NUTRIENT REMOVAL FROM ANAEROBIC DIGESTER SIDE-STREAM AT THE BLUE PLAINS AWTP Dimitrios Katehis *, Sudhir Murthy **. Bernhard Wett ***, Edward Locke * and Walter Bailey ** * Metcalf and Eddy, Inc. 5000 Overlook Ave., SW, Washington, DC 20032 ** DC Water and Sewer Authority, 5000 Overlook Ave., SW, Washington, DC 20032 ABSTRACT *** University of Innsbruck, Institute of Environmental Engineering, Technikerstr.13, A-6020 Innsbruck, Austria Implementation of a novel sidestream treatment configuration will provide the Blue Plains AWTP sidestream treatment process the flexibility to operate in four different process modes including the MAUREEN, SHARON, STRASS and DEMON configurations. Each process confers specific benefits to the AWTP that will need to achieve stringent nutrient limits to support ongoing efforts within the Chesapeake Bay area. In order to maximize the operating cost savings, while also assuring that treatment goals are achieved, the MAUREEN process configuration can be deployed in two seasonal modes, providing differing levels of preferential nitritation and bioaugmentation. To maximize operating cost savings, when performance is not limited by temperature, configurations such as SHARON, STRASS or DEMON would be utilized. KEYWORDS Nitritation, bioaugmentation, sidestream treatment, seeding BACKGROUND Wastewater treatment plants needing to remove nitrogen are increasingly considering sidestream treatment for anaerobic digester dewatering flows as both, an economical means to reduce capital and operating costs, as well as a mechanism to fortify the performance of the main plant. The Blue Plains AWTP, is a two sludge plant, with chemically enhanced primary treatment, followed by a high rate BOD removal stage, and a separate nitrification/denitrification stage. With a design flow of 370 mgd the Blue Plains AWTP will need to process 1.5 mgd of anaerobic digester dewatering flow. This flow will contain as much as 1500 mg/l NH 4 -N representing as much as one third of the biological process nitrogen load for nitrification and denitrification. Nitrogen removal is achieved by first nitrifying ammonia to nitrite or nitrate followed by denitrification of nitrite or nitrate to nitrogen gas. The air requirements for nitrification and the carbon requirements for denitrification are substantially reduced if the process is controlled to stop at nitrite. The HRT and SRT requirements are also reduced by 5215
operating the process at higher temperatures close to 35ºC, as is typically present in anaerobic digester dewatering recycles. Sidestream treatment processes for nitrogen removal can be generally grouped into two categories 1. Nitrogen elimination with bioaugmentation of the main plant process 2. Nitrogen elimination with no bioaugmentation of the main plant process The preferred process configuration would be dependent on the combination of loadings, treatment performance requirements and economics of treatment that will vary from plant to plant, and has been observed to vary within a given plant on a seasonal basis. Some of the modes described can be operated in parallel using multiple trains for treatment thereby maintaining overall treatment process robustness provided by the bioaugmentation processes, while also tapping into the enhanced treatment economics more readily obtained through the processes that do not provide bioaugmentation. To be able to adjust to this range of operating requirements, the Blue Plains AWTP is actively considering four methods to reduce nitrogen processing costs by removing it from its side-stream before it reaches the mainstream process. These methods are: 1. Bioaugmentation and enrichment of sidestream recycle (MAUREEN process) 2. Nitritation/ denitritation of sidestream recycle using chemostat (SHARON process) 3. Intermittent nitritation/ denitritation of sidestream recycle (STRASS process) 4. Intermittent deammonification of sidestream recycle (DEMON process) This paper will show how selection of the proper performance assumptions, particularly with respect to bioaugmentation and enrichment processes is key in the development of the economic model that will be used to select the appropriate process operating mode. A detailed description of each process option, energy, oxygen and carbon balance, SRT and HRT requirements, and pros and cons is discussed. The overall treatment strategy would allow DCWASA to operate two of these four process configurations in parallel using multiple treatment trains. BIOAUGMENTATION USING THE MAUREEN PROCESS DCWASA and its program manager partners developed the MAUREEN process which stands for Mainstream Autotrophic Recycle Enabling Enhanced N-removal to allow for integration of the state of the art in sidestream treatment within a single, highly flexible process configuration. The MAUREEN process is shown integrated within the Blue Plains AWTP in Figure 1. This process configuration includes a side-stream bioreactor to allow for nitrification and denitrification of the centrate stream. The process exhibits significant flexibility when applied to a two sludge system such as the Blue Plains AWTP. 5216
