Achieving Low Effluent Total Nitrogen Concentration

WEFTEC 2013 Achieving Low Effluent Total Nitrogen Concentration H.D. Stensel1*,J.B. Neethling2, D. Clark2, Tsuchihashi3, R. J. Sandino4, A. Pramanik5 1 Univ. of Washington, 2 HDR Inc., 3 AECOM, 4 CH2M Hill, 5 WERF *Email: [email protected] ABSTRACT Full scale performance data and special surveys have been done to evaluate effluent total nitrogen performance in state of the art biological nitrogen removal facilities, and the removal of individual nitrogen species during treatment. Long term removal of ammonia N and particulate effluent nitrogen can be reliably done to effluent concentrations of less than 0.20 mg/L. Effluent nitrate and nitrite N provide a greater contribution to effluent total N and concentrations of less than 1.0 mg/L of oxidized N are possible. Tertiary denitrification filters with external carbon addition produce minimum effluent oxidized N. The effluent soluble organic nitrogen is the largest contributor to effluent total nitrogen and its concentration can be quite variable from site to site. Adding external carbon for denitrification and solids separation increases the greenhouse gas (GHG) emissions from nitrogen treatment to achieve low effluent total nitrogen concentrations. At high levels of nitrogen removal, the GHG production/N removed ratio increase by orders of magnitude. KEYWORDS: Nitrogen removal, treatment performance, nitrogen species, soluble organic nitrogen, biological nutrient removal INTRODUCTION Nitrogen removal in wastewater treatment plants (WWTPs) has been accomplished in various process designs since the 1970s to meet different removal efficiencies as determined by site specific surface water quality and discharge permit levels. A common effluent discharge limit for the Chesapeake Bay area and coastal estuaries with impaired surface waters from excessive nutrient loads and eutrophication is an effluent total nitrogen (TN) limit of 3.0 mg/L. In a few locations in the U.S. discharge limits of 2.0 and 1.0 mg/L of TN have been required. In-stream nutrient criteria set by EPA require very low total nitrogen concentration in all water bodies, which are expected to lead to lower effluent limits in general for point discharges. The Water Environment Research Foundation (WERF) Nutrient Removal Challenge Program has supported direct and collaborative research projects that address the need to produce low effluent TN concentrations from WWTPs. Projects have evaluated process designs and nutrient removal treatment performance, the variability in nitrogen removal performance for real-world WWTPs, the use and selection of alternative external carbon sources to minimize oxidized nitrogen in biological nitrogen removal (BNR) system effluents, the concentration of nitrogen species in effluents from BNR systems designed and operated to achieve a minimal effluent TN concentration, sidestream treatment for nitrogen removal, the amount and persistence of soluble organic nitrogen (SON) in BNR effluents, and the characteristics and bioavailability of effluent soluble organic nitrogen to algae. This paper summarizes the results of these studies with regard to the nitrogen species in wastewater effluents, the ability of treatment processes to remove these species, the effect of process designs on meeting low effluent TN concentrations, the reliability Copyright ©2013 Water Environment Federation. All Rights Reserved. 409 WEFTEC 2013 of performance, and the implications for water quality and sustainability. Nitrogen Species and Removal Table 1 shows a summary of commonly used parameters to describe effluent nitrogen and nitrogen species. The various measurement methods and parameter calculation are also given in the table. Effluent permits may specify effluent nitrogen limits as one or more of the following constituents: ammonia-N, nitrate-N, oxidize nitrogen (NO3-N plus NO2-N), total inorganic nitrogen (TIN), and total nitrogen (TN). The term ammonia-N refers to total ammonia nitrogen (TAN) concentration, which is the sum of free NH3-N and ionized NH4-N in a sample. For typical effluent pH values in wastewater effluents, over 98% of the TAN is in the ionized NH4-N form. The TN concentration includes all forms of effluent nitrogen which are (1) ammonia-N, (2) NO3-N, (3) NO2-N, (4) soluble organic nitrogen (SON), and (5) particulate organic nitrogen (pON). Soluble vs. Particulate. The analytical