University of Southern California Dissertations and Theses

Integrated technologies, blending schemes, and reuse practices to address contaminant and energy challenges in water reclamation

(USC Thesis Other)

Integrated technologies, blending schemes, and reuse practices to address contaminant and energy challenges in water reclamation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Integrated technologies, blending schemes, and reuse practices to address contaminant and energy challenges in water reclamation by Sophia Lauren Plata A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY ENGINEERING (ENVIRONMENTAL ENGINEERING) December 2021 Copyright 2021 Sophia Lauren Plata ii TABLE OF CONTENTS List of Tables…………...….....…………………...….....………………….…………...…......iv List of Figures……………….....…………………...….....…………………...…......………...v Abstract……………….....…………………...….....…………………...…......…………..…..vii Chapter 1: Introduction…………….……...….....…………………...…......………….……...1 Chapter 2: Limiting power density in pressure-retarded osmosis: Observation and . . implications...……………...….....………………..…...…......…………….……..5 2.1 Introduction………...….....……………….…..…........…………….……...6 2.2 Materials and Methods....……………….…..…........…………….……....8 2.2.1 Membranes....……………….…..…........…………….………....8 2.2.2 Membrane Module....……………….…..…........…….………....9 2.2.3 PRO Bench-scale System....……………….…..….........……..10 2.2.4 Burst Pressure Testing....……………….…..…........…………11 2.2.5 Fouling Experiments.……………….…..…........………...……11 2.2.6 Experimental Performance Parameters.…………………..….13 2.3 Results and Discussion.……………….…..…........……….....………….13 2.3.1 Burst Pressure.……………….…..…........…………........…….13 2.3.2 Water Flux…………….…..…………........…………........…….15 2.3.3 Power Density …………….…..…………........…………..........17 2.4 Conclusions and Implications …………….…..…………........………..19 Chapter 3: Minimizing N-nitrosodimethylamine formation during disinfection a . . . wastewater-seawater blend for potable reuse......…...…......….……….…….21 3.1 Introduction......…...…......….……….…………………………………….21 3.1.1 Pre-disinfection of seawater and wastewater......…...…........24 3.1.2 Pre-disinfection of a blended seawater/wastewater stream…26 3.2 Methods and Materials......…...…......….……….………………………..27 3.2.1 Modeling of chloramine and bromamine kinetics......…...…....27 3.2.2 Bench-scale Experiments......…...…......….……….………….28 3.3 Results and Discussion......…...…......….……….……………………….30 3.3.1 Modeling Results......…...…......….……….……………………30 3.3.2 Experimental Results ......…...…......….……….………………35 3.4 Conclusions......…...…......….……….……………………………………41 Chapter 4: The role of zero-liquid discharge on water reuse practices in the power . . sector: power sector review and natural gas case study analysis.........……..43 4.1 Introduction......…...…......….……….…………………………………….43 4.1.1 Water reuse practices in the power sector ......…...…......…..47 4.1.2 Zero-liquid discharge approaches......…...…......……………..48 4.1.3 Natural gas combined-cycle power facilities......…...…..........53 4.1.4 Objectives......…...…......….……….……………………………55 iii 4.2 Methods......…...…......….……….……………………………………..…56 4.2.1 Selection of case-study facilities......………………………..…56 4.2.2 Water Techno-economic Assessment Pipe-Parity Platform (Water-TAP3) model......…...…......….……….……………………………………….57 4.3 Results and Discussion......…...…......….……….………………………60 4.3.1 Existing Cooling Water Systems......…...…......….………...…60 4.3.2 Water-TAP3 Model Scenarios......…...…......….………...……66 4.4 Conclusion......…...…......….……….……………………………………..78 Chapter 5: Conclusions...……………...….....………………..…...…......…………….……80 References.....…...…......….………....…...…......….……….……......…...…......………....82 iv List of Tables Table 1. Membranes tested in PRO bench-scale system. Intrinsic transport parameters determined using the FO method*……………………………………………………………..9 Table 2 Chemical compositions of draw and feed solutions for fouling experiments…….12 Table 3 Burst pressure results for all membranes………………………………………….14 Table 4. Water flux for three stages of fouling experiments……………………………….15 Table 5. Composition of secondary treated wastewater and seawater…………………...27 Table 6. Composition of synthetic treated wastewater and seawater for bench-scale experiments……………….……………………………………………………………………30 Table 7. Composition of secondary treated wastewater and seawater used in realistic bench-scale experiments……………….……………………………………………………..38 Table 8. Summary of Water-TAP3 model results for Cherokee Generating Station…….67 Table 9. Summary of Water-TAP3 model results for Gila River Power Station…………..68 v List of Figures Figure 1 Schematic of desalination facility under (a) brine blending conditions and (b) feed blending conditions (direct potable reuse conditions)…………………………………2 Figure 2. Schematic of PRO bench-scale system…………………………………………10 Figure 3 Fouling protocol for PRO bench-scale experiments…………………………….12 Figure 4 Burst pressure results for Membrane 4. Operating pressure is shown by triangles with values reported on the primary vertical axis. Water flux is shown by circles with values reported on the secondary vertical axis. Feed solution conductivity is shown by squares with values reported on the secondary vertical axis………………………….14 Figure 5 Experimental operating water flux for Membranes 1, 2, 3300, 3180, and 4 for Stages I, II, and III………………………………………………………………………….….15 Figure 6 Experimental operating water flux and power density values for Membranes 1, 2, 3300, 3180, and 4. Power density is shown as red circles with values reported on the primary vertical axis. Water flux is shown as blue bars with values reported on the secondary vertical axis. ………………………………………………………………………18 Figure 7. Treatment trains of a (a) seawater desalination facility and (b) potable reuse facility……………………………………………………………………………………………23 Figure 8. Blending/disinfecting scenarios where, (a) wastewater is disinfected alone before the addition of seawater, (b) the blended stream is disinfected at the onset of blending, and (c) seawater disinfected alone before the addition of wastewater…….…27 Figure 9. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios A, B, and C at WW:SW = 1:1………………….31 Figure 10. Modelled (a) monochloramine, (b) dichloramine, and (c) bromochloramine formation of blending scenarios A, B, and C at WW:SW = 1:1 …………………………...31 Figure 11. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios A, B, and C at WW:SW = 1:1………………………………………32 Figure 12. Modelled (a) monobromamine and (b) dibromamine formation of blending scenarios A, B, and C at WW:SW = 1:1…………………………………………………….32 Figure 13. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios A WW:SW = 1:3, 1:1, and 3:1…………………33 Figure 14. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios B WW:SW = 1:3, 1:1, and 3:1…………………34 Figure 15. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios C WW:SW = 1:3, 1:1, and 3:1…………………34 Figure 16. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios A at WW:SW = 1:3, 1:1, and 3:1…………………………………..35 Figure 17. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios B at WW:SW = 1:3, 1:1, and 3:1…………………………………..35 vi Figure 18. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios C at WW:SW = 1:3, 1:1, and 3:1……………………………….…35 Figure 19. NDMA formation for blending scenarios A, B, and C at a WW:SW = 1:1 in a (a) clean and (b) realistic systems …………………………………………………………...37 Figure 20. Modelled (a) cumulative exposure of HOBr and (b) HOBr formation over time for scenarios A, B, anc C at WW:SW = 1:1…………………………………………………37 Figure 21. NDMA formation for blending scenarios A at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system…………………………………………………………..38 Figure 22. NDMA formation for blending scenarios B at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system…………………………………………………………..39 Figure 23. NDMA formation for blending scenarios C at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system…………………………………………………………..40 Figure 24. NDMA formation for blending scenarios A, B, C at a WW:SW = 1:1 in a clean system with (a) DMBzA and (b) DMA……………………………………………………….40 Figure 25. 2018 (a) water withdrawal intensity and (b) water consumption intensity from thermoelectric power plants (Gallons/MWh) …………………………………………….…44 Figure 26. Annual (a) water withdrawn and (b) water consumed by the power sector in 2018……………………………………………………………………………………………..45 Figure 27. Zero liquid discharge (ZLD) treatment scheme with (i) pretreatment, (ii) pre- concentration, and (iii) crystallization/evaporation…………………………………………52 Figure 28. Distribution of natural-gas-fired power facilities in the continental U.S. in 2016 …………………………………………………………………………………………….54 Figure 29. Natural-gas combined-cycle power facility…………………………………….55 Figure 30. Current Cherokee Generating Station cooling water flow diagram…………63 Figure 31. Proposed Cherokee Generating Station cooling water flow diagram with ZLD system…………………………………………………………………………………………..63 Figure 32. Gila River Power Station process flow diagram………………………………66 Figure 33. Water-TAP 3 results broken down by unit for Cherokee Generating Station.73 Figure 34. Water-TAP 3 results broken down by unit for Gila River Power Station…….75 Figure 35. (a, e) Relationship between water recovery and evaporation pond area, and sensitivity analysis evaluating impact of evaporation pond area on (b, f) LCOW, (c, g) percent LCOW attributed to electricity, and (d, h) electricity intensity of Cherokee Generating Station and Gila River Power Station………………………………………….77 vii Abstract Global water stress has prompted recent interest in desalination systems for treatment of saline and/or otherwise impaired water sources for potable use. These desalination systems typically rely on reverse osmosis (RO) as the main treatment process. RO, however, can have high energy demands and produces a high-salinity waste brine that must be disposed of or further treated. Seawater RO (SWRO) facilities may be co-located with wastewater treatment facilities that also discharge a stream to the ocean. Currently, the only synergistic use of these waste streams is to use the treated wastewater to dilute the SWRO brine stream prior to discharge. However, discharge of treated wastewater to the ocean could be considered a waste of the water resource within this stream and has no impact on the high energy demands of SWRO. The main goal of this dissertation is to provide solutions to increase potable water supply, reduce undesirable impacts associated with alternative water treatment, and improve water reuse practices at power facilities by evaluating the monetary, environmental, and public health costs associated with: (i) a membrane-based salinity gradient power process, (ii) a new concept for blending reclaimed wastewater and seawater upstream of disinfection, and (iii) zero-liquid discharge (ZLD) technologies employed at natural gas power facilities. For the first time, this work determined: (i) a power density phenomenon that limits osmotic power production for SWRO, (ii) an optimal wastewater-seawater blending and disinfection scheme to limit disinfection by-product formation, and (iii) the levelized cost of water for employing ZLD technologies at power facilities used to increase on-site water reuse. viii Chapter 1 1. Introduction It is estimated that by 2025, 1.8 billion people will live in areas plagued by water scarcity [1]. California, which has experienced severe droughts in recent years, is confronting water scarcity by securing alternative water supplies (e.g., wastewater and saline waters) that can be treated for potable use. Desalination facilities in Southern California currently provide 53 million gallons of water