mode and negligible ramping from the integral mode in four dead times. Polymerization and specialty chemical reactors with a heat release from an exothermic reaction that can exceed the cooling rate can develop a runaway response for an increase in temperature. Determine the type of process response and choose the tuning rules accordingly. Make step changes to the process to identify the total dead time, open loop gain, and secondary time constant. If the process continues to ramp or accelerate in this time frame, the process response is near-integrating, integrating, or runaway. The open loop gain for these processes is an integrating process gain (change in ramp rate divided by change in controller output) and integrating process tuning rules are used. The PID structure and tuning settings must provide immediate action overdriving the PID output past its final resting value by more proportional than integral action. For surge tank level control discussed in a future post, we will see that the controller gain is moderated to maximize the absorption of variability. However, the PID output must still be driven past its final resting value (balance point where manipulated flow equals net load).

The post, What is Truly Important for Measurements and Valves , first appeared on ControlGlobal.com's Control Talk blog. Control is all about dealing with change whether it is disturbances or moving setpoints. The controller can only do as well as the changes it sees and the changes it is able to make. Here we look at what is truly important realizing what is on the specification sheet may not be addressing what we really need. First let’s digress to get a perspective. If there were no change, we could set valves to give us the flows on a Process Flow Diagram (PFD). This steady state type of mentality is sometimes propagated from chemical engineering courses focusing on process design. When I first taught a course on process control to chemical engineering students, the question asked was what is the big deal with all this stuff you are talking about in terms of dynamic responses? You just need to set the flows per the PFD. Correspondingly dynamic modeling fidelity is most often defined to be how well the process flows in the model match the flows in the plant at a given operating point (process setpoint), a carryover in thinking from steady state modeling. For me, dynamic modeling fidelity is how well the change in the process variable, change in measurement and change in valve position match the changes in the plant in terms dead time, time constant and gain (see the last blog for a greater understanding of these terms). While seeking to match the dynamic response of the model to the plant should be the objective and criteria for fidelity, this is kind of rare in first principle models and even missing in many neural network models. Model Predictive Control (MPC) realizes this is important using powerful identification techniques to model the changes observed for the changes made. This is one of the many reasons why MPC has been successful. Even if you are going to do PID control, I would use the identification software for MPC to get the open loop process gain, primary time constant, secondary time constant and total loop dead time. The change is best noted and denoted by the details that define the response. Here we first look at problems in the response that can be corrected, such as offset, drift, nonlinearity, span errors, and deadband. While valve and measurement offset is not nice, these effects can largely be eliminated by calibration adjustments and the changes in the setpoint either manually or automatically by the feedback action of upper loops. Drift can be considered to be an offset if slow enough. Smart measurements and valves have reduced offset and drift by an order of magnitude or more. Nonlinearity is more problematic but can be dealt with by internal compensation in transmitters (e.g. thermocouple and RTD sensor matching), signal characterization (e.g., titration curve and installed valve characteristic) and adaptive control. Span errors such as pH electrodes efficiency can be eliminated by calibration and smart instrumentation. Remaining span errors can result in a change in measurement gain but this effect is usually quite small compared to other sources of nonlinearity and could be theoretically corrected based on the size of the change in measurement seen. Deadband does not exist in non-mechanical sensors or variable frequency drives but may be introduced in the configuration setup. Deadband is a common problem in valves due to backlash in connections and linkages. Deadband can be minimized by control valve design and by configuration settings. Remaining valve deadband can be compensated for by adding a delta equal to the deadband to the change in PID output to jump through the deadband whenever the PID output changes direction by an amount greater than a noise band. What is left that cannot be readily corrected in the control system design in terms of its response (R) are the five Rs of measurements and valves - resolution, repeatability, 86% response time, rangeability, and reliability. Not commonly recognized is that the effect of Rs besides reliability can be reduced to some extent for measurements by middle signal selection of three separate sensors and transmitters. Middle signal selection can inherently protect against a single failure of any type and does a great job of largely ignoring a slow sensor. Resolution is the smallest change that a measurement can detect or a valve can respond to. Once the change occurs, the response is a step whose size is the resolution limit. For a threshold sensitivity limit, the change would match the final change whereas with a resolution limit there is a stair step response that almost by definition would not match the final change. Valve slip equal to valve stick is essentially a resolution limit. Repeatability is the difference in final responses for the same change in the process variable for measurements and for the same change in controller output for valves. Noise can appear to be a repeatability error. 