ASM Science Journal: Tracking Water Quality with Advanced Control Charts

This study explored whether statistical control charts, which are normally used in manufacturing, can help monitor water quality at a treatment plant in Kota Kinabalu. Four indicators were tracked: turbidity, pH, dissolved oxygen, and iron levels (Fe). Both univariate charts (one variable at a time) and multivariate charts (multiple variables together), including Hotelling’s T² and Multivariate Cumulative Sum (MCUSUM), were tested.

Because the data did not follow a normal distribution, the charts gave mixed and sometimes conflicting signals. Turbidity and iron occasionally exceeded limits, while pH and DO stayed stable. MCUSUM became overly sensitive, flagging many warnings, unlike Hotelling’s T². The study concludes that traditional charts may be unreliable here and suggests using nonparametric methods instead.

Visit the ASM Science Journal to read the full article.