| Contents: |
|
Product variability refers to the variations or non-uniformity in product quality variables. In the pulp and paper industry product variability refers to parameters such as basis weight, sheet smoothness, pulp brightness, pulp strength, white liquor strength, steam pressure and temperature.
Product variability can cause problems for the end user and have an impact on internal production rates and costs. For example, paper variability can result in poor registration, warped boxes, uneven stacking and poor printing properties. The internal process variability that is responsible for the product variability can also result in reduced paper production due to sheet breaks and cull. Additionally high levels of process variability tend to result in the over application of the expensive chemicals and fiber.
The product variability refers to the spread in the product quality parameter. The most common measure of product variability is 2 times the standard deviation (2 sigma). In a normally distributed data set, 95% of all data points will fall within ± 2 standard deviations of the mean.
The spectral content of the variability is measured by the power spectrum analysis. This analysis helps to define the period (cycles/seconds) and amplitude of the cycles in the product. It is an important diagnostic tool since the product cycles can be matched to the process sources.
Product variability originates with process variability. Some process variables are tightly coupled to the product. Variability in these processes can have an important impact on product variability. On a paper machine, headbox pressure and machine chest consistency variability will generate basis weight and moisture variability. At the other extreme, even high amounts of variability in the high density chest level will have minimal impact on product variability.
There are many sources of process variability. Variability in the raw material input streams will cause process variability. However it is typical that most of the process variability is generated internally by the process equipment and by the control loops. Specific sources of variability in the pulp and paper industry include:
The process variability is present in all parameters and is not stationary. It varies with grade and operating conditions. It propagates throughout the process via stock, fluids, steam and air streams.
The process design, process equipment selection and process control all affect the impact of process variability on process performance.
There are many forms of process variability. Some common types are:
Reduced process variability results in increased operating efficiency, improved product quality and reduced operating costs.
The first step in reducing process variability is identifying its source. Poor controller tuning, bad control valve performance and inadequate mixing are examples of potential sources.
The identification of the variability can be difficult without appropriate tools. Many distributed control systems (DCS) are not capable of accurately displaying the variability. In addition, excessive filtering applied to the field sensor or at the DCS can mask the variability. With the variability accurately identified, tests can then be conducted on the process and process variables to identify the source.
The open loop bump test is an important test that can be conducted to identify process, equipment and process control problems. The test involves stepping the controller output in manual mode. The process variable is monitored to identify the process dynamics as well as control valve, measurement and tuning problems. Additionally the responses of interacting processes are monitored to identify process equipment problems and process and process control interactions.
In many cases the identified variability can be resolved with improved tuning or control valve maintenance resulting in significant process and product variability improvements.
LAMBDA CONTROLS, 17 Erbsville Court (RR# 3), Waterloo, ON, Canada N2J 3Z4
Tel.: 519-884-6998, Fax: 519-884-1089
E-mail: inquiries@lambdacontrols.com, Web site: www.lambdacontrols.com