86% response time is the time for a measurement or valve to reach 86% of its final response. For a linear first order approximation, this corresponds to the sum of the dead time and two time constants. The 86% response time is the response criteria per the ISA Standard for Valve Response Testing and should be the standard for sensors particularly pH electrodes because this is the response time of greatest interest in terms of the correction by the controller and because there is a long protracted and variable time to reach a 95% or 98% response time due to various non-ideal effects. Measurement rangeability is the ratio of the maximum to minimum process variable where the response at the minimum process variable has the same resolution, repeatability, 86% response time and reliability. Valve rangeability is the ratio of the maximum to minimum flow where the response at the minimum flow has the same resolution, repeatability, 86% response time and reliability. What can I say about reliability except it is a metric of availability of the measurement or valve (e.g., time to failure)? I would take an expanded view of failure as the inability of a measurement or valve to continue to respond with about the same resolution, repeatability, rangeability, and 86% response time as it normally does. While you should mind your Ps and Qs (Process Variables and Quality Variables) to achieve the best values of these, you need to pay attention to the 5 Rs in the response of measurements and valves to change (resolution, repeatability, 86% response time, rangeability, and reliability). Unfortunately, these rarely appear on specification sheets.

The post How to Improve Automation by Understanding Self-Regulating Process Dynamics first appeared on the ISA Interchange blog site. The following insights are part of an occasional series authored by Greg McMillan , industry consultant, author of numerous process control books and a retired Senior Fellow from Monsanto. This is the first part of a two-part series. The three types of process responses encountered in industry are defined based on an open loop test where the PID is in manual or remote output so there is no response of the PID to the process (no closed loop response) . A step change is made in the controller output. The process response is observed until the process can be identified. During the test, there should be no disturbances so that the process response seen is entirely the result of the step change in PID output. The three types of processes are self-regulating, integrating, and runaway. A self-regulating process will decelerate to a new steady state operating point (see the figure) due to negative feedback within the process. An integrating process will continually ramp from the lack of feedback within the process. A runaway process will accelerate until hitting a relief or interlock setting due to positive feedback within the process. As the degree of negative feedback within the process (degree of self-regulation) decreases, more negative feedback action is needed from the PID controller (more proportional action by a higher gain setting). For estimating loop performance and tuning settings the parameters used to identify each type of response are gain, time constant and dead time. The definition of the parameters depends upon the type of response. The terms have alternate names in industry. For example, “lag” is used for time constant, “delay” is used for dead time, and “sensitivity” is used for gain. The response observed in these tests includes the response of the analog output, final control element (e.g., control valve or variable frequency drive), process, sensor, transmitter, analog input, and the process variable ( PV ) input to the PID. The observed response includes the effect of velocity limits, dead times, time constants, and gains in the automation system. Better terminology would be “open loop response” than “process response” because the observed response includes almost everything in the loop response. Also, the source of the individual parameters that create the particular dynamic in the response should precede the term (e.g., valve dead time and measurement dead time). The best controller tuning settings can be determined by identifying key dynamic variables that characterize the process response. See the recently extensively updated Good Tuning: A Pocket Guide 4 th Edition for concise guidance on this topic and everything else you need to know to get the most out of your PID controller. All processes have a dead time that is the time interval between the step change in output and the first recognizable change in process. Noise can delay the recognition until the excursion is beyond the noise band creating a longer dead time. The observed dead time is frequently called the process dead time. The observed dead time is really a total loop dead time ( θ o ) that is the sum of all the pure dead times and the equivalent dead times from all time constants smaller than the largest time constant in the loop for a first