per day, serving over 400,000 people [2]. Wastewater reclamation facilities currently provide 2% of Los Angeles’ water supply, however, this figure is projected to increase to 35% by way of Mayor Eric Garcetti’s vow to recycle 100% of LA’s wastewater by 2040 [3]. Both desalination and wastewater reclamation facilities typically rely on reverse osmosis (RO) as the main treatment process [4]–[7]. RO is a pressure-driven membrane process that removes a range of components, from suspended solids to monovalent ions [8]–[11]. RO is operated at a net hydraulic pressure (DP) greater than the net osmotic pressure (Dp) [5], [8], [10]. RO systems typically operate between 5 – 120 bar with desalination applications on the higher end and wastewater reclamation on the lower end because of their respective salt concentrations [6], [10]. Although RO is an effective treatment process for seawater desalination (i.e., seawater RO (SWRO)), the high-pressure required can result in high energy demand [10], [12]. Additionally, RO produces a brine stream that must be disposed of or further treated [8], [13], [14]. Resolving or improving upon these undesirable consequences is 1 ix foundational to increasing reliance on SWRO as freshwater scarcity continues to grow scarce. Thus, to address the challenge of SWRO brine disposal, blending of saline waters with advanced treated wastewater has been proposed. Depending on the blending scenario, either at the inlet or outlet of the desalination system, the blending scheme could be considered potable reuse, which is regulated at the state level (not at the federal level by USEPA). Figure 1 depicts a (a) desalination facility under brine blending conditions where the brine is blended with treated wastewater before disposal to the ocean, and a (b) desalination facility under feed blending conditions (i.e., potable reuse conditions) where the feed stream is diluted with treated wastewater before it is sent to the desalination facility. Potable reuse can be classified as either indirect potable reuse (IPR) or direct potable reuse (DPR). In IPR, advanced treated wastewater is introduced into an environmental buffer (e.g., surface water body or groundwater aquifer) before being withdrawn for treatment at a drinking water treatment facility for potable purposes. In DPR, wastewater is recycled by treating it at an advanced water treatment facility before it is either directly sent to the municipal water system or is discharged upstream of a drinking water facility [15]–[17]. Figure 1 Schematic of desalination facility under (a) brine blending conditions and (b) feed blending conditions (direct potable reuse conditions) 2 x This work began with the more conservative form of blending, brine blending (Figure 1b), through the use of a novel osmotic process known as pressure-retarded osmosis (PRO). PRO was used to reduce system energy consumption and dilute the brine to meet discharge regulations. PRO relies on the difference in chemical potential across a dense, semi-permeable membrane to drive the movement of water molecules from a low-salinity feed solution to a high-salinity draw solution. A hydraulic pressure less than the osmotic pressure (Δ𝑃 <Δ𝜋) is applied to the draw solution so that water flux is still in the direction from the feed solution to the draw solution. Volume and pressure thus accumulate in the draw solution chamber which can later be released through a hydroturbine or pressure exchanger to generate energy. In this work, a PRO study was conducted on four commercially available membranes using a synthetic treated wastewater feed solution and seawater RO brine draw solution to understand the effects of membrane fouling and limiting flux. Extending the blending analysis from the more conservative brine blending scheme to blending of seawater with wastewater upstream of the RO process, an analysis of the specific energy consumption and formation of disinfectant byproducts from blending seawater and treated wastewater for potable reuse applications is being performed. The value of mutual dilution is two-part: (1) reduction of energy demand in the RO process by diluting seawater with treated wastewater effluent, and (2) increased potable water supply where seawater can be used to augment reclaimed wastewater. In both blending schemes, brine blending and feed blending, the relationship between water and energy played a critical role. To expand upon this relationship, the water and water reuse practices in the power industry were reviewed. Power facilities 3 xi are facing increasingly stringent discharge regulations that are requiring more facilities to adhere to zero-liquid discharge (ZLD) practices. This can be costly as ZLD systems typically rely on high recovery technologies that have high maintenance and electricity costs. However, an added benefit of adhering to regulations is increased on-site water reuse, which minimizes water withdrawals and environmental footprint. Thus, to understand the value and cost of ZLD implementation, two natural gas power facilities with different ZLD technologies were analyzed as case studies. 4 xii Chapter 2 2. Limiting power density in pressure-retarded osmosis: Observation and implications This chapter is published in Desalination [18] Abstract Pressure-retarded osmosis (PRO) shows promise for mitigating energy and environmental concerns associated with seawater reverse osmosis (RO) desalination. A systematic investigation of PRO membrane fouling and limiting flux was conducted for the RO-PRO scenario, where increasingly complex feed and draw solutions were used to represent treated wastewater and seawater RO brine, respectively. Four membranes were tested for water flux and power density. When there were no foulants in the feed solution, the membranes had distinctly different water flux and power density values. When there were foulants in the feed and draw solution, all membranes reached a similar water flux and power density regardless of intrinsic transport parameters and operating pressure. For the first time, this study identified limiting water flux for multiple membranes in PRO and also, a limiting power density. To increase water flux and power density in PRO, it was suggested to first focus efforts on mitigating fouling on the feed side (support layer) of the membrane. If that can be done, then improvement of membrane transport parameters would play a vital role in increasing the limiting water flux and power density. 5 xiii 2.1 Introduction Global water stress has increased interest in desalination systems for treatment of saline and/or otherwise impaired water sources for potable use. Desalination systems typically rely on reverse osmosis (RO) as the main treatment process [4], [5], [19]. RO, however, can have high energy demands and produces a high-salinity brine stream that must be disposed of or further treated. To mitigate the energy and environmental costs of RO desalination, emerging processes have been proposed. For example, pressure- retarded osmosis (PRO) has been suggested in a RO-PRO scenario as a way of diluting RO brine and, at the same time, harvesting energy from the osmotic pressure difference between the RO brine (serving as a high-salinity draw solution) and an impaired water stream (or low-salinity feed solution) [20]–[24]. In PRO, water molecules pass from the feed solution, through a semipermeable membrane, to the draw solution due to the difference in chemical potential between the feed and draw solutions. While water molecules pass through the membrane by osmosis, hydraulic pressure is applied to the draw solution; thus, osmosis is “retarded” by hydraulic pressure. Volume and pressure accumulate in the draw solution, which can then be released through a hydroturbine or pressure exchanger to produce energy [20], [25]–[30]. The literature suggests that under ideal conditions, PRO operation at a hydraulic pressure equal to half the osmotic pressure achieves the theoretical global maximum power density (e.g., [31], [32]). However, operation of PRO systems at hydraulic pressures less than half the osmotic pressure have also been discussed/mentioned in the literature for achieving maximum power density under non-ideal conditions (e.g., [20], [27]). 6 xiv Ideally, PRO membranes have high water permeability (A), low salt permeability (B), low structural parameter (S), and resistance to fouling [28]. These general characteristics are also sought in forward osmosis (FO) and RO membranes [33]. FO membranes do not require high mechanical strength, so optimization can focus on increasing water permeability (high A) while maintaining high rejection (low B) [34], [35]; RO membranes require much higher mechanical strength to withstand higher pressures [36], [37]. In PRO, with hydraulic pressure opposing flux, higher A and greater mechanical strength are required compared to FO membranes, while a relatively low S must be maintained to reduce internal concentration polarization in both FO and PRO and to achieve high energy recoveries in PRO [33]–[35]. Currently, there are no commercially available membranes specifically manufactured for PRO; however, FO membranes are typically used for PRO testing because these membranes also operate under an osmotic pressure driving force [33]. The effects of membrane characteristics on PRO performance have been studied at length in bench-scale systems using FO membranes under ideal conditions (e.g., [20], [22], [30], [34], [38]). Fewer studies have examined these effects under fouling conditions (e.g., [39]–[42]). She et al. [40], [42], [43] introduced the concept of a critical flux under fouling conditions in PRO. Critical flux is defined as the water flux of a single membrane, at a single operating pressure, at which there is no decline due to membrane fouling – but above which, fouling occurs. She et al. [44] later expanded upon the critical flux concept by demonstrating a pressure-independent limiting flux for a single membrane and developing a novel PRO limiting-flux model. This phenomenon of several membranes having a similar steady-state water flux upon exposure to a fouling 7 xv feed solution has also been observed in RO, nanofiltration (NF) (e.g., [45], [46]), and FO (e.g., [40], [42], [43]), but was only recently referred to as a limiting flux in Morrow [47]. In this study, FO membranes were operated under fouling conditions in a bench-scale system and it was suggested that membrane transport parameters (i.e., A and B) play a minimal role, if any, in system performance when membrane fouling occurs. The limiting flux was defined as the steady-state water flux reached by several FO membranes under fouling conditions regardless of increased driving force (i.e., osmotic pressure). The objective of the current study is to conduct a systematic investigation of PRO membrane fouling and limiting flux using increasingly complex feed and draw solutions with application-specific solution chemistries. In order to isolate fouling effects on both sides of the membrane, solution complexity was increased stepwise via sequential addition of salts and foulants, eventually resulting in solution chemistries representing seawater RO brine as the draw solution and treated wastewater as the feed solution. Several FO membranes were tested to confirm the presence of a membrane-independent limiting flux in PRO. Also, one membrane was tested at two hydraulic pressures to support earlier work that showed limiting flux is independent of operating pressure. Power densities were calculated and compared for fouling and non-fouling conditions, leading to the development of a new concept: limiting power density. 