pid-loop-tuning-tips-pocket-guide.pdf

PID Loop Tuning Pocket Guide Version 3 Printing Notes for 12-page booklet: 1. Print one single-sided copy of this this PDF file. You will get this page of instructions plus six 8-1/2 x 11 sheets. 2. The six sheets should be arranged as follows: 

1st sheet, side 1 = spread of first & last pages (pages 12 & 1) Reverse side of 1st sheet = spread of pages 2 & 11



2nd sheet, side 1 = spread of pages 10 & 3 Reverse side of 2nd sheet = spread of pages 4 & 9

LOOP OPTIMIZATION SOFTWARE

 

INTUNE+ Loop Optimization Software is a plant management and process control predictive maintenance tool that combines process control monitoring with powerful diagnostic and PID tuning and optimization tools, including:  Comprehensive performance analysis Customized diagnostics for plant-specific indices  Web-based reporting and customized event notification Automatic, adaptive, model-based, and offline tuning

INTUNE+ can report on any Key Performance Indicator (KPI) an end user requires. You can champion your process control strategy based on: energy reduction; increasing productivity; more consistent product quality; increasing plant engineering productivity; and supporting continuous improvement.

PID Loop Tuning Pocket Guide Version 3

With INTUNE+, you can finally complete your process control predictive maintenance program by going beyond the device level. Differentiate between hardware and tuning problems with functional, flexible, and customizable tools that truly put you in control. This revolutionary new software is an integral part of ControlSoft’s PROCESS NAVIGATOR. INTUNE+ software is the foundation of ControlSoft’s partnership to improve plant operations, maintenance, and engineering productivity.

INTUNE Loop Optimization Software 2006 Engineers’ Choice Product of the Year in the process control category

INTUNE T UNING T OOLS To help correct loop quality problems, INTUNE provides robust tuning tools including its auto tuning, heuristic-based adaptive tuning, offline data tuning, and offline model-based tuning. Also included is a PID comparison tool, along with a variety of online and offline process and PID simulation features.

EMAIL [email protected] WEB www.controlsoftinc.com PHONE 440-443-3900 FAX 440-443-0249 5387 Avion Park Drive

To learn more about these powerful problem correction tools, request a copy of SL 106. Visit www.controlsoftinc.com or phone 440-443-3900.

Highland Heights OH 44143

DS405-03/08 -- www.controlsoftinc.com


Page 2 -- PID Loop Tuning Pocket Guide, Version 3

PID Loop Tuning Pocket Guide, Version 3 -- Page 11

CONTENTS

page 3 Description of Processes

MANTRA is a state-of-the-art advanced control system. It is a fully functional, configurable control system with more than 90 function blocks, configurable faceplates, and very easy-to-use, intuitive, drag-and-drop type of function block programming.

page 4 Description of PID Units

Advanced control functions in MANTRA include our patented model-based control suite—IMC (Internal Model Control), CC (Coordinated Control), and MMC (Modular Multivariable Control).

page 5 Closed Loop Tuning Procedure pages 6-7 Open Loop Tuning Procedure pages 8-10 Reference List of Common Controllers page 11 MANTRA & Services & Training page 12 INTUNE+ Loop Optimization Software & INTUNE Tuning Tools

MANTRA allows users to build all necessary logic on top of the modelpredictive control functions to take care of any operation or control needs. It also allows users to switch the control back and forth between MANTRA and their existing control system. Process applications where MANTRA can be used Chemical Reactors Pasteurization Extruders Autoclaves / Retorts Bio-Reactors Cooking Stations Fermentation Gas Purification Sterilizers pH & ORP Chlorine Injection CONSULTING SERVICES & LOOP TUNING

Our experienced controls engineering professionals perform control loop diagnostics and loop performance optimization services. Our process control experts observe and analyze the relevant processes in a plant and then recommend changes to improve control system performance. ADVANCED CONTROLS ENGINEERING

We have developed and applied advanced process control strategies since 1985. Our advanced process control services typically reduce process loop variation well within the desired limits, often exceeding the expectations of our clients. TRAINING: PID LOOP TUNING & PROCESS CONTROL STRATEGIES