definition of particulate (or filterable) is dependant on the pore size of the filter. Typically, a 0.45 um filter size is used to distinguish between particulate (filterable) and soluble (dissolved or non-filterable) species. Colloidal particles can sometimes pass through the filtration process and be measured as a soluble. With the exception of organic nitrogen contained in influent wastewater suspended and colloidal solids, and in effluent suspended and colloidal solids nitrogen is in the soluble form as SON or TIN, which includes ammonia-N, NO3-N, and NO2-N. Systems designed to achieve a minimal effluent TN concentration include tertiary filtration, tertiary membrane separation, or membrane bioreactors (MBRs) to remove most all of the particulate nitrogen. Inorganic Nitrogen. About 60 to 70 percent of the influent wastewater nitrogen is as ammonia-N and ammonia-N is further produced by the breakdown of influent particulate and soluble organic nitrogen. Biological treatment methods are the most common and cost effective means to remove the inorganic nitrogen. Biological oxidation of ammonia-N (nitrification) converts the ammoniaN to oxidized inorganic nitrogen (NOx), which includes NO3-N, and NO2-N. In the absence of oxygen and with organic substrate available heterotrophic bacteria carry out biological denitrification to reduce NOx to nitrogen gas. A schematic of nitrogen species contained in the effluent TN from BNR facilities designed to maximize nitrogen removal is shown in Figure 1. The key components contributing to the effluent total nitrogen concentration are (1) NOx-N, (2) ammonia-N, (3) pON, and (4) SON. The NOx-N concentration is minimized by using an external carbon source to maximize denitrification efficiency in a postanoxic reactor in the BNR system. The NH3-N concentration is minimized by having a long enough aerobic reactor SRT in the BNR system to maximize ammonia oxidation. The effluent pON nitrogen concentration is minimized by having a tertiary filter or membrane separation. The effluent SON concentration appears to be a product of the wastewater and treatment system as there is presently no design or operating strategy identified to maximize SON removal in a biological nitrification/denitrification process. As shown in Figure 2, the effluent SON can represent a large fraction of the effluent TN concentration for systems aiming to produce low effluent TN concentrations. Copyright ©2013 Water Environment Federation. All Rights Reserved. 410 WEFTEC 2013 Table 1. Nitrogen species terms and common measurement parameters. Common Measurement TN fTN TKN fTKN NH3-N1 NO3-N NO2-N Symbol (1) (2) (3) (4) (5) (6) (7) Calculate TN X (1) TN X X X (3)+(6)+(7) Total Inorganic N TIN X X X (5)+(6)+(7) Oxidized N NOx X X (6)+(7) Soluble Organic N2 SON2 X X X X (2)-(5)+(6)+(7) SON X X (4)-(5) Particulate Organic N pON X X (1)-(2) pON X X (3)-(4) Total Organic N TON X X X X (1)-(5)+(6)+(7) TON X X (3)-(5) TN-from persulfate digestion or chemiluminescence method, f-filtered. 1 Standard Methods refers to as ammonia-N but is total ammonia-N (TAN), the sum of NH3-N and NH4-N in sample, 2soluble organic N (SON) has also been referred to as dissolved organic N (DON) Nitrogen Term Total N Figure 1. Example of effluent nitrogen species in a biological nitrogen removal system effluent designed to achieve a low effluent TN concentration. Copyright ©2013 Water Environment Federation. All Rights Reserved. 411 WEFTEC 2013 Types of Treatment Processes for Low Effluent Nitrogen Removal A variety of BNR system designs have been used to achieve low effluent TN concentration. Schematics of different types or processes that are categorized in terms of (a) separate stage nitrification/denitrification, (b) combined nitrification/denitrification with membrane separation or secondary clarification, and (c) multistage denitrification. And separate stage nitrification/denitrification, the BOD removal, nitrification, and denitrification steps are all separate and sequential. The final step in Figure 2(a) involves external carbon addition to a denitrification fluidized-bed reactor prior to a sand filtration polishing step. This process game is employed at the Truckee Meadows water reclamation facility (TMWRF). The combined nitrification/denitrification systems shown in Figure 2(b) are basically Bardenpho process type designs with preanoxic and postanoxic reactors in a single sludge