Progress 10/24/13 to 08/31/16
Outputs
Target Audience:Targeted audiences of the proposed work include environmental research community, environmental consultants, personnel involved in the design of wastewater treatment plants, coordinators of biosolids land application program, and treatment plant operators. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?This project is being primarily executed by a Ph.D.-level graduate student (in Biological Systems Engineering), Ms. Rania Bashar. How have the results been disseminated to communities of interest?Sampling activities and data collection were facilitated by the personnel at the Nine Springs WWTP serving Madison, WI. We have discussed the results of Biowin WWTP modeling work and whole plant mass-balance analysis with appropriate personnel at the Nine Springs WWTP. Projects results are being presented at relevant technical conferences and are also being prepared for submission to high-impact, peer-reviewed scientific journals. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Objectives: 1) Technical and economic evaluation of high-performing (effluent TP ≤ 0.05 mg/L) P removal/recovery processes that have been implemented on full-scale systems. 2) Assessment of the potential for full-scale implementation of a novel mainstream anaerobic wastewater treatment scheme. a) Demonstrate that the proposed anaerobic configuration can effectively meet a typical effluent criteria (TCOD < 50 mg/L, TN < 10 mg-Ntot/L, TP < 1 mg/L); b) Compare the anaerobic configuration with a conventional aerobic treatment scheme to determine how it impacts: (i) effluent quality (i.e., CBOD, TSS, N removal), (ii) sludge production, and (iii) operational cost and energy. Methods: The treatment configurations were simulated under identical influent flowrate and characteristics to facilitate direct comparisons. All the simulations were performed in Windows-based process model simulating software BioWin v. 5.0 (EnviroSim Associates Ltd., Canada). It was assumed that the facility is a mid-sized U.S. treatment WWTP with a capacity of 5 MGD (average dry weather flow). To establish wastewater load characteristics, monthly average data from January, 2010 to November 2015 were collected from the local Nine Springs Wastewater Treatment Plant (Madison, WI). Objective 1: The following six treatment scenarios were selected for evaluation: Scenario 1 (S1): Preliminary Treatment + Primary Treatment + MUCT BNR + Mesophilic Anaerobic Digestion + Energy Recovery Scenario 2 (S2): Preliminary Treatment + Primary Treatment + Bardenpho 5-stage BNR + Mesophilic Anaerobic Digestion + Energy Recovery Scenario 3 (S3): Preliminary Treatment + Primary Treatment + MUCT BNR + Membrane Bioreactor + Chemical P removal + Mesophilic Anaerobic Digestion + Energy Recovery Scenario 4 (S4): Preliminary Treatment + Primary Treatment + IFAS-EBPR + Mesophilic Anaerobic Digestion + Energy Recovery Scenario 5 (S5): Preliminary Treatment + Primary Treatment + MUCT BNR + Mesophilic Anaerobic Digestion + Energy Recovery + P recovery as struvite Scenario 6 (S6): Preliminary Treatment + Primary Treatment + Bardenpho 5-stage BNR + Mesophilic Anaerobic Digestion + Energy Recovery + P recovery as struvite Results: All the treatment scenarios achieved more than 93% of COD removal and 90% of BOD removal. Scenarios 2 (94.31%), 3 (94.21%) and 6 (94.61%) were the most efficient in removing COD. All the treatment scenarios achieved more than 98% of influent TSS removal. Scenario 3 was the most efficient in solids removal. The filtration mechanism of membrane bioreactors was beneficial to keep a higher TSS removal. The inclusion of struvite recovery in Scenarios 5 & 6 increased TN removal by about 2% than Scenarios 1 & 2, respectively. Scenario 3 was able to achieve the target nitrogen removal goal (TN < 10 mg/L) at a much lower SRT (8 days) than the other scenarios. All treatment scenarios needed lime addition for alkalinity correction in the nitrification process. Other than Scenario 6, all the other configurations also required methanol addition for heterotrophic denitrification process. In general, all the treatment scenarios achieved very low concentrations of NH4-N (< 1 mg/L) and nitrite (≤ 0.2 mg/L) in the effluent. The highest TP removal efficiency was observed in Scenario 5 (87.5%), where MUCT P removal was paired with side stream P recovery. In general, MUCT process proved to be much more efficient in P removal due to