2.2 Materials and Methods 2.2.1 Membranes Four semi-permeable FO membranes (Table 1) were tested. All membranes were flat-sheet and either thin-film composite (TFC) or cellulose triacetate (CTA). 8 xvi Table 1. Membranes tested in PRO bench-scale system. Intrinsic transport parameters determined using the FO method* Membrane Material Company Water Permeability Coefficient A (L m -2 h -1 bar -1 ) Salt Permeability Coefficient B (L m -2 h -1 ) 1 TFC HTI (Albany, OR, USA) 2.78 ± 0.21 1.36 ± 0.14 2 TFC Toray (Chūō, TKY, JP) 2.29 ± 0.34 0.30 ± 0.08 3 CTA FTS (Albany, OR, USA) 0.77 ± 0.08 0.29 ± 0.04 4 TFC Porifera (San Leandro, CA, USA) 3.35 ± 0.51 1.18 ± 0.18 *Characterization data from reference [47] 2.2.2 Membrane Module A stainless-steel membrane module providing an active membrane surface area of 0.012 m 2 was used. The feed solution channel had a depth of 1.7 mm and the draw solution channel had a depth of 0.55 mm. All membranes were installed with the active layer facing the draw solution (AL-DS), which is typical for PRO [29]. Spacers were used on either side of the membrane to enhance solution mixing and provide structural support. Two spacers were used in the feed solution channel: a 0.5-mm tricot warp-knit spacer (Hornwood Inc., Lilesville, NC, USA) and a 1.20-mm diamond-mesh spacer (Sterlitech Corporation, Kent, WA, USA). Only the tricot warp-knit spacer was used in the draw solution channel. The edges of the spacers were taped onto the membrane module to prevent the spacers’ edges from puncturing the membrane [30]. The module also had rounded edges and offset channels to prevent puncturing or tearing of the membrane, which would limit the maximum pressure that could be applied. 9 xvii 2.2.3 PRO Bench-scale System A schematic of the PRO bench-scale system is shown in Figure 2. The feed and draw solutions were circulated at a flow rate of 0.5 L/min with a low-pressure, variable- speed gear pump (Cole-Parmer, Vernon Hills, IL, USA) for the feed solution and a high- pressure positive displacement pump (Wanner Engineering Inc. Minneapolis, MN, USA) for the draw solution. Hydraulic pressure on the draw side of the membrane was controlled by a needle valve on the outlet of the module. The pressures of the feed and draw solutions were monitored with pressure transducers at the inlets and outlets of the module. Figure 2. Schematic of PRO bench-scale system As water passed through the membrane, feed solution from a 4-L reservoir continuously replenished the feed solution tank, which had a float valve to maintain constant volume. Water flux was calculated from the change in weight of the feed solution reservoir that sat on a scale. Solution conductivity was monitored with conductivity probes in the feed and draw solution tanks. To maintain a constant osmotic pressure driving force across the membrane, the draw solution was maintained at a constant conductivity. When 10 xviii draw solution conductivity dropped below a pre-set value, a peristaltic pump (Cole- Parmer, Vernon Hills, IL, USA) was triggered to pump three-times concentrated draw solution from the reservoir to the draw solution tank. Data from the pressure transducers, scale, and conductivity probes were recorded with data acquisition software (National Instruments, Austin, TX, USA). 2.2.4 Burst Pressure Testing Membranes were loaded in the module in the AL-DS configuration. Hydraulic pressure was gradually increased on the draw solution side at a rate of 5 psi/min. Water flux and feed solution conductivity were measured. This continued until a sharp drop in water flux and sharp increase in feed solution conductivity were observed, indicating a breach in membrane integrity. The pressure at which the membrane was compromised was defined as the burst pressure. 2.2.5 Fouling Experiments In order to isolate fouling effects on each side of the membrane, salts and foulants were incrementally added to the draw and feed solutions in three 10-hour stages over a three-day period (Figure 3). A hydraulic pressure of 300 psi was applied on Membranes 1, 2, and 3 but not Membrane 4, which had a manufacturer-specified maximum operating pressure of 180 psi. Membrane 3 was also tested at 180 psi to compare performance with Membrane 4. The net osmotic pressures shown in Figure 3 were calculated using OLI Stream Analyzer 5.0 (OLI Systems Inc., Cedar Knolls, NJ, USA). Net osmotic pressure decreased slightly in each stage due to the changes in salt composition discussed below. 11 xix Figure 3 Fouling protocol for PRO bench-scale experiments In Stage I, a baseline was established with 70 g/L NaCl as the draw solution, which approximated the salinity of seawater RO brine, and DI water as the feed solution. In Stage II, the draw solution complexity was increased to include other salts and foulants characteristic of seawater RO brine. In Stage III, the feed solution complexity was increased to include salts and foulants characteristic of treated wastewater. The draw and feed solution chemistries used in Stages II and III of the fouling experiments are shown in Table 2. Table 2 Chemical compositions of draw and feed solutions for fouling experiments SWRO Brine Draw Solution* Treated Wastewater Feed Solution * Salt mM Salt mM Sodium sulfate 56.2 Sodium citrate 3.32 Magnesium chloride 123 Ammonium chloride 1.88 Sodium bicarbonate 3.78 Monopotassium phosphate 0.9 Calcium chloride 23.2 Calcium chloride 1 12 xx Sodium chloride 832 Sodium bicarbonate 1 sodium chloride 4 Magnesium sulfate 1.2 Foulant mg/L Foulant mg/L Sodium alginate 200 Sodium alginate 50 Bovine serum albumin 50 Natural organic matter 50 *Treated wastewater and SWRO brine recipes are from reference [48] 2.2.6 Experimental Performance Parameters Water flux and power density were used to quantify membrane performance. Water flux (Jw in m/s) was determined from [49]: 𝐽 ! = " # ! ∗∆& (1) where V is volume (L) of water flowing across the membrane, Am is membrane area (m 2 ), and Dt is elapsed time (hr). Power density (W in W/m 2 ) was determined from: 𝑊 =𝐽 ! ∗∆𝑃 (2) where ∆P is change in applied hydraulic pressure across the membrane (Pa). 2.3 Results and Discussion 2.3.1 Burst Pressure Results from the burst pressure experiment for Membrane 4 is shown in Figure 4. Results for all membranes are shown in Table 3. As operating pressure increased, water flux decreased, and feed solution conductivity remained fairly constant. The membrane was considered compromised when the slope in water flux between two consecutive time intervals changed by 100% and was confirmed by a subsequent sharp increase in feed conductivity. The operating pressure at the midpoint of the time interval corresponding to the sharp flux decline (when the membrane was considered 13 xxi compromised) was taken as the burst pressure. Burst pressure is shown in Figure 4 with a dashed line and results for all membranes are shown in Table 3. The higher- pressure membranes, Membranes 1, 2, and 3 reached the system’s maximum operating pressure (550 psi) without failing and are thus indicated as having burst pressures of 550+ psi. The lower pressure membrane, Membrane 4, had a burst pressure of 220 psi, which was above the maximum manufacturer specified operating pressure of 180 psi. Figure 4 Burst pressure results for Membrane 4. Operating pressure is shown by triangles with values reported on the primary vertical axis. Water flux is shown by circles with values reported on the secondary vertical axis. Feed solution conductivity is shown by squares with values reported on the secondary vertical axis. Table 3 Burst pressure results for all membranes Membrane Company Burst Pressure (psi) 1 HTI 550+ 2 Toray 550+ 3 FTS 550+ 4 Porifera 220 14 xxii 2.3.2 Water Flux Graphs of water flux as a function of time for Membranes 1, 2, and 3 are shown in Figure 5. The system operating water flux was taken as the average over the last four hours of each stage. Operating water flux values for each stage are reported in Table 4 for Membranes 1, 2, and 3. Figure 5 Experimental operating water flux for Membranes 1, 2, 3300, 3180, and 4 for Stages I, II, and III Table 4. Water flux for three stages of fouling experiments Membrane Stage I Stage II Stage III 1 25.6 ± 1.0 24.3 ± 0.5 5.7 ± 0.8 2 21.0 ± 0.7 15.8 ± 0.6 5.6 ± 0.6 3 17.2 ± 0.2 12.0 ± 0.2 7.5 ± 0.7 Average 21.3 ± 3.5 17.3 ± 5.2 6.2 ± 1.1 In Stages I and II of the membrane fouling experiments, the operating water flux values of the Membranes 1, 2, and 3 decreased with decreasing A values (Table 1). For all membranes, Stage II operating water fluxes were lower than Stage I, likely due to the reduced osmotic pressure driving force. Although only a small decrease, the compounding factors of increased reverse solute diffusion, external concentration polarization, and solute/foulant interactions may increase the negative effect on water 15 xxiii flux [27], [50], [51]). Solute/foulant interactions on the draw side of the membrane may also have hindered water flux if foulants deposited on the active layer; however, the majority of these deposits would likely be lifted off of the surface because water flux is in the direction of the feed to the draw solution [48], [52]; this effect is analogous to osmotic backwashing [41], [53]–[58]. From Stage II to Stage III, the operating water fluxes of Membranes 1, 2, and 3 decreased by 77, 65, and 38%, respectively. This was a notably sharper decrease in operating water flux as compared to the decrease from Stage I to Stage II (5, 25, and 30%). The decreased water fluxes in Stage III were expected due to the addition of foulants to the feed solution. However, the fact that the operating water fluxes decreased to a similar value of 6.2 L/m 2 hr was not expected, especially given the distinctly different operating water fluxes of the membranes in Stages I and II and their distinctly different intrinsic transport parameters (Table 1). The relatively low standard deviation of the average operating water flux in Stage III (± 1.1 L/m 2 h) strongly contrasts the much higher standard deviations reported for the average operating water flux values for Stages I (± 3.5 L/m 2 h) and II (± 5.2 L/m 2 h). The phenomenon of similar water flux values upon exposure to a fouling feed solution in PRO was previously observed in work by She et al. From Stage II to Stage III, the operating water fluxes of Membranes 1, 2, and 3 decreased by 77, 65, and 38%, respectively. This was a notably sharper decrease in operating water flux as compared to the decrease from Stage I to Stage II (5, 25, and 30%). The decreased water fluxes in Stage III were expected due to the addition of foulants to the feed solution. However, the fact that the operating water fluxes decreased to a similar value of 6.2 L/m 2 hr was not expected, especially given the distinctly different operating water fluxes of the 16 xxiv membranes in Stages I and II and their distinctly different intrinsic transport parameters (Table 1). The relatively low standard deviation of the average operating water flux in Stage III (± 1.1 L/m 2 h) strongly contrasts the much higher standard deviations reported for the average operating water flux values for Stages I (± 3.5 L/m 2 h) and II (± 5.2 L/m 2 h). The phenomenon of similar water flux values upon exposure to a fouling feed solution in PRO was previously observed in work by She et al. [40], [42], [44] who defined “limiting flux” in PRO for a single membrane operated under a range of operating pressures [44]. The current work extends this pressure-independent limiting flux concept by demonstrating that a limiting flux exists amongst several membranes with different intrinsic transport parameters. 