Unique features of this process configuration include: Preferential bioagumentation of ammonia oxidizing bacteria from the second to the first stage via the sidestream reactor Oxidation of reject centate preferentially to nitrite in the enrichment reactor, resulting in reduced energy (aeration) and chemical (organic substrate) consumption The ability to fortify the second stage system with a combination of primarily ammonia oxidizers and anoxic methanol degrading bacteria produced in the sidestream reactor under conditions that limit the presence of nitrite oxidizing bacteria and general heterotrophic bacteria. The ability to supernate the effluent from sidestream process and use the supernatant for odor and corrosion control in the headworks or within process streams at the plant. Main Plant Treatment Train External Carbon Wastewater First Stage Aeration Tank First Stage Clarifiers Second Stage Nite/Denite Tanks Second Stage Clarifiers First Stage Return Activated Sludge Second Stage Return Activated Sludge External Carbon Alkalinity Bioaugmentation Sludge Sidestream Bioaugmentation and Enrichment Reactor High Strength Digester Reject Water (Centrate) Figure 1. MAUREEN process: Sidestream treatment and mainstream bioaugmentation in a two-sludge process. Key to the success of the process is the physical configuration and selection of operating conditions of the sidestream reactor. This would include utilization of a plug flow reactor with the ability to introduce supplemental alkalinity in multiple locations along the length of the reactor and flexibility to modify the aerated fraction of the reactor in response to seasonal temperature changes and influent loading conditions. Parts of the MAUREEN process configuration and operating details are provided in Constantine et al. (2005). 5217
Baseline Operating Mode Bioaugmentation to the First Stage The baseline operating mode allows the Blue Plains AWTP to nitrify the centrate load in the sidestream reactor, and remove the produced nitrite and nitrate in anoxic zones within the first stage aeration tanks using the main plant s influent organics (BOD). Furthermore this process configuration promotes nitrification in the first stage aeration tanks, where nitrification would normally not be available, as the first stage aeration tanks operate at a solids retention time of approximately one day. Waste activated sludge (WAS) from the first (BOD removal) or second stage (Nitrification/Denitrification) is directed to the side-stream system, thereby seeding the sidestream system with ammonia oxidizing bacteria. A key feature of the MAUREEN process is the ability, when operated in this mode, to preferentially oxidize ammonia to nitrite, rather than nitrate, in the sidestream reactor. The sidestream reactor would be operated aerobically at DO and ph levels that would enhance the preferential production of nitrite (Anthonisen, 1973; Wett, 1998; Katehis, 2003; Wett, 2003). The temperature in the reactor is a function of the temperatures and flows of the WAS and dewatering centrate introduced into the reactor. Previous experience has shown that minimal changes in temperature occur across the reactor under typical full scale conditions. The operating ph of the reactor is maintained at above neutral levels (7.5-7.8) in the front portion of the reactor, where ammonia concentrations are high, to maintain elevated free unionized ammonia levels (NH 3 ). The ph is monitored at multiple points along the plug flow reactor and is allowed to drop to below neutral levels in the latter portion of the reactor, where nitrite levels are elevated, thereby resulting in elevated concentrations of free nitrous acid (HNO 2 ). Coupled with the induced bicarbonate limitation in the tail end of the reactor (Wett, 2003) the combination of selective mechanisms applied results in the oxidation of 50-80% of the ammonia to nitrite, the balance being converted to nitrate. Care must be exercised to avoid a significant ph depression in the back end of the plug flow reactor as this would lead to conversion of the residual carbonate alkalinity to CO 2 and thus loss of the CO2 through stripping, posing challenges to cost effective ph control within the reactor using common inexpensive alkali such as lime. However the presence of the heterotrophic biomass, which continues to respire endogenously, helps provide a low level of CO 2 to the system, countering the stripping effect, and thus allow for enhanced stability of the process. The treated centrate stream from the side-stream system, consisting primarily of ammonia and nitrite oxidizing autotrophs, would then be used to seed a portion of the first stage aeration tanks (Constantine et al., 2005). The seed would promote selective nitritation in the first stage aeration tanks as practiced in facilities such as the New York City Department of Environmental Protection (NYCDEP) 26 th Ward Water Pollution Control Plant (Katehis, 2002; Carrio, 2003; Ramalingam; 2005) for over a decade. Typical data from the NYC sidestream reactor at the 26 th Ward Water Pollution Control Plant, that has been operating since 1994 is shown in Figure 2. 5218
