Bioremediation of oil-contaminated soil : a South African case study.
In 1990, an oil recycling plant situated in Hammarsdale, South Africa, was decommissioned and a decision was taken by management to rehabilitate the site in preparation for resale. The heavily impacted area covered over two hecatares and oil contamination penetrated soil to depths in excess of three metres, making excavation and removal of the soil very expensive. The options for remediation of the site were limited. No facility for incineration of contaminated soil exists in South Africa, and landfilling was not permitted. The emphasis in developing a remediation strategy, therefore, focussed upon the possibility of in situ remediation with minimal excavation of soil. This study, the first of its kind in South Africa, was subsequently initiated to assess the feasibility of this approach, the results of which would underpin a full-scale cleanup programme. The development of such a strategy involved four key stages of work : (1) a comprehensive site investigation to evaluate and fully understand the particular problems at the site; (2) treatability studies to determine the potential for biological treatment of the contaminated soil and the optimisation of such treatments, particularly in terms of time and cost; (3) the testing of some of the more effective treatments on a pilot-scale; and (4) recommendations for full-scale bioremediation of the contaminated site. various conditions unique to South Africa had to be considered at each stage viz. the lack of funds and remediation experience, which created numerous problems and emphasised the requirement for a simple, "low-tech" approach. Site investigations revealed that in situ remediation may be possible due to the high permeability of the sandy soils and low concentrations of heavy metals. Laboratory experiments also showed that a mixed association of indigenous microorganisms was present which, once stimulated by nutrient supplementation at C:N:P, ratios of between 10:1:1 and 20:1:1, was capable of degrading total petroleum hydrocarbons at an average rate of 11% week -1. Further experimentation, aimed at reducing the cost of remediation and improving the soil quality, focussed on the efficacy of oil solubilisers, a soil ameliorant (composted pine-bark), indigenous fungi and higher plants in the remedial process. Three commercial surfactants (Arkopal N-050, N-060 and E2491) and one natural solubiliser (soybean lecithin) were tested for their biotoxicity, solubilisation and biodegradability at various concentrations (0.01 - 1.0%). Formulation E2491 was able to support a microbial population and was selected as the preferred commercial surfactant if soil washing was to be recommended; however, lecithin was considered to be more useful in situ because of its localised solubilising effect, biological origin and nutritional contribution. The use of fungi was of particular interest in addressing the persistent organic compounds, such as the heavy fractions of oil, for which bacterial remediation methods have been slow or ineffective. While it was not possible, however, to demonstrate in the laboratory that the indigenous fungi contributed significantly towards the degradation of the contaminating oil, the basic trends revealed that the fungal component of the indigenous microbial population was readily stimulated by the addition of nutrient supplements. The bulking-up process was also a success and additional exploratory work was proposed in the form of a larger scale composting design. Finally, the potential for using higher plants and 20% (v / v) composted pinebark (in addition to nutrients) to increase the microbial degradation of the contamination was investigated in both greenhouse and field plot studies. Greenhouse investigations employed soybeans which were postulated to have soil quality and cost benefits. However, although the soybeans were found to significantly enhance the remedial process, the complex soil-contaminant- plant interactions gave rise to strange nutritional effects and, in some cases, severe stunting. In contrast, the field studies employed grasses that had previously established on the site and which ultimately demonstrated a better tolerance for the contaminated conditions. Scanning electron microscopy revealed that there were considerable differences between the root tips of soybean plants which had been grown in contaminated soil and those which had been grown in uncontaminated soil. It was concluded that toxicity symptoms, which are readily observed in the root, could be used as an early indicator for determining the suitability of vegetation for remediation purposes. In both instances, despite the differences, the addition of composted pine-bark and nutrients (nitrogen and phosphorus) resulted in total petroleum hydrocarbon reductions of >85%, illustrating the benefits of plant establishment and oxygen availability. The need to link results from laboratory or pilot-scale experiments to achieve reliable predictions of field-scale behaviour was an essential component of this research. The results of the field study provided evidence, similar to that found in the pot trial, of the accelerated disappearance of organic compounds in the rhizosphere. All experiments incorporated parallel measurements of hydrocarbon residues, microbial activity and pH changes in the contaminated soil, the results of which strongly supported the argument that biodegradation was the dominant component of the remediation process. Thus, after consideration of the significant interactions which dominated the study (time-contaminant-nutrient; time-contaminant-pine-bark; and time-contaminant- pine-bark-plant), it was clear that, aside from these limiting factors, little should preclude the in situ bioremediation of the impacted soil.