Characterizing transient and seasonal responses of bioreactors for subsurface drains

The quality of Earth's water is a most pressing environmental issue. Nutrient runoff from agricultural fields into water bodies is of particular concern in the Midwest, where increased nitrate leaching into the Mississippi River has been identified as a major contributor to growing hypoxia in the Gulf of Mexico. Scientists agree that solutions to the hypoxia problem lie in the use of a combination of practices to reduce nitrates and phosphates from subsurface (tile)-drained farmlands. The use of fixed-bed, in-field tile bioreactors is one such practice being investigated. In-field bioreactors are trenches filled with carbon material (usually wood chips) which serve as a medium for denitrifying bacteria to grow, thus reducing the amount of nitrate that enters water bodies from tile drains. Bioreactors provide many advantages that make them a strong candidate for implementation as a best management practice across the Midwest: they use proven technology, require no modification of current practices, require little to no maintenance and last for up to 20 years. This study served to characterize the response of three in-field, wood chip tile bioreactors to various maintained hydraulic retention times (HRTs) and seasonal temperature variations over time. Controlled daily field experiments were conducted roughly one to two times per week from July to December 2012, during which time samples of bioreactor influent and effluent waters were collected and analyzed for a variety of water quality parameters. Changes in water quality parameters as a result of bioreactor treatment were observed, and the potential for these changes to be harmful to the local environment was assessed. Predictive models were developed which related the independent variables of HRT and influent water temperature to bioreactor performance. The average load reduction achieved over all runs was 63.1%, with average minimum and maximum load reductions of 20.1 and 97.5%, respectively. Efficiency of nitrate load removal decreased as flow rate increased. In cases when nitrate-N influent concentrations were not limiting, average removal rate was determined to be 11.6 g NO3-N m-3 d-1, with minimum and maximum removal rates of 5 and 30 g NO3-N m-3 d-1, respectively. Both percent nitrate-N load reduction and removal rate decreased over time as temperatures decreased, regardless of HRT. During initial runs, effluent concentrations of dissolved reactive phosphorus (DRP) and dissolved organic carbon (DOC) were exponentially greater than influent concentrations. However, effluent concentrations of DRP and DOC stabilized to near influent concentrations within one month of bioreactor operation. Methylation of mercury occurred during the first month of bioreactor operation, when nitrate-N concentrations were nearly zero. However, no methylation occurred when influent nitrate concentrations were above 10 mg L-1, even when sulfate reduction was occurring. Due to the limited number of methylmercury data collected during this study, more bioreactor studies which incorporate methylmercury data and are carried out over a longer period of time are recommended in order to determine whether methylation of mercury in bioreactors should remain a topic of concern. A robust exponential multiple regression model was developed to predict percent nitrate-N load reduction and removal rate. The percent nitrate-N load reduction model was used to develop a method to designate monthly target HRT values for any site, which could then be integrated into an existing interactive routine for the sizing of subsurface bioreactors. The inclusion of monthly HRT values, as opposed to one annual value, allows the program to more accurately size bioreactors, as well as more accurately predict annual nitrate-N load reduction. However, more studies relating bioreactor input parameters to bioreactor performance should be carried out throughout various locations in order to more accurately predict bioreactor performance. This study serves as one of few studies that: relates multiple bioreactor input parameters for the development of a model to predict bioreactor performance, reports the transient response of bioreactors from initial startup, and records methylmercury production and sulfate reduction in field-scale bioreactors. These results demonstrate that bioreactors to treat subsurface tile drainage are an effective means to reduce nitrate loads from agricultural fields, while producing minimal adverse impacts on water quality