Performance and Microbial Characteristics Of A Novel Continuous Flow Nitrogen Removing Biofilter For Onsite Wastewater Treatment

Background:
Onsite wastewater is the major source of excess nitrogen loading in Long Island and other coastal areas, contaminating the groundwater and threatening the human health and aquatic ecosystem balance. Conventional soil-based passive biofiltration system (e.g. conventional nitrogen removing biofilters) can provide an economically feasible option for nitrogen removal from onsite wastewater. However, it requires large footprint to accommodate the design loading rate (0.75 gallons d-1 ft-2), therefore limited its treatment capacity. It is important to upgrade the system design to accommodate various hydraulic loadings while still achieves optimal nitrogen removal performance. Continuous flow nitrogen removing biofilter (CF-NRB) is proposed in this study, with optional pre-aeration pattern, variable internal recirculation, and continuous flow pattern. Our pilot-scale tests have demonstrated CF-NRB be a feasible solution to efficiently removal nitrogen from onsite wastewater with a smaller footprint.
Objectives:
Identify the optimum operational conditions including hydraulic loading rate (HLR), nitrogen loading rate (NLR) and recycle ratio for CF-NRB to efficiently remove nitrogen from onsite wastewater to the targeted level (10 mg N L-1);
Investigate the distribution and abundance change of functional species involved in nitrogen transformations in the CF-NRB at different operational conditions.
Methodology: A pilot-scale CF-NRB was constructed in the Water Research Innovative Facility (WRIF) at New York State Center for Clean Water Technology (NYSCCWT) (Figure 1). The CF-NRB consists of a pump tank, two replicate vertical flow nitrification columns, two replicate intermediate denitrification columns, and a final denitrification column. Both nitrification columns were packed with sand and all denitrification columns were packed with 100% woodchips. Septic tank effluent (STE) was mixed with the recycled effluent from intermediate denitrification columns in the pump tank and was continuously dispensed to the surface of nitrification columns. The intermediate and final denitrification columns were fully saturated with an up-flow hydraulic pattern. The intermediate denitrification effluent was recirculated to the pump tank by a peristaltic pump and the final denitrification effluent was discharged to the wet well for disposal. Water samples were collected from STE, pump tank effluent (PTE), two parallel nitrification columns effluent (NE), two intermediate denitrification columns effluent (IDE), and the final effluent (FE) during the experiment. Each sample was analyzed for the following parameters: - Dissolved oxygen (DO), pH, and temperature were measured onsite by a Sension+ MM150 multi-parameter meter. - Alkalinity was measured according to the standard method (U.S. EPA Method 310.1). - NH4+, NO2, NOx and total Kjeldahl nitrogen (TKN) were analyzed by a Lachat QuikChem 8500 autoanalyzer. - Chemical oxygen demand (COD) was measured by a Hach DR 6000 spectrophotometer. At the meantime, solid samples in both sand nitrification columns and woodchip denitrification columns were collected using a push-core sampler. Then genomic DNA was extracted from the sand and woodchip samples. The overall biomass (16S rRNA), two nitrifying functional genes (amoA for AOA and amoA for AOB) and three denitrifying functional genes (nirS, nirK, and nosZ) were measured via quantitative PCR (qPCR).
Status: The pilot-scale CF-NRB was started on October 2019 and has been successfully operated for over 2 years. During the entire experimental period, the CF-NRB was continuously fed with STE from nearby residential houses at various NLRs (1.1 - 8.6 g N m-2 d-1), HLRs (0.75 - 3 gallons d-1 ft-2) and recycle ratios (2 - 3). The optimum operational conditions for CF-NRB was identified (Table 1). In addition, the system was shut down twice (December 25th 2019 to January 11th 2020, and January 1st 2021 to March 6th 2021) to test the potential of CF-NRB treatment performance to recover from a long-term idle condition. Findings: The results of our study demonstrated that, the proposed CF-NRB, with a continuous flow pattern and internal recirculation, was able to achieve efficient total nitrogen (TN) removal (80.1% - 97.5%) from STE at various HLRs (0.75 - 3 gallons d-1 ft-2), nitrogen loadings (1.1 - 8.6 g N m-2 d-1) and recycle ratios (2 - 3) with low effluent TN levels (0.7 - 13.6 mg N L-1) (Table 1 and Figure 2). The prominent nitrogen removal performance and the small footprint make it an ideal candidate for onsite wastewater treatment at locations with shallow groundwater tables and/or limited space. The nitrification performance of CF-NRB was less impacted by the HLR and NLR change, while the overall denitrification efficiency was greatly impacted by the pump tank TN removal performance at high HLRs and carbon availability at high NLRs. Recycle ratio increase enhanced NH4+ removal performance, while limited impact was observed on TN removal. In nitrification columns, the overall biomass (16S rRNA) and nitrifying microorganisms (AOA and AOB) decreased as the depth increased. However, in denitrification columns, homogeneous distributions of biomass (16S rRNA) and denitrifying microorganisms (nirS, nirK and nosZ) were observed. Significance: This study proposed a novel onsite wastewater treatment system (OWTS) with unique continuous flow pattern and recirculation to achieve both efficient N removal performance and small footprint. At the meantime, the module-based design allows CF-NRB to be flexibly implemented in most environmental conditions with affordable cost. Compared with conventional OWTSs such as recirculating sand filters (RSFs), constructed wetlands (CWs), conventional nitrogen removing biofilters (NRBs) and the hybrid adsorption and biological treatment system (HABTs), CF-NRB showed higher TN removal efficiency at high HLRs (Christopherson et al., 2005; Fan et al., 2013; Rodriguez-Gonzalez et al., 2020; Gobler et al., 2021) (Table 2). Conventional NRBs showed stable and efficient nitrogen and biochemical oxygen demand removal performance. However, the low designed HLR required a large footprint and it substantially enhanced the construction cost (Gobler et al., 2021). Although RSFs and CWs can accommodate slightly higher HLR than the NRBs, the wide application was also limited by the insufficient aeration capacity, insufficient TN removal efficiency, and high maintenance frequency (Saeed and Sun, 2012). Furthermore, the treatment performance of CWs and RSF was significantly restricted by high HLR, only 40 - 60% of TN removal efficiency were achieved with HLR ranged from 2 to 4 gallons d-1 ft-2 (Christopherson et al., 2005; Saeed and Sun, 2012). The HABTs could achieve stable nitrogen removal performance by using elemental sulfur as an external electron donor for denitrification at high HLR (Rodriguez-Gonzalez et al., 2020). However, the release of SO42- in final effluent (20 - 60 mg L-1) may pose a potential pollution to the surface or groundwater.