Industrial Production of Gypsum: Quality Control Charts

Abstract

The production of gypsum (marketed if it accomplishes the required specifications) occurs during the process of flue gas desulphurization in a Portuguese Coal Thermoelectric Central. Important variables in this process are statistically analyzed in the chemical laboratory and quality control charts are implemented to monitor the entire process. In this study individuals and moving range charts of the variable "density of gypsum slurry" are compared with the "more efficient" ones obtained after a Box-Cox transformation. This transformation is used to normalize the data, because its observations come from non-normal models-where classical control charts are considered less appropriate, since they usually exhibit rates of false alarms different from what would be expected. Although it is important to consider different statistical approaches for quality control charts, during the monitoring of an industrial process, in this case study the achieved results lead us, essentially, to similar conclusions.

Robust control charts in industrial production of olive oil

“…Wheeler [11,12,13] consider that "the transformation of the data to achieve statistical properties is simply a complex way of distorting both the data and the truth" and a normal distribution is not required in the calculation of control limits, which made Shewhart chartsa robust tool. On the other hand, when wehave highly skewed data and heavy tails the false alarm rate can be quite high and a Box-Cox transformation or bootstrap control charts are recommended to obtain efficient and robust estimators [3,4,5,6].In this particularstudy (where we may consider light tails, moderate outliers and asymmetry) seems unnecessary once we reach similar conclusions. The results obtained are also relatively easier to interpret and very useful in practice, especially for nonstatisticians.…”