Multivariate Analysis in the Investigation of Dust Pollution from Quarry Activities

Dust emission from quarrying can lead to conflict between quarry operators and the surrounding local community. Quarry activities involve disturbing the ground to remove rock composed of limestone, sandstone or granite, followed by the subsequent transport, dumping, crushing, and processing of this material The activities of all quarrying stages represent a potential for dust production. Quarry operators must ensure that dust does not adversely affect local environments, and that new extensions or new quarry sites, which are increasingly likely to be located in environmentally sensitive areas, meet acceptable environmental standards. In this study, multivariate statistical techniques such as principal component analysis (PCA), reliability analysis (RA), and multiple regression analysis (MRA), were applied to develop new empirical formulae to calculate the dust deposition rate surrounding quarry area. These statistical techniques were employed to interpret and evaluate variations of large and complex dust deposition air quality data sets, which were generated from five years (2000-2005) of monitoring 13 parameters at 17 quarry sites in Malaysia (4264 observations). Meteorological data such as wind speed, wind direction, and rain amount were also gathered from each quarry site via nearby meteorological department stations and were associated with corresponding analytical data. PCA was conducted on six emission variables to identify possible pollutant sources. Finally, three emission variables were observed in two components, representing 74.2% of the total variance of the data. The highest loading, at 0.918, was contributed by drilling activity. Another varimax rotated PCA was carried out on eight dispersion variables; this PCA included only five variables drawn in two components, representing 95.3% of the total variance. The highest loading in this PCA, at 0.989, was contributed by wind speed. A new empirical model was developed to estimate the dust deposition based solely on the PCA results, followed by stepwise MRA of the RA findings. The adequately high correlation coefficient value of R = 0.732 (R2 = 0.535) indicated that the seven parameters included in the model (wind speed, receptor distance from source, percent and number of wind blown to receptor, blasting frequency, total drill hole, number of vehicles and quarry road distance) are responsible for the dust deposition pattern in and around the quarry site. This study illustrates the usefulness of multivariate statistical techniques for analysing and interpreting large/complex data air quality sets, for source apportionment, and for understanding spatial/temporal variations for effective air quality management.