order (one time constant) plus dead time approximation. For a second order plus dead time approximation that includes a secondary time constant, all time constants smaller the largest time constants create an equivalent dead time. The secondary time constant creates the bend in the initial response right after the dead time. The equivalent dead time increases from 30% to 99% of the time constant as the ratio of the time constant to the largest time constant gets smaller. Time constants small compared to the largest time constants are summed as being essentially 100% dead time. We will see in a future post on loop performance that ultimate limit to the peak and integrated errors are proportional to the dead time and dead time square, respectively. If there was no dead time and no noise, perfect control would be possible. Not seen in these open loop responses is the additional delay experienced in closed loop operation from the time it takes for the output signal to pass through dead band (backlash), threshold sensitivity (stiction), and resolution limits of the final control element (e.g., control valve). The test uses a step change larger than these limits. In closed loop operation, a step change in output can occur for a step change in set point but subsequent closed loop action to recover from overshoot or disturbances involves gradual changes in the output. The additional dead time from these limits can be approximated as the limit divided by the rate of change of the signal (e.g., valve dead band divided by the rate of change of the PID output). Self-Regulating Process Response The open loop time constant ( τ o ) is the largest time constant (i.e., primary time constant) plus any the portion of smaller time constants not taken as a secondary time constant or as effective dead time. While often called the process time constant, the largest time constant can occur anywhere in the loop. For liquid pressure and flow loops, the largest time constant is usually somewhere in the automation system (e.g., valve, sensor, transmitter, or DCS). Ideally, the largest time constant is the primary time constant of the process downstream of where the disturbances and manipulated flow enter the process. We will see in the section on loop performance, the effect of such a primary time constant is beneficial. The ultimate limit of the peak and integrated error is inversely proportional to this primary time constant. The primary process time constant slows down an excursion from a disturbance giving time for the PID to catch up with it. If the largest time constant is in the measurement, the trend chart oscillation may look better because the amplitude is attenuated by the filtering effect of the measurement time constant. In an open loop test, you cannot discern the location of the largest time constant. The second largest time constant is called a secondary time constant ( τ s ). The secondary time constant can be quite large due to heat transfer surface and thermowell lags in temperature processes and mixing or electrode lags in concentration or pH processes. The open loop gain ( K o ) is the product of the final control element, process, and measurement gain. Consider a loop with a control valve. The final control element gain (change in flow divided by the change in % PID output) is the slope of the valve’s installed characteristic curve. The process gain (change in process variable divided by the change in valve flow) is the slope of a plot of the process variable versus valve flow. The measurement gain (change in % PID input divided by the change in process variable) is the 100% divided by the measurement span. Since the PID algorithm uses % signals, calculations of the open loop gain must involve % signals despite the fact that the PID block and graphics show the PID process variable and in some DCS the PID output in engineering units. Self-regulating processes are defined by a total loop dead time, an open loop time constant, and an open loop gain called a steady state gain (see the figure). A secondary time constant may be used to describe the initial lag (bend) in the response. The open loop gain is a steady state gain that is the % change in PID input from its initial to its final value divided by the % change in PID output giving a dimensionless gain. Liquid pressure and flow loops have a self-regulating response. Continuous composition, pH, and temperature loops have a self-regulating process response but the process time constant for large well mixed vessels is so large that in the time frame of interest for the PID (4 dead times), the response resembles the ramp of an integrating process. For tuning and analysis it is useful to treat self-regulating processes with a time constant much larger than the total loop dead time as “near integrating.” The next post will detail the dynamic response of integrating processes and runaway processes. Determine the type of process response and choose the tuning rules accordingly. Make step changes to the process to identify the total dead time, open loop gain, and secondary time constant. If the process decelerates within the time frame of the major part of the PID response (four dead times after start of response), the process can be classified as self-regulating. For this process response, a primary time constant needs to also be identified and self-regulating process tuning rules are used taking care to use PID structures and tuning to minimize large abrupt changes in the controller output. In these processes, there may be no need to overdrive the PID output past its final resting value.

