Statistical process controls

Processes must be controlled for quality purposes. Most major chemical processes have automatic computerized data logging and many parameters are automatically recorded and available for study. These can include set points, process measurements such as flow rate, temperature, pressure, stirring speed, and so forth or product parameters such as composition, density, clarity, and so forth, especially when the product is measured on-line. 

(Mark each of a-d as True or False) Design of Experiments. .. 

Examine the process data below. The data is listed sequentially and represents 20 samples taken every eight hours and the resultant specific gravity results. 

Statistical process control (SPC) is a quality tool based on the principles of statistical mathematics and applied to a process to control product quality. The theory of SPC is based on some complex mathematics, but you need not be a mathematician to understand how to use the system. 

In any process, a certain amount of variability occurs. Variability s defined as the tendency to vary or change. For example, process equipment does not heat up to 450.5°F and stay at exactly 450.5°F. Instead, it tends to move a bit lower and higher. These variations occur with temperature, pressure, flows, and levels. Each process has its own variability and ability to tolerate change. SPC identifies this variability and enhances an operator s ability to control the process by setting limits on the variability. 

Customers identify the key target setpoints for the products they require. In-house engineers identify other process variables, such as temperatures, flows, levels, and pressures, that support the ability of the process to produce the desired products. 

SPC allows normal equipment and process fluctuations to be considered over a longer period of time. To warrant adjustment, sample results must demonstrate a downward or upward trend away from the key target setpoints. 

Most companies use a product directive or recipe for each product they make. Product directives detail operating parameters and control points such as  

Another reason for taking control action only when the process is out of control is because there is assumed to be a cost associated with control action. This is often true in the discrete manufacturing industries where it may be necessary to shut down the assembly line in order to make adjustments to equipment. Often this is not true in the chemical process industries where corrective action may be free , as in the adjustment of the temperature of a polymerization reactor. 

If the hypothesis is made that the points on the Shewhart chart are normally distributed, the probability of a point falling outside the control limits can be calculated, and is quite small. The likelihood of various other occurrences (nine points in a row all below [or above] the mean, 14 points in a row alternating up and down, etc.) can also be calculated. If the likelihood of these patterns is slight, they can be used as additional rules to determine when the process is out of control. Such pattern-based rules are easily implemented by a chemical plant operator. 

In a batch polymerization, each batch may take up to a full day to process. In this case, it is not possible to take multiple samples and calculate the average for an x-bar chart. (It is possible to take multiple samples from each batch, but the variance will reflect only the variance of the sampling and analysis techniques, and not batch-to-batch variations.) In this situation an individuals chart is used instead [47]. The individuals chart is plotted in a manner similar to the x-bar chart. The mean and standard deviations of the process are 

in continuous polymerization, the samples are sufficiently infrequent that the process settles between samples, and if the samples are not autocorrelated, the procedures outlined above can be used. This amounts to manual steady-state control with the need for control identified by the control chart. If the process is fast compared with the sampling interval, but the samples are autocorrelated (as is often the case), a controller can be developed which specifies the correction to be made to the process. A process model is needed for a process which is fast compared with the sampling interval, this can be simply a gain between the manipulated input and the process output. A disturbance model is also necessary a simple but effective model for continuous process systems is that of an integrated white noise sequence. With these assumptions, a minimum variance controller may be derived [45]. This takes the form of the following discrete pure integral controller  

Here m, is the manipulation to be made at time t, g is the process gain, Yj is the deviation of the measurement from setpoint at time j, and jS is a tuning parameter. Many continuous polymerization processes fit the assumptions used to derive the above controller (the process is fast with respect to the sampling interval, but the samples are still autocorrelated, and an integrated white noise model is appropriate). Other noise models may be used. If the process is not fast with respect to the sampling time, the process model must capture the dynamics of the process. Similar controllers can be derived for these situations. If the process is slow with respect to the sampling time, then conventional process control is appropriate. 

Raw material characterization. A simple, rapid single-point melt index (MI) test is used (see Chapter 10). Although the MI does not completely characterize a resin, for purposes of SPC it does not need to. Also used is a simple time-dependent sampling method. The resin supplier should sequentially number each box/container in the preliminary runs. A sample from each box is measured, and results are plotted using a standard control chart format. 

Implementation, The complete support and commitment of management are required, as SPC initially requires significant funds. It cannot be done quickly data gathering requires time. SPC may not be practical for every product because of the high cost in time and personnel. New products being prepared for production are often excellent candidates. 