500 450 NO2-N NOx-N 400 NO 2 -N, NO x -N, mg/l 350 300 250 200 150 100 50 0 01/2003 02/2003 03/2003 04/2003 05/2003 06/2003 07/2003 08/2003 09/2003 Figure 2. Preferential Nitritation in the Sidestream Reactor at the 26 th Ward WPCP(Adapted from Fillos et al, 2006). Similar performance can be obtained if WAS from the second stage (Nitrification/Denitrification) is utilized to continuously seed the sidestream reactor. Although the preferential nitritation efficiency may not be as high in this mode, it would be necessary to have flexibility to seed from the second stage system in instances where the nitrifier concentration in the first stage sludge is inadequate to maintain nitrification efficiency in the sidestream reactor (see design criteria below). Although process modeling can be utilized to define the amount of WAS needed and the operating conditions in the sidestream reactor, caution must be utilized as pilot and full scale demonstration work has shown that the reaction rates and the seeding efficiency within the sidestream reactor are lower than would be expected if conventional mass balance calculation are used (Katehis, 2002; Zimmerman, 2004; Bowden, 2006). The increased free ammonia and TDS levels, along with the higher operating temperature in the sidestream reactor have been hypothesized to contribute to the observed reduction in ammonia oxidation activities. Nitrification efficiency below expected levels has been observed particularly during the winter, when the temperature difference between the sidestream and main reactors is greatest (Zimmerman, 2004; Bowden, 2006). However, as in the case of the sidestream reactor, care must be exercized in assessing the impacts of the introduced seed into the first stage aeration tanks. Assessment of the efficiency of the seeding process into the secondary plant under full scale conditions has shown that a portion of the expected increase in nitrification efficiency is realized in the portion of the plant treating the main flow. Research is ongoing to identify causes and develop mitigation strategies (Bowden, 2006), including usage of an internal recycle within the plug flow reactor to minimize ammonia concentration and thus free ammonia levels, as well as increasing the relative mass of seed (WAS in this instance) to the sidestream reactor. 5219
At the Blue Plains AWTP this could be implemented by feeding the sidestream reactor with WAS from a dedicated second stage clarifier. The net result of implementation of the MAUREEN process is increased process capacity at the Blue Plains AWTP. One approach to visualizing the effect of the MAUREEN process is summarized in Figure 4, where the reduction in second stage nitrification volume requirements over the expected range of seeding efficiency in the first stage process is shown at a temperature of 16ºC, with MLSS levels maintained equal throughout the simulations. In effect, implementation of the MAUREEN process in this mode enhances the safety factor of the AWTP allowing the plant to successfully operate with less reactor volume on-line, facilitating maintenance, while also reducing the impact of a process upset in the second stage on overall nitrogen removal. E PI WPL ES_WPL E PI Ferric East PI East Primary ES_P-1 ES_P-2 ES_P-3 ES_P-4 Caustic Ferric Nit Odd Outfall 001 East Primary Sludge East MLSS Mxr East SE Nit_P-1A Nit_P-1B Nit_P-2 Nit_P-3A Nitr FE Inf. MLSS Nitr FE Filter Nit_P-3B Nit_P-4 Nit_P-5A Nit_P-5B Nit_ML Channel SWW Chlorine Kill BackWash Water Plant EFF ES_RAS MeOH-4 GT Nit_Underflow Gravity Thickener Underflow Nitrification WAS West Primary Sludge WS_WPL All secondary waste DAF WS_P-1 WS_P-2 WS_P-3 WS_P-4 West SE Blend 3 W PI WPL WS_RAS W PI Ferric West Primary NaOH-R1 NaOH-R2 NaOH-R3 Blend 1 Dig. Inf Dewatering Centrate MAUREEN Rxr 1 MAUREEN Rxr 2 MAUREEN Rxr 3 MAUREEN Rxr 4 West PI Seeding Correction Figure 3. Whole Plant Simulator for the Blue Plains AWTP Implemented in BioWin. 5220