2.3.3 Power Density Power density values for all membranes in all stages are shown in Figure 6 along with operating water flux values. Power density was calculated using the operating water flux values and average applied hydraulic pressure taken over the last four hours of each stage. 17 xxv Figure 6 Experimental operating water flux and power density values for Membranes 1, 2, 3300, 3180, and 4. Power density is shown as red circles with values reported on the primary vertical axis. Water flux is shown as blue bars with values reported on the secondary vertical axis. In addition to the experiments conducted on Membranes 1, 2, and 3 at 300 psi, Membrane 3 was also tested at 180 psi for comparison with the high-pressure experiments and for comparison with Membrane 4. In Stages I and II, Membrane 1 had the highest operating water flux and power density; Membrane 4 also had a high water flux but a low power density. Given its high water flux, the low power density of Membrane 4 is due to the lower applied hydraulic pressure. In Stage I, the average of the operating water fluxes for Membrane 3 at 300 and 180 psi was 18.5 ±1.5 L/m 2 hr. In Stage II, the average operating water flux for Membrane 3 at 300 and 180 psi decreased to 12.9 ± 1.0 L/m 2 hr. In Stage III, the average operating water flux decreased further to 7.5 ± 0.5 L/m 2 hr; the similarity in water flux at different applied pressures agreed with the pressure- independent limiting flux observed by She et al. [44]. The average power density was 3.5 ± 0.9 W/m 2 . Thus, operation of Membrane 3 at 180 psi has clear energetic benefits – less pressure is required to achieve a similar power density as can be achieved at 300 psi. 18 xxvi This was further exemplified by Membrane 4, which had the highest water flux rate but a comparable power density to Membranes 1, 2, 3300, and 3180 (3.5 ± 0.6 W/m 2 ). In fact, in Stage III, the average operating power density of Membranes 1, 2, 3300, 3180, and 4 (3.4 W/m 2 ) had a standard deviation of ± 0.7 W/m 2 whereas the standard deviations in Stages I and II were ± 2.9 W/m 2 and ± 3.2 W/m 2 , respectively. This demonstrates the existence of a limiting power density in PRO under fouling conditions. Regardless of membrane transport parameters and operating pressure, Membranes 1, 2, 3300, 3180, and 4 reached a similar limiting power density within 1 W/m 2 . 2.4 Conclusions and Implications The concept of limiting water flux in PRO was demonstrated for several membranes and the concept of a limiting power density was introduced for the first time. The observation of a limiting power density that is independent of membrane transport parameters has direct implications for recommendations in the literature to operate membranes under PRO conditions at hydraulic pressures close to half the osmotic pressure [20], [27]. In the case demonstrated here (i.e., with synthetic treated wastewater as the feed solution and synthetic SWRO brine as the draw solution), operating at half the osmotic pressure would mean operating at 407 psi. The membranes in this study were tested at 300 psi due to system limitations; however, based on comparison of power densities achievable at 300 and 180 psi, it is unlikely that operation at 407 psi would be more beneficial. Under fouling conditions, it appears that membranes can be operated at lower hydraulic pressures and still produce a similar power density to systems operating at higher hydraulic pressures. It should also be noted that the average limiting power density of 3.4 ± 0.7 W/m 2 in this study, and the 19 xxvii power density of each individual membrane (Membranes 1, 2, 3300, 3180, and 4) were all slightly under 5 W/m 2 , which was previously set as a benchmark minimum for PRO to be economically feasible [22], [59]. Based on Stage I and II results, these membranes appear to exceed this benchmark; however, based on Stage III results this was not the case; all membranes reached a limiting power density below the 5 W/m 2 benchmark. Based on the limiting power density values found in the current investigation, it is clear that it is not enough to develop membrane materials with high power densities under non-fouling conditions, but instead, to consider ways to mitigate membrane fouling and increase the limiting water flux and power density under fouling conditions, particularly on the feed side (support layer) of the membrane in PRO. Until then, given the burst pressure of the membranes, it could be suggested that it is not necessary for membrane manufacturers to expend efforts in increasing the durability of these membranes (e.g., to reach burst pressures > 550 psi). Acknowledgements This study was supported by the California Department of Water Resources Water Desalination Grant Agreement 4600011018 and the University of Southern California’s Viterbi School of Engineering Fellowship. Additionally, the authors would also like to acknowledge Toray Industries Inc., Porifera Inc., and Fluid Technology Solutions (FTS) Inc. for providing their membranes. 20 xxviii Chapter 3 3. Minimizing N-nitrosodimethylamine formation during disinfection a wastewater-seawater blend for potable reuse 3.1 Introduction Similar to peak oil, the concept of peak water demonstrates that sustainably managed water is becoming scarce [60]. This is particularly true for groundwater aquifers with slow recharge rates that are over pumped. Thus, analogous to renewables that are used as a backstop for oil, alternatives waters (e.g., wastewaters or saline waters) can be used as a backstop for freshwaters [60]. Although water is a finite source, it can become practically unlimited at a price. Global desalination capacity has increased approximately 7% annually from 2010-2019 with seawater desalination capacity increasing 813% from 1990-2019 [61]. Desalination facilities are often built in response to drought, however, facilities run the risk of becoming obsolete if drought conditions subside. For example, Santa Barbara Desalination Facility, which was built in the 1980s in response to severe drought, was mothballed in 1992 when the region experienced consistent, heavy rainfall [62], [63]. For this reason, cities often elect to make desalinated water a permanent part of their water supply portfolio to ensure use of their investment [63]. According to a 2016 study by the Pacific Institute, the levelized cost of water for large seawater desalination projects was $2,100/acre-ft, while the cost of large indirect potable reuse projects was $1,800/acre-foot [64]. Thus, in the future, as public health concerns associated with potable reuse decrease due to advances in science and technology, perhaps reclaimed water can replace seawater as a permanent part of the water supply portfolio to decrease water costs. Moreover, seawater can be 21 xxix periodically added to the water supply as a peak load to meet demand. In this way, reclaimed water would be considered a part of the baseload of the water supply portfolio and seawater as peak load. Not only does this reduce water costs to meet demand but it also aligns with Southern California’s Green New Deal that aims to recycle 100% of all wastewater for beneficial use by 2035 and to locally source 70% of Los Angeles’ water by 2035 [65]. Wastewater reclamation first began in practice as de-facto reuse of municipal wastewater effluent discharged upstream of drinking water treatment facilities [66], [67]. As water availability became increasingly limited, water-stressed regions (i.e., California and Texas) began leading the way in potable reuse projects as early as the 1960s [68]– [70]. Currently, there are roughly 150 potable reuse projects throughout the US [71], [72] with the largest in Southern California at Orange County Water District’s (OCWD) Groundwater Replenishment System (GWRS) [73]. OCWD’s GWRS produces 100 million gallons/day (MGD) of potable water to recharge a large groundwater basin in North and Central Orange County [73]. Seawater intrusion that occurs when groundwater levels are lowered due to over drafting is a challenge facing OCWD’s GWRS and many other coastal aquifers in the US [74]. Seawater intrusion and infiltration has also been observed at coastal wastewater treatment facilities as a result of aging infrastructure, corroded pipes, and leaky gates [75]. While the high total dissolved solids (TDS) seawater (i.e., 35,000 mg/L) can be considered a challenge to treat, the additional seawater has the benefit of recharging groundwater aquifers to meet potable water demands. The additional seawater can be considered a peak load to meet demand; however, there has been limited research into the chemical 22 xxx repercussions of blending and disinfecting a wastewater/seawater stream for potable use. The blending of reclaimed water and seawater for potable reuse has not been extensively studied in the literature. Only recently, during an energy feasibility study performed by Wei et al., was significant overlap between desalination and potable reuse treatment trains identified [76]. Nearly identical treatment schemes are used for desalination and potable reuse, except for the type of pre-disinfectant chemical and use of biological activated carbon and post-oxidation step in the potable reuse treatment scheme (Figure 7). Thus, there can be significant energetic benefits, without the capital cost, footprint, and other concerns associated with building a new facility, that could be obtained by blending treated wastewater with the intake seawater and treating the blended stream with the appropriate processes. Figure 7. Treatment trains of a (a) seawater desalination facility and (b) potable reuse facility The benefits of blending treated wastewater with intake seawater upstream of reverse osmosis (RO) are three-fold: (i) reduced pumping energy required for the RO 23 xxxi process, (ii) reduced environmental impact associated with the discharge of a high- saline seawater RO (SWRO) brine, and (iii) increased volume of potable water produced. The first two benefits have been moderately studied in the literature from the perspective of augmenting seawater with wastewater to reduce cost and environmental impacts [76], [77], while the third benefit has not been explored, and in particular, from the perspective of augmenting wastewater with seawater to increased potable water produced. Although these blending scenarios have several potential benefits, there are still significant concerns associated with its direct potable reuse (DPR) applications (e.g., disinfection byproducts (DBPs)). In this study the chemical ramifications of blending and disinfection seawater and wastewater for potable reuse purposes was explored. 