the two-step anoxic treatment. Modeling results from S5 & S6 suggest that up to 45% of the influent P can be recovered as struvite from the side stream. About 49% reduction of P content in recycle stream was observed due to P recovery, which resulted in considerable reduction in aeration requirement. Furthermore, the introduction of P recovery process resulted in about 50% reduction in chemical requirement for P removal in both the MUCT and Bardenpho processes. There was a 50% reduction in the mass of biosolids after the introduction of struvite recovery in S5 and S6 compared to S1 and S2. Since only small amount of P was leaving the system (effluent) in S1 and S2, the only significant outlet for P was in the biosolids prior to struvite recovery. So, struvite recovery resulted in about 45% reduction in biosolids P content. The results from this study showed that harvesting P from the sidestream of a treatment process could have a secondary benefit from the standpoint of biosolids management by yielding a more agronomically-balanced product. The reduction in N & P content in recycle stream due to struvite recovery in S5 and S6 resulted in 15-20% reduction in aeration requirement compared to S1 and S2. The sludge disposal cost also was reduced by about 50% in these two scenarios. Although S3 was very efficient in removing organics and suspended solids, the aeration requirement was much higher compared to all the other treatment scenarios. In addition, 33% of the digester off-gas was converted to power via the combined heat and power (CHP) engine. The option to use recovered power for treatment was specified in the simulator, so this credit offsets the total power demand of the treatment process. P recovery did not appear to have any negative impact on biogas generation. Since Scenario 4 was run at a lower SRT, the sludge disposal cost was lower than the other biological P removal configurations. Objective 2: The mainstream anaerobic treatment configuration (Configuration 1, C1) proposed in this study consists of fine screen pretreatment, anaerobic membrane bioreactor (AnMBR) for BOD and TSS removal, reactive filter media for adsorptive P recovery and Anammox process for N removal. The AnMBR system consisted of an upflow anaerobic sludge blanket (UASB) reactor connected to two submerged membrane tanks. It also included a sludge dewatering system for conditioning the resulting sludge, a degassing membrane for capturing the dissolved methane in the effluent, and a CHP system enabling energy to be recovered from methane. Moving bed biofilm reactor (MBBR) was used for effective retention and prolonged SRT of Anammox biomass in our study. For adsorptive P recovery, it was assumed that the steel slag filter media would be used for reactive filtration. Configuration 2 (C2) involved MUCT process for nutrient removal (anaerobic, primary anoxic, secondary anoxic and aerobic) in conjunction with sidestream sludge treatment. In the sludge line, waste activated sludge (WAS) and primary sludge (PS) were thickened and dewatered in a belt filter. This process was coupled with mesophilic anaerobic digestion for solids stabilization and energy recovery via biogas combustion. Results: Results show that the proposed anaerobic treatment scheme could reliably achieve a typical effluent quality (TCOD < 50 mg/L; TN < 10 mg-Ntot/L, and TP< 1 mg/L). The removal efficiencies of TN, TP and TCOD in the effluent were 93%, 86% and 98%, respectively, comparable to a high performing BNR plant. A comparison of results from C1 and C2 shows that Configuration 1 performed remarkably well in removing COD and TSS due to the presence of the membrane bioreactors. AnMBR removed 74% of the influent COD, which provided a very promising stream for the subsequent Anammox process. The total biogas production in C1 was 1516 m3/d which was about 8% higher than that in C2. In BNR treatment scheme, a significant portion of the incoming COD was consumed during heterotrophic denitrification, which resulted in less COD availability for energy recovery. Configuration 2 also required chemical addition in the form of methanol and lime for denitrification. Anammox being an autotrophic process, did not require any organics addition for N removal. Configuration1 resulted in 73% reduction in energy requirement compared to the aerobic treatment scheme. However, there was a distinct increase in energy demand in both configurations (C1 and C2) for heating purposes.