Designed for engineers and technicians, classes cover basic terminology and building blocks of process control as well as fundamentals and advanced process control techniques using our fully automated, awardwinning control solutions.  Each class is taught by instructors with many years of experience in the industry, to insure maximum content.  Professionals who have attended these classes have given us rave reviews on both technical content and method. 


include: Ore Processing Inline Blending Glass Furnaces Dryers & Evaporators Kilns

Class outlines at www.controlsoftinc.com or phone 440-443-3900.




continued from page 9

DESCRIPTION OF PROCESSES MANTRA PID B LOCKS PID Block Series P Gain (Gain * 1) I Time Constant (Sec * 1) D Time Constant (Sec * 1) PID Block Parallel P Gain (Gain * 1) I Reset Rate (Repeats/sec * 1) D Gain (Gain * 1) PID Block Non-Interacting P Gain (Gain * 1) I Time Constant (Sec * 1) D Time Constant (Sec * 1)

M ODICON PID BLOCKS KPID Block P Gain (Gain * 1) I Time Constant (Sec * 1) D Time Constant (Sec * 1) PID Block P Gain (Gain * 1) I Time Constant (Sec * 1) D Time Constant (Sec * 1) PID2 Block P Band (Band * 1) I Reset Rate (Repeats/min * 100) D Time Constant (Min * 100)

S IEMENS PID BLOCKS TI PID Block P Gain (Gain * 1) I Time Constant (Min * 1) D Time Constant (Min * 1) S7 controller C non-interacting P Gain (Gain * 1) I Time Constant (Sec * 1) D Time Constant (Sec * 1)

S QUARE D PID BLOCKS PID Block P Gain (Gain * 100) I Reset Rate (Repeats/min * 100) D Time Constant (Min * 6000)

W ESTINGHOUSE PID BLOCKS WFDP PID Block P Gain (Gain * 100) I Time Constant (Sec * 1) D Gain (Gain * 1) Ovation PID Block P Gain (Gain *1) I Time Constant (Sec *1) D Gain (Gain *1)

Fast Loops (flow, pressure) P - Little [Too much will cause cycling] I - More D - Not needed Slow Loops (temperature) P - More I - Some [Too much will cause cycling] D - Some Integrating (level & insulated temp) P - More I - Little [Will cause cycling] D- Must [If D is not used, the loop will cycle.] Noisy Loops (any PID loop where the measurement is constantly changing) P - Low [Will cause cycling] I - Most [Accumulated error] D - Off [Will cause cycling]

O MRON PID BLOCKS PID Block P Gain (Gain * 1) I Time Constant (Sec * 10) D Time Constant (Sec * 10)





DESCRIPTION OF PID UNITS

Proportional Term is the amount added to the output based on the current error. Proportional Gain is a Multiplier. If the error is 10 and the Gain is .8, then output will move 8%. Proportional Band is a Divider as a percentage. If the error is 10 and the Band is 125%, then output is (10*(100/125))=8%.

B AILEY I NFI 90 PID BLOCKS 156,0 PID Block P Gain (Gain *1) I Reset Rate (Repeats/min * 1) D Time Constant (Min * 1) 156,1 PID Block P Gain (Gain * 1) I Reset Rate (Repeats/min * 1) D Time Constant (Min * 1)

Conversion between P-Gain and P-Band: P-Band = 100/P-Gain

156,2 PID Block P Gain (Gain * 1) I Reset Rate (Repeats/min * 1) D Time Constant (Min * 1)

Integral Term is the amount added to the output based on the sum of the error.

156,3 PID Block P Gain (Gain * 1) I Time Constant (Min * 1) D Time Constant (Min * 1)

Time Constant is the time for one full repeat of P-Term. If the P-Term is 8% and the Time Constant is 10 seconds, then the output will ramp up 8% every 10 seconds. Reset Rate is amount the output will move in one second. If the P-Term is 8% and the Reset Rate is .1 repeat/sec, then the output will move .1*8 every second and take 10 seconds for the full repeat of the P-Term of 8%. Integral Gain is the same as Reset Rate multiplied by P-Gain. Conversion between Time and Reset Rate: Reset Rate = 1/ Time Constant I-Gain = (1/Time Constant)*P-Gain Derivative Term is the amount subtracted from the output based on the rate of change of the error.