3.1.1 Pre-disinfection of seawater and wastewater The most common pre-disinfectant in seawater desalination is free chlorine (i.e., HOCl/OCl - ) [78], [79] and the most common pre-disinfectant in advanced water treatment is chloramines (i.e., NH2Cl/NHCl2/NCl3) [80], [81]. Free chlorine is less costly and effectively controls biofouling, however, it has been known to damage membranes [82], [83] and produce higher levels of regulated DBPs such as trihalomethanes (THMs) and halo acetic acids (HAAs) [84]–[87]. Chloramines, on the other hand, form lower concentrations of THMs and HAAs, but have been known to produce other, more carcinogenic, DBPs like N-nitrosodimethylamine (NDMA) [87]–[90]. Additionally, in waters with high concentrations of bromide (e.g., seawater), chloramines produce bromamines, which can lead to brominated DBPs that are more cytotoxic and genotoxic 24 xxxii than THMs and HAAs [91]–[93]. When free chlorine is introduced into a process water that is rich in ammonia, the chlorine (HOCl) oxidizes the ammonia (NH3) rapidly, forming chloramines. Monochloramine (NH2Cl) is formed first, followed by dichloramine (NHCl2) and trichloramine (Cl3) in a process known as in-situ chloramination: 2NH ' +H2OCl →2NH ( Cl+2H ( O (1) 2NH ( Cl+2HOCl→2NHCl ( +2H ( O (2) 3NHCl ( +3HOCl→3NCl ' +3H ( O (3) Thus, due to the high levels of ammonia in wastewater, in-situ chloramination will occur in the disinfection of the blended seawater-wastewater streams. The use of free chlorine has been extensively studied in both seawater desalination [94], [95] and advanced water treatment [96], [97]. Chloramines, however, have only been extensively studied in the context of advanced water treatment [98]–[101]. There are no previous studies, to the best of our knowledge, that evaluate the chemical kinetics and speciation of chloramine disinfection of seawater. Aside from free chlorine and chloramine, there are several other disinfectants for disinfection that could be considered for pre-treatment of the blended stream. Ultraviolet (UV) radiation, for instance, is a strong oxidation processes that effectively break down organic compounds from water; however, it has no effect on the biological stability of seawater and can have higher costs when compared to other pre-disinfection processes [102], [103]. Ozone is another strong oxidant that effectively kills microorganisms; however, it is known to react with bromide to form bromate (BrO3-), a DBP, or hypobromous acid (HOBr), a DBP precursor [104]–[106]. Thus, the high levels of 25 xxxiii bromide in seawater make the selection of an appropriate disinfectant/oxidant more complex than if wastewater or seawater were treated independently. 3.1.2 Pre-disinfection of a blended seawater/wastewater stream One critical consideration in the disinfection of a blended seawater/wastewater stream is the speciation of the oxidants formed during chlorination. Wastewater chlorination produces chloramines by ammonia oxidation, which has been associated with the formation of disinfection byproducts (DBPs) like NDMA [100], [107]. The currently accepted reaction pathway for NDMA formation during chloramination occurs when dichloramine (NH2Cl) reacts with an amine precursor (e.g., dimethylamine (DMA), ranitidine (RNA), and dimethylbenzylamine (DMBzA)) to form chlorinated unsymmetrical dimethylhydrazine (UDMH-Cl). UDMH-Cl then reacts with dissolved oxygen in water to form NDMA [100]. Seawater chlorination, on the other hand, produces free bromine by bromide oxidation, which when blended with ammonia-rich wastewater produces bromamines that are also associated with NDMA formation [107]–[110]. Bromamine formation and speciation between monobromamine (NH2Br), dibromamine (NHBr2), and bromochloramine (NHBrCl) is expected to be a function of which stream is chlorinated, and whether the disinfection happens before or after blending. In this work, for the first time, the chemical kinetics and speciation of blending and disinfecting a seawater/wastewater stream for potable reuse applications was studied. Various blending and disinfection scenarios (e.g., disinfecting seawater before blending with wastewater or disinfecting wastewater before blending with seawater) were first modeled with chemical kinetics software to determine the formation and speciation of chloramines and bromamines. These were followed by bench-scale experiments in 26 xxxiv clean and realistic systems with the overall goal of determining an optimal blending/disinfection scenario that would reduce NDMA formation. 3.2 Methods and Materials 3.2.1 Modeling of chloramine and bromamine kinetics The extent of bromamine formation and speciation between monobromamine (NH2Br), dibromamine (NHBr2), and bromochloramine (NHBrCl) was expected to be a function of stream composition (Table 5), chlorine dose, and contact time. Thus, three different blending scenarios were evaluated (Figure 8) with the overall goal of minimizing the formation of the bromamine species principally responsible for byproduct formation (i.e., dibromamine and bromochloramine). Table 5. Composition of secondary treated wastewater and seawater Figure 8. Blending/disinfecting scenarios where, (a) wastewater is disinfected alone before the addition of seawater, (b) the blended stream is disinfected at the onset of blending, and (c) seawater disinfected alone before the addition of wastewater Stream Chlorine (M) Bromide (M) Chloride (M) CT, PO4 (M) CT, NH3 (M) µ (M) pH Wastewater 7.05E-05 1.25E-06 4.65E-03 0.01 5.88E-04 0.01 7.50 Blend 4.23E-05 3.95E-04 2.70E-01 0.01 2.94E-04 0.36 7.88 Seawater 1.41E-05 7.88E-04 5.36E-01 0.01 1.18E-07 0.70 8.10 27 xxxv Chemical kinetics of bromamine formation was modeled as a function of blending scenario using Kintecus software [111]. Reaction constants were obtained from the literature and the model was validated using results from previously validated experiments (i.e., Luh and Mariñas [112]). To evaluate the byproduct formation associated with the treatment of the three blending scenarios (e.g., nitrosamines), the model was used to identify which blending scenario was associated with the lowest formation of dibromamine and bromochloramine. Since bromamines cannot be directly measured in real blended waters, the model was used to quantify bromamine formation and speciation. 3.2.2 Bench-scale Experiments To track NDMA formation potential, NDMA formation from chlorination as a function of blending scenario was measured in synthetic blended waters with the addition of either 50 nM DMBzA or DMA (model DBP precursors) in the wastewater solution. The formation of NDMA was also measured during the chlorination of real blended waters collected from the influent of a local water recycling facility (Orange County Water District, Fountain Valley, CA) and desalination facility (Carlsbad Desalination Facility, Carlsbad, CA). Bench-scale experiments were conducted headspace-free in triplicate in 23.5 mL amber borosilicate glass vials and were capped with PTFE-faced septa. Synthetic wastewater and seawater (Table 6) were made in 23.5 mL of 10 mM phosphate buffer at pH 7. Real wastewater and seawater (SI Table A) were buffered at pH 7 with 1 M phosphate buffer at 1% volume. The wastewater and seawater solutions were disinfected 28 xxxvi so that the final concentration of the blended solutions was 5 mg/L NaOCl. The experiments were conducted in the dark for 60 minutes with the blending event occurring at 30 mins for scenarios A and C. Three different blending ratios were used: (i) 25% wastewater and 75% seawater, (ii) 50% wastewater and 50% seawater, and (iii) 75% wastewater and 25% seawater. Solutions were quenched with 200 mM of ascorbic acid after 60 minutes to halt the reaction. DBP formation was quantified using gas chromatography and tandem mass spectrometry (Agilent 7010, Santa Clara, CA) after liquid-liquid extraction (NDMA). Samples were extracted with 10% volume dichloromethane and were further dried with 1 g of magnesium sulfate. Table 6. Composition of synthetic treated wastewater and seawater for bench-scale experiments Constituent Wastewater (M) Seawater (M) NH4Cl 5.87E-04 5.87E-07 NaBr 1.25E-06 7.88E-04 NaCl 4.65E-03 5.36E-01 3.3 Results and Discussion 3.3.1 Modeling Results 3.3.1.1 Effect of blending scenario on chloramine and bromamine formation during disinfection Disinfectant chemistry was modeled under blending scenarios A (chlorinate wastewater), B (blend then chlorinate), and C (chlorinate seawater) to determine the concentrations of chloramine and bromamine species formed during disinfection. The data in Figure 9 shows the cumulative monochloramine (9a), dichloramine (9b), and bromochloramine (9c) exposure (i.e., concentration ! time) for blending scenarios A, B, and C at a wastewater (WW) to seawater (SW) ratios of 1:1. Experiments were 29 30 xxxvii conducted for 60 minutes with the blending event occurring at t = 0 min for scenario B, and t = 30 minutes for scenarios A and C. Because DMBzA (model NDMA precursor) was only present in the simulated wastewater, Figure 9 shows the cumulative chloramine exposure for the time that DMBzA was exposed to halamines (i.e., t = 0-60 minutes for scenarios A and B, and T = 30-60 minutes for scenario C). Monochloramine and dichloramine exposure was greatest in scenario A where the wastewater was disinfected. The ammonia present in the wastewater at high concentrations was oxidized to monochloramine and subsequently dichloramine in a linear fashion (Figure 10) [88], [113]. It was not until seawater was introduced at t = 30 min that monochloramine began to rapidly oxidize the bromide present in seawater to form bromochloramine [114]. The rate of bromide oxidization by monochloramine was much faster than the speciation shift from monochloramine to dichloramine once seawater was introduced (Figure 10). Scenario B yielded the highest bromochloramine exposure, which is likely because it had the longest contact time with both the high ammonia and bromide concentrations in the wastewater and seawater samples. This was also observed in scenario C, where there was a negligible amount of monochloramine and dichloramine exposure in comparison to scenarios A and B. This is likely because most, if not all, of the chlorine oxidized the high levels of bromide present in seawater, forming free bromine, before it came in contact with the high levels of ammonia present in wastewater to form bromamines. Overall, monochloramine formation was two orders of magnitude larger than the dichloramine formation. This is likely because of the presence of bromide, which favored the conversion to bromochloramine instead of dichloramine. 31 xxxviii Figure 9. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios A, B, and C at WW:SW = 1:1 Figure 10. Modelled (a) monochloramine, (b) dichloramine, and (c) bromochloramine formation of blending scenarios A, B, and C at WW:SW = 1:1 Cumulative exposure to monobromamine and dibromamine, on the other hand, were negligible for scenario A in comparison to scenarios B and C (Figure 11). This is likely because the chlorine oxidized the ammonia present in the wastewater (kapp = 1.3 x 10 4 M -1 s -1 [115]) before it came in contact with the high levels of bromide present in the seawater. Moreover, bromide oxidation by monochloramine and dichloramine favored conversion to bromochloramine (Figure 10c) as opposed to monobromamine or dibromamine (Figure 11). Scenario B yielded monobromamine primarily at the onset of blending and disinfection (Figure 12) and a negligible amount of dibromamine. And as expected, scenario C yielded the largest amounts of both monobromamine and dibromamine in comparison to scenarios A and B. The majority of this occurred after wastewater was 31 xxxix introduced at t = 30 min (Figure 12). The high levels of ammonia in wastewater were likely oxidized by the HOBr that was formed during seawater disinfection (i.e., t = 0 – 30 mins), leading to the formation of monobromamine and subsequently dibromamine. Both monobromamine and dibromamine, which are a more favorable leaving group than their chloramine counterparts, were both likely hydrolyzed leading to the depletion seen in Figure 12. Figure 11. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios A, B, and C at WW:SW = 1:1 Figure 12. Modelled (a) monobromamine and (b) dibromamine formation of blending scenarios A, B, and C at WW:SW = 1:1 3.3.1.2 Effect of blending ratio on chloramine and bromamine formation during disinfection In addition to the WW:SW = 1:1, two additional WW:SW ratios were evaluated that were representative of wastewater-augmented seawater desalination (WW:SW = 1:3) and seawater-augmented wastewater reclamation (WW:SW = 3:1). For scenario A, 32 xl both the monochloramine and dichloramine cumulative exposure decreased with increasing WW:SW ratio (Figure 13). This is likely because the WW:SW ratio of 1:3 had the highest Cl:N mass ratio (0.7) in comparison to 1:1 (0.3) and 3:1 (0.2), which led to more monochloramine formation and subsequent dichloramine formation [88], [113]. Bromochloramine concentrations also decreased with increasing wastewater fractions. This is likely because of the limited contact that scenario A had with high bromide concentrations in the seawater. Figure 13. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios A WW:SW = 1:3, 1:1, and 3:1 In scenario B, cumulative monochloramine exposure increased with increasing wastewater fractions, while cumulative dichloramine exposure decreased with increasing wastewater fraction (Figure 14). However, for scenario B, in which chlorine was exposed to high levels of bromide at t = 0 min, monochloramine exposure decreased as the Cl:N mass ratio increased. However, for scenario B, in which chlorine was exposed to high levels of bromide at t = 0 min, monochloramine exposure decreased as the Cl:N mass ratio increased. Monochloramine formation increased with increasing wastewater fraction, while bromochloramine decreased with decreasing wastewater fraction in the blended stream. Thus, more monochloramine oxidized bromide to form bromochloramine once blended with the seawater. 33 xli Figure 14. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios B WW:SW = 1:3, 1:1, and 3:1 In scenario C, similar trends to scenario A were observed for cumulative exposure of monochloramine, dichloramine, and bromochloramine as exposure decreased with increasing wastewater fraction (Figure 15). These were all negligible, however, in comparison to the cumulative chloramine exposure of scenarios A and B. Figure 15. Modelled cumulative exposure of (a) monochloramine, (b) dichloramine, and bromochloramine of blending scenarios C WW:SW = 1:3, 1:1, and 3:1 In scenario A, cumulative bromamine exposure decreased with increased wastewater fraction (Figure 16). This is likely because the bromide concentration of the blended stream decreased as the wastewater fraction increased. Less HOBr was available to oxidize the monochloramine formed pre-blending (t = 0-30 mins). 34 xlii Figure 16. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios A at WW:SW = 1:3, 1:1, and 3:1 Monobromamine formation increased with increasing wastewater volume in scenario C (Figure 18), similarly to how monochloramine formation increased with increasing wastewater volume in scenario B (Figure 17). Dibromamine, decreased with increasing wastewater volume in scenario C, similarly to how dichloramine formation decreased with increasing wastewater volume for scenario B. Figure 17. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios B at WW:SW = 1:3, 1:1, and 3:1 Figure 18. Modelled cumulative exposure of (a) monobromamine and (b) dibromamine for blending scenarios C at WW:SW = 1:3, 1:1, and 3:1 35 xliii 3.3.2 Experimental Results 3.3.2.1 Effect of blending scenario on NDMA formation during disinfection To understand the role of chloramine and bromamine species formed during chlorination of blended waters on NDMA formation potential, bench-scale experiments were conducted in buffered deionized water, and with actual seawater and secondary wastewater effluent. Water quality data from the wastewater and seawater samples are shown in Table 7. NDMA formation was two orders of magnitude greater in the clean system (Figure 19a) than the realistic system (Figure 19b), however, both systems followed similar trends between blending scenarios (Figure 19). Scenario A yielded the most NDMA formation, while scenario C yielded the least. Based on the model and experimental results (Figures 9, 11, and 19), it seems that the monochloramine and dichloramine were principally responsible for NDMA formation as scenario A yielded the most monochloramine, dichloramine, and NDMA formation. This was unexpected as bromochloramine, which was largest in scenario B, was expected to produce more NDMA than monochloramine as bromochloramine’s higher electronegativity would favor nucleophilic substitution with DMBzA [108]. Scenario C, which yielded the largest amount of monobromamine and dibromamine in the model, yielded the smallest amount of NDMA in both the clean and realistic systems. This agrees with the literature that reported suppressed NDMA formation in the presence of bromide at neutral pH [108], [109] and low monochloramine concentrations [116]. This is likely attributed to the pKa of DMBzA (8.93) [117], which is lower than the pKa of protonated ammonia (9.25). Less ammonia will be available to be oxidize by HOBr formed in scenario C (Figure 20) to form bromamines that will lead NDMA formation. Moreover, in the realistic system, the 36 xliv HOBr formed in scenario C is likely oxidizing the natural organic matter (NOM) in the seawater, forming other byproducts that compete with NDMA formation. Scenario C, which yielded negligible amounts of monochloramine, yielded the smallest amount of NDMA in comparison to scenarios A and B. Thus, the dominant bromamine species present in scenario C (monobromamine and dibromamine) were not responsible for NDMA formation. Figure 19. NDMA formation for blending scenarios A, B, and C at a WW:SW = 1:1 in a (a) clean and (b) realistic systems Figure 20. Modelled (a) cumulative exposure of HOBr and (b) HOBr formation over time for scenarios A, B, and C at WW:SW = 1:1 37 xlv Table 7. Composition of secondary treated wastewater and seawater used in realistic bench-scale experiments Constituent Wastewater Seawater Unit pH 7.13 7.93 - NO2 - 0.87 <0.05 mg/L NO3 - 7.68 <1.00 mg/L NH3 0.81 <0.50 mg/L TOC 11.1 1.25 mg/L 3.3.2.2 Effect of blending ratio on NDMA formation during disinfection In accordance with the model experiments, three blending ratios were tested experimentally at the bench-scale (WW:SW = 1:3, 1:1, and 3:1). For scenario A, in both the clean and realistic systems, NDMA formation increased with increasing wastewater fraction (i.e., increasing NDMA yield with increasing ammonia concentration and decreasing bromide concentration) (Figure 21). Scenario B followed a similar trend where NDMA formation increased with increasing wastewater fraction (Figure 22). This is likely because of the monochloramine species that were formed. Figure 21. NDMA formation for blending scenarios A at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system 38 xlvi Figure 22. NDMA formation for blending scenarios B at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system In scenario C, differing trends between the clean and realistic systems were observed where NDMA formation decreased with increasing wastewater fraction in the clean system and increased with increasing wastewater fraction in the realistic system (Figure 23). This could possibly be due to the presence and complexity of other constituents in the real samples such as natural organic matter (NOM). Additionally, the distribution of NDMA yield between blending ratios was more significant in the clean system than the realistic system for all three blending scenarios. This is likely because of the NDMA precursor chosen, DMBzA, which has shown a high conversion to NDMA of 40% [118], [119]. Thus, clean experiments were also conducted with DMA as the model precursor and found that there was less of a dramatic distribution between blending ratios and drop in NDMA concentration for scenario C (Figure 24). 39 xlvii Figure 23. NDMA formation for blending scenarios C at a WW:SW = 1:3, 1:1, and 3:1 in a (a) clean and (b) realistic system Figure 24. NDMA formation for blending scenarios A, B, C at a WW:SW = 1:1 in a clean system with (a) DMBzA and (b) DMA. Based on these results, to minimize NDMA formation of a blended wastewater- seawater stream, disinfection scenario C in which seawater is disinfected prior to blending with wastewater is ideal. This is representative of wastewater augmented seawater desalination, which reduces the energetic and environmental impacts of desalination. To increase wastewater reclamation by replacing seawater desalination as 40 xlviii part of a city’s water supply portfolio, scenario B, where disinfection occurs at the onset of blending, is ideal for minimizing NDMA formation. 3.4 Conclusions As climate change and freshwater scarcity worsen, demand grows to secure new reliable sources of potable water, including desalinated seawater and recycled wastewater. During a recent techno-economic assessment, Wei et al. identified significant overlaps between desalination and potable reuse treatment trains [76]. Significant energetic benefits, without the capital cost, footprint, and other concerns associated with building a new plant, could be obtained by blending seawater and wastewater upstream of potable reuse treatment plants. In this work, for the first time, the chemical ramifications of pre-treatment blending of seawater and treated wastewater for potable reuse applications was explored. The value of diluting either stream with the other is two-fold: (1) diluting seawater with wastewater to reduce specific energy demand and environmental impacts associated with the reverse osmosis process, and (2) increasing potable water supply by permanently including reclaimed water as a permanent part of the water supply portfolio and augmenting it with seawater when needed. One critical consideration of the blended stream is the speciation of the oxidants formed during chlorine addition to prevent biofouling. Wastewater chlorination produces chloramines by ammonia oxidation, and seawater chlorination produces free bromine by bromide oxidation. Blending the two produces bromamines, which are associated with the formation of certain toxic byproducts such as NDMA. Bromamine formation is expected to be a 41 xlix function of which stream is chlorinated, and whether the disinfection happens before or after blending. Three blending/disinfection scenarios were evaluated: wastewater disinfection (scenario A), blended disinfection (scenario B), and seawater disinfection (scenario C). The chloramine and bromamine species formed during disinfection were first modeled using chemical kinetics software and found that wastewater disinfection produced the most monochloramine and dichloramine, blended disinfection produced the most bromochloramine, and seawater disinfection produced the most monobromamine and dibromamine. Clean and realistic bench-scale experiments were then conducted to quantify the NDMA formation associated with these blending/disinfection scenarios and to draw links between NDMA formation and the chloramine and bromamine species reported in the model. Wastewater disinfection produced the most NDMA formation in both the clean and realistic systems, and seawater disinfection produced the least. Based on the modeling results, blended disinfection was expected to produce the most NDMA as it produced the most bromochloramine, which has a higher electronegativity than monochloramine and would favor nucleophilic substitution with DMBzA. However, wastewater disinfection produced the most NDMA, which was associated with the most monochloramine formation based on the model. This is likely because of the longer contact times that were observed for scenarios B and C with the high concentrations of bromide found in seawater. Competitive reactions with bromide likely lead to the formation of other byproducts instead of NDMA. 42 l Chapter 4 4. The role of zero-liquid discharge on water reuse practices in the power sector: power sector review and natural gas case study analysis 4.1 Introduction Natural gas surpassed coal as the predominant electricity generation source in 2016 [120]. This transition resulted in decreased water withdrawals by thermoelectric power plants because natural gas facilities require