Each problem will have its own solution or solutions (Fig. 2-17). Simplified guides to troubleshooting granulators, conveying equipment, metering/pro-portioning equipment, chillers, and dehumidifiers are available (1-3, 6-8, 33-35). 

Recognize that no two similar machines (from one or more suppliers) will operate in exactly the same manner, and that plastics do not melt or soften as perfect blends, but they do all operate within certain limits. 

There are several ways to determine the efficiency of the melt. One method is to observe the screw drive pressure it should be about 75 percent of maximum. If it is less than that, lower the rear zone heat until the drive pressure starts to rise. With melt quality changing, raise the center zone to restore quality to what is required. Heat changes should be accomplished in 10 to 15 degree increments, with 10 to 15 min. of stabilization time allowed prior to the next change. 

In industrial plants, large numbers of process variables must be maintained within specified limits in order for the plant to operate properly. Excursions of key variables beyond these limits can have significant consequences for plant safety, the environment, product quality and plant profitability. Statistical process control (SPC), also called statistical quality control (SQC), involves the application of statistical techniques to determine whether a process is operating normally or abnormally. Thus, SPC is a process monitoring technique that relies on quality control charts to monitor measured variables, especially product quality. 

8-45 Diagram for selection of optimization techniques with algebraic constraints and objective function. 

An example of the most common control chart, the Shewhart chart, is shown in Fig. 8-46. It merely consists of measurements plotted versus sample number with control limits that indicate the range for normal process operation. The plotted data are either an individual measurement x or the sample mean x if more than one sample is measured at each sampling instant. The sample mean for k samples is cal- 

The Shewhart chart in Fig. 8-46 has a target (T), an upper control limit (UCL), and a lower control limit (LCL). The target (or centerline) is the 

8-46 Shewhart chart for sample mean x. (Source Seborg et al., Process Dynamics and Control, 2d ed., Wiley, New York, 2004.) 

The major objective in SPC is to use process data and statistical techniques to determine whether the process operation is normal or abnormal. The SPC methodology is based on the fundamental assumption that normal process operation can be characterized by random variations around a mean value. The random variability is caused by the cumulative effects of a number of largely unavoidable phenomena such as electrical measurement noise, turbulence, and random fluctuations in feedstock or catalyst preparation. If this situation exists, the process is said to be in a state of statistical control (or in control), and the control chart measurements tend to be normally distributed about the mean value. By contrast, frequent control chart violations would indicate abnormal process behavior or an out-of-control situation. Then a search would be initiated to attempt to identify the assignable cause or the special cause of the abnormal behavior 

One approach for using DOE on more complex processes is to do the majority of the process development on smaller, representative sections of material, such as test panels, rather than on full-scale parts, and then to scale up with a more limited experimental matrix. There is no guarantee that experience on small-scale test panels will directly translate to large parts because dimensions and thickness of the part are important variables in their own right. Another way to save on costs is to start with a satisfactory process and to continue, via careful monitoring of process variations and results, to extend the range of experience. This method is variously called statistical process control or statistical quality control. 

SPC or statistical quality control (SQC) is similar to DOE in that it is a statistical, rather than mechanistic, method. Both SPC/SQC and DOE rely on the theory that there is a direct relationship between variations in process controls and resulting changes in product quality. In SPC, however, the experiments are not forced on the process like they are in DOE. The variations in product quality and the random process variations are traced over time instead. The variations in end product are then correlated, if possible, with changes in the process that have occurred during that time. SPC techniques are usually applied to the process after some baseline process has been established by other methods. 

The advantage of SPC is its relatively low cost. Many of the parameters that should be tracked for SPC are already part of any good quality assurance and postprocess inspection program. The codes necessary to do analysis are commercially available and do not require additional experts for interpretation. SPC can be used to react to quality drift by adjusting the process. Another advantage is that SPC may pick up unanticipated effectors, such as the time of day or operator performance. 

SPC does require statistical quantities of product and automated data tracking. The best processes for SPC are those which are used to make large quantities of inexpensive parts. SPC is also difficult to apply to processes where the number of independent variables is large. Automated data acquisition is a must for SPC, but this is becoming inexpensive and common in the workplace. SPC is also a delayed control method. Many defective parts may be made before SPC corrects the process. The longer it takes to evaluate the results of the process, the more delay in the ability to react to process changes. Another important requirement of SPC is attention to detail on the part of the operator and/or process engineer. No battery of QC tests will detect every variation either in materials and process or in final quality. 

