2000 Hydrophobic Soils: Site Characterization and Tests of Hypotheses Concerning their Formation

This study aims to improve ability to predict and prevent the occurrence of soil hydrophobicity, and to guide the development of effective remediation solutions for hydrophobic sites. It was predominantly a field investigation, but extensive laboratory analysis of field samples was needed to meet study objectives.

In Phase I we developed an inventory of hydrophobic sites, complemented it with as much case history information as possible, and used this information to select sites to inspect. A total of 26 locations were made available. One site was a landfarm and we divided it into two sites, yielding a total of 27 sites. These were inspected from May to July 1999. The site locations were in an area extending from Calgary northeast to Cold Lake Alberta. In Phase II (July & August 1999), we returned to 12 of the 27 sites to obtain more detailed soil, vegetation and landscape information and collect soil samples for laboratory analyses. Grid soil samples were sampled at two layer depths on a grid pattern that covered both hydrophobic and non-hydrophobic areas. The grid samples were used together with information from detailed soil profile descriptions and profile soil samples taken from pits excavated in hydrophobic and adjacent non-hydrophobic areas. Phase II was completed during July and August 1999 and yielded 672 grid samples and 116 profile samples. Phase III analyses of soil samples were completed from September 1999 to April 2000.

Our conclusions are: I) Indigenous soil organic matter contributes to Dichloromethane Extractable Organics (DEO). In pristine soil with MED = 0 M, about 72% of the variability in DEO was predictable from Total Organic C (TOC) content. The equation is DEO (mg/kg) = 17.9 ´ TOC (g/kg) – 0.16, so approximately 1.8 % of soil TOC is extractable by dichloromethane in these samples. II) There is a significant positive correlation between soil hydrophobicity index (Molarity of Ethanol Droplet; MED) and DEO. The correlation between DEO and MED was found by independent analyses of profile sample data from 12 sites and grid sample data from four sites. III) Samples with DEO < 630 mg/kg can be expected to have an MED = 0 M. IV) It appears we can reject the hypothesis that vapor-phase sorption of volatile organics is necessary for the development of soil hydrophobicity because we found no consistent indication that oil contamination remains or was ever present underneath hydrophobic surface soil. V) The frequent occurrence of mottling, which indicates periodic anaerobic conditions or a shallow water table, might be related to development of soil hydrophobicity. VI) Soil hydrophobicity is a complex phenomenon and its expression should not be reduced to a few simplistic generalizations. For example, vegetation establishment in hydrophobic soils varies from vegetated to completely barren, texture varies from sand to silty clay, surface soil structure varies from powdery and fine platy to granular and blocky, and hydrophobicity can occur both at the surface and the subsurface.

This is the final report for Phases I to III.  It contains the background information relating to the problem, research plan, a summary of results, appendices and the project management structure.