The post Getting Automation Started by Selection of Control Action and Modes first appeared on the ISA Interchange blog site. The following insights are part of an occasional series authored by Greg McMillan , industry consultant, author of numerous process control books and a retired Senior Fellow from Monsanto. Pulp and paper plant studies found 75 percent of the loops caused more variability in the automatic mode than in the manual mode. A third of them oscillated as a result of nonlinearities such as valve dead band. Another third oscillated because of poor controller tuning. The remaining loops oscillated because of deficiencies in the control strategy. A well-designed control loop with proper tuning and a responsive control valve can minimize this variability. Because this means you can operate closer to constraints, good tuning can translate into increased production and profitability. Actions Speak Louder than Words The very first settings that must be right are the controller and valve actions. If these actions are not right, nothing else matters. The controller output will run off scale in the wrong direction regardless of the tuning settings. The controller action sets the direction of a change in controller output from its proportional mode for every change in the controller’s process variable (feedback measurement). If you choose direct action, an increase in process variable ( PV ) measurement will cause an increase in controller output that is proportional to its gain setting. Since the controller action must be the opposite of process action to provide feedback correction, you should use a direct-acting controller for a reverse-acting process except as noted later in this guide. Correspondingly, you should select reverse control action for a direct acting process so an increase in process variable measurement will cause a decrease in controller output that is proportional to its gain setting, except as noted later. A direct -acting process is one in which the direction of the change in the process variable is the same as the direction of the change in the manipulated variable. A reverse -acting process is one in which the direction of the change in the process variable is opposite the direction of the change in the manipulated variable. The manipulated variable is most frequently the flow through a control valve, but it can also be the set point of a slave loop for a cascade control system or variable speed drive. The valve action sets the display. For example, it determines whether a 100 percent output signal corresponds to a wide open or a fully closed valve. It also determines the direction of a change in the actual signal to the control valve when there is a change in the controller’s output. In some analog controllers developed in the 1970s, such as the Fisher AC 2 , the valve action affected only the display of the valve signal, not the actual signal. To compensate for this lack of signal reversal for a reverse-acting valve (i.e., an increase-to-close or fail-open valve), the control action had to be the opposite of the action that would normally be appropriate based on process action alone. Fortunately, the valve action corrects both the display and the actual valve signal in modern controllers, so the control action can be based solely on process action. However, the user should verify this before commissioning any loops. In control systems that use fieldbus blocks, the valve action should be set in the analog output (AO) block rather than in the PID controller block. This ensures that the “back-calculate” feature is operational for any function blocks (split range, characterization, and signal selection) that are connected between the PID and AO blocks. The signal can also be reversed in the current-to-pneumatic transducer (I/P) or in the positioner for a control valve. Before the advent of the smart positioner, it was preferable for the sake of visibility and maintainability that any reversal be done in the control room rather than at the valve. Reversal of signal in the control room is still advisable to prevent the wrong fail action for a scenario where you lose the field signal to the positioner but not the power to the positioner. In this case, the positioner could possibly reverse the signal to the actuator causing an “Increase-Close” valve to close. It is important to standardize on the location of the signal reversal to ensure that it is done and done only once. Table 1 summarizes how the controller action depends upon both the process and valve actions and on the signal reversal. Table 1: Controller Action   Which brings us to a rule of thumb: The controller action should be the opposite of the process action unless there is an increase-to-close (fail-open) control valve for which there is no reversal of the valve signal. This means that you should use reverse and direct-acting controllers for direct and reverse-acting processes, respectively. The valve signal can be reversed for a fail-open valve at many places, but it is best done in