CCB 3072PROCESS INSTRUMENTION AND CONTROL LABMAY 2014TITLE: RATIO CONTROL

GROUP: 7

AZHAR BIN ZAWAWI 0015596

MOHAMMAD FAIZAL BIN SUHAIMI 0015675

MOHAMMAD AZIM TARMIZI BIN YA'CCOB 0015403

NUR ASIAH BINTI MOHD FAUZI 0015369

NUR FATIN DARIAH BINTI MOHAMAD DAUD 0015668

MEMBERS:

LAB INSTRUCTOR : FAEZAH ISA

DATE OF EXPERIMENT: 26TH JUNE 2014DATE OF SUBMISSION: 3RD JULY 2014

OBJECTIVE(i)To demonstrate the characteristic of Proportional Band, Integral Action and Derivative Action on a flow process control loop.(ii)To demonstrate the characteristic of ratio control.INTRODUCTIONThe experimental is designed for the ratio control in relation to a single loop flow control. The tank, pump instrumentation and valves are strategically located for easy access. For safety cause, the control panel shall be protected against water splashes. The process piping shall be made of industrial pipes.THEORYOn/OffOn/Off control generally both the simplest and the least expensive type of process control and has wide application in industry. A process controlled by an on/off controller almost always has some error in it, in fact the controller turns on or off only at those times there is no error in the measurement, when the measurement crosses the set point on its way from one extreme error to another. The valve goes either fully open or closed depending on the direction of the error.No attempt is made to balanced the inflow with the outflow. The energy or material supplied to the process is always either too much or not enough. The measured variable cycles continuously.When on/off control is applied to the right type of process, the effect of the cycling is small and acceptable. On/off controller best applied to a large capacity process that has relatively little dead time and small mass or energy inflow with respect to the capacity of the system.The cycling is illustrated in Figure 1 which shows the relationship between the temperature and the action of the manipulated variable. A typical application for on/off control is the temperature of a large tank or bath.

Figure 1 System Response to a Process Upset with ON/OFF ControlPIDEach of the three basic control modes and the combinations discussed so far, proportional (P), proportional plus integral (PI) have limitations which may not be significant if the process and controller are carefully matched. However some processes are so difficult to control or so critical.Loop tuningThe closed loop control system attempts to achieve a balance between supply and demand by comparing the controlled variable to the set point and regulating the supply to an amount which will maintain the desired balance. Tuning the controller adjusts it so it can achieve that balance as quickly as possible. This is done when instrument is first put in service and later on a periodic basis as part of preventive maintenance. When tuning remember that each controller is part of a closed loop. All the parts of the loop are interactive, behaviour of other devices in that loop. The controller response must be matched to that of the process. There are several procedures for doing this, some mathematical most using trial and error.A simple three step method for tuning most three mode controllers follows. Batch controllers and one through processes are special cases discussed after the three mode and two mode controllers) . This three steps procedure is based on a simple test to determine the nature period of oscillation of the process.Step 1 : Set the integral time of the controller at its maximum and the derivative time at its minimum, thereby providing proportional only control. Then reduce the proportional band until oscillation begins. Measure the period of this oscillation (also called the natural period) as the time between two Successive crests or valleys (Figure 2).

Figure 2 Period of oscillation with proportional only controller after first tuning stepStep 2 : Set the derivative time at 0.15 times the natural period and the integral time at 0.4 times the natural period. Observe the new period of oscillation there should be a 25 percent decrease(Figure 3). If the new period of oscillation is shorter than this reduce the derivative time, if period is longer, increase the integral time.

Figure 3 Period of oscillation for correctly tuned PID controller after second tuning stepStep 3: Finally the proportional band to achieve the desired degree of damping (the amount of correction to a process upset which when too much or too little shows up as either overshoot or sluggishness respectively)When adjusting a two mode PI controller a slightly different method should be used since integral mode introduces phase lag that is not counteracted by derivative. The procedure follows:Step 1 : Set the integral time of the two mode controller at its maximum and the derivative time at its minimum, providing proportional only control just as with the three mode controller. Then reduce the proportional band until oscillation begins and measure this period.Step 2 : Set the integral time to the natural period. The period of oscillation should increase about 40 percent (ideally 43%). If the period is longer than this, increase the integral time (Figure 4)

Figure 4 Period of oscillation for correctly tuned PI controller after second tuning step

Step 3: Finally adjust the desired degree of damping is achieved. Adding integral will always increase the proportional band required for stable control.Some consideration must be given to processes with variable dynamic characteristics. Once through processes such as tubular heat exchangers exhibit a natural period that varies inversely with flow in such situations. One combination of controller settings cannot be ideal for all flow rates. Integral time should be set according to the lowest anticipated flow rate and the derivative of time accordingly to the highest.Some batch controllers because of their mechanical arrangement will become unstable if equal values of integral and derivative time are used. Always keep their integral time at least twice the derivative time.

