USING FACTOR ANALYSIS TO IDENTIFY SOIL CONTAMINANTS FROM THE PALMERTON, PA, ZINC SMELTERS

It is estimated that during their 90 years of operation the Palmerton, PA, zinc smelters (now an EPA superfund site) emitted 3,740 st Cd, 7,560 st Pb, and 286,000 st Zn. The ore mineral feed to the smelters consisted of sphalerite, galena, willemite, franklinite, and zincite. Twenty-four soil samples were collected from the area surrounding the Palmerton zinc smelters. Samples of ore minerals that made up the primary feedstock to the smelters were obtained from museum and private collections (the mines are closed). Instrumental Neutron Activation Analysis (INAA) was used to determine the concentrations of 39 elements in the soils and ore minerals.

One group of soil samples was from Lehigh Gap, close to the smelters, and the other group was from a mountaintop traverse along an east-west trending ridge line just south of the smelters. Plots of element concentrations versus distance from the smelters showed that As, Sb, Se, Cd, Hg, Zn, Fe, Cr, Co, and Sc co-varied. Factor analysis was applied to the data set and two factors explained 81% of the variance. The first factor accounted for 45% of the variance and consisted of As, Sb, Se, Cd, and Hg. Based on the ore mineral chemistry, this factor reflects the contributions of the sulfide ore minerals, sphalerite and galena. The second factor accounted for 36% of the variance and consisted of As, Fe, Cr, Co, and Sc. This factor was related to the oxide ore minerals, franklinite and zincite. The samples collected in Lehigh Gap, the remediated area proximal to the smelters, exhibited differences in elemental patterns when compared to those collected along the mountaintop traverse. We hypothesize that these remediated samples are largely influenced by the mineral component franklinite which has relatively high amounts of Fe and strong positive correlations with Sc and Cr; whereas the mountaintop samples are largely influenced by the sulfide minerals which exhibit strong positive correlations between Hg, Sb, and Se.

The combination of standard element-element plots and factor analysis is a powerful approach for identifying the source(s) of contaminants. Given an adequate data set, this approach can be used not only for soil contamination, but also for air and water contamination. The derived information should prove useful in developing remediation strategies and identifying pollution sources.