Anda di halaman 1dari 36

Handbook for Pulp & Technologists

Chapter 24: Process Control

Introduction
All industrial processes manifest inherent variability which must be minimized it the plant operation is to yield uniform, high-quality product with efficient utilization of raw materials, manpower, and energy. The objective of process control is to maintain the process (and each element of the process) within well-defined limits of variation. A time plot for a process variable (shown in Figure 24-1) illustrates acceptable control, except for one period of operation that is "out of control". For a process under manual control, the operator watches the trend of the data points, and makes adjustments to keep the process within limits. But sometimes, depending on the degree of process stability, accuracy of the data points, and narrowness of the limit lines, the operator may only be able to take corrective action after the variable has strayed outside the process limits. Very few processes or process "elements" within the pulp and paper mill are stilt under manual control. Most processing steps are automatically controlled utilizing digital or analog equipment. Increasingly, computers are utilized to coordinate control and record-keeping aspects for the entire manufacturing sequence.

Introduction

I. Measurement and Control


A typical automatic control loop consists of three basic components: sensor (with transmitter), controller, and control element, The interaction of these components is shown schematically in Figure 24-2. An example of a control loop common to all pulp and paper operations is the consistency control loop shown in Figure 24-3. Here the sensor measures consistency in the stock line and transmits an appropriate signal to the controller; the controller compares the incoming signal to the set point and transmits an error signal to the dilution valve; the dilution valve opening is then attenuated in the direction of correcting the error. Perhaps the most common control application with respect to all process industries is temperature regulation utilizing a steam valve as the control element. Other common applications are control of flow rate and tank level using a flow control valve and control of gas pressure using a pressure relief valve.

I. Measurement and Control

I. Measurement and Control

I.1 Sensor
The sensor is at the heart of any control loop. Automation cannot be properly implemented without a reliable indication of dependent variable behavior; and some operations within the pulp and paper industry have only recently become amenable to automatic control by virtue of new measurement techniques. Some control strategies must still rely on manual testing results for input into the control system. In general, few variables can be measured directly with process instruments and dependence must be placed on measurement of related properties. For example, the moisture content of paper can only be measured directly by laboratory methods; but it can be accurately inferred from measurement of such related properties as: resistance to electric current, dielectric resistance, absorption of microwave energy, absorption of infrared energy. The principles of measurement for some of the variables commonly monitored or controlled in the pulp and paper industry are summarized in Table 24-1.

I.1 Sensor

I.2 Controller
In almost every process, certain variables must be rigidly controlled at specified values in order to maintain efficient operation. The controllers are given these desired values in the form of set points. For a basic control loop, the set point may be keyed in at a control console, or entered by means of a set point knob on the controller itself. In either case, the setting corresponds to a dc electrical signal within the 4-20 mA range (or to a pneumatic signal within the 3-15 psig range). The measured value is also converted to a 4-20 mA DC (or 3-15 psig) signal and is transmitted from the sensor to the controller. Comparison of the set point and input signals results in a difference, known as the error, which is the key actuating signal in the controller. The responding control action will depend on the direction and magnitude of the error, and in some cases on how fast the error is changing. The controller acts to reduce the error and bring the actual value toward the desired value by means of the output signal. The output signal is generally of the same type as the set point and input signals, namely 4-20 mA do (or 3-15 psig).

I.3 Analog model


If the Controller is to carry out its function to automatically regulate a process variable at a desired value, the control loop must be adapted and tuned to the characteristics of the process. Three common modes of control are used in analog systems for the majority of process applications depending on response time and lag-lead requirements. The respective adjustable settings are known as proportional band or gain (P), integral or reset (I), and derivative or rate (D). Proportional hand is defined as the range of values of the process variable in which there is proportional control action, expressed as a percentage of the full scale range of the controller. For example, if the full scale range of a temperature controller is 0 to 200 C and the proportional control action is over a 50 C range, the proportional hand is then 25%. If the proportional control action is over a 2000 C range, the proportional hand is then 100%. The width of the proportional band determines the amount of valve motion for any given change in the controlled variable.

I.3 Analog model


A process may stabilize at a point other than the set point when using a simple proportional controller. The difference between the stabilization point and the set point is known as offset. Sometimes the offset is small and can be tolerated. When offset is undesirable, proportional plus reset (two-mode control) is required to modify the output signal and eliminate the offset. Reset action can be visualized as a shifting of the proportional hand. The reset adjustment (tuning) is in the form of timing, and the reset rate is commonly expressed as repeats per minute. Reset is proportional to the magnitude of error and to the duration of error; its action accumulates as long as there is an error, but ceases when the error is reduced to zero. Occasionally, a particular process will have a dead time and transfer lag in such a way as to cause unsatisfactory control with single or two-mode control. It is then necessary to add the third mode of control known as derivative or rate action. This third mode provides a continuous relationship between the rate of change of the process variable and the output signal. It provides an initially large over-correction to compensate for an unfavorable process lag. Derivative time is the interval by which derivative action advances the time of proportional action, and the derivative mode adjustment (tuning) is in a timing form.