Ratio ControlRatio control provides a means a blending two or more variables in adjustable proportions to obtain a desire mixture. The measurement of the load (wild or uncontrolled flow) is the set point to the ratio controller which adjusts the flow of the controlled variable. A preset ratio regulates the flow of the controlled variable for example if the ratio is 2 to 1 for every gallon of the uncontrolled variable flowing, two galloons of the controlled variable is allowed to flow.Ratio control is applied to processes where the flow of one variable fluctuates but the desired blend can be maintained satisfactorily by adjusting the related variable proportionally.Ratio control can be accomplished by taking the signal from the uncontrolled flow as the set point for the ratio controller. This value is multiplied by an adjustable factor (the ratio setting) . The measurement to the controller is the flow of the controlled variable.The ratio can be set to a desired value by a calibrated dial on the controller. The ratio concerns only the flow signals, not the actual amount of the flows, since the measurement ranges of the two flow are not necessarily the same.The ratio controller can be used with any combination of suitably related process variables. Control action is usually proportional plus integral. The response of a system with ratio control to process upsets is the same as the response of the basic control mode used.

Figure 5 :Single Ratio Control System

PROCEDUREA. PID Flow Control:1. The values: PB = 200, I = 6 s and D = 1 s are entered.2. The control loop is put into manual mode and the set point is adjusted to 50 LPM.3. The output is tuned gradually so that the measurement matches the set point of 50 LPM flow.4. The recorder is turned on then the control loop is put into auto mode.5. Load changes are stimulated by closing HV537 for 3 seconds and then it is returned to its original position.6. Once the measurement stabilize, the recorder is turned off. The control loop is put back into manual mode.7. The output is tuned gradually so that the temperature measurement matches the set point of 50 LPM.8. The recorder is turned on and then the control loop is put into auto mode.9. Set point changes are stimulated by increasing the set point to 75 LPM.10. Once the measurement stabilize, the recorder is turned off. The control loop is put back into manual mode.11. The previous PB and I values are retained. The control system specifications for this equipment accept D values in the range of 1-10. Based on the engineering knowledge that we have acquired in CCB3013, we have selected values of D which are 5s and 10s. Step 2 to step 10 are repeated.

B. PID Flow Control Loop Tuning1. The control loop is put into manual mode and the output is adjusted gradually so that temperature matches the set point of 50 LPM.2. The following values: PB = 1000, I = 1000 s and D = 0 s are entered.3. The recorder is turned on then the control loop is put into auto mode.4. Load changes are stimulated by closing HV537 for 3 seconds and then it is returned to its original position.5. The I and D values are maintained, FIVE (5) PB values in decreasingorder in the range between 1-1000 which are 550, 150, 15 and 4 are selected. Step 3 and step 4 are repeated with these reduced PB until the measurement oscillates about the set point.6. The natural period is determined.7. The PB value is maintained at which the measurement oscillates about the set point. I is set to natural period. Step 2 and 4 are repeated and the response is observed. A 40% decrease occurred in this period. If the new period of the oscillation is longer than this, the interval time is increased if the period is shorter decrease integral time.8. The PB is adjusted increasingly until the desire degree of damping is achieved.

C. Ratio Control1. The control loop is put into manual mode.2. Selector switch is used to select ratio control mode.3. FT520 is adjusted to about 40 LPM using HV533.4. The recorder is turned on. The control loop is put into Auto mode.5. A change is stimulated by adjusting FT520 to 50 LPM using HV533.