activated sludge system. External carbon is added to the postanoxic tank to maximize nitrogen removal performance. A tertiary filter is used after the system with secondary clarification to maximize the removal of pON. The multistage denitrification system in Figure 2c has nitrification/denitrification in the suspended growth process, such as an oxidation ditch or anoxic/aerobic (MLE process) activated sludge system followed by a denitrification filter with external carbon addition to polish the effluent inorganic nitrogen to minimal concentrations and to filter pON. Figure 2. Types of nitrogen removal processes evaluated. Performance and Reliability The evaluation of the performance capability of BNR processes must consider the fact that for a given design and operation the effluent concentration of nutrients and other substances from WWTPs are not constant and vary due to a number of factors. Variations in effluent Copyright ©2013 Water Environment Federation. All Rights Reserved. 412 WEFTEC 2013 concentrations can be due to variable influent loading conditions, external impacts due to weather changes, industrial discharges and other inputs, and internal impacts from process changes, equipment failures, construction activities and other impacts. Neethling et al. (2007) presented a statistical approach to quantify these factors. Bott and Parker (2011) adopted this approach to develop technology performance statistics (TPS) values to describe process performance. Effluent data is analyzed on a probability scale or by rank to provide a quantifiable measure of reliability. Error! Reference source not found. illustrates the data analysis of the reliability of achieving low effluent ammonia-N concentration from analyzing 3 years of performance data. The statistical presentation of performance also proves a measure of the reliability to meet permit. For example, the process performance median concentration (TPS-50%) is indicative of the average performance; statistically, this level of performance is exceeded 50% of the time. However, a treatment facility operating under an annual permit limit must do better than average. Otherwise it also has a 50% chance of failure. On average, every two years it would exceed the effluent limit. Similar, while maximum month TPS concentration is represented by the 91.7th percentile (11/12 = 91.7%), a higher reliability is required to meet permit consistently. 10.00 1.00 mg/L 0.17 0.10 95% 0.05 50% 0.01 0.007 3.84% 0.00 0.1% 1% 10% 25% 50% 75% 90% 99% 99.9% Percent of values less than of equal to indicated value BNR NH4 Figure 3. Example statistical analysis of ammonia-N data illustrating the performance variability and reliability of 50th and 95th percentiles. This facility produced a median ammonia-N concentration of 0.05 mg/L. Two key statistics has been used to represent reliable treatment: the 80th percentile would be representative of the concentration that can be achieved on an annual basis with a risk of exceeding it once in a 5-year period (20% of 5 annual values); the 95th percentile is indicative of a monthly concentration with a risk of exceeding it three times in a 5-year period (5% of 60 Copyright ©2013 Water Environment Federation. All Rights Reserved. 413 WEFTEC 2013 monthly values) (Neethling and Stensel, 2013; Jimenez, et al., 2007). These statistics show that the reliable performance for monthly and annual ammonia limits for this facility is 51 and 140 ug SRP/L respectively. The statistical data can also be used to determine the reliability of achieving a certain performance. For example, the data in Figure 3 indicates that a concentration below 70 ug/L is achieved 85% of the time or exceeded 15% of the time. On an annual permit basis, 70 ug/L would potentially be exceeded 0.75 times in a 5 year period (15% of 5 times = 0.75 times); i.e. less than once in a 5 year period. For a monthly permit, the same performance would risk exceeding a monthly permit 9 times (15% of 60 months) in 5 years. This means that the level of performance may be acceptable for meeting an annual permit limit, but risky to meet a monthly permit limit. Environmental Sustainability Sustainable practices involved choosing alternatives to meet a necessary goal that maximizes the amount of resources left for future generations and minimizes negative impacts on the environment that can affect the quality of life for future generations. Wastewater