Additional Volume Required (MG). 2 1 0-1 -2-3 -4-5 -6-7 -8 Scenarios 1- No Sidestream Treatment 2- Sidestream Treatment with No Seeding 3- Sidestream Treatment with Partial Seeding (50%) 4- Sidestream Treatment with Full Seeding (100%) 1 2 3 4 Treatment Scenario Figure 4. Reduction in Second Stage (Nitrification/Denitrification) Reactor Volume Requirements by Offloading Influent Nitrogen Using the First Stage at Different Levels of Seeding Efficiency Key elements of the operating costs of the baseline process configuration are the oxygen and supplemental alkalinity utilized for nitrification. The operating costs are offset by the savings in aeration in the main plant s first stage (BOD removal) through the use of nitrite/nitrate to oxidize the influent wastewater organics. Alternatively, the effluent from the sidestream reactor can be allowed to settle and the supernatant, rich in nitrite and nitrate, can be redirected to the head of the plant for odor control purposes. The settled biomass would be directed to the first stage. Under this scenario, the presence of the nitrite and nitrate in the influent (up to 6 mgn/l in the instance of the Blue Plains AWTP) would serve to mitigate odors in the headworks, and primary clarifier area, with the residual nitrite/nitrate being denitrified in the first stage anoxic zones. In order to develop a cost effective design for the overall facility, that is compatible with the range of operating configurations discussed in this paper it was necessary to define specific design constraints around which the process would be configured. From a design perspective, the approach would entail assuming that under the outlined base configuration the dewatering centrate is completely nitrified to nitrate, and nitrification activity induced within the first stage aeration tanks also results in the production of primarily nitrate. Key design criteria for the system would include: Soluble TKN to Nitrifier Mass Ratio (mgn/mgvss) of no greater than 5:1 Soluble TKN Loading Rate of 0.3-0.4 kgn/m 3 -d 5221
Actual Hydraulic Retention Time (Seed Biomass Flow + Dewatering Centrate Flow) of 1.5 days Minimum Sidestream Reactor Temperature of 17ºC Aeration System sized for complete ammonia oxidation to nitrate Four point ph control system Multipoint automated supplemental alkalinity addition system Note that the above criteria are based on operational experience in pilot and full scale developed primarily via operations at the NYCDEP facilities rather than process modeling. A design basis developed using typical process modeling coefficients as applied for example in the BioWin model, could result in the development of a much smaller sidestream reactor. The commercially available process models do not consider the reduced nitrification efficiency observed in the sidestream reactor, as the causes are not adequately defined to permit inclusion into the models equation/parameter set. Care must also be exercized to ensure that the reactor is not subject to oxygen transfer limitations particulary when high strength dewatering liquors are being treated, as oxygen uptake rates of 150-200 mg/l-hr are difficult to sustain with conventional fine bubble aeration systems. Bioaugmentation to the Second Stage Modification of the baseline sidestream treatment process shown in Figure 1, to include implementation of unaerated denitrification zones and methanol addition within the sidestream reactor, coupled with diversion of the sidestream reactor effluent to the second stage (Nitrification/ Denitrification) aeration tanks, rather than the first stage, will permit augmentation of the second sludge s nitrification and denitrification capacity. Under these conditions, the sidestream reactor can produce both nitrifiers and specialized denitrifiers that utilize methanol. The efficiency of the bioaugmentation step is difficult to ascertain, however, a range of mitigation strategies to minimize the observed reduction in nitrification efficiency in the sidestream reactor, as is the case for the baseline operating mode, would be implemented. The bioaugmentation process provides for enhanced process performance in the second sludge system, however the seeding efficiency of the process would not be expected to be at the levels indicated by a calculated mass balance (Katehis, 2003). The level of seeding efficiency imparted by the process can significantly alter the main stream plant s performance. An analysis of the nitrification and denitrification volume requirements in the second stage process configuration was conducted using a BioWin whole plant process model calibrated to the Blue Plains AWTP (Figure 5). The analysis (data not shown) indicated that seeding will impact process performance in a similar manner to that shown in Figure 4 for the baseline operating mode. 5222
