Nitrate contamination in groundwater and bank filtrate is becoming serious in Korea and thus the needs for reliable and cost-effective treatment technology are present. In this study, biological autotrophic denitrification was implemented into the permeable reactive barrier (PRB) system to reduce the nitrate concentration in groundwater and bank filtrate. The biological PRB consisted of sulfur granules as an electron donor and autotrophic sulfer-oxidizing bacteria as a biological component. Limestone was also used to provide alkalinity. The objective of this study was to evaluate the potential applicability of an in situ biological PRB system to treat nitrate-contaminated groundwater and bank filtrate by laboratory scale tests and semi-pilot scale test.

  A mathematical model of the biological PRB was developed and evaluated by using a one-dimensional advection-dispersion equation and a half-order kinetic model. From the model simulation, the effective barrier thickness was found to be more dependent on the autotrophic denitrification kinetic constant than the influent nitrate concentration. The model simulation presented an autotrophic denitrifying reactive barrier design for removing nitrate.

  Laboratory scale tests were conducted to investigate the applicability of autotrophic sulfer-oxidizers in PRB system to remove nitrate. The results showed that the autotrophic sulfer oxidizers successfully colonized on the surfaces of the sulfer particles and removed nitrate efficiently. When influent nitrate concentrations were 30, 40, and 60 mg-N/L, half-order autotrophic denitrification reaction rate constants were 31.73×10-3, 33.3×10-3, and 36.4×10-3 (mg/L)^0.5min, respectively. Our data on the nitrate distribution profile along the column suggested that an effective wall thickness of the reactive barrier for autotrophic denitrification is about 30 cm when influent nitrate concentration is less than 60 mg-N/L.

  A microbial community in an autotrophic denitrification column was analyzed using molecular tools based on 16S rDNA. The autotrophic denitrifiers, Thiobacillus denitrificans-like bacteria, enriched in the column with time and were abundant at the bottom part with highest denitrification activity in the column. A dynamic change of a microbial community involved in the change of nitrate and sulfate concentration profiles in the biological PRB.

  The present study also described the effects of initial alkalinity and various solid alkalinity sources on nitrate removal in a sulfer-oxidizing autotrophic denitrification system. The results showed that denitrification rate increased as the initial alkalinity (i.e., about two times of theoretically required alkalinity) is needed to obtain a desirable sulfer-based denitrification reaction. Only 1.6 and 5% of initial nitrate remaind for oyster shell and calcite in seven days, respectively, but about 15% was still detected when dolomite was used.

  The long-term performance of the autotrophic denitrifying reactive barrier was investigated during 500 days of operation period. The 60 mg-N/L of influent nitrate was completely removed in the area with 30-50 cm from the bottom of the column and average of 81-95 % of nitrogen was recovered as nitrogen gas. Nitrate removal efficiency was stabel regardless of the change of the hydrodynamic characteristics in this study during the biological reactive barrier column operation.

  The results from the semi-pilot test showed that limestone as well as oyster shell can sufficiently provide a biological PRB with alkalinity, and moreover, indigenous microorganisms in oyster shell were found to have good denitrifying activity. In addition, an appropriate hydraulic residence time should be determined for better removal performance.

Keywords : Alkalinity, Autotrophic denitrification, Biological permeable reactive barrier, Mathematical model, Microbial community analysis, Nitrate, Thiobacillus denitrificans