treatment designs to maximize nitrogen removal results in a dilemma between protecting and improving water quality and having a negative impact on sustainability because of the use of external carbon and more energy to meet higher treatment goals. Falk et al. (2010) evaluated the impacts of greenhouse gas (GHG) emissions, as a surrogate for sustainability, on increasingly more stringent nitrogen removal limits for a typical 37,900 m3/d (10 mgd) WWTP. The study accounted for operational GHG emissions from energy, chemical addition, solids hauling and disposal, nitrous oxide emission, and credit for cogeneration. Bioavailability of Effluent Nitrogen In view of the larger fraction of SON in BNR facilities design to achieve a low effluent TN concentration, a WERF study was carried out to investigate the bioavailability of effluent SON to algae (Sedlak et al., 2013). An algal bioassay procedure was carried out using the freshwater alga Pseudokirchneriella subcapitata (formerly Selenastrum capricornutum) for inocula to shaker flasks with BNR WWTP effluent samples and maintained under light for a 14-day incubation period at 250C. The algae growth was correlated to the bioavailable nitrogen concentration in the effluent samples. A resin extraction technique was developed that separated the effluent SON into a hydrophilic or hydrophobic fraction. The hydrophobic fraction was not bioavailable to algae. RESULTS Effluent Total Nitrogen Concentration Performance Effluent TN performance is reviewed for WWTPs designed and operated to meet low effluent TN permit levels. The effluent TN performance shown in Table 2 is from a WERF study to evaluate long term performance statistics for nutrient removal (Bott and Parker, 2010). The plants selected were screened on the ability to provide three years of performance data. The plants shown in Table 2 are for those with stringent permit levels for effluent TN, ranging from 2.0 to 4.0 mg/L. All of the facilities had tertiary denitrification in denitrification filters, a fluidized bed reactor or activated sludge and clarifier system. Four of the facilities had an effluent suspended solids polishing step which further removed particulate nitrogen. Methanol Copyright ©2013 Water Environment Federation. All Rights Reserved. 414 WEFTEC 2013 was the external carbon source for all the facilities. The results show a wide range in effluent TN performance and the 50 percentile (media value) suggests that for some of the plants there were many days in the 3-year period with effluent TN concentrations below 1.0 to 2.5 mg/L. The 95 percentile data shows that there are also some days for which the effluent TN concentration can exceed 2.7 to 4.2 mg/L. Table 2. Effluent performance statistics for WWTPs with low effluent TN concentration permit (T-warm or cold temperature, DF-denitrification filter, FBR-fluidized bed denite filter, A.S.-activated sludge). WWTP Fiesta Village River Oaks Truckee Meadows Western Branch Tahoe Truckee Scituate T W W C C C C N Removal Process Secondary Tertiary Final Denite Denite Filter Oxid. Ditch DF DF Yes FBR Yes A.S. Yes DF Yes DF Effluent TN, mg/L Permit Performance Statistic Limit 50% 95% 3 1.03 2.71 3 1.45 2.92 2 1.57 2.85 3 1.47 3.20 3 2.50 3.37 4 2.37 4.22 Though 3 years of performance data was not possible, the data shown in Figure 4 shown similar effluent TN performance for a Bardenpho-MBR facility (RU) as that for the Truckee Meadows facility. For this more recent set of data for the TM facility, the 50 percentile effluent TN concentration is 1.7 and 2.2 mg/L, respectively, while their 95 percentile values are similar at about 3.4 mg/L. The RU facility has a similar climate as the TM facility but does not yet use methanol as it is operating at about 70 percent of its design capacity. The aeration is cycle on/off between 0.0 and 0.50 mg/L to accomplish significant nitrogen removal before the downstream anoxic and aerobic membrane separation zones. 7.0 TN Concentration, mg/ L 6.0 5.0 4.0 3.0 B-MBR 2.0 TM 1.0 0.0 0 20 40 60 80 100 Percent of values equal to or less than Figure 4. Statistical performance comparison of a Bardenpho-MBR (B-MBR) WWTP and Truckee Meadows (TM) WWTP for effluent total nitrogen concentration. Copyright ©2013 Water Environment Federation. All Rights Reserved. 