Modeling to improve crop yields; however, excess nitrogen

        Modeling and Validation of
Denitrification Bioreactors for Agricultural


We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now



nitrogen contamination is an increasingly worldwide threat to terrestrial,
aquatic and atmospheric environments. (Galloway et al., 2003, 2008). Nitrogen
fertilizers have been benefited in modern and intensive agriculture in order to
improve crop yields; however, excess nitrogen (N) cannot be taken up by the
roots of plants. Throughout the poorly-managed agricultural practices, hence, excess
amount of nitrogen is transferred from subsurface (tile) drainage systems to
local surface water bodies. The most widely approach to define the nitrate
attenuation process in the subsurface flows is the heterotrophic denitrification,
which includes the conversion of nitrate to nitrogen gases with the help of
carbon (C) source as the electron donor (Rivett et al., 2008). Thus, the
availability of degradable carbon source makes the denitrification reaction
possible on the pathway of the subsurface drainage flows before it enters
aquatic environments. At this point, the most promising solution against
nitrate loads in tile drainage systems is denitrifying bioreactors. They are an
edge-of-field treatment process reducing nitrate concentration in runoff and
tile drainage flows as a part of agricultural best management practices.


Review of Literature


nitrate attenuation performance of a denitrifying bioreactors depends on a
number of parameters including hydraulic retention time in the reactor,
reaction kinetics, temperature, carbon media, influent nitrate concentrations.
Based on one of the most recent studies conducted by Christianson et al.
(2012), the hydraulic retention time which is greatly subject to the flow rate
of subsurface drainage, porosity of the carbon media and bioreactor size ratio
is likely to have a crucial impact on nitrate removal rates. Estimating optimum
operating conditions on the basis of retention time for a bioreactor is
significant since low retention times may not be adequate so as to decrease the
concentration of dissolved oxygen of influent drainage to a certain level which
enables the denitrification process to proceed (Christianson & Helmers,
2011; Robertson and Merkley, 2009). Quite higher hydraulic retention times, on
the other hand, can contribute to remarkable NO3? removal
rate, but also excessive retention times have been correlated to undesirable side
effects like sulfate reduction and the methylation of mercury (Van Driel et
al., 2006; Hudson and Cooke, 2011).

field and lab studies related to nitrate attenuation reactor kinetics have
suggested that the chemical mechanism of denitrification process can fit both
zero and first order kinetics depends upon the nitrate concentration of tile
drainage outflow. By and large, it is assumed that denitrification reaction obeys
zero-order kinetics under the conditions in which influent nitrate
concentration is between 1 mg NO3-N/L and 20 mg NO3-N/L
(Van Driel et al., 2006). Elevation in nitrate removal rates is predominantly
attributed to tile drainage water temperature. Denitrification reactors are
claimed to be operated properly for the conversion of nitrate into nitrogen gas
when bed temperatures vary from 2 and 20 degrees C (Schipper et al., 2010). Not
surprisingly, many studies suggest that nitrate removal rates are directly
proportional to the temperature of drainage water entering a bioreactor within
approximately the temperature range noted above. (Cameron and Schipper, 2010; Van Driel et al;
2006; Christianson, 2011). The performance of a variety of denitrification fill
material including wood chips, corn cobs, wheat and barley straw has been
analyzed successfully throughout past field tests and laboratory trials. Considering
the field application of denitrifying bioreactors, woody chipped media has been
the most widely utilized type of carbon source due to its longevity (5–15
years) with relatively constant
removal and minimum maintenance requirement. (Schipper et al., 2005,
2010; Robertson et al., 2000). Compared to smaller sized woody particle media
(i.e. Sawdust) with lower porosity, a larger sized carbon type (such as wood
chips) can provide greater nitrate removal performance with the help of longer
retention times. (Ranaivoson et al., 2012).



major objectives of the project will be simulating subsurface drainage effluent
entering the bioreactor in terms of volumetric flow rates and modeling nitrate
removal performance throughout the monitoring period beginning from March to
May 2018 in 2018. Drainage effluent flow criteria can be estimated through a
hydrological model which has been created to simulate subsurface flows. The
simulated tile outflow data could be derived with the widely used field-scale
deterministic model DRAINMOD (Skaggs, 1978,1982). For the study area, the
essential input data so as to proceed DRAINMOD could be mainly categorized into
five groups: soil, drainage system, crop, weather and trafficability
parameters. Site-specific soil hydraulic parameters, especially, have a profound
impact on the accuracy of the model including soil water characteristics curve,
hydraulic conductivity, and soil water content at saturation and lower limit
(Skaggs et al., 2012). Estimating field-based soil-water characteristics calls
for long, difficult and expensive measurements. Thereby, hydrology experts are
prone to take advantage of Rosetta (Schaap et al., 2001) computer software
based upon pedotransfer functions (PTFs). As an alternative way to predict
required soil input parameters for DRAINMOD, Rosetta model requires users to utilize
web-based soil survey tools in order to obtain soil textural and bulk density

results of subsurface drainage flow rate which is the final product of DRAINMOD
software are planning to be utilized as an input for the development of
bioreactor nitrate removal model. To date, many studies have investigated that denitrification
reaction is expected to obey Michaelis-Menten kinetics (Schipper et al., 2010).
Robertson (2010) indicates that rising NO3-N concentrations (3.1–49 mgL?1)
in drainage effluent do not lead to increased nitrate attenuation. This founding
can be considered as an indicator of the approach that denitrification reaction
is supposed to be controlled by an independent parameter regardless of bioreactor
influent nitrate concentration. Therefore, under the conditions stated above
nitrate transformation could be effectively modeled based upon zero-order
kinetics (Schipper et al., 2010; Robertson et al., 2000).

4. Field Measurements and

the evaluation of nitrate removal model, a denitrification drainage bioreactor
was previously constructed at the edge of an agricultural field tile drainage
system located in the Oregon State Experimental Dairy Farm. Water samples will
be collected from inlet and outlet control structure of bioreactor periodically
(once a week) by ISCO samplers (Model 2900) from March through May 2018 when
tile drainage is supposed to be at its highest. Following grab sampling, to
estimate NO3? removal efficacy of the bioreactor, water
samples taken from influent and effluent
this period will be tested for nitrate-nitrogen concentration. The NO3?
removal performance based on the difference in the nitrate composition of
inflow and outflow field samples monitored, could be utilized for the
validation of nitrate removal model. Simulated nitrate attenuation
results and the corresponding removal rates estimated based on field monitoring
measurements, can be used to evaluate the overall performance of the proposed
model. In the study, two widely used statistical metrics (Nash Sutcliffe
Efficiency and Root Mean Square Error) will be used to ensure accuracy and
reliability of the model.