An example of successful use of SPC comes from the injection molding industry where part counts are high and part values are generally low. The material in reject parts can sometimes be recycled. The process is very rapid, with few control variables and the major, often the only, criteria for quality is reproducible shape. For this situation, SPC is an excellent 

In many cases of batch and semi-batch polymerization control there are no on-line measurements of polymer quality (for example, polymer composition, molecular weight) during the batch and these measures of end-use properties are only available at the end of the batch. In this case recipe modifications from one run to the next are common. The minimal information needed to carry out this type of mn-to-run control is a static model relating the manipulated variable to the quality variables at the end of the batch. As pointed out in Ref 7, this model can be as simple as a steady-state (constant) gain relationship or a nonlinear model that includes the effects of different initial conditions and the batch time. The philosophy of statistical process control can be very useful in this case, since the polymer quality variable (for example the Mooney viscosity in elastomer manufacture) can be plotted for each successive batch on a Shewhart (%-bar) chart with the upper and lower control limits placed at three standard deviations above and below the target. 

A point outside the control limits means that the batch is out of control and the batch recipe and possibly the sequence logic must be adjusted for the next batch. 

If the quality variable for the batch is within the control limits, no control action is taken to prevent manipulations of the batch process based on stochastic variations within it. 

Very often in DCS-operated batch polymer reactors the primary process variables such as pressure, temperature, level, and flow (Section 12.2.1-12.2.4) are recorded during the batch as well as the quality variables at the end of the batch. However, it may be very difficult to obtain a kinetic model of the polymerization process due to the complexity of the reaction mechanism, which is frequently encountered in the batch manufacture of specialty polymers. In this case it is possible to use advanced statistical techniques such as multi-way principal component analysis (PCA) and multi-way partial least squares (PLS), along with an historical database of past successful batches to construct an empirical model of the batch [8, 58, 59]. This empirical model is used to monitor the evolution of future batch runs. Subsequent unusual events in the future can be detected during the course of the batch by referencing the measured process behavior against this in ontrol model and its statistical properties. It may therefore be possible to detect a potentially bad batch before the run is over and to take corrective action during the batch in order to bring it on aim. 

---

Manufacturing processes have been improved by use of on-line computer control and statistical process control leading to more uniform final products. Production methods now include inverse (water-in-oil) suspension polymerization, inverse emulsion polymerization, and continuous aqueous solution polymerization on moving belts. Conventional azo, peroxy, redox, and gamma-ray initiators are used in batch and continuous processes. Recent patents describe processes for preparing transparent and stable microlatexes by inverse microemulsion polymerization. New methods have also been described for reducing residual acrylamide monomer in finished products. 

QA = quahty assurance QC = quaUty control SQC = statistical quaUty control SPC = statistical process control CIM = computer-integrated manufacturing. 

Quality in Japan. Japanese economic prowess has been attributed variously to such quahty improvement activities as quahty circles, statistical process control (SPG), just-in-time dehvery (JIT), and zero defects (ZD). However, the real key to success hes in the apphcation of numerous quahty improvement tools as part of a management philosophy called Kaizen, which means continuous improvement (10). 

Statistical Process Control. A properly miming production process is characterized by the random variation of the process parameters for a series of lots or measurements. The SPG approach is a statistical technique used to monitor variation in a process. If the variation is not random, action is taken to locate and eliminate the cause of the lack of randomness, returning the process or measurement to a state of statistical control, ie, of exhibiting only random variation. 

Statistical Control. Statistical quahty control (SQC) is the apphcation of statistical techniques to analytical data. Statistical process control (SPC) is the real-time apphcation of statistics to process or equipment performance. Apphed to QC lab instmmentation or methods, SPC can demonstrate the stabihty and precision of the measurement technique. The SQC of lot data can be used to show the stabihty of the production process. Without such evidence of statistical control, the quahty of the lab data is unknown and can result in production challenging adverse test results. Also, without control, measurement bias cannot be determined and the results derived from different labs cannot be compared (27). 

While the single-loop PID controller is satisfactoiy in many process apphcations, it does not perform well for processes with slow dynamics, time delays, frequent disturbances, or multivariable interactions. We discuss several advanced control methods hereafter that can be implemented via computer control, namely feedforward control, cascade control, time-delay compensation, selective and override control, adaptive control, fuzzy logic control, and statistical process control. 