415 WEFTEC 2013 Effluent Nitrogen Species Concentration Performance Statistical performance data for effluent nitrogen species is for the Fiesta Village WWTP are shown in Figure 5. The effluent ammonia-N concentration is generally very low with 75% below 0.10 mg/L and 95% below 0.20 mg/L. The curve is steeper above the 95% value to indicate excursion in effluent ammonia-N concentration, with possible causes a significant change in plant operation, equipment issues, or a large influent load variation. The effluent NOxN concentration is also low for a large portion of the effluent data with 75% below 0.10 mg/L and 90% below 0.50 mg/L. Above the 90 percentile the curve is steeper to again indicate some excursion in performance from normal operation. The effluent SON concentration (most of the effluent organic-N can be assumed to be soluble in view of the tertiary filtration step) behaves differently with a flatter curve and perhaps showing a performance that is less sensitive to plant operating changes. The SON concentration is very close to the effluent TN concentration and the curves tend to track together. The statistical performance graph (Figure 6) for the Bardenpho-MBR WWTP shows a similar pattern for effluent NOx-N concentration as that for Fiesta Village but at a higher concentration, which may be due less reliance on methanol addition. The 50% values is about 0.70 mg/L with an increase to about 1.2 mg/L at the 90% statistic. The curve is steeper above the 90% point, which again indicates a small frequency of excursion in effluent NOx-N concentrations. The effluent ammonia-N concentration is not plotted due to less data points available. The data in general showed very low effluent ammonia-N concentrations near 0.10 mg/L. Again the effluent SON concentration curve is flatter and again shows no major excursions as was seen for the effluent NOx-N concentrations. A comparison of effluent nitrogen species concentrations are shown in Table 3 for WWTPs achieving low effluent TN concentrations. The effluent ammonia-N concentration can be very low and represent a small fraction of the effluent TN. Suspended growth technologies, in particular those with multiple-stage reactors or those operating in warm weather, are able to reliably achieve very low ammonia concentrations. Table 3. Comparison of effluent nitrogen species concentrations at the 50 percentile statistic for WWTPs achieving low effluent TN concentrations (the sum of the N species concentrations may not equal the TN concentration due to not all components sampled on the same day). Effluent concentration, mg/L WWTP Fiesta Village Truckee Meadows Western Branch Tahoe Truckee B-MBR NH3-N 0.01 0.05 0.04 0.28 0.20 NOx-N 0.03 0.11 0.64 0.43 0.70 SON 0.9 1.3 0.7 1.7 1.2 TN 1.0 1.6 1.5 2.5 2.2 The effluent NOx-N concentration can also be very low and for all the WWTPs shown here it is well below the effluent SON concentration. These data show that the facilities without tertiary denitrification filters (Western Branch and the Bardenpho-MBR) had the highest effluent NOx-N Copyright ©2013 Water Environment Federation. All Rights Reserved. 416 WEFTEC 2013 N Species (mg N/L) 1 0.1 NH3-N Org N NOx-N TN Ln-Normal Values 0.01 0.001 0.01 0.1 1 10 30 50 70 90 99 99.9 99.99 % of Values Less Than or Equal to Indicated Value Figure 5. Statistical performance summary of effluent TN and nitrogen species for Fiesta Village, Florida 6.0 TN DON Nitrogen Concentration, mg/ L 5.0 NOx 4.0 3.0 2.0 1.0 0.0 0 20 40 60 80 Percent of values eq ual t o or less th an 100 Figure 6. Statistical performance summary of nitrogen species concentrations in the Bardenpho-MBR WWTP effluent. concentrations. Facilities that achieve very low nitrate concentrations all use carbon addition and typically employ tertiary denitrification processes. The tertiary denitrification process with carbon addition has an added reliability feature with the ability to adjust chemical dose. Copyright ©2013 Water Environment Federation. All Rights Reserved. 