Mobility and Persistence of Petroleum Hydrocarbons in Peat Soils of Southeastern Mexico

Abstract

Spilled crude petroleum from oil wells contains numerous hydrocarbons, some of which are toxic and threaten life. We have studied the mobility and persistence of hydrocarbons in waterlogged soils that contain large proportions of fermented organic matter (Histosols) and large concentrations of dissolved organic carbon (DOC) in the State of Tabasco, Mexico. We sampled soil and phreatic water at sites polluted by oil spills for several decades, as well as at sites that had only recently (few weeks) been polluted, and compared their hydrocarbon contents with those of unaffected sites in the same area. Samples were analyzed for 16 non-alkylated polyaromatic hydrocarbons (PAHs) and n-alkanes from nC9 to nC34. The spilled hydrocarbons had remained predominantly in the organic surface horizons of the soil where spillage occurred; there was little evidence of movement within the soil. The fraction of low molecular weight compounds was larger at sites of recent spills than where spills happened several decades ago. Nevertheless, sites of old spills still contained large concentrations of hydrocarbons, among which those of low molecular weight represented from 30 to 49% of total PAHs and from 50 to 84% of total n-alkanes, indicating that volatilization or microbial degradation is slow in these soils. In the peat horizons the measured organic carbon partition coefficients (K _oc ) for the higher molecular weight PAHs were consistently smaller than those estimated by empirical equations by up to two orders of magnitude. The dissolved organic carbon of these peat soils seems to influence this behavior. At sites of old spills, partition coefficients for the PAHs were larger than at sites of recent spills.

Keywords:

• Soil contamination
• polycyclic aromatic hydrocarbons
• alkanes
• partition coefficients
• Histosols.

Acknowledgements

This investigation was financed by PEMEX (Petroleos Mexicanos) and was part of a large interdisciplinary research project performed within the alliance of the Instituto Mexicano del Petróleo (IMP), the Universidad Nacional Autónoma de México (UNAM) and Battelle. We thank Guillermo Calderón from IMP for coordinating the project, Dr. Calvin Ainsworth and Dr. Charles Brandt from Battelle for their valuable contributions to the sampling strategy and the discussion of results, Kumiko Shimada, Carmen Curiel, and Pedro Avilés from the Instituto de Geología (UNAM) for the routine analyses on soil and water samples, and Luis Alberto Olvera Vargas, Daniel Pinales Bravo and Manuel Hernández Quiroz for their support during field work. We want to also thank both reviewers; their appropriate comments helped us to substantially improve this manuscript.

Notes

∗Data for the horizon at which the actual water table was located. These data were used to calculate the Kd values.

∗Data for the horizon at which the actual water table was located. These data were used to calculate the Kd values.

∗Equation (4) yields negative values, since estimated K _oc by Equation (2) <K _oc-app ; these values were not considered in the regressions presented in Figure 6.

^a Kostecki et al., 2001

^b Cancino et al., 1998

^c Bulman, 1992.