the AO block of the control system. Controller à la Mode The names for the operational modes of the PID vary from manufacturer to manufacturer. Thus, the Foundation™ Fieldbus modes listed next provide a uniformity that can be appreciated by all. Auto (automatic) – The operator locally sets the set point. PID action is active (closed loop). In older systems, this mode is also known as the local mode . Cas (cascade) – The set point comes from another loop. PID action is active (closed loop). This is also known in older systems as the remote mode or remote set point ( RSP ). LO (local override) – PID action is suspended. The controller output tracks an external signal to position the valve. This mode is typically used for auto tuning or to coordinate the loop with interlocks. In older systems, it is also known as output tracking . Man (manual) – The operator manually sets the output. PID action is suspended (open loop). IMan (initialization manual) – PID action is suspended because of an interruption in the forward path of the controller output. This is typically caused by a downstream block that is not in the cascade mode. The controller output is back-calculated to provide bumpless transfer. RCas (remote cascade) – The set point is remotely set, often by another computer. PID action is active (closed loop). This mode is also known in older systems as the supervisory mode. Nothing else matters if you don’t get the control action right. The control action depends upon knowing the process action and if the PID output signal is reversed in controller or in the valve’s current to pneumatic converter (I/P) or positioner for a fail-open (increase-to-close) valve. Knowing the choice and function of the various PID controller modes is also essential. See the recently extensively updated Good Tuning: A Pocket Guide 4th Edition for concise guidance on this topic and everything else you need to know to get the most out of your PID controller. Anyone setting up a PID controller must select the correct control action and enable the proper choices of control modes including the right normal mode and initial mode. To make sure the PID setup is correct, one must verify by field tests and conferring with instrument design and maintenance, operators, and the process engineer whether the process is reverse acting or direct acting and whether the valve is fail open (increase-to-close) or fail close (increase-to-open). One must also find out from the person who configured and tested the PID and control valve I/P and positioner if the valve signal is reversed anywhere. The signal should only be reversed for an increase-to-close valve and the reversal should be done at one point only. Provided there is no improper or missing signal reversal, the control action should be the opposite of the process action to give the negative feedback action necessary to achieve stability. In general, the manual mode should never be taken away from the operator because there are too many unforeseen possibilities and the operator is the last defense for safety and stability. There is nothing worse than an operator feeling powerless. The initial mode is chosen based on what mode is appropriate for a total download and startup. The normal mode is the mode desired for best performance. Secondary PID controllers must have the Cascade mode permitted in order for the primary PID controller to manipulate the secondary controller setpoint. For intelligently prepositioning of the PID output for sequences and to preemptively deal with situations and optimization of the output for setpoint changes, the remote output mode should be permitted. Finally, when PID setpoint is being intelligently adjusted or optimized (e.g., override control, valve position control, model predictive control, or supervisory control), the remote cascade mode is permitted.

The post, Understanding Terminology to Advance Yourself and the Automation Profession , first appear on the ControlGlobal.com Control Talk blog. The automation profession suffers from the lack of a common understanding of the terms used and their units. By taking a closer look at key terms for plant dynamics, we can have more intelligent discussions and better recognition of the contributing factors so that we can find what part of the loop is the leading cause of poor performance whether it be excessive nonlinearity, delay, or deception. I recommend every user get a copy of The Automation System and Instrumentation Dictionary 4 th Edition published by International Society of Automation (ISA). Here I give a perspective of the role played and offer some descriptive words preceding the common terms to create more useable and descriptive terms seen in my publications and presentations but not much elsewhere. Let’s start with general terms and once we have that under our belts using a Texas size belt buckle if necessary, move on to specific terms. You can also use your cowboy boots to kick all that bull you now see out of your trail to success. General Terms The 3 key general terms