less water on average than coal-fired power facilities [121], [122]. As of 2020, natural gas generation accounted for 40.3% of total US electricity generation [123]. Although natural gas generation is expected to decrease in the coming years due to increasing natural gas prices and significant growth in electricity generation from renewable energy sources [124], natural gas is still projected to occupy the majority of electricity production through 2050 [125]. Although water withdrawals have decreased in recent years, thermoelectric power plants account for over 40% of total fresh water use in the US [126], [127]; in 2015, thermoelectric power plants accounted for 133 Bgal/day of total water withdrawals. This is the lowest water withdrawal value reported since before 1970 [127]. The reduction in water withdrawals, primarily attributed to the shift in electricity generation mix, was also observed by the US Energy Information Administration (EIA), which reported a decline in water withdrawal intensity (water withdrawal per unit of net electricity generated) from 15.1 gal/kWh in 2014 to 13.0 gal/kWh in 2017 [122]. In general, facilities in the western US, where recirculating cooling systems are more prevalent, withdraw less water 43 li (Figure 25) [122], [128]. Recirculating cooling systems require less water withdrawals to cool heat exchangers because the cooling water is reused; once-through cooling systems require greater water withdrawals because the cooling water is used in a single pass and discharged back to the source [129], [130]. Once-through cooling systems consume little water whereas recirculating cooling systems consume water through evaporation [130]. Based on 2018 EIA data, Figure 26 shows the annual water withdrawn and consumed by cooling system type for the various types of cooling systems used by recirculating and once-through cooling systems at power facilities in the US (dry, hybrid, and mixed cooling systems are also shown). Once-through cooling systems account for 73% of total water withdrawals and recirculating systems account for 94% of total water consumption by the power sector. Figure 25. 2018 (a) water withdrawal intensity and (b) water consumption intensity from thermoelectric power plants (Gallons/MWh) 44 lii Figure 26. Annual (a) water withdrawn and (b) water consumed by the power sector in 2018 [128] Water withdrawals are substantially influenced by water availability, water cost, and discharge regulations [131], [132]. Water availability, which may be limited by physical scarcity or restrictive water rights, is a major issue for power facilities throughout the US. Although an existing facility may have secured a legal right to water through riparian or prior appropriation rights [133], there is still risk of not meeting demand based on physical availability of the water [134]. This is more of a concern for facilities in arid and semi-arid regions that struggle with water scarcity [134]–[136], however, it also impacts other regions that are subject to changes in water rights as water demands from other sectors increase [137]. Competition for water is dictated by the type of the water right (i.e., riparian or appropriative rights and in some cases federal reserved, adjudicated, and Pueblo rights) and its effective date [138]–[140]. Thus, facilities may be subject to water supply reductions or reallocations by regulators based on the needs of the region [137], [141]. For example, the Colorado River Compact, which allocates specific quantities of the Colorado River for agricultural irrigation, municipal uses, industrial uses, recreational uses, fish and wildlife, and power production is based on average Colorado River flows from 1905 to 1922 [137], [142]. These historic average Water Withdrawal and Consumption Mixed (Other) Dry Hybrid Once Through Recirculating 1% 0% 0% 73% 26% 1% 0% 0% 5% 94% a) b) 45 liii flows are greater than current average flows, which are only expected to decrease as the effects of climate change worsen. The Colorado River Compact does not directly address shortage sharing, which has made it difficult for stakeholders to agree on a long-term solution for meeting water demand [137]. New facilities are typically located where water supply is high, competition for water is low, and water prices are low [143]. For existing facilities, increasing water cost can incentivize a power facility to reduce its water withdrawals. The average cost of water in the US is $0.30/100 gallons with a range of $0.05/100 gallons in Florida to $0.79/100 gallons in Alaska [144]. Moreover, if water costs increase too much in a region, dry cooling systems can be employed, which use about 95% less water than wet cooling systems [145]. Currently, there are at least 70 dry-cooling systems throughout the US, with 40% in California, Nevada, New York, and Virginia. Dry and hybrid cooling facilities consume an average of two gal/MW-hr of water according to a 2012 study by Macknick et al [146] but account for only 3% of U.S. thermoelectric generating capacity according to an EIA 2018 report [145]. Although dry cooling systems offer a valuable solution for water scarce regions, they have high capital costs and are inefficient when ambient air temperatures increase [145], [147], [148]. In 2015, saline surface water and groundwater (i.e., water with 1,000 mg/L or more of dissolved solids) accounted for 28.4% and reclaimed municipal wastewater accounted for 0.15% of total water use by thermoelectric power facilities [127]. Finally, discharge regulations have also driven reductions in water withdrawals; as facilities increase on-site water reuse practices to limit discharge of wastewater, reduced water withdrawals are required. In September 2015, the Environmental 46 liv Protection Agency (EPA) issued its most stringent discharge regulation for the power sector since 1982 [132]. A new rule amending 40 CFR Part 423 under the Clean Water Act set new effluent guidelines for the Steam Electric Power Generating category that called for zero discharge of pollutants from the fly ash and bottom ash waste streams [149], [150]. In August 2020, the new rule was updated to include flue gas desulfurization wastewater and bottom ash transport water to limit levels of toxic metals in wastewater that could be discharged from power plants [151], [152]. The EPA estimates that 75 of the 914 facilities may incur compliance costs under the final 2020 rule [33]. These discharge regulations have forced the power industry to take leadership in zero liquid discharge (ZLD) implementation. Facilities affected by the regulations, the majority of which are in the western U.S. [153], have begun implementing ZLD approaches that have the added benefit of increasing water reuse practices. Historically, water reuse in power facilities has focused on increasing cycles of concentration (COCs); however, new opportunities for water reuse are being considered to address increasingly stringent discharge regulations and limited water availability. 4.1.1 Water reuse practices in the power sector On-site water reuse is inherent in power facilities in the concentration cycling that occurs for cooling towers and steam boilers [154]. As the number of concentration cycles increases, freshwater consumption and wastewater discharge decrease [155]. Water reuse in power facilities can also occur in a “cascading” manner, where lower quality water from upstream processes can be used (with or without treatment) in downstream processes. For example, high-purity boiler water blowdown can be used as 47 lv cooling tower makeup water [156]; cooling tower blowdown can be recycled for flue-gas desulfurization (FGD) makeup [157], [158]; FGD blowdown can be recycled for ash sluicing; and ash pond run-off can be recycled for dust control [159]. Other on-site water reuse practices include the reuse of decant water from sludge ponds or clarifiers for spray dryers; cooling tower blowdown for brine concentrators; and ash system water for FGD makeup. For example, Coronado Generation Station (St Johns, AZ), Salt River Project Gila River Power Station (Gila Bend, AZ), and SRP Springerville Generating Station (Springerville, AZ) use cooling tower blowdown as makeup water for ash system surge tanks, brine concentrators, and FGD scrubber reagents, respectively. Water reuse can also occur in an “ascending” manner, where lower quality water from downstream processes is treated and reused in upstream processes. For example, SRP Gila River Power Station (Gila Bend, AZ) and APS Redhawk Power Plant (Arlington, AZ) use crystallizer distillate for cooling tower and steam boiler water makeup. Aside from on-site reuse applications, water from power facilities can also be reused for peripheral services (e.g., road dust suppression). 4.1.2 Zero-liquid discharge approaches ZLD is a wastewater management strategy where no liquid wastewater is discharged, water recovery is maximized, and wastewater discharge is eliminated. At power facilities, ZLD typically refers to the elimination of liquid waste leaving the property, not the elimination of liquid waste leaving the facility; thus, ZLD at power facilities is often achieved using high-recovery treatment processes followed by discharging waste streams to an evaporation pond or injecting waste streams in a deep well [160]. Use of evaporation ponds and deep wells help minimize the size and cost of 48 lvi the high-recovery treatment processes; however, the water in the waste stream is lost either to the atmosphere or the subsurface and is not available for reuse. Evaporation ponds are a simple, low-technology process to dispose of high- salinity waste streams. Solar energy evaporates the water and salts accumulate at the bottom of the pond; the accumulated salts are typically disposed of in landfills. Evaporation ponds are primarily designed based on the flow rate of water wasted to the pond and also by the regional evaporation rate, with high flow rates and/or low evaporation rates requiring larger evaporation pond area. Although evaporation ponds can have high capital costs (e.g., land area, liner materials) [161], the low electrical energy costs, low operating costs, and simplicity of operation give them treatment primacy over separation processes that recover the water, instead of evaporating it [162]. Evaporation ponds are more common? in arid and semi-arid climates where evaporation rates are high [163]. In other parts of the US that experience heavier rainfall, deep-well injection is a low-technology process that is often used to dispose of high salinity waste streams. In deep-well injection, the high-salinity stream is injected into subsurface porous rock formations. Deep-well injection is desirable because it is not as limited by concentrate flow rate [164], [165], which means that facilities can limit their electrical supply costs by operating at lower recoveries. Challenges associated with deep-well injection include scaling, corrosion, and possible pollution of the groundwater [165]–[167]. The most significant drawback of deep-well injection is that the water resource is lost to the deep subsurface. Power facilities have generally greater value in using evaporation ponds or deep- well injection to dispose of high-salinity streams than in implementing processes (e.g., 49 lvii brine concentrators and/or crystallizers) that recover the water resource for subsequent reuse. However, with recent advancements in high-recovery RO (HRRO) processes (e.g., closed-circuit RO (CCRO) and flow-reversal RO) water can be recovered with relatively low energy consumption [168]. For example, the average power required by an HRRO system for a typical 300-gpm system is an order of magnitude less