Feasibility of phytoremediation for common soil and groundwater pollutants

Abstract

During the past two to three decades numerous studies reporting highly efficient remediation of contaminated soil and groundwater by plants have been published. The promises of phytoremediation has been great but till now the technology has not been widely applied and recognized, commercially and in a regulatory context, on par with other conventional soil and groundwater remediation technologies. This thesis elucidates the field of phytoremediation and addresses the lack of recognition of the technology. It aims to assesses the overall feasibility of phytoremediation and identify obstacles within the field. Further, it provides examples and suggestions of how to overcome these obstacles. The first part of the thesis scrutinizes the literature for data and experiences regarding application of phytoremediation and uncovers potential barriers and where the existing knowledge is insufficient. Further, it considers phytoremediation from a more technical perspective, setting up a mass balance for a generic plant-soil system. On basis of the data review and the mass balance application, an initial assessment of the feasibility of phytoremediation for common soil and groundwater pollutants is conducted. Several knowledge gaps and limitations were identified. Phytoremediation is restricted by phytotoxicity, root depth and long remediation time. Further, there is an evident lack of information about fate of contaminants in plants and the plant processes affecting uptake, excretion and metabolism of contaminants. These data are needed for mass balance modelling purposes. The mass balance revealed that if contaminants are degraded (kdeg ~ 0.01/day or higher) the governing loss of contaminant mass in the soil matrix is by aerobic microbial degradation in the rhizosphere and not by plant uptake. If the total removal of contaminant mass by degradation, direct volatilization and other unknown processes (not including plant uptake and leaching) is negligible, the governing removal process is plant uptake. However, removal by plant uptake is slow compared to rhizo-degradation and it is controlled by the soil matrix volume and the soil/water distribution coefficient (Kd). For compounds with high Kd (e.g. lead) phytoremediation may take immensely long time (>80,000 years). Extraction efficiency of plants is proportional to the soil volume which explains why pot experiments tend to under estimate the remediation time. The feasibility of phytoremediation for organic compounds with limited aerobic degradation (e.g. many chlorinated solvents) and low Kd (~10 L/kg) is more difficult to assess due to lack of knowledge of their fate. The second part of the thesis proposes two flow charts for assessment of the applicability of phytoremediation and applicability of phytoremediation techniques, respectively. They can be seen as a check list of things to consider before implementation of phytoremediation and are meant to ease the decision-support process. Further, they emphasize the need for phytotoxicity data, fate, metabolism- and mass balance studies. It is shown that existing phytotoxicity tests with higher terrestrial plants, not including seed germination, lack standardization and reporting requirements for easy comparison and transparency. The willow tree acute toxicity test by Trapp et al. (2000) is suggested as a candidate for a standardized test as it allows combined uptake, metabolism and toxicity studies. An overview of phytotoxicity data generated by the willow tree acute toxicity test is provided and compared to results generated by the standardized phytotoxicity tests on freshwater green algae and duckweed. No species could be classified as the most sensitive. It is proposed to establish a phytotoxicity database for standardized tests on terrestrial plants, not including seed germination. The fate and metabolism of trichloroethylene (TCE) in plants were evaluated by the willow tree acute toxicity test with chloride as indicator of dehalogenation. Willows and willows inoculated with TCE co-metabolizing strains of the plant endophyte Burkholderia cepacia were continuously exposed to TCE during a tree week period. Approximately 96% of the added TCE evaporated from solution and only 4% was taken up by the plants. Less than 3% of the added TCE was mineralized. To assess the feasibility of phytoremediation at a sodium fluoride contaminated site, the uptake of fluoride in plants were determined by the willow tree acute toxicity test. The mass balance revealed that the willows were capable of enzymatic removal of fluoride from the roots at low external concentrations. At high external concentrations the enzyme system collapsed and fluoride was taken up leading to further toxic effects. Phytoremediation was not assessed to be feasible at the site. The work encompassed in this thesis underlines that phytoremediation may occasionally be feasible. However, there are several limitations which strongly inhibit the applicability. Phytoremediation is assessed to be feasible for nutrients and organic pollutants which can be degraded aerobically. Most trace elements are too strongly bound to the soil matrix to be available for plant extraction and are thus not feasible for phytoremediation. For phytoremediation to become an accepted and recognized technology it must be proven by providing examples of field-scale applications where the soil quality criteria are reached within acceptable timeframes.