used for describing all types of dynamics are dead time, time constant, and gain. There are two types of time constants. In almost every case we are talking about a negative feedback time constant. An important and not well understood case for a positive feedback time constant is a runaway reactor with extraordinary performance and safety considerations. A system is nonlinear if the dead time, time constant, or gain changes. Nearly all loops are nonlinear. The most common linear loops are flow or level loops (constant diameter vessel) manipulating the speed of a pump with low static head. Composition, pH, pressure, and temperature loops are nonlinear. Dead time (delay) (sec) is the time after a change in PID output till there is a recognizable change in PID input. Dead time is detrimental wherever it exists because it prevents any recognition of an impending change in a process variable. For an unmeasured disturbance the peak and integrated errors are proportional to dead time and dead time squared, respectively. An increase in dead time makes a loop closer to instability unless retuned. Dead time slows down the setpoint response. Time constant (lag) (sec) is the time for a variable to reach 63% of the final value once the variable starts to change (after the dead time has expired). For runaway processes it is a positive feedback time constant that is the time for a variable to reach 172% of the process value predicted by process gain. A time constant in the process is beneficial in terms of slowing down the effect of an unmeasured disturbance upstream of the time constant (most prevalent case is a disturbance at process input instead of process output as shown in control theory text books). A time constant in the automation system or path of the manipulated flow to the point of entry of the disturbance slows down the recognition and correction. A time constant anywhere slows down the setpoint response. An increase in the largest time constant can reduce the peak and integrated errors observed if the PID is retuned. If the time constant is in the measurement instead of the process, the actual process errors are greater than the observed errors due to filtering action. An increase in any of smaller time constants increase the dead time in a first order plus dead time (single time constant and dead time) approximation. Gain is the change in output divided by the change in input to a part of the process or component in the automation system. The units depend upon the part or component. For the PID whose internal algorithm uses % inputs and % outputs, the gains used to reflect what the plant does and how the PID reacts are dimensionless (%/%). For a steady state response, the gain is a final change in output divided by a given change in the input. The gain here is the slope on the plot of the output versus the input. For an integrating response, the gain is a ramp rate divided by the given change. For the PID, an integrating gain reflecting the response of a process is 1/sec (%/sec per %). A high control valve gain (steep slope of installed characteristic) can magnify signal error, deadband, and resolution or threshold sensitivity limits. A low control valve gain (flat slope of the installed characteristic) can cause wandering and excessive wear of the valve. A high process gain is beneficial in terms of increasing sensitivity to detection of disturbances but if too high like in pH can cause excessive variability from valve deadband or resolution or threshold sensitivity limits and measurement noise. In general a high measurement gain is desirable because it corresponds to a smaller span and better accuracy when error is % of span. Specific Terms Total loop dead time (total delay) (sec) is the sum of all the sources of dead time in the loop that can originate in the process and automation system including the PID controller (termed the process dead time in most of the literature). A fraction of all time constants smaller than the largest are converted to dead time. The fraction approaches one as the ratio of the small to large time constant approaches zero. For practical purposes, small time constants are simply summed with all pure delays to estimate the total loop dead time since we usually underestimate the total loop dead time. Dead times from transportation delays are inversely proportional to flow and can dramatically increase at low production rates. The delay from discontinuous devices is ½ the cycle time plus the latency (time after that start of the cycle until the result is available). Thus, the delay from digital controllers is ½ the module execution time and for digital transmitters is ½ the update time. For chromatographs and most at-line analyzers the delay is 1.5 times the cycle time plus any delay from sample transportation and multiplexing. Primary time constant (primary lag) (sec) is the largest time constant in the loop. Hopefully it is in the process but it can be in the automation system (termed the open loop time constant in my publications and