E PI Ferric E PI WPL ES_WPL Dewatering Centrate NaOH-R1 NaOH-R2 NaOH-R3 MeOH-S MAUREEN Rxr 1 MAUREEN Rxr 2 MAUREEN Rxr 3 MAUREEN Rxr 4 Seeding Correction East PI East Primary ES_P-1 ES_P-2 ES_P-3 ES_P-4 Caustic Ferric Nit Odd Outfall 001 East Primary Sludge East MLSS Mxr East SE Nit_P-1A Nit_P-1B Nit_P-2 Nit_P-3A Nitr FE Inf. MLSS Nitr FE Filter ES_Underflow MeOH-4 Nit_P-3B Nit_P-4 Nit_P-5A Nit_P-5B Nit_ML Channel Nit_Underflow SWW Chlorine Kill BackWash Water Plant EFF GT Gravity Thickener Underflow Nitrification WAS West Primary Sludge WS_WPL All secondary waste DAF WS_P-1 WS_P-2 WS_P-3 WS_P-4 West SE Blend 3 W PI WPL WS_Underflow W PI Ferric West Primary Blend 1 Dig. Inf West PI Figure 5. Whole Plant Simulator for the Blue Plains AWTP Implemented in BioWin Bioaugmentation to the Second ((Nitrification/Denitrification) Stage. NITRITATION AND DENITRITATION OF SIDESTREAM RECYCLE IN A CHEMOSTAT PROCESS (SHARON) The Blue Plains AWTP is considering providing the ability for its sidestream reactors to operate in the Single reactor System for High activity Ammonia Removal Over Nitrite (SHARON) mode as a default mode when operating without bioaugmentation (Hellinga et al., 1998). The SHARON process will be operated as a flow through process (chemostat) without clarification. The process will be operated in the nitritation and denitritation using methanol mode. Without bioaugmentation capabilities, the SHARON process configuration will be deployed when direct enhancement of the main plant s second sludge nitrogen removal performance is not necessary, or if organic loadings to the main plant s first stage (BOD removal stage) significantly exceed design capacities, resulting in aeration limitations in that stage. Under these conditions use of DCWASA s MAUREEN process would not be beneficial. The Blue Plains AWTP currently has supplemental carbon storage and dispensing facilities which support the second stage (Nitrification/Denitrification) system. Coupled with the reduced aeration requirements of the SHARON process, relative to the baseline operating mode, implementation of the process would consist primarily of modification of the reactor 5223
operating strategy. Cooling of the process liquor may be necessary for implementation of the process, necessitating the addition of a heat exchanger facility using plant effluent cooling water. The main disadvantage of the SHARON process is the inability to supernate the recycle water and separate it from the sludge stream for alternate treatment if desired. INTERMITTENT NITRITATION AND DENITRITIATION (STRASS PROCESS) The STRASS process, originally developed in Austria in the early 1990 s (Wett, 1998) is being considered as an alternate to SHARON for default operations without bioaugmentation. The process uses a high sludge age sequencing batch reactor to oxidize ammonia to nitrite (nitritation), followed by reduction of the produced nitrite to nitrogen gas (denitritation). A supplemental carbon source, such as primary sludge is used to drive the denitritation process. Unlike the SHARON process, the sludge age in the reactor is relatively high (greater than 20 days), but also the solids concentration in the reactor allowing for improved denitritation rates at reduced reactor volume. The key feature of the STRASS process is the innovative process control strategy. A simple, yet highly effective, ph based control mechanism is utilized to control the intermittent aeration system (Wett, 1998). During an aeration interval acidification due to nitritation occurs. When the lower phsetpoint is reached the aeration stops and alkalinity/ph recovers. At the upper ph-setpoint aeration is switched on again resulting in a characteristical sawtooth profile of the ph-course. Following this control strategy frequency and length of aeration intervals is self.adjusting to the feed-rate and concentration of reject water. In concert with the operating conditions employed: Operating temperatures of 25-35ºC Near neutral ph Intermittent aeration at low DO concentrations The STRASS process is flexible and robust as demonstrated by close to a decade of operation in Austria (applied both to lime- and polymer comditioned sludge liquors), and represents a viable alternative to the SHARON process if cost effective treatment of the sludge dewatering centrate is the primary process objective. However it does have additional processing requirments including, the need to use a pre-sedimentation process to remove solids from the centrate stream prior to sidestream treatment and careful ph control. These cons are somewhat mitigated by having the ability to supernate the treated centrate and using it for odor and corrosion control in the headworks and primary clarifiers. NITRITATION/DEAMMONIFICATION (DEMON PROCESS) While the SHARON and STRASS processes are being considered for default operations, the preferred mode for operation for nitrogen elimination without bioaugmentation for the Blue Plains AWTP is the DEMON deammonification process that is operated in a combined 5224