I.4 Digital Controller


In the 1970's, instrument loops requiring the PID control function were best served by analog equipment. Computer control techniques such as time sharing and point sampling introduced deadtime and phase lag into the PID function which diminished the responsiveness and stability of the control system. These problems were overcome in the distributed control systems of the 1950's, through development of new algorithms and nonlinear compensation methods. The majority of controllers today are micro-processor based. Modern digital controllers provide some significant advantages over analog systems. Firstly, the same control functions can he implemented with less hardware and with more reliable components. Perhaps the greatest advantage is that the control functions are not tied to hardware, but can be configured to implement a variety of functions. More sophisticated and better control strategies can be implemented at little cost once the basic controller hardware is installed. Also, the high accuracy and drift-free nature of digital control allows the user to operate his process nearer to its limits.

I.5 Smart Sensors


In the context of microprocessor-based controllers, the term "smart sensor" is often used. Generally a product with the designation "smart" will have certain capabilities of self calibration, internal diagnostics, setting of parameters from remote locations. compensation for variations in other variables, linearization, and simplified tuning of coefficients.

I.6 Programmable Logic Controller (PLC)


A programmable controller is a digitally operating electronic apparatus that uses a programmable memory for the internal storage of instructions that implement specific functions such as logic, sequence, timing, counting. and arithmetic to control machines and processes. This type of controller was developed in the 1970's and has a wide application in certain manufacturing operations. In the pulp and paper industry, these controllers are utilized mainly for discrete logic, safety interlocks and other sequencing operations. While modern PLC's have the capability to control certain analog loops, they tend to interface poorly with computers because their design emphasis has not included algorithmic and configuration requirements. The inability to integrate with computer software has thus far severely limited their application to process control in the pulp and paper industry.

I.7 Feedback and Feed-Forward Control


A typical control loop utilizes feedback control. With this node of control, the controller cannot act until an error has developed. In most control applications, the elapsed time between controller action and feedback signal is short, and satisfactory control is achieved. For slow processes with long time delays, change in load or composition may show up too late to enable the controller to hold the controlled variable within the desired limits. By adding feed-forward control, a correction can be made for anticipate,) changes of load or composition, thereby minimizing the transient error that would otherwise result, Feed-forward control by itself is functionally limited, and must be used in conjunction with a feedback loop to provide full control.

I.8 Control Valves


In the majority of control applications, an automatic valve is employed as the control element. The control valve is essentially a variable orifice used to regulate the flow of a process fluid in accordance with the requirements of the process. Proper selection of the valve with respect to design, size and materials of construction is critical to satisfactory performance and service. Three features of control valves should be evaluated to ensure correct control response: capacity, characteristic and rangeability. The characteristic is the flow response to valve opening (percent of stem travel) over the full range. Different valve characteristics are illustrated in Figure 24-4. For most control applications, a linear-type response is desired. The rangeability is the ratio of flows (high to low) through which the control valve can give stable operation: a relatively low figure indicates that the valve characteristic is non-linear.

I.8 Control Valves

I.8 Control Valves


It is also important to select the proper type and structure of valve to avoid such service-related problems as corrosion, flashing, plugging, cavitation, noise, vibration and seat leakage. Some common types of valves are depicted in Figure 24-5. A typical automatic flow control valve with pneumatic actuator is shown in Figure 24-6. The actuator on a control valve is normally an air-operated diaphragm or piston, which translates it signal from the controller into stem or vane movement. Springs oppose the force of air pressure to hold the plug or vane in position against the forces of fluid flow, and act to return the valve to a closed position when the air pressure is reduced.

I.8 Control Valves

I.8 Control Valves

I.9 Advanced Control Concepts


More complex control systems often use two or more measurements in a control loop. The basic form for multiple-variable control is cascade control. Other common systems are ratio control and cutback control. With a cascade arrangement, the objective is to improve control of the primary variable by interlocking the primary controller with the controller for a secondary variable which influences the value of the primary variable, The output from the primary controller adjusts the set point of the secondary controller, thus helping to manipulate the primary variable. The secondary loop is typically introduced to improve control of processes marked by upsets and long time constants, and which cannot be satisfactorily controlled by it single three-mode controller. The secondary variable is not controlled in the usual sense, but is manipulated.

