ABSTRACT
Underground oil spills leaking into navigable waterways at Baltimore, Maryland, and Hilo, Hawaii, were investigated using soil-gas sampling methods. Gas extraction methods were used in Hilo with mixed but favorable results. Component adsorption methods were used in Baltimore with highly favorable results. In comparing the two soil-gas sampling methods and results, the component adsorption method was found to be superior in detecting the presence or absence of underground diesel oil. The gas extraction method produced results that were somewhat limited in use because of lack of sensitivity to the components of diesel oil. Interpretation of data from both methods was found to be complicated by the presence of surface spill contamination leading to false indications of underground oil.