417 WEFTEC 2013 The data summary clearly shows that effluent SON is a major contributor to effluent TN concentration for systems meeting low effluent TN concentrations. For these data the effluent SON concentration ranged from 47 to 90 percent, with the highest being for Fiesta Village, which had very low effluent ammonia-N can NOx-N concentrations. The SON values in this study are within the range of values seen for other BNR facilities. A survey compiled by Stensel (2008) showed that SON values can range between 0.7 and 2.5 mg/L in nutrient removal WWTP effluent. The effluent SON concentrations do not appear to be related to the BNR process design and may in fact be a function of influent SON substances (Sedlak et al., 2013). Characteristics of Effluent SON In view of the significant contribution of SON to effluent TN concentrations for BNR systems intent on maximizing nitrogen removal performance, the bioavailability of SON to algae in receiving water was of interest. Bioassay tests were used to evaluate SON uptake by algae with effluent samples from ten WWTPs (Sedlak et al., 2013). Developments in the testing procedure also led to the finding that most of the effluent SON is hydrophilic, which was completely bioavailable. The hydrophobic fraction ranged from 12 to 28 percent and was not bioavailable in the algal assay tests. Evaluation of the carbon to nitrogen ratios for the hydrophobic DON suggested that humic substances may be associated with the hydrophobic DON but not the hydrophilic DON. GHG Emissions and Achieving Low Effluent TN Concentrations Falk et al. (2010) calculated the equivalent GHG emissions for 5 treatment levels of nitrogen removal. Level one for reference is wastewater treatment for BOD removal and no nitrogen removal. Subsequent levels represent increased treatment to meet lower effluent TN concentration goals, with effluent TN concentrations less than 8, 6, 3, and 2 mg/L for Levels , 2, 3, 4, and 5, respectively. Increased energy is needed for nitrification and an external carbon source is assumed for Levels 3, 4, and 5. Level 5, however, must deal with methods to removal effluent SON, and technologies considered for that based on today's treatment process knowledge are ultrafiltration or reverse osmosis with deep well injection of the spent brine. This analysis did not account for the water loss due to brine discharge. Figure 2 shows the metric ton equivalence per year for the 10 mgd treatment plant as the nitrogen removal requirements increase. Figure 3 provides the incremental GHG emission as ratio of the mass N removed as the treatment level change. As the treatment level increases, the incremental GHG emissions ratio rises rapidly. The rapid rise seen in this figure relates to the fact that the mass of N removed gets smaller and smaller, while the cost of treatment increases higher and higher as treatment level increases. The cost (GHG emission) to benefit (reduced N) ratio changes more than an order of magnitude for every level of nutrient removal applied. Copyright ©2013 Water Environment Federation. All Rights Reserved. 418 WEFTEC 2013 CO2 equivalent tonnes/yr 18,000 N2O Emissions (w/Data Range as Bars) 16,000 Aeration 14,000 12,000 Pumping/ 10,000 8,000 Chemicals Mixing Biosolids Hauling and CH4 Emissions Deep Well Injection Miscellaneous Cogeneration 6,000 4,000 2,000 0 -2,000 Incremental GHG Emissions per Nutrient Load Removed (CO2 eq kg/N or P kg) Figure 2. GHG distribution for increasingly more stringent nitrogen removal levels (adapted from Falk et al., 2012). 25 20 20 15 10 6 5 3 5 0 Level 1 to 2 Level 1 to 3 Level 1 to 4 Level 1 to 5 Incremental GHG Emissions per N Load Removed Figure 3. Incremental GHG Emission/mass N eliminated as N removal Treatment Level Increase (adapted from Falk et al., 2012). Copyright ©2013 Water Environment Federation. All Rights Reserved. 419 WEFTEC 2013 SUMMARY AND CONCLUSIONS Relative Importance of Effluent Nitrogen Species in Achieving Low Effluent TN Concentrations Effluent total nitrogen consists of four major components: (1) ammonia-N, (2) particulate organic N (pON), (3) oxidized N (NOx-N), and (4) soluble organic N (SON). Of these, ammonia and pON are the N species that can most reliably be removed to very low effluent concentrations by using efficient and reliable designs for biological nitrification and using membrane separation or tertiary filtration for particulate removal. It is not possible to biologically reduce NOx-N to as low a concentration as that for ammonia-N and pON. Based on current observations on high-performing biological nitrogen removal facilities, an effluent NOxN concentration in the range of 0.5 to 1.5 mg/L may reasonably be expected. The ability to achieve lower effluent NOx-N concentration is dependent on a more factors and with more uncertainty than that for ammonia-N and pON removal. Achieving very low effluent NOx-N concentrations requires a tertiary denitrification treatment step and external