A hydrogen-based membrane biofilm reactor （MBfR） using H2 as electron donor was investigated to remove nitrate from groundwater. When nitrate was first introduced to the MBfR, denitrification took place on the shell side of the membranes immediately, and the effluent concentration of nitrate continuously decreased with 100% removal rate on day 45 under the influent nitrate concentration of 5 mg NO3^--N/L, which described the acclimating and enriching process of autohydrogenotrophic denitrification bacteria. A series of short-term experiments were applied to investigate the effects of hydrogen pressures and nitrate loadings on deniWification. The results showed that nitrate reduction rate improved as H2 pressure increasing, and over 97% of total nitrogen removal rate was achieved when the nitrate loading increased from 0.17 to 0.34 g NO3^--N/（m^2.day） without nitrite accumulation. The maximum deniwification rate was 384 g N/（m^3.day）. Partial sulfate reduction, which occurred in parallel to nitrate reduction, was inhibited by denitrififcation due to the competition for H2. This research showed that MBfR is effective for removing nitrate from the contaminated groundwater.
Siqing XiaFohua ZhongYanhao ZhangHaixiang LiXin Yang
A laboratory trial was conducted for evaluating the capability of a continuously stirred hydrogen-based membrane biofllm reactor to simultaneously reduce nitrate （NO3--N）, sulfate （SO42-）, bromate （BrO3-）, hexavalent chromium （Cr（VI）） and para- chloronitrobenzene （p-CNB）. The reactor contained two bundles of hollow fiber membranes functioning as an autotrophic biofiim carder and hydrogen pipe as well. On the condition that hydrogen was supplied as electron donor and diffused into water through membrane pores, autohydrogenotrophic bacteria were capable of reducing contaminants to forms with lower toxicity. Reduction occurred within 1 day and removal fluxes for NO3--N, SO42-, BrO3-, Cr（VI）, and p-CNB reached 0.641, 2.396, 0.008, 0.016 and 0.031 g/（day.m2）, respectively after 112 days of continuous operation. Except for the fact that sulfate was 37% removed under high surface loading, the other four contaminants were reduced by over 95 %. The removal flux comparison between phases varying in surface loading and 1-12 pressure showed that decreasing surface loading or increasing 1-12 pressure would promote removal flux. Competition for electrons occurred among the five contaminants. Electron-equivalent flux analysis showed that the amount of utilized hydrogen was mainly controlled by NO3--N and SO42- reduction, which accounted for over 99% of the electron flux altogether. It also indicated the electron acceptor order, showing that nitrate was the most prior electron acceptor while sulfate was the second of the five contaminants.