partial nitritation/autotrophic denitritation mode. This configuration uses a sequential batch reactor and a robust control for intermittent aeration in order to avoid a build-up of toxic nitrite concentration. This ph-based control system determines the length of aeration intervals depending on the production of H + ions or nitrite (Wett, 2005). During the subsequent aeration break the nitrite produced is depleted autotrophically. The process is operated at an SRT of at least 25 days to prevent wash-out of these slowly growing anaerobic ammonia oxidisers. The partial nitritation of ammonia (approximately half of the ammonia is oxidized to nitrite) reduces the aeration costs considerably (Figure 6). Subsequent deammonification using autotrophic organisms eliminates the need for an external substrate such as methanol. After a successful full scale demonstration in the Strass WWTP in Strass, Austria, this novel process will be piloted and refined over the next two years in the United States by a research consortium sponsored by DCWASA, NYCDEP and the Alexandria Sanitation Authority (ASA), prior to full scale implementation. Elements of the research program will include development of rapid enrichment techniques and assessment of process reliability under variable loading and influent centrate characteristics, as is typical in centralized dewatering facilities serving multiple WWTPs. 600 NH4-load elim.[kg N d-1] spec.energy [kwh (kg N)-1] 20.00 18.00 NH4-load eliminated [kg N d -1 ] 500 400 300 200 100 Dec/Jan 2003/04 19771 kwh Dec/Jan 7692 kwh 2.20 kwh (kg N) -1 2004/05 0.79 kwh (kg N) -1 0 11.11.03 12.12.03 12.01.04 12.02.04 14.03.04 14.04.04 15.05.04 15.06.04 Change-over to DEMON 16.07.04 16.08.04 16.09.04 17.10.04 17.11.04 18.12.04 18.01.05 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 specific energy [kwh (kg N) -1 ] 0.00 Figure 6. Decreasing specific energy demand (related to the eliminated nitrogen load) before and after changeover from the STRASS to the DEMON strategy (50 day average values shown in boxes). COMPARATIVE COST ANALYSIS In the case of the Blue Plains AWT, the sidestream treatment facility will have the capability of operating in all of the modes outlined, with a focus on balancing operating costs with the need to maximize nitrogen removal performance from the plant s raw wastewater. Thus 5225
although there is a price premium for the high level of flexibility to be provided by the Blue Plains system, development of comparative capital costs for the various operating modes is not relevant in this instance. In applications where the capital cost must be considered in the analysis the following cost elements should be included: Tankage Costs for the Sidestream Treatment Facility At a hydraulic retention time of 3 days (including tankage redundancy to facilitate concurrent operation in different process modes) the reactor volume Tankage Cost Offset for Main Plant Process The sidestream treatment facility will impart increased process capacity, translating to a reduced volume requirement for the main plant s biological reactors and/or clarification facilities. This represents a credit to the cost of the sidestream facility. Major Process Mechanical Equipment Costs Items such as Process Blowers and their ancillary facilities represent the major process equipment costs. However there may be opportunities to develop the design criteria so as to create synergies in equipment utilization, reducing the effective cost of the new facilities. As described in the paper the operating cost avoidance is a common characteristic of all the sidestream treatment options consideredd for the Blue Plains AWTP. The different operating modes impart specific benefits to the process but also have different operating costs associated with treatment. Definition of the operating costs for each scenario is poses challenges as there is limited practical experience in key components that will have a significant impact on the net operating cost of the system. However the combination of operating cost savings and enhancement of the capacity of the main plant s nutrient removal system under limiting conditions provides the major driver for implementation of this novel approach at the Blue Plains AWTP. REFERENCES Constantine, T., Murthy, S., Bailey, W., Benson, L., Sadick, T., Daigger, G. (2005) Alternatives for treating high nitrogen liquor from advanced anaerobic digestion at the Blue Plains AWTP. Proc. WEFTEC, 78th Annual Conference & Exhibition, Washington, DC, USA, 2005 Fillos, J., Ramalingam, K., Carrio, L.A. (2005) Integrating Separate Centrate Treatment with Conventional Plant Operations For Nitrogen Removal Hellinga, C, Schellen, A.A.J.C., Mulder, J.W., van Loosdrecht, M.C.M., Heijnen, J.J. (1998). The SHARON process: An innovative method for nitrogen removal from ammonium-rich wastewater, Water Science and Technology, 43(11), 135-142. Katehis, D., Stinson, B., Anderson, J. Gopalakrishnan, K., Carrio, L., Pawar A. (2002) Enhancement of Nitrogen Removal Thru Innovative Integration of Centrate Treatment. Proc. WEFTEC, 75th Annual Conference & Exhibition, Chicago, Illinois, USA, 2002 Wett, B.: Solved scaling problems for implementing deammonification of rejection water. (2005) Proc. IWA Special. Conf. on Nutrient Management, Krakow 2005 Wett, B.; Rostek, R.; Rauch, W.; Ingerle, K. (1998): ph-controlled reject water treatment. Water Science & Technology 137/12, p.165-172. 5226