I.9 Advanced Control Concepts


Ratio control is used when it is desired to control one variable in a fixed ratio to a second variable. In practice, a change in set point for a primary controller will actuate the ratio mechanism in the second controller, which adjusts the set point of the secondary variable to maintain the fixed ratio. This control principle is commonly used for paper mill stock proportioning systems. Cutback (or auto selector) control is used when a single, final control element is to be manipulated to prevent a process variable from exceeding a preset limit. This control mode may be provided to ensure operator safety or protect equipment in case of process system malfunction.

I.10 Electronic vs. Pneumatic Instrumentation


Electronic and pneumatic instrumentation systems are both widely used, and each type provides good performance. Electronic systems cover virtually all industrial processes. Pneumatic systems can measure most variables, with a few exceptions such as pH and chemical composition. In many applications, there appears to be little difference in sensitivity and reliability between the two types of systems. However, electronic systems have the advantage of instantaneous response; and there is no limit to the distance that an electronic signal can be transmitted. Generally, costs for electronic instrumentation are higher than for the pneumatic counterpart. For computer and microprocessor-based systems, electronic signals are required. While a pneumatic signal can be easily converted to an electrical signal using a transducer (and vice versa), the conversion imposes additional limits on control accuracy.

II. Process Control Computers


In the 1960's the digital computer was introduced into process control, but was slow to gain acceptance in pulp and paper mills because of high capital investment and because measuring techniques were not available to provide many of the inputs required for overall control strategy. The utilization of computers accelerated in the 1970's when their relative cost declined markedly at the same time that many new sensing techniques were introduced. In the final analysis, the major reason for installing a computer is to be able to operate the process more efficiently and more profitably. Other benefit, are the centralized interface which increases operator awareness of the process, and the provision of reporting documentation for management purposes. Today's process computers and microprocessors utilize concepts of logic and handle complex control schemes including niultiloop cascade system, compensation for time delays, and non-linear and variable instrument parameters. They also provide for data acquisition and display, statistical analysis and establishment of mathematical models.

II.1 Evolution of Development


The first computer technology available for process control was the large main frame computer. To justify installation of this expensive unit, extensive (and often unrealistic) project objectives were developed. The purchasers were often unaware of the time and effort that was necessary to program their computer. The earliest installations were used primarily to gather and record process data (i.e., monitor the process and act as data loggers), set off alarms (if necessary) and make calculations. Initially, these computers were not used to control process instruments directly, but they supplied the operator with the information lie needed to properly adjust the set points of various control devices.

II.2 Supervisory Digital Control (SDC)


Supervisory control was a logical progressive step for a mill that was fully automated with analog control loops. In this concept, the computer receives the same sensor signals as the controllers and is able to provide a picture of the current status of the process variables. The computer periodically calculates (on the basis of input signals and programmed instructions) what the optimum values should be for selected variables (e.g., to produce maximum yield of product at the required quality level). It then changes the set points of the controllers to bring the process operation in line with the newly calculated conditions, Without the computer, these calculations were handled manually, usually with less frequency and with far less precision. In a mill with existing controllers that are capable of accepting remote setpoints, SDC may still be it cost-effective approach to adding it computer control system. However, if the existing controllers must be replaced. a distributed control system will typically be the most viable option.

II.3 Direct Digital Control (DDC)


The next development was direct digital control where the computer replaced the functions of the analog controllers in the control loops, and added a level of supervisory control. With DDC, the loop concept is retained, but all loop variables feed to the computer. In each case, the computer calculates the necessary corrective action and transmits it directly to the control element. The major problem with DDC is total dependency on the computer, and it is necessary to provide backup options to ensure continuity of operation during servicing of the computer. The most satisfactory strategy has been to employ two digital computers, each with the full facility to control the process should a failure occur in the other. The popularity of DDC was short-lived, and this concept is now seldom used. Dual computers are costly and quickly require upgrading or the addition of more dual computers as the control and programming tasks increase.

II.4 Minicomputer
With the development of smaller, lower-cost computers in the early 1970's, it was no longer necessary to put together an extensive control project to justify, the purchase and installation of a system. The minicomputer was economically more suited than the mainframe to simple control strategies involving a few key loops. The special-purpose computer along with improved sensor technology was utilized most successfully around the paper machine. In spite of shortcomings, packaged control systems were successful because they concentrated on the key variables of any papermaking operation and controlled them. However, by the late 1970's, the dedicated minicomputer approach was already at its limits. As the mini became more powerful, functions and features were added; and this brought to the minicomputers the problems of complexity and unreliability formerly associated with large computer projects, Expansions and retrofits were complicated, and intersystem communications were difficult and costly.

