Process benefits of ozone/BAC as pretreatment to membrane-based advanced treatment for direct potable reuse

California has large agencies, serving 20 million people, that are planning and developing large-scale potable reuse projects that will be regulated as direct potable reuse (DPR). Significant investments will be made in new treatment processes and pipelines that will total more than $20 billion US. To support the development of these projects, California regulators have been working over the past decade to develop DPR regulations. One of the most significant requirements of the DPR regulatory criteria is a requirement to include ozone and BAC as pretreatment to the membrane-based full advanced treatment (FAT) train, commonly comprised of membrane filtration (MF), reverse osmosis (RO), and ultraviolet light (UV) advanced oxidation process (AOP), that is already in operation at more than a half dozen facilities in California with more than 8 m³/s (>180 million gallons per day (MGD)) production capacity today. California is currently the second state after Colorado with DPR regulations. In Colorado, ozone/BAC are optional treatment processes in contrast to California but Colorado also requires specific pathogen removal goals. Ozone is specifically identified as an alternative treatment for disinfection and ozone/BAC as an example of an alternative filtration process. Both disinfection processes, whether by ozone or UV, and filtration are required in Colorado for DPR. Other states such as Texas have implemented a case-by-case approval of DPR projects that rely on a blending of purified water with other waters to achieve regulatory approvals. Other past studies considered ozone/BAC as a carbon-based advanced treatment as an alternative to the RO-based treatment trains to avoid generation of RO concentrate waste streams that are challenging for disposal inland and to reduce treatment costs (Gerrity et al. 2011; Gerrity et al. 2014; Sundaram et al. 2020; Vaidya et al. 2020; Hogard et al. 2021; Teel et al. 2022). The growing interest in ozone/BAC can be explained by the successful demonstration of its ability to provide a robust organics removal barrier for a variety of compounds present in wastewater, including compounds of emerging concern (CEC). The organics removal occurs by both ozone and biological filtration, oxidation by ozone, and biological removal of bulk organic matter (Bourgin et al. 2018; Sari et al. 2020; Sundaram & Pagilla 2020; Zhiteneva et al. 2021; Hogard et al. 2023).

The requirement for the inclusion of ozone/BAC in California for DPR projects is to enhance the protection of public health. This is based on past research projects that showed these two processes help address a variety of concerns besides potential improvements in the treatability of feed water to the FAT train. For example, the work carried out by the authors during the Water Research Foundation (WRF) Project 4765/Reuse-14-12 resulted in a quantitative microbial risk assessment that demonstrated that a full-scale DPR treatment train employing ozone/BAC-MF-RO-UV/AOP treatment train can reliably meet performance goals and produce water that provides public health protection equivalent to, or greater than, conventional drinking water supplies (Trussell et al. 2018). This project utilized routine performance monitoring and performed chemical and viral surrogate challenge tests that assessed the benefits of the enhanced treatment train. This project demonstrated how a combination of treatment redundancy, robustness, reliability, and resilience can ensure DPR safety (Pecson et al. 2017; Pecson et al. 2018; Tackaert et al. 2019). The WRF Project 4991, ‘Defining Potential Chemical Peaks and Management Options,’ was another recently completed research project that evaluated data from the world largest indirect potable reuse (IPR) facility–Orange County Water District's Ground Water Replenishment System (100 MGD or 4.4 m³/s) as well as Singapore's Public Utility Board that operates several NEWater (Chew et al. 2010; Lefebvre 2018) advanced treatment reuse facilities (combined capacity of 175 MGD or 7.7 m³/s), as well as results from Pure Water San Diego's demonstration scale facility (1 MGD or 0.044 m³/s) that has been operating the California regulator proposed DPR train with ozone/BAC (Debroux et al. 2021) since 2015. Data from these facilities were used to assess removals of various chemicals by conventional wastewater treatment and advanced treatment to evaluate occurrence of chemical peaks and treatment options. Ozone/BAC was found to provide significant benefits in attenuation of certain types of organics, especially those with lower molecular weight including N-nitrosodimethylamine (NDMA), acetone, formaldehyde, and 1,4-dioxane as documented by Tackaert et al. (2019). Debroux et al. (2021) summarized four key strategies for utilities to ‘average’ chemical peaks and recommended the use of a balanced approach that includes two or more of these strategies: source control, monitoring, treatment, and blending. Another recent WRF study, project 4832—’Evaluation of CEC Removal by ozone/BAF Treatment in Potable Reuse Applications’ evaluated the effectiveness of ozone with biologically active filtration (O₃/BAF)-based treatment trains to address CECs in order to meet compliance with removal performance-based regulations, and to identify and address knowledge gaps with respect to public health (Robinson et al. 2023). This project highlighted ozone/BAC as being such an effective treatment to address CECs that non-membrane treatment trains can be safely used for IPR if supplemented by additional treatment steps such as granular activated carbon to meet drinking water standards.