Statistical Process Control Statistical process control (SPG), also called statistical quahty control (SQC), involves the apphcation of statistical concepts to determine whether a process is operating satisfactorily The ideas involved in statistical quahty control are over fifty years old, but only recently with the growing worldwide focus on increased productivity have applications of SPG become widespread. If a process is operating satisfactorily (or in control ), then the variation of product quahty tails within acceptable bounds, usually the minimum and maximum values of a specified composition or property (product specification). 

There are important economic consequences of a process being out of control for example, product waste and customer dissatisfaction. Hence, statistical process control does provide a way to continuously monitor process performance and improve product quahty. A typic process may go out of control due to several reasons, including... 

More details on statistical process control are available in several textbooks (Grant and Leavenworth, Statistical Quality Control, McGraw-HiU, New York, 1980 Montgomery, Introduction to Statistical Quality Control, Wiley, New York, 1985). 

Statistical Process Control (SPC) The use of statistical techniques (such as control charts) to analyze a process and take appropriate action to maintain statistical control and improve process capability. 

Fig. 4.12. Statistical process control chart for TXRF measurement systems. The sensitivity of the system can be controlled by daily calibration with an...

Failure modes analysis Statistical process control Measurement systems analysis Employee motivation On-the-job training Efficiency will increase through common application of requirements for Continuous improvement in cost Continuous improvement in productivity Employee motivation On-the-job training... 

Several methods have evolved to achieve, sustain, and improve quality, they are quality control, quality improvement, and quality assurance, which collectively are known as quality management. This trilogy is illustrated in Figure 2.1. Techniques such as quality planning, quality costs, Just-in-time , and statistical process control are all elements of... 

You should review the contract and the detail specifications to identify whether your existing controls will regulate quality within the limits required. You may need to change the limits, the standards, the techniques, the methods, the environment, and the instruments used to measure quality characteristics. One technique may be to introduce Just-in-time as a means of overcoming storage problems and eliminating receipt inspection. Another technique may be Statistical Process Control as a means of increasing the process yield. The introduction of these techniques needs to be planned and carefully implemented. 

The action taken during process monitoring (see Part 2 Chapter 9) can be considered preventive action when corrections are made to the process ahead of occurring nonconformities. Hence Statistical Process Control is a technique which serves nonconformity prevention as well as detection. 

Fundamental statistical process control reference manual (GM, Ford, Chrysler)... 

Guidelines to Statistical process control Parts 1 2 (SMMT)... 

It should be noted that, even with optimal PIFs, errors are still possible. There are two reasons for this. Even in the optimal case, some random variability in performance will remain. These random variations correspond to the "common causes" of process variability considered in statistical process control. Variations in PIFs correspond to the "special causes" of variability considered within the same framework. 

The alternative is hexane, which because of the explosion hazard requires a more expensive type of extractor construction. After the extraction the product is dull gray. The continuos sheet is slit to the final width according to customer requirements, searched by fully automatic detectors for any pinholes, wound into rolls of about 1 m diameter (corresponding to a length of 900-1000 m), and packed for shipping. Such a continuous production process is excellently suited for supervision by modern quality assurance systems, such as statistical process control (SPC). Figures 7-9 give a schematic picture of the production process for microporous polyethylene separators. 

Statistical process control (SPC) is an important on-line method in real time by which a production process can be monitored and control plans can be initiated to keep quality standards within acceptable limits. Statistical quality control (SQC) provides off-line analysis of the big picture such as what was the impact of previous improvements. It is important to understand how SPC and SQC operate. 

In petrochemical and bulk commodity chemical manufacture, real-time process control has been a fact of life for many years. There is considerable understanding of processes and control of process parameters is usually maintained within tight specifications to ensure statistical process control to within six sigma, or the occurrence of one defect in a million. This has been enabled through the use of real-time analytical capability that works with programmable logic circuits to make small changes to various process inputs and physical parameters as required. 

In batch chemical operations, this level of real-time process control is rarely achieved, although there are increasing attempts in recent years to achieve greater statistical process control, the industry is generally only able to operate at about three or occasionally four sigma, or one defect in 1000-10 000. 

Metrics for this might include number of excursions from statistical process control, but one very useful metric for controllability is process capability, or more accurately, process capability indices. Process capability compares the output of an in-control process to the specification limits by using capability indices. The comparison is made by forming the ratio of the spread between the process specifications (the specification width ) to the spread of the process values. In a six-sigma environment, this is measured by six standard deviation units for the process (the process width ). A process under control is one where almost all the measurements fall inside the specification limits. The general formula for process capability index is ... 

---