II.5 Distributed Control Systems (DCS)


Today's preferred mode of computer control is by deployment of numerous smaller computers in place of the large general-purpose computer. The key to DCS is to spread computing and controlling functions throughout the plant. This distribution puts the controlling "intelligence" close to the unit operations. Generally, smaller computers or microprocessors (designed and programmed for specific tasks) are easier to justify both technically and economically. Many distributed systems utilize a hierarchy of controllers from singleloop microprocessor-based controllers tip to multi-loop or interactive microprocessor-based controllers. This ability to select the appropriate sophistication of control technique for each specific process unit provides for the functional distribution of control throughout the entire process.

II.5 Distributed Control Systems (DCS)


In a truly distributed system, the various process control units (which may be geographically separated) are linked together on it data bus or data highway, allowing speedy, efficient transfer of data back and forth between local computers (and with a central-operations computer, if used). This communication technique usually results in, significant reduction in field wiring casts (compared to it central-control computer) since a single coaxial cable can link together as many as four or five hundred single control loops. In DCS, redundant communications are common, allowing for a back-up communications link to take over should the primary one fail.

II.5 Distributed Control Systems (DCS)


While some weatherproof housing is required for the distributed controllers, time environmental conditions can be significantly more extreme than those required for the computer itself. If the computer is lost for any reason, the "intelligent controllers continue to control the process. Thus, it redundant computer does not have to be purchased. Obviously, should a microprocessor-based controller fail, the degree of failure in terms of the number of loops affected is dependent on the design of the control system. Usually, those controllers on the most critical loops of the distributed system are backed-up by secondary units (redundant controllers). Although a distributed system can be configured, if necessary, as it supervisory or direct digital control system, these control concepts are not applicable in the majority of cases because of the sophistication available within the microprocessor-based systern at the single loop level. Essentially, each microprocessor-based controller has all the input, output and computing sub-systems as the computer itself.

II.6 Mill-Wide Control


While mini-computers and microprocessors have taken over many of the control functions, interest has re-focused on the central computer as an information and guidance system, especially for it paper mill or integrated operation. The objective would be to interconnect all the digital systems into a mill-wide network to provide effective management decision support tools. This integration of process control and decision making is sometimes referred to as the pulp and paper industry's realization of computer-aided manufacture. In view of the complexity of pulp and paper mill systems and the constant efforts to improve overall mill performance, a sophisticated information and analytical system can be a valuable tool for both management and operating personnel. Typically, the central computer (as the final element of a "hierarchical" control structure) will utilize a high-speed printer to produce a number of reports on daily, weekly and monthly bases. While graphic display of an individual process (e.g., digester, lime kiln) is handled in the respective dedicated computer, the central computer usually provides overview displays. It also carries out material and energy balances, produces time-based curves (current and historical), and calculates production optimization strategies.

II.6 Mill-Wide Control

II.6 Mill-Wide Control


There are many concepts of millwide control. One conceptualization is illustrated in Figure 24-7 wherein the objectives, implementation responsibility, and typical response times are laid out (I). The basic building blocks at the bottom of the pyramid (level I) are the instrumentation systems. Without a well-designed and maintained instrumentation/electrical system, any attempt at millwide control would be foolhardy. At the second level is unit process control and supervision of quality. The third level is production and energy management, which is usually referred to as the "rnillwide control" level. Finally, the highest level in this particular model is production scheduling. Here, the major objective is to schedule orders in a manner to achieve maximum utilization of the production facilities. In spite of extensive lobbying by the proponents of mill-wide control, relatively few mill-wide systems had been implemented in North America by the early 1990s, perhaps because other concerns were perceived as being more urgent of attention.

II.7 Control room


The move away from conventional analog instrumentation to computer control has revolutionized the control room. The most visible aspect is the video-based operator station. In the modern control room, the video screen (CRT) is the operator's window to the process, wherein the logical and functional presentations of process information permit a rapid and efficient two-way communication. Although the information on the screen is much the same as was formerly available on analog instruments, the method of depiction, the amount of information, and the accuracy are all enhanced. Typically, the operator runs the process from an air-conditioned and humidityregulated control center. Ideally the work station should be situated in an environment where the operator can concentrate on plant status. grid not be distracted by high foot traffic, insufficient work space, or disorganized reference materials (2). All the displays and controls should be located at a central point within the reach and viewing range of the operator. A console should consist of four to eight CRT's installed in a cabinet configured for use by a seated operator (for example, see Figure 24-8). The console is usually semicircular in shape with one or two tiers of CRT's. The keyboard is also a key part of the operator interface. By appropriate input, the operator changes the set points on the basic control loops, and pages through the different groups while sitting at the console, analogous to moving along a panel monitoring various analog instruments. Although most keyboards are standard units, customized keyboards may be designed for specific operator functions to simplify the operating system and provide greater security.

II.7 Control room

Anda mungkin juga menyukai