Current DPR projects being pursued in California are most commonly a combined project with some water being delivered as DPR water and other applications defined as IPR with different regulatory requirements because the advanced treated water is not delivered directly to the customers but will first reside in either a large reservoir, or groundwater basin, which serves as an environmental buffer that provides response time and attenuation of any deviations in treatment, or water quality. With the confidence established from the industry's success with IPR facilities, a significant concern facing the water industry today is ‘over treating’ the water for IPR applications, when the facility is a combined IPR/DPR facility. While many configurations are being considered to provide the additional treatment requirements described by California's DPR regulations, there is currently a gap in understanding the impacts of ozone/BAC pretreatment on the additional cost over the life cycle of these facilities.

The goal of this paper is to provide engineers and researchers with a balanced perspective on the cost implications of incorporating ozone/BAC as pretreatment to a membrane-based potable reuse train. The paper is organized into two parts. Part 1 starts with providing an overview of improvements to feed water quality, enhancements in the performance of membrane systems, optimized control of biological fouling, and overall reductions in operating costs based on previously conducted studies. Part 2 considers a case study of a potable reuse facility currently in a design phase to assess life-cycle costs and compares costs of a membrane-based treatment train with and without ozone/BAC pretreatment. Cost comparison includes a discussion of the differences between capital costs and operation and maintenance (O&M) costs.

Test conditions: To evaluate exactly what concentrations of chloramines are necessary when ozone/BAC provide membrane pretreatment, over a course of several years, various operating conditions were evaluated in series: (1) variable flux 30–60 gfd (51–102 LMH), no ozone/BAC, chloramines dose of 3 mg/L as Cl₂; (2) testing with ozone/BAC pretreatment, chloramines at 3 mg/L as Cl₂ over flux conditions of 30–72 gfd (51–122 LMH); (3) 60 gfd (102 LMH) flux, ozone/BAC, and chloramines at a reduced dose chloramines dose of 1.2 mg/L; (4) 60 gfd (102 LMH) flux, ozone/BAC with enhanced flux maintenance (EFM) cleans on the ultrafiltration system and intermittent chloramine spikes (up to approximately 4.5 mg/L as Cl₂ for variable duration of 1–8 h) on the RO system. A membrane flux of 60 GFD (102 LMH), a flux well above the typical industry design point, was used for testing the UF system to demonstrate the difference in performance between test conditions 1 and 2. Test conditions 3 and 4 were used to assess whether the operating costs (energy and chemical) can be further optimized as compared to conditions 1 and 2. Operational data were collected daily to track the performance for each treatment process, including power and chemical use.

Table 1 provides analysis results for feed water quality that was collected monthly for a period of 1 year to develop a comprehensive understanding of seasonal variability as well as ozone/BAC performance operating continuously on tertiary treated wastewater. The tertiary effluent was well nitrified and limited nitrification occurred in the BAC filter over this test period.

The example provided in Figure 5 shows that with the right pretreatment, there is potential to optimize biological fouling control as well as to minimize organic fouling. Based on the results of this study, it was shown that operating with ozone/BAC and no chloramines was feasible for the UF with more than 30 days of operation without requiring a shutdown for a chemical clean. After several consecutive runs, the study found that lower transmembrane pressure could be achieved by introducing weekly maintenance cleans as preventive maintenance. Maintenance cleans are common practice in the water reuse industry and entail a brief recirculation step and soak of 20–30 min with 500 mg/L as Cl₂. The maintenance cleans with sodium hypochlorite provided even better performance than with the use of continuous chloramines and ozone/BAC at a reduced cost. With ozone/BAC pretreatment, the UF operation used 17.5% less power and 30% less chemicals.

The results of these tests have two significant implications. The first is that ozone/BAC allows for a membrane design and operation at a significantly higher flux, which reduces capital costs and footprint. The second is that the ozone/BAC pretreatment allows for a lower operating cost, due to a decreased rate of fouling (lower pumping costs) and less frequent cleaning (lower chemical costs).

Past studies reported positive benefits of pre ozonation and filtration by BAC for reducing fouling rates of reverse osmosis membranes (Stanford et al. 2011; Pramanik et al. 2015; Yin et al. 2020). Here the RO system was evaluated for 3-month test periods at two test conditions to evaluate the power and chemical requirements for controlling organic and biological fouling. Operating without continuous chlorination was again determined to be not only feasible but a significant process enhancement. Chloramines were intermittently used at a frequency of twice a week, once a week, once every 2 weeks, and even once every 4 weeks. Dosing chloramines twice a week for 5–6 h resulted in the lowest power and chemical consumption as compared to the conventional continuous chloramines strategy. However, with ozone/BAC both power and chemical use were 15% lower.

