Abstract
Potable water reuse typically relies on treated wastewater from a conventional water resource recovery facility (WRRF) with further treatment through an advanced water treatment facility with chemical oxidation followed by membrane filtration. Conventional activated sludge (CAS) is used at WRRFs for bulk chemical oxygen demand (COD) and nutrient removal. Chemical oxidation and membrane filtration are employed to remove micropollutants and achieve additional pathogen inactivation. Although CAS is commonly used, it is not without challenges, such as sludge production, long detention times, susceptibility to treatment upsets, and lack of complete micropollutant removal2. In contrast, chemical oxidation can remove COD, making it suitable not only for advanced water treatment, but possibly for bulk COD removal as well1,3,5,6. It would be beneficial to evaluate the efficacy of replacing a sequence of CAS, chemical oxidation and membrane treatment with a simpler sequence of chemical oxidation (i.e., ozonation) followed by membrane treatment (i.e., reverse osmosis (RO)). This proposed system may achieve secondary treatment goals while also providing micropollutant removal and pathogen inactivation for potable reuse in a simpler process. However, no R&D reports were found that advance the proof-of-concept to eliminate CAS when chemical oxidation and membrane treatment are employed for municipal wastewater reuse. Currently, the scenario to eliminate CAS is at technology readiness level (TLR) 2 (concept and/or application formulated), whereas more work remains to move to TLR 3 (invention and research moving to proof of concept) and beyond4. In this study, bench-scale research was performed to prove the concept of ozonation followed by RO as a technology for municipal wastewater reuse (ozone-RO). Municipal wastewater primary effluent COD removal kinetics during ozonation were determined and compared to removal observed from CAS at a full-scale WRRF. Additionally, COD removal and nutrient fate and removal across ozone-RO were determined. Results were compared to COD and nutrient removal values achieved at a full-scale CAS followed by RO (CAS-RO). To determine ozonation COD removal kinetics, municipal primary effluent (PE) 24-h composite samples were collected from a WRRF (Fox River WRRF, Brookfield, WI). The CAS system fully nitrifies, with alum added to the secondary clarifiers for phosphorus removal. PE was ozonated in a 3-L batch ozone contactor at three different temperatures (7 degrees C, 22 degrees C and 34 degrees C). Ozonated gas was generated using a nozone generator (TG-10 Ozone Generator, Ozone Solutions, Hull, IA). A total applied ozone dose of 33 g O3/LWater over 3 h was used (45 g/m3 ozone gas with a 184 mg/L-min ozone application rate). To determine nutrient fate and removal, experiments were carried out using PE in the ozone contactor at 22 degrees C for 3 h. After ozonation, the water was sparged with nitrogen gas to remove residual ozone before a flat sheet RO membrane (FilmTec BW30XFRLE brackish water membrane) held in a cross-flow membrane cell (CF-106, Sterlitech). Pressure was held constant at 250 psi for 24 h. The RO process was then repeated with the same operating parameters but treating a 24-h composite sample of CAS effluent. COD, total and reactive phosphorus, total nitrogen, ammonia, nitrate, nitrite, and organic nitrogen concentrations were determined using commercial kits (Hach Company). Batch COD removal curves were fitted to the data using a first-order model with respect to COD concentration (Fig. 1) and the first-order rate constant (k) values were 0.0043 + 0.00037, 0.0046 + 0.00042, and 0.0067 + 0.0011 at 7 degrees C, 22 degrees C and 34 degrees C respectively. An ANNOVA test indicated that these k values were statistically different (p=0.0126). The resulting Arrhenius temperature coefficient value (ï±ï€© was 1.077. The CAS system had a 7-h average hydraulic residence time (HRT) before 1-hr secondary clarification compared to the ozone HRT of 3-h. Both COD and sCOD removal values with ozone and CAS were similar, and both treatment systems resulted in final effluent COD <5mg/L after RO treatment. Ozonation removed some phosphorus and converted some nonreactive phosphorus to reactive. It is hypothesized that phosphorus removal occurred by precipitation since a white precipitate was observed in the ozone contactor. Precipitation may have occurred due to a pH increased from 7.7 to 8.4 during ozone contact. Future experiments are planned to determine if this is the case. Final phosphorus and nitrogen concentrations from ozone-RO and CAS-RO systems were relatively low and similar. For example, total phosphorus concentrations for both effluents were below detection (<0.15 mg/L PO43--P) after RO. Although ozonation did not result in nitrogen removal, it did oxidize ammonia to nitrate (Fig, 4), which was subsequently removed by RO. This is significant since ammonia is not effectively removed by RO. No nitrite was detected (<0.6 mg/L NO2--N) in any samples. The concept of ozonation followed by membrane treatment was verified for municipal wastewater potable reuse and advanced from TLR 2 to TLR 3. Ozone rapidly oxidized PE COD under the conditions studied. In addition, quasi first-order kinetic constants and an Arrhenius temperature coefficient was established for ozonation of PE. Additionally, ozone-RO achieved final COD and nutrient concentrations similar to CAS-RO treatment. These data can inform system scale-up for R&D.
This study investigates potable water reuse, comparing ozonation followed by reverse osmosis (O3-RO) with conventional activated sludge followed by reverse osmosis (CAS-RO). Bench-scale tests show that O3-RO significantly reduces chemical oxygen demand (COD) to 1.0 +/- 0.88 mg/L, outperforming CAS-RO (3.7 +/- 1.9 mg/L, p=0.031). While ozone does not remove nitrogen, it converts Kjeldahl nitrogen to nitrate, enabling RO removal. These results advance O3-RO technology from readiness level 2 to 3.
Author(s)Booton, Alex, Mayer, Brooke, Zitomer, Daniel, Loftis, Payton
Author(s)A. Booton1, B. Mayer2, D. Zitomer2, P. Loftis1
Author affiliation(s)1Marquette University, WI, 2Marquette University Haggerty Eng Hall, WI
SourceProceedings of the Water Environment Federation
Document typeConference Paper
Print publication date Oct 2024
DOI10.2175/193864718825159621
Volume / Issue
Content sourceWEFTEC
Copyright2024
Word count10