Publications
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2017
Citation:
Bashar, R., Karthikeyan, K.G. Energy positive municipal wastewater treatment: Potential for full-scale implementation of a novel anaerobic scheme. ASCE-EWRI Congress, Sacramento, CA, May 2017.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2016
Citation:
Bashar, R., Karthikeyan, K.G. Potential for full-scale implementation of a novel mainstream anaerobic wastewater treatment scheme. Presented at the Water Connection symposium (Riveredge Nature Center, WI) on Nov. 12, 2016.
- Type:
Journal Articles
Status:
Other
Year Published:
2017
Citation:
Bashar, R., Gungor, K., Karthikeyan, K.G. Cost of Phosphorus Removal in Municipal Wastewater Treatment (to be submitted to Water Science and Technology by Feb. 2017).
- Type:
Journal Articles
Status:
Other
Year Published:
2017
Citation:
Bashar, R., Noguera, D.R., Karthikeyan, K.G. Assessment of the potential for full-scale implementation of mainstream anaerobic wastewater treatment scheme (to be submitted to Water Research by Mar. 2017).
|
Progress 10/01/14 to 09/30/15
Outputs
Target Audience:Targeted audiences of the proposed work include environmental research community, environmental consultants, personnel involved in the design of wastewater treatment plants, coordinators of biosolids land application program, and treatment plant operators. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?This project is being primarily executed by a Ph.D.-level graduate student (in the department of Biological Systems Engineering), Ms. Rania Bashar. How have the results been disseminated to communities of interest?Sampling activities are facilitated by the personnel at the Nine Springs WWTP serving Madison, WI. We have discussed the results of Biowin WWTP modeling work and whole plant mass-balance analysis with appropriate personnel at the Nine Springs WWTP. What do you plan to do during the next reporting period to accomplish the goals?In the next reporting period, we will continue activities around the original three goals.
Impacts
What was accomplished under these goals?
Goal: The major goal was to perform a model-based technical evaluation of established/emerging phosphorus (P) recovery/removal technologies, using design parameters that are typical for a wastewater treatment plant (WWTP) in the U.S. A broad assessment involving a variety of treatment combinations and conditions was conducted. A mass balance was performed for various treatment scenarios and process modifications were appropriately selected to meet a very low effluent P level (total P, TP ≤ 0.05 mg/L) as well as efficient side-stream P recovery. Methods: All the simulations were performed in Windows-based process model simulating software BioWin v. 4.1 (EnviroSim Associates Ltd., Canada). Integrating struvite recovery processes into Biowin software: The first step was integrating P recovery processes into the WWTP modeling software. To simulate struvite precipitation, Biowin's default chemical precipitation module was calibrated and validated using literature data from full- and/or pilot-scale systems. Three independent sets of experimental data were used for calibration and three additional datasets were used for validation. The module with the calibrated precipitation rate constant parameter minimized the Mean Squared Error (MSE) based on measurements of orthophosphate removal. An iterative approach for model calibration was adopted until the MSE value could not be further minimized. Selection of treatment processes: The following treatment technologies were selected for simulation: I. Biological Treatment: (i) Anaerobic/Anoxic/Oxic (A2O), (ii) Phoredox (Anaerobic/Oxic), (iii) Bardenpho process (five stage), (iv) University of Cape Town (UCT) process, (v) Modified University of Cape Town (MUCT) process, and (vi) Johannesburg (JHB) Process. II. Physico-Chemical Treatment: (i) Chemical Precipitation (alum addition, iron salt addition), (ii) Media Filtration, and (iii) Membrane Bioreactor (MBR). III. Fixed-film Systems: Integrated Fixed-Film Activated Sludge Systems with Biological Phosphorus removal (IFAS-EBPR). Simulations: The following