presentations but is termed process time constant in most of the control literature). This time constant is defined by making a test with the controller in manual (open loop). Secondary time constant (secondary lag) (sec) is the second largest time constant in the loop. Hopefully it is in the process but it can be in the automation system. This time constant is defined by making a test with the controller in manual (open loop). For integrating processes (e.g. batch) and runaway processes (e.g. highly exothermic reactors), the secondary time constant has a huge effect that can be compensated by derivative action (PID rate time = secondary time constant). Open loop self-regulating (steady state) process gain (%/%) is the final change in the PID input in % divided by a fixed change in the PID output in % for a self-regulating process (termed the process gain in most of the control literature) with no disturbances. An open loop gain is the product of the manipulated variable gain (e.g. valve or variable speed drive gain), process gain, and measurement gain. For self-regulating processes the time units of a manipulated flow are cancelled by time units in the process gain. This is not the case for an open loop integrating process gain. Open loop integrating process gain (1/sec) is the ramp rate of the PID input in %/sec divided by a fixed change in the PID output in %. In the control literature, the integrating process gain may have engineering units because the academic is looking only at the process. The integrating process gain is the product of the manipulated variable gain (e.g. valve or variable speed drive gain), process gain, and measurement gain. The units of 1/sec result from time units from the manipulated flow not cancelled out by the units in the process gain. For level the manipulated gain of kg/sec per % PID output are not cancelled by the process gain units of meters level per kg or the measurement gain units of % PID input per meters level. Manipulated variable gain (e.u./%) is the change in manipulated variable per % change in PID output. For loops directly affecting the process, the manipulated variable is flow. The manipulated process input is almost always flow (exceptions are power for electrical heaters or speed for agitators). Thus, the manipulated variable gain is the slope of the installed characteristic of the valve, damper, guide vanes or variable speed pump, fan, or compressor. Here, I use the more specific term valve gain or variable speed drive gain in my publications and presentations. For upper level loops in cascade control or model predictive control manipulating the setpoint of the secondary loop directly affecting the process, the manipulated variable gain takes on the units of the process variable of the secondary loop. Here the manipulated variable gain is the secondary PID process variable scale span in engineering units divided by 100% PID output. Thus a change in secondary PID measurement calibration, valve trim and valve to system pressure drop ratio or variable speed drive capacity and static head will affect tuning. Measurement gain (%/e.u.) is the change in % change in PID input for a change in the process variable in engineering units. Since nearly all smart measurements are linear, the measurement gain is 100% divided by the PID process variable scale span in engineering units. Thus a change in secondary PID measurement calibration will affect tuning. Process gain (e.u./e.u.) is the change in the process variable in engineering units divided by the change in the manipulated variable in engineering units. The most common manipulated variable is flow. The process gain for continuous composition and temperature control is the slope of a plot of the change in process variable versus a change in the ratio of the manipulated flow to feed flow. Since the slope is nonlinear often getting steeper at low ratios and the required abscissa being a ratio to feed flow, the process gain tends to increase dramatically at low production rates. Response Time (sec) is meaningless unless the % of final response is specified. It is used to quantify the dynamic response of an automation system component or steady state process. It cannot be used to describe an integrating or runaway process. 63% Response Time (T 63 ) is the time to reach 63% of the final response (total loop dead time plus primary time constant in a first order plus dead time model). 86% Response Time (T 63 ) is the time to reach 86% of the final response (total loop dead time plus 2 times the primary time constant). 95% Response Time (T 95 ) is the time to reach 86% of the final response (total loop dead time plus 3 times the primary time constant). 98% Response Time (T 98 ) is the time to reach 98% of the final response (total loop dead time plus 4 times the primary time constant). Finally, take heart that while sometimes like this last week for me you may be called the “tail wagging the dog”, by more intelligent discussion you can graduate from the tail of the dog to the eye of the dog.