F OXBORO PID BLOCKS I/A Series PID, PIDA, PIDE, PIDX, PIDXE P Band (Band * 1) I Time Constant (Min * 1) D Time Constant (Min * 1)

GE PID BLOCKS Series 90/30/70 PID Block ISA P Gain (Gain * 100) I Reset Rate (Repeats/sec * 1000) D Time Constant (Sec * 100) Series 90/30/70 PID Block Independent P Gain (Gain * 100) I Reset Rate (Repeats/sec * 1000) D Gain (Gain * 100) RX3i/RX7i PID Blocks PID Block Parallel P Gain (Gain * 1) I Reset Rate (Repeats/Min * 1) D Gain (Gain * 1) RX3i/RX7i PID Blocks PID Block Series P Gain (Gain * 1) I Reset Rate (Repeats/Min * 1) D Time Constant (Min * 1) RX3i/RX7i PID Blocks ADV_PID Block Parallel P Gain (Gain * 1) I Reset Rate (Repeats/Min * 1) D Gain (Gain * 1) RX3i/RX7i PID Blocks ADV_PID Block Series P Gain (Gain * 1) I Reset Rate (Repeats/Min * 1) D Time Constant (Min * 1)

continued on page 10

Time Constant is the amount of time the controller will look forward. Derivative Gain is the amount of time the controller looks forward multiplied by the P-Gain.





REFERENCE

CLOSED LOOP TUNING PROCEDURE

Following is a summary of many of the more common controllers available on the market.

CLOSED LOOP STEP 1: KNOW THE PROCESS

For units of other controllers, please contact ControlSoft (see page 1 for contact information). AB PID BLOCKS PLC5 Integer Block ISA P Gain (Gain * 100) I Time Constant (Min * 100) D Time Constant (Min * 100) PLC5 Integer Block Independent Gains P Gain (Gain * 100) I Reset Rate (Repeats/sec * 1000) D Gain (Gain * 100) PLC5 PD Block ISA P (Ga in *1 ) I Time Constant (Min * 1) D Time Constant (Min * 1) PLC5 PD Block Independent Gains P Gain (Gain * 1) I Reset Rate (Repeats/sec * 1) D Gain (Gain * 1) PLC5 PCO Block Dependent Gains P Gain (Gain * 1) I Time Constant (Min * 1) D Time Constant (Min * 1) PLC5 PCO Block Independent Gains P Gain (Gain * 1) I Reset Rate (Repeats/sec * 1) D Gain (Gain * 1)

SLC PID Block with RG bit on P Gain (Gain *100) I Time Constant (Min * 100) D Time Constant (Min * 100) SLC PID Block with RG bit off P Gain (Gain *10) I Time Constant (Min * 10) D Time Constant (Min * 100) ControlLogix PID Block Independent P Gain (Gain * 1) I Reset Rate (sec * 1) D Gain (Gain * 1) ControlLogix PID Block Dependent P Gain (Gain * 1) I Time Constant (Min * 1) D Time Constant (Min * 1) ControlLogix PIDE Block Independent P Gain (Gain * 1) I Reset Rate (Min * 1) D Gain (Gain * 1) ControlLogix PIDE Block Dependent P Gain (Gain * 1) I Time Constant (Min * 1) D Time Constant (Min * 1)


Identify the loop you intend to tune and determine the speed of the loop. A rough categorization is as follows: Fast loop has response time from less than one second to about 10 seconds, such as a flow loop. Use of PI controller is sufficient. Medium loop has response time of several seconds up to about 30 seconds, such as flow, temperature, and pressure. Use either PI or PID controller Slow loop has response time of more than 30 seconds, such as many temperature loops and level loops. Use of PID controller is recommended.