carbon addition. The effluent NOx-N concentration achievable is thus subject to impacts of variable nitrogen loadings and plant operational skills and methods. The largest factor limiting the ability to achieve much lower effluent TN concentrations than currently demonstrated is the effluent SON concentration. Biological nitrogen removal processes do not appear to have mechanisms that can be manipulated to improve effluent SON removal. A collaborative WERF project with Gdansk University found that a residual SON concentration persists in biological nutrient removal processes and can increase with longer activated sludge SRT (Czerwionka et al., 2012). Longer SRTs may be needed for more efficient ammonia-N and NOx-N removal. Methods for additional SON removal must consider physical and chemical processes alone or in combination and perhaps with biological processes. In view the EPA in-stream nutrient criteria there may be a need to lower effluent TN concentrations to well below the more common TN removal value of 3.0 mg/l. In that case there is a great need to find a cost effective method to reduce effluent SON as an alternative to reverse osmosis or ultrafiltration with brine disposal. The latter technologies are very costly and as shown in this paper have a very negative impact on sustainability and greenhouse gas emissions. Demonstrated Effluent TN Concentrations Based on the demonstrated performance for meeting low effluent TN concentrations, Figure 9 shows an estimate of the range of effluent N concentrations that may be reliably expected for well designed facilities with careful operation. Effluent pON concentrations of less than 0.10 mg/L are possible with membrane separation or tertiary filtration. Ammonia-N concentrations of less than 0.20 mg/L are possible. A summary of effluent SON concentrations in the WERF compendium on effluent SON (referred to as DON in the document) showed a large variation between WWTPs with values ranging from 0.5 to 2.5 mg/L for 33 facilities (Stensel, 2008). The effluent SON concentration possible for a given BNR facility may be very site specific. Copyright ©2013 Water Environment Federation. All Rights Reserved. 420 WEFTEC 2013 Effluent N Concentration, mg/L 5.0 4.0 3.0 2.0 1.0 0.0 pON NH3-N NOx-N SON TN Nitrogen Species Figure 9. Estimated reliable range of effluent nitrogen species concentrations. NEXT STEPS The findings to date have answered many questions about nitrogen removal efficiency, reliability, and sustainable impacts. Based on the findings some challenges remain.       Incorporate nitrogen species into water quality models to determine impacts on receiving waters. Evaluate the fate of hydrophobic organic nitrogen in surface waters. Understand the source and composition of effluent soluble organic nitrogen. Develop more cost effective methods to remove effluent soluble organic nitrogen. Relate statistical performance to setting of effluent permit limits. Advance in plant sensors and automatic process control methods to minimize external carbon requirements and minimize effluent nitrate-N concentration. REFERENCES Bott, C.B, and Parker, D.S. (2011), “Nutrient Management Volume II: Removal Technology Performance & Reliability” WERF Nutrient Removal Challenge project NUTR1R06k. Czerwionka, K., J. Makinia, K. Pagilla, and H. D. Stensel (2012) "Characteristics and Fate of Organic Nitrogen in Municipal Biological Nutrient Removal Wastewater Treatment Plants," Water Research, 46, 7, 2057-2066. Falk, M. W.; Neethling, J. B.; Reardon, D. J. (2011) Striking the Balance Between Nutrient Removal in Wastewater Treatment and Sustainability. Report NUTR106n; Water Copyright ©2013 Water Environment Federation. All Rights Reserved. 421 WEFTEC 2013 Environment Research Foundation: Alexandria, Virginia Neethling, J.B., David Stensel, Denny Parker, Charles Bott, Sudhir Murthy, Amit Pramanik, Dave Clark (2009) “What is the Limit of Technology (LOT)? A Rational and Quantitative Approach” WEFTEC 2009. Stensel, H.D. (2008). Dissolved Organic Nitrogen (DON) in Biological Nutrient Removal Wastewater Treatment Processes. Compendium from WERF Nutrient Removal Challenge project NUTR1R06 (2008) Sedlak, D.L., Jeong, J., and Liu, H. (2013) “Uptake Of Dissolved Organic Nitrogen From Bnr Treatment Plant Effluents By Algae” WERF Nutrient Removal Challenge project NUTR1R06e. Copyright ©2013 Water Environment Federation. All Rights Reserved. 422