It was estimated that for every 1 mg/L as Cl₂ of chloramines residual present in RO permeate an additional 0.05 kWh/kgal (0.0132 kWh/m³) was needed to maintain the same level of UV dose to achieve disinfection and advanced oxidation process based on modeled removal of 1,4-dioxane at various concentration of monochloramine. Thus, when operating without continuous chloramines, UV/AOP power consumption will be reduced by 55%. Incorporating the intermittent chloramine strategy, the total power reduction is still significant at 41% on average.

Table 4 provides unit cost assumptions that were used to calculate operating and maintenance (O&M) costs for IPR and DPR trains based on the flows and chemical dosing provided in Table 3. Table 5 provides an estimate of the total annual operating and maintenance O&M costs for the two treatment trains. As discussed in the first part of the paper, the DPR treatment relies on a lower average chloramine dose, while MF and RO membrane systems operate at lower energy, higher feed water recoveries, and lower cleaning frequencies. Due to reduced background chloramines, the power use by UV/AOP is dramatically lower as well as benefits from upstream BAC that lowers the overall TOC in the product water as well as constituents such as NDMA that allows UV/AOP to target a lower UV dose as noted in Table 3.

As may be expected, the capital costs are higher for the DPR treatment train. However, the cost per acre-foot is comparable for the IPR and DPR trains when the benefits to water quality are considered in the operating costs for the downstream membrane and UV systems. Although for a hypothetical 10 MGD (26.3 m³/min) facility the combined construction costs were estimated to be approximately $29 million higher for the DPR treatment train as shown in Table 6, the O&M costs were more than $1.4 million less for the DPR configuration due to the water quality enhancements on downstream process performance. The capital costs included both construction costs and non-construction costs. Construction costs included equipment costs, electrical, mechanical, buildings, and any additional concrete structures (e.g. ozone contactor) and 25% contingency. Non-construction costs include estimates of typically required liability insurance, builder's risk insurance, contractor overhead, trade contractor profit and bond. Total capital costs were annualized over a 30-year period, which is the typical minimum life expectancy for major system components.

The lower O&M costs of $1.4 million per year for the DPR train are counterintuitive and the savings are high enough for this size of the facility to off-set the annualized increased capital cost associated with a DPR train. This finding highlights that the potential implications for the water industry are significant. First, this analysis illustrates that having ozone/BAC as part of a DPR train while adding significant capital and some additional O&M cost actually provides some enhancements to the water quality that reduce the capital costs of the downstream membrane and UV systems, allowing them to be more compact while producing the same volume of water and further enhancing water quality. Secondly, the downstream processes run more efficiently with lower average pressures and much fewer chemical cleanings. As a result, there is a reduction in chemical consumption due to the effective reduction in organics concentrations provided by the Ozone/BAC pretreatment that reduces the potential for biological fouling, and organic fouling, and adds to our treatment capabilities for low molecular weight compounds that have challenged membrane advanced treatment facilities in the past, such as NDMA. What the cost analysis did not consider is the increased reliability for public health protection that is provided by the addition of ozone/BAC to this treatment train. With ozone/BAC, the DPR treatment train's robustness and redundancy is enhanced with new treatment processes that address toxic organic chemicals and provide significant attenuation for a wide range of pathogens (e.g. 6-log virus, 6-log Giardia, 1-log Cryptosporidium). Considering the public health benefits and improvements in water quality for both the product water and RO concentrate, ozone/BAC pretreatment offers the potential to transform conventional potable reuse treatment approaches and based on this analysis, an ozone/BAC based DPR train is as cost-effective as a conventional RO-AOP-based IPR train due to the improved water quality that reduces the overall O&M costs.

Chloramines have a proven track record of providing biological fouling control of membrane filtration, however, increased oxidative damage of RO membranes, shortening the useful life of these membranes, and increasing salt passage accompany their use. In addition, chloramines reduce the UV transmittance of the UV/AOP influent, increasing the energy use of the AOP system. The use of ozone/BAC pretreatment offers significant improvements in water quality by eliminating membrane foulants (complex organics and manganese) and reducing the biofouling potential of the water improves the efficacy of the MF and RO membranes, as well as the UV/AOP process currently employed for indirect potable reuse projects. Based on the presented capital and O&M costs over a period of 30 years, the DPR treatment train has a similar water cost and this finding is significant because it also provides a greater degree of public health protection for potable reuse as well as a significant enhancement to the RO concentrate water quality, which is a growing regulatory concern as all these RO concentrate streams will ultimately be discharged to the Pacific Ocean. The water quality benefits offered by ozone/BAC provide significant performance enhancements that close the gap between the higher capital cost with lower O&M costs over the project life and this is currently poorly understood and underappreciated by the industry.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.