CLOSED LOOP STEP 2: KNOW THE CONTROLLER Identify the units of your PID controller, Proportional term (P-Term) is either a Proportional Gain (P-Gain) or Proportional Band (P-Band). Integral term (I-Term) can be a Time Constant (in minutes or seconds), Reset Rate (1/ second or 1/minute), or I-Gain (Reset rate * Proportional gain). Derivative term (D-Term) can be Time Constant (in seconds or minutes) or Derivative Gain (derivative time constant * proportional gain). In this document, Proportional Gain, Integral Reset Rate, and Derivative Gain are assumed. Convert back to your controller units if necessary. The Reference section is a summary of many of the more common controllers available on the market. For units of other controllers, please contact ControlSoft.

CLOSED LOOP STEP 3: WATCH THE RESPONSE Make a small change of setpoint (say 5%) or wait for a disturbance in the process. Then watch for process variable (PV) and control output (CO) responses. — If no visible instantaneous change of CO upon the change of setpoint or no apparent overshoot (over damped), increase your proportional gain by 50%. — If the PV is unstable or has sustained oscillation, with overshoot greater than 25%, reduce Proportional Gain by 50% and reduce Integral Reset Rate by 50%. — If PV oscillation persists with tolerable overshoot, redu ce Proportional Gain by 20% and reduce Integral Reset Rate by 50%. — If 3 or more consecutive peaks occur upon the change of setpoint, reduce Integral Reset Rate by 30% and increase Derivative Gain by 50%. — If PV stays fairly flat and below (or above) the setpoint for a long time, after change of setpoint or beginning of disturbance (long tail scenario), increase Integral Reset Rate by 100%. Repeat step 3 until the closed-loop response is satisfactory to you.




OPEN LOOP TUNING PROCEDURE OPEN LOOP STEP 1:

OPEN LOOP STEP 3:

KNOW THE PROCESS

OPEN LOOP TEST FOR NON-INTEGRATING LOOPS

Identify the loop you intend to tune and determine the speed of the loop. A rough categorization is as follows: Fast loop has response time from less than one second to about 10 seconds, such as a flow loop. Use of PI controller is sufficient. Medium loop has response time of sever al seconds up to about 30 seconds, such as flow, temperature, and pressure. Use either PI or PID controller Slow loop has response time of more than 30 seconds, such as many temperature loops and level loops. Use of PID

a. Put loop in manual control mode, keep control o utput constant, and wait for process to stabilize.

controller is recommended.

b. Make a small step change on control output (say 5 or 10%), and watch the response. c. Estimate process model according to Figure 1: --Model gain = PV change / CO change. --Deadtime = time lapse between change of CO and observable changes on PV. --Time constant = the time it takes for PV to reach about 63% of total changes.

d. Select initial PID values such that P = 2 / Model gain I = (Deadtime + Time constant) D = (Deadtime / 3) or (Time constant / 6) For slow loop, select whichever is greater. For fast loop, select smaller value. e. These initial PID values should give you a reasonable closed-loop response. Fine-tune the controller by using the closed-loop tuning method (see step 3 on page 5) to get your desired response.

OPEN LOOP STEP 2: KNOW THE CONTROLLER Identify the units of your PID controller, Proportional term (P-Term) is either a Proportional Gain (PGain) or Proportional Band (P-Band). Integral term (I-Term) can be a Time Constant (in minutes or seconds), Reset Rate (1/second or 1/minute), or I-Gain (Reset rate * Proportional gain). Derivative term (D-Term) can be Time Constant (in seconds or minutes) or Derivative Gain (derivative time constant * proportional gain). In this document, Proportional Gain, Integral Time Constant (in seconds), and Derivative Time Constant (in seconds) are assumed. Convert back to your controller units if necessary. The Reference section (starting on page 8) is a summary of many of the more common controllers available on the market. For units of other controllers, please contact ControlSoft.

Controller Output (CO) Process Input

CO Change

Time

Deadtime

Time Constant

Process Value (PV) Process Output

63% of PV Change PV Change

Noise Band

Figure 1. First Order Non-Integrating Model Calculation

Time

Gain = PV_Change

/ CO_Change

pid-loop-tuning-tips-pocket-guide.pdf

Recommend Documents