four treatment scenarios were chosen for evaluation: Scenario 1: Baseline P removal treatment process configuration. Scenario 2: Scenario 1 with an effluent performance criterion of TP ≤ 0.05 mg-P/L. Critical operational parameters that have significant impact on P removal were adjusted as needed. The biological P removal method was combined with tertiary chemical precipitation to reach the target effluent P level. Scenario 3: Mainstream P removal configurations from Scenario 1 coupled with a side-stream P recovery process. Scenario 4: Scenario 3 with an effluent performance criterion of TP ≤ 0.05 mg-P/L. The biological P removal method was combined with tertiary chemical precipitation to reach the target effluent P level. All the treatment configurations were simulated under similar influent flowrate and characteristics to facilitate direct comparison: Flow: 5 mgd; BOD: 165 mg/L; TKN: 24 mg/L; TSS: 188 mg/L; VSS: 132 mg/L; TP: 5.7 mg/L; Alkalinity: 200 mg/L as CaCO3; pH: 6.8. It was assumed that the facility is a mid-sized U.S. WWTP with a capacity of 5 MGD. Rather than establishing site-specific wastewater load characteristics, we used the generic population based equivalent values to calculate the concentrations of nutrients and other constituents. Tank sizes and process parameters, such as, hydraulic retention time (HRT), solids retention time (SRT), were established according to the standards set forth in the WEF manual. The mainstream treatment processes were paired with a sludge treatment process comprising anaerobic digestion and subsequent dewatering of primary and thickened waste activated sludge (WAS). A side-by-side comparison of the simulation results for the Modified University of Cape Town (MUCT) process is presented in Table 1 (below). The results suggest that about 45% of the influent P can be recovered as struvite from the side-stream. The addition of iron salt in Scenario 4 reduces soluble P availability. Hence, the struvite production decreased slightly in Scenario 4. The introduction of P recovery process in Scenario 3 decreased the effluent TP significantly compared to Scenario 1, due to lower P concentration in the recycle stream. This step resulted in about 50% reductions in chemical requirement (for Scenario 3) for precipitation compared to Scenario 2. Furthermore, modeling results suggest that implementation of side-stream P recovery could produce biosolids with a N:P ratio that is more attractive for agricultural land application, as there is a substantial reduction in biosolids P content compared to that in conventional P removal process simulated in Scenarios 1 and 2. Table 1. Simulation results for proposed treatment scenarios for Modified University of Cape Town (MUCT) Process. Total carbonaceous BOD: Scenario 1: 1.57 mg/L Scenario 2: 1.66 mg/L Scenario 3: 1.60 mg/L Scenario 4: 1.62 mg/L Total suspended solids (TSS): Scenario 1: 1.58 mg/L Scenario 2: 1.92 mg/L Scenario 3: 1.70 mg/L Scenario 4: 1.72 mg/L Total inorganic nitrogen (TIN): Scenario 1: 4.73 mg/L Scenario 2: 5.10 mg/L Scenario 3: 4.20 mg/L Scenario 4: 3.1 mg/L Total phosphorus (TP): Scenario 1: 1.74 mg/L Scenario 2: 0.05 mg/L Scenario 3: 0.51 mg/L Scenario 4: 0.05 mg/L pH Scenario 1: 6.98 Scenario 2: 7.01 Scenario 3: 7.01 Scenario 4: 7.02 Iron addition: Scenario 1: - Scenario 2: 507 ppd Scenario 3: - Scenario 4: 253.41 ppd Overall p Mass Balance: Influent, ppd: Scenario 1: 237.88 ppd Scenario 2: 237.88 ppd Scenario 3: 237.88 ppd Scenario 4: 237.88 ppd Effluent, ppd: Scenario 1: 71.18 (29%) Scenario 2: 3.50 (1.5%) Scenario 3: 20.43 (8.6%) Scenario 4: 3.50 (1.5%) Dewatered biosolids, ppd: Scenario 1: 165.44 (70%) Scenario 2: 233.06 (98%) Scenario 3; 102.87 (43.2%) Scenario 4: 130.50 (55%) P in struvite, ppd: Scenario 1: - Scenario 2: - Scenario 3: 113.32 (47%) Scenario 4: 102.8 (43%) ppd= pounds per day
Publications
- Type:
Conference Papers and Presentations
Status:
Under Review
Year Published:
2016
Citation:
Bashar, R., Gungor, K., and K.G. Karthikeyan. Model-based Evaluation of Implementing Phosphorus Recovery in a Municipal Wastewater Treatment Plant. Water Environment Federation (WEFTEC Meeting), New Orleans, LA, Sep. 2016.