RESULTS 1 ) PID FLOW CONTROL

1.1.1.1 D=1s1.1.1.2 D=5s1.1.1.3 D=10s

2 ) PID FLOW CONTROL LOOP TUNING1.1.1.4 TB=1000s, 600s, 200s

1.1.1.5 TB=20s, 5s

3) RATIO CONTROL

DISCUSSION

PID FLOW CONTROL

For this experiment, we would like to determine the effect of the varied Derivative Action on the process stability. Process stability is occurred when the variable tallies with the prior set value. For this part there I two value that we have kept constant which is The Proportional Band has been set to 200s and Integral Action are set to constant 6s. Derivative Action is varying with changing at 1s, 5s and 10s. First set point was set to 50 LPM with Derivative Action initially at 1 second and disturbance is introduced for 3 seconds. After the graph has turn to normal response the response has been stopped and the set point has been change to 75 LPM. As we can see that the spreading of the graph for both 70 and 50 LPM is a little bit longer which show that the response of process to the disturbance and to change back to set point is a little bit late. As we increase the set point, the response curve is a little bit shorter which show quick response compare to 50 LPM. However as we increase the Derivative Action to 5s the responds is more faster compare to 1 s where the spreading of the response of curve is a little shorter which show the quick response to change back to set point. But as compare to 50 LPM and 70 LPM the response of the curve for 70 LPM is a little bit shorter. This result is same to first part where Derivative Action is 1 s.Hence as we increase to Derivative Action to 10s the result is better where the response is faster. Hence for this experiment, we can see that as we increase the Derivative Action, the response is better where the quick response will get. This can cause the process will not be effect with the disturbances and back to set value. This is due to the derivative action in a PID controller functions to ensure that the controller output proportionate the rate of change or error. We also can see that as we increase the set point, the effect of the derivative action to higher set point is quicker. It is useful when sudden changes in measured variable occur. This explains why with higher derivative action value, the graph reaches stability faster upon the introduction of disturbances.

PID FLOW CONTROL LOOP TUNING

For this experiment, we would like to determine the tuning value by setting the PB=1000s, 600s, 200s, 20s, and 5s. The integrative value and derivative value is set to 1000s and 0s respectively. As we running the experiment, from 1000s to 200s the graphs only produce a curve graph when the disturbance is introduced. This response show that we cannot use this value as it not produce sinusoidal curve for us to calculate the distance of peak-to-peat to calculate the natural period. But as we see when the PB is set from 20s to 5 s, the sinusoidal curve is produce. However both produce inconsistent sinusoidal where the tuning value is also cant be used in for this value to calculate the natural period as to calculate natural period, the consistent sinusoidal curve is needed to measure peak-to-peak value. However for our part as the 20s produce almost consistent graph, we use the response for PB=20s to calculate the natural period. The result is as shown below. Sample calculation is provided in Appendix.

Proportional band, PBD (mm)Natural period, T (s)

202.57.5

Table 1: Natural Period for PB=20s

Hence we conclude that the best tuning value for our experiment is when PB is 20s.

RATIO CONTROL

For this experiment we would like to determine response of the system when the set point is set manually. Ratio control provides a means a blending two or more variables in adjustable proportions to obtain a desire mixture. First we have set the flow rate to be 40 LPM manually. As we waited for the response of the process, the value show in the board does not come to 40 LPM where large fluctuation occurs. Same result occurs when we set the value to 50 LPM. This because the tuning parameter in PID controller is not good where it cannot cause the process to response as same the flow rate that has been set manually.

ERROR AND RECOMMENDATIONERRORS1. The eye level is not the same level with the scale2. Valve is not fully open when making the disturbance3. The period of time for disturbance is not consistent for 3 seconds

RECOMMENDATIONS1. Make sure that the eye level is perpendicular to the scale2. Push the valve until its limit to make sure that it is fully open3. Count the period of time of 3 seconds using stop watch or normal watch

CONCLUSION

As conclusion for this experiment we can conclude that, as the Derivative Action is increase, the response of process to disturbance also become quicker. As the set point increase, the response of controller to disturbances also becomes quicker. Hence we can conclude that it is the best for PID controller by using the higher value of the derivative action where it can help the process to eliminate error that causes by the disturbance and come back to set point value.For second part, we can conclude that the best tuning value for this experiment is when PB=20s where it produce almost consistent sinusoidal curve where the peak-to-peak value can be measure to calculate the natural period.For the third part, as out tuning value is not good, the result we get is not good where the response on the board is not same as we set manually. Hence better tuning value is needed to help the process to response to disturbance that has been introduced. PB is stand for Proportional Band, I is Integration and D is Derivative. As for first part for PID controller only changes the derivative value to know the effect of derivative to PID controller where the best value of derivative can be taken to get the better result for the response. For second part we set I=1000s where the maximum value and D=0 so that the controller only become a Proportional controller. In this part also we determine the best tuning value for PB which is 20 s where the peak-to-peak produce can be used to calculate natural period since it almost constant sinusoidal graph.

APPENDIXCALCULATION Natural period is calculated using the formula:Natural period, T = 60 minWhere D = distance in mm between successive crests or valleysSample of calculation: For PB = 20 s, D = 2.5 mm, Trend Speed = 1200 mm/h (from graph),Natural period, T= 2.5 mm / 1200 mm/h 60 min/h = 7.5 sGRAPH FROM EXPERIMENT

REFERENCESLecture Notes (Chemical Process Dynamics Instrumentation & Control)Seborg D.E, T.F Edgar and D.A Melliechamp, Process Dynamics and Control, John Wiley and Sons, New York, 1989, pp 116-118.