The post How to Achieve Tight Process Recirculation Temperature Control first appeared on the ISA Interchange blog site. The following insights are part of an occasional series authored by Greg McMillan , industry consultant, author of numerous process control books and a retired Senior Fellow from Monsanto. Editor’s Note : This is Part 4 of a four-part blog series on smart automation of vessel heating and cooling. Click these links to read Part 1 or Part 2 or Part 3 . In this post I look at how the confusing unusual dynamics of a heat exchangers temperature response in a recirculation line can be addressed by tuning and by the manipulation of a bypass flow. A simple additional PID is also offered to reduce heat exchanger fouling. Heat exchangers in a process recirculation stream are used for vessel temperature control. The high velocities on the process side in the exchanger increase the heat transfer coefficient and reduce fouling. For cascade control the vessel temperature PID output is the setpoint of the exchanger outlet temperature (Figure 1). For the exchanger to do its job the recirculation flow must be high and a wide range of returning process temperature must be tolerated. Figure 1 – Vessel process recirculation heat exchangers have a two stage temperature response that can confuse open loop tuning tests and users. The response of the exchanger outlet temperature has a two stage response. The first stage is a response relatively fast self-regulating response of the exchanger. The second stage response is a protracted integrating response from the recycle effect of the process fluid in the vessel changing temperature. For large vessel volumes, the second stage response ramp rate is slow enough to allow fast exchanger temperature loop tuning based on the initial first stage response. Open loop tests that wait on the exchanger outlet temperature to settle out at a new operating point may be confused by the ramping from the second stage response. Practitioners may inadvertently tune for the second stage response rather than concentrating on having the PID quickly deal with the first stage response. Insight: The goal is to make initial self-regulating first stage response much faster than the protracted integrating second stage response of the process recirculation temperature control loop. The speed of response of control of heat exchanger outlet temperature can be significantly increased by throttling a bypass around the exchanger and keeping the utility flow constant (Figure 2). This mode of operation bypasses the thermal lag of the heat exchanger. The response is as fast as the blending of the streams bypassing and going through the exchanger. Due to the change in heat transfer coefficient with velocity, there is an additional response from the change in heat transfer. PID tuning for the initial response and manipulation of a heat exchanger bypass including a valve position controller to optimize the bypass valve position can eliminate the poor control from getting into the slow integrating process response and heat exchanger fouling and frosting. See Greg McMillan’s new ISA book Advances in Reactor Measurement and Control for an extensive view of practical opportunities for designing control strategies to achieve product quality and maximize yield and capacity in different types of fermenters, bioreactors, and chemical reactors. Consider the response of the heat exchanger with cold water. An increase in cooling demand will cause a decrease in bypass flow and an immediate decrease in exchanger outlet temperature. The higher velocity through the exchanger will increase the heat transfer rate to the chilled water, making a further, slower reduction in temperature. If the measurement and the valve are fast enough, the PID can be tuned for a faster rate of response from blending providing further separation between the first stage self-regulating response and the second stage integrating response. Process recirculation heat exchanger bypass control provides a much faster secondary loop for vessel temperature control. Insight: The response of a process recirculation heat exchanger outlet temperature can be significantly faster by throttling a bypass around the exchanger and keeping the utility flow constant. Faster tuning of the secondary loop can compensate for utility temperature and pressure disturbances before they affect the vessel temperature loop and can help to prevent the protracted integrating response from recirculation. A faster secondary loop also makes the primary loop ultimate period faster, enabling more aggressive vessel temperature control and better compensation of process disturbances (e.g. feed temperature and concentration). Insight: Faster tuning of the secondary temperature loop make both secondary loop and the primary vessel loop faster and eliminates the complications of the protracted integrating response from recirculation. The throttling of a heat exchanger bypass enhances this opportunity. Figure 2 – The throttling of a bypass flow around the process heat exchanger provides a much faster initial response that provides separation of the first stage self-regulating and the second stage integrating responses. At low cooling or heating requirements, the hot or cold utility valve can be throttled to extend the rangeability of the exchanger. A valve position controller (VPC) can be used to prevent the bypass valve from going wide open by cutting back on the utility flow (Figure 3). A VPC can increase the turndown of the heat exchanger bypass control and prevent the process velocity through the exchanger from going below a minimum, thus reducing the degree of fouling and frosting. This minimum flow also keeps the process side heat transfer coefficient from dropping too low. A low output limit on the VPC can be used to prevent fouling and deterioration of the heat transfer coefficient on the utility side. The VPC should be the enhanced PID to minimize interactions by enabling the use of directional move suppression and eliminating the limit cycles from valve backlash. Insight: A valve position controller can increase the turndown and efficiency of heat exchanger bypass control. Figure 3 – A valve position controller can reduce fouling on the process side and extend the turndown of the temperature loop. Tune the PID for a fast reaction to the initial self-regulating response of the heat exchanger in a process recirculation stream. Manipulate a heat exchanger bypass flow to eliminate the heat transfer lags of the heat exchanger giving a much faster response from a simple blend of bypass and heat exchanger process flows. Use a valve position controller to push a utility valve to a minimum throttle position increasing the velocity of the process through the exchanger, decreasing the fouling on the process side.