Removal of nitrate in raw water using a vertical roughing filter with an external carbon source

Abstract

High accumulation of nitrate above the recommended maximum guideline value has become a common problem in most water supply sources. According to the World Health Organization (WHO), about 30% of water supply sources in the world exceed the maximum nitrate contamination level of 11 mg/L-N / 50 mg/L-NO3. Consumption of water with high nitrate concentration poses health hazards to both humans and livestock. Several technologies such as reverse osmosis and electrodialysis, have been adopted in removing nitrate from raw water. However, they have drawbacks that include the production of high strength residual brine and low efficiency. Nonetheless, biological denitrification has proved to be an effective technology for nitrate removal and the process can be enhanced by adding an external carbon source. Denitrification in roughing filters has not been widely studied, except in bio-filters and slow sand filters. This research aimed to investigate the efficacy of roughing filters enhanced by an external carbon source in removing nitrate in raw water. Two upward vertical roughing filters in series (UVRFs) were used, one was a vertical roughing filter with ethanol as a carbon source (VRFwt) and the other was a vertical roughing filter without a carbon source (VRFwo). The inflow and outflow of nitrate and other physicochemical parameters were monitored to evaluate their influence on a roughing filter’s performance in removing nitrate in raw water. The carbon: nitrogen ratios (C/N ratios) of 1.05, 1.08 and 1.1, were investigated, coupled with a nitrate removal kinetic model. Furthermore, filter design parameters and the effect of biomass on flow rate were also studied. The average nitrate removal efficiency in a vertical roughing filter with a carbon source was 88%, 70%, and 83%, for carbon: nitrogen ratios (C/N ratios) of 1.05, 1.08, and 1.1, respectively. The drop-in flow rate was 27% for a vertical roughing filter with a carbon source (VRFwt) and was attributed to the biological layer growth, whereas a 15% decline was observed in the vertical roughing filter without a carbon source (VRFwo). The decrease in flow rate was evident at 30-35 days from the start of the filter operation. The removal efficiency was 75%, 43%, and 46% at C/N ratios of 1.05,1.08 and 1.1, respectively. The residual ethanol measured as chemical oxygen demand (COD) in the filter with an external carbon source (VRFwt) ranged between 85 mg/L to 632 mg/L during the filter run. The average residual ethanol measured as COD during the filter rest period ranged between 41 mg/L and 561 mg/L with a removal efficiency of 88%, 49% and 53% at C/N ratios of 1.05,1.08 and 1.1, respectively. The overall average reduction of dissolved oxygen (DO) in the VRFwt at C/N ratios of 1.05, 1.08 and 1.1 was 42%, 54%, and 51% respectively, while DO reduction in the VRFwo was 17% 18% and 17%, respectively. A decline in DO was profound in the VRFwt compared to the VRFwo.

The VRFwt showed a high potential for removing nitrate in raw water for potable use. Therefore, when the VRFwt is applied at large scale, it will increase access to water sources that were initially rendered unsuitable to many water utilities due to high nitrate concentrations; thereby increasing their water supply. Importantly, the lack of nitrate in potable water would minimize water-related diseases induced by the use of high nitrate-rich water. Again, the reaction rate order (n) and reaction rate constant (k) determined from the nitrate removal kinetic model can help in assessing the total nitrate removal rate and efficiency in a vertical roughing filter, without the need to operate the filter, thus saving time and money.