|
Progress 10/24/13 to 09/30/14
Outputs
Target Audience: Targeted audiences of the proposed work include environmental research community, environmental consultants, personnel involved in the design of wastewater treatment plants, coordinators of biosolids land application program, and treatment plant operators. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided? This project is being primarily executed by a Ph.D.-level graduate student (in Biological Systems Engineering), Ms. Rania Bashar. How have the results been disseminated to communities of interest? Sampling activities are facilitated by the personnel at the Nine Springs WWTP. We have discussed the results of initial Biowin WWTP modeling work and preliminary whole plant mass-balance analysis with appropriate personnel at the Nine Springs WWTP. What do you plan to do during the next reporting period to accomplish the goals? We will continue all activities as originally proposed and approved.
Impacts
What was accomplished under these goals?
Data collection from Nine Springs Wastewater Treatment Plant (WWTP), Madison WI: For the purpose of mitigating maintenance costs associated with struvite scaling in anaerobic digesters, reducing phosphorus (P) content in biosolids, and reusing nutrients, the Nine Springs WWTP added a struvite recovery process in late 2013. To better understand the P recovery process, a whole-plant modeling on the plant operation is being done using Biowin v. 4.0 (EnviroSim Associates Ltd., Canada). A sample collection campaign was carried out in Sep. 2014 at the Nine Springs WWTP. A total of 16 samples were collected from the solid and liquid streams that covered the activated sludge, struvite recovery, and anaerobic digestion processes. The samples were analyzed for total and soluble components of Mg2+, Ca2+, Fe, K+, Na, PO43-, and NH4+. A mass balance check was also done on a load basis (flow*concentration) for Ca, Mg, Fe, K and P. The collected data will be used to calibrate the Biowin model. Additional samples will be collected in Spring 2015 for calibration and validation of the model. Brushite Precipitation Modeling: The objective here is to set up a model to simulate the chemical precipitation of dicalcium phosphate dihydrate (DCPD), known as brushite (CaHPO4. 2H2O), in organic acid digest upon the addition of Ca(OH)2. Acid phase digester is the first phase in a standard multi-phase anaerobic digestion system in which acidogenic bacteria convert organic matter into soluble compounds and volatile fatty acids. The supernatant from the organic acid digester is high in soluble phosphate. The addition of base raises the pH value and facilitates controlled precipitation of phosphate as brushite. The simulation of brushite precipitation was performed using BioWin. The composition of acid phase digestate from a pilot setup at the Nine Springs WWTP was used as the input stream to the model. Biowin’s user-defined variable option was used to integrate brushite mineral into the database. A custom module including brushite precipitation reaction was developed using the “model builder” feature of BioWin. This module and the default models in Biowin were used together to predict the amount of brushite precipitation. A separate module for struvite precipitation will also be developed to compare the two recovery processes in terms of cost and efficiency. To validate the results, a similar model for brushite precipitation was set up in Visual Minteq. Visual Minteq is a chemical equilibrium computer program that has an extensive thermodynamic database facilitating speciation calculations in an aqueous solution. The model was set to output the equilibrium concentrations of Mg2+, Ca2+, K+, PO43-, and NH4+, ionic strength of the solution and the mineral identity and amount of precipitates formed at the end of the reaction. It was determined from the model that approximately 1370 mg/L of Ca(OH)2 needs to be added to a typical acid digestate with dissolved Ca and PO43- concentrations of 110 and 1100 mg/L, respectively, to form 3.32 x 10-2 M brushite . The model also estimated that 66.7 % Ca2+ and 98.8 % of PO43- was precipitated in the form of brushite.
Publications
|