than that of a brine concentrator or crystallizer (160 kW for HRRO versus 1,200 kW for a brine concentrator or 3,600 for a crystallizer) [169]. HRRO processes offer recoveries of 90% or higher [170] and may provide high enough recoveries that brine concentrators and crystallizers are not necessary. The first implementation of HRRO technology in the power generation industry occurred in 2017 when Southern California Edison implemented CCRO systems at five of its gas-fired combustion turbine facilities [171]. At that time, it was expected that implementation of CCRO would save 44 million gallons of water per year, improve reliability, and cut annual water operating costs by more than $1 million in savings per plant (i.e., 85% of annual water operating costs). In particular, implementation of CCRO reduces water disposal costs because the CCRO brine can be disposed of at a lower-cost brine disposal facility [171]. In 2019, Global Water Intelligence stated that emerging RO systems, such as CCRO, are “encroaching on the part of the treatment train for brine concentration that an evaporator would traditionally operate in.” Although other alternatives to brine concentrators have entered the market over the last ten to fifteen years (e.g., humidification-dehumidification, forward osmosis, and membrane distillation), the emerging RO solutions have the advantage that they are based on mature, familiar RO desalination technology [172]. Additionally, avoiding 50 lviii brine concentrators and crystallizers is desirable because these processes are known to be operationally complex and have low reliability due to frequent breakdowns [162]. As of 2016, there were 72 power facilities in the US that employed ZLD systems with a total combined capacity of 119,000 m 3 /day [173]. Currently, Arlington Valley Power Station, a natural gas combined-cycle facility, employs the largest ZLD system in the US [173]. This facility utilizes an HRRO system to treat 19,600 m 3 of water per day [174]. According to a report by Mordor Intelligence LLP [57?], ZLD systems are expected to register a compound annual growth rate of over 9% from 2020-2025 [175]. The power industry is expected to occupy the majority of the ZLD market as wastewater disposal costs and demand for freshwater resources are expected to rise [175]. While ZLD systems are often unique systems that are supplied by different manufacturers and depend on water quality and discharge volumes/flowrates, they consist of three general steps (Figure 27): (i) pretreatment, (ii) pre-concentration, and (iii) crystallization/evaporation [176]–[179]. In pretreatment, suspended solids, metals, hardness, and silica are typically filtered and/or precipitated out. Pre-concentration typically involves a high-pressure membrane processes (typically reverse osmosis (RO)) and/or brine concentrator to concentrate the stream even further, usually recovering 60-80% of the water [177], [179]. Finally, crystallization/evaporation generates a solid through an evaporative or thermal process and the evaporated water can often be reused [177], [179]. High purity solid salts resulting from the brine crystallizer and/or evaporation pond can be reused or sold as industrial materials, however, mixed solid salts cannot be reused and are often disposed as hazardous waste [136], [180]. 51 lix Figure 27. Zero liquid discharge (ZLD) treatment scheme with (i) pretreatment, (ii) pre-concentration, and (iii) crystallization/evaporation ZLD operation is not just the coupling of five-to-six unit operations; it is a holistic philosophy that affects how an entire power facility operates. Power facilities do not choose to implement ZLD – they do so only if required by regulation. Although the key challenge with ZLD systems it thought to be high cost and intensive energy [181], experiences of power facility operators and managers suggests that operational complexity and frequent breakdowns can be more a limiting factor than energy cost. Additionally, the processes in ZLD systems (in particular, the brin concentrators and crystallizers) are known to be challenging to operate and to suffer frequent break downs which can create an unwanted domino effect on other processes of the system. Third, the brine concentrator and crystallizer are often dependent on the steam from a facility’s boiler system and are thus subject to facility pauses and shutdowns [160]. ZLD Treatment Train Wastewater Pretreatment Reverse Osmosis Brine Concentrator Brine Crystallizer Solids Recovery Solids Disposal Evaporation Pond Treated Water for Reuse (i) (ii) (iii) 52 lx A positive corollary of ZLD implementation is increased on-site water reuse. In addition to reuse of the distillate from the brine crystallizers and concentrators, facilities that convert to ZLD operation also implement more upstream and downstream recycling to extract water from waste streams throughout the plant that would otherwise be discharged [182], [183]. As on-site reuse practices increase and water withdrawals decrease as the result of ZLD implementation, pipe-parity metrics can be used to compare and quantify the value of ZLD technology implementation. Pipe parity is defined by the US Department of Energy (DOE) as, “technology solutions that are cost competitive with existing water sources and end-use applications” [184]. This means that a water source (e.g., on-site reuse water) has achieved pipe parity once it is considered equally desirable when compared to its marginal water source (e.g., municipal water or saline water). In addition to cost, pipe parity can be applied to other metrics (e.g., levelized cost of water and energy intensity) that drive technology investments. For example, decisionmakers may pay more for a water source that is more abundantly available throughout the year than a water that is cheaper but fluctuates in volume over time. In this work, pipe parity metrics will be used to quantify the value of on-site water reuse practices driven by ZLD implementation within the natural gas power industry. 4.1.3 Natural gas combined-cycle power facilities There are three main types of natural gas-fired technologies: steam turbines (17% of total natural gas generator capacity), combustion turbines (28%), and combined-cycle (53%) The map in Figure 28 shows this data geographically [120], with natural gas combined-cycle (NGCC) facilities comprising the majority. 53 lxi Figure 28. Distribution of natural-gas-fired power facilities in the continental U.S. in 2016 [120] NGCC facilities differ from those operating with steam or combustion turbines because NGCC facilities capture wasted heat to increase the facility’s electrical output (Figure 29). In the NGCC process, compressed air and fuel are mixed and ignited in a combustion chamber. The hot air/fuel mixture spins the gas turbine blades and drives the generator that converts mechanical work into electricity. The heat recovery steam generator captures some of the exhaust heat from the gas turbine that would otherwise escape through the exhaust stack. The heat recovery steam generator then generates a high-pressure steam from the captured exhaust heat and delivers it to the steam turbine which operates similarly to the gas turbine to produce additional electricity via the 54 lxii generator. The steam-liquid mixture that leaves the steam turbine is converted to a liquid water through the condenser, which is then fed to the cooling tower. Figure 29. Natural-gas combined-cycle power facility 4.1.4 Objectives The objective of this review and analysis is to provide a baseline assessment of on-site water reuse at NGCC power facilities and to consider additional water reuse opportunities associated with implementation of ZLD operations. In addition to water reuse examples from NGCC facilities across the US, detailed operating information from two representative case-study facilities are used to: (i) determine the changes in on-site water reuse of a non-ZLD facility that is retrofitted to include ZLD operations, (ii) establish evaporation pond area as a critical metric for quantifying ZLD costs, (iii) use pipe-parity metrics to quantify the energetic and monetary cost of retrofitting a non-ZLD facility to include ZLD operations, and (iv) compare the energetic and monetary costs of NGCC Compressor Combustion Chamber Turbine Generator HRSG Stack Steam Turbine Generator Condenser Cooling Tower Pump 55 lxiii the two ZLD systems with different concentration technologies (i.e., brine concentrators vs closed-circuit RO). Future technology research needs to enable more energy- efficient desalination and water reuse are also considered. 4.2 Methods 4.2.1 Selection of case-study facilities Two NGCC power facilities with on-site water reuse practices common to the US power sector were selected as the case-study facilities. Both facilities are located in semi-arid/arid regions [185], [186] and face challenges with water availability, water cost, and discharge regulations. Cherokee Generating Station is an 886-MW natural-gas-fired power facility with 576 MW produced from an NGCC unit commissioned in 2015 and 310 MW produced from a natural gas steam turbine that was converted from a coal-fired unit in 2017. The NGCC facility has two combustion engines and two heat recovery steam generators supplying one steam turbine generator. In 2003, Cherokee Generating Station began using secondary treated wastewater from Metro Wastewater Reclamation (Denver, CO) (2059 acre-feet) for cooling tower makeup in addition to their allocation of water from Clear Creek (596 acre-feet) and the Platte River (52 acre-feet). Cherokee Generating Station discharges 0.9 MGD of blowdown water to the Platte River and has historically used clarification prior to discharge. Due to a 2017 permit, the facility has the following discharge limits for the Platte River: total inorganic nitrogen (TIN) at 10 mg/L daily maximum by 2022; chloride at 250 mg/L over a 30-day average by 2023; and sulfate at 553 mg/L over a 30-day average by 2023. For this reason, Cherokee Generating 56 lxiv Station will begin operating as a ZLD facility in 2021. Transition of the facility to ZLD operations is requiring numerous changes to water treatment operations; in addition to adding HRRO, other upstream and/or downstream processes have been modified/retrofitted. For example, an additional filter press has been added prior to landfill disposal. These additions and modification will result in increased on-site water reuse at the facility, which is beneficial because increase reuse results in less makeup water required from Clear Creek, Platte River, and Metro Wastewater Reclamation. Currently, the only reuse practice at Cherokee Generating Station is from the RO blowdown which is recycled for cooling tower makeup. Gila River Power Station is a 2,200-MW NGCC power facility located in Gila Bend, AZ. The facility, which has been in operation since 2003, is owned by the Salt River Project and Tucson Electric Power; the facility has been operated by the Salt River Project since mid-2018. Gila River Power Station has four power blocks, each comprised of two combustion turbine generators, two triple-pressure heat-recovery steam generators, and one steam turbine generator. The facility uses groundwater from an on-site well for cooling tower makeup, service to the boiler and ROs, and fire protection. The Gila River Power Station was designed and is operated as a ZLD facility with three thermal evaporator brine concentrators and three evaporation ponds (with 16 acres each). 4.2.2 Water Techno-economic Assessment Pipe-Parity Platform (Water-TAP3) model Pipe parity metrics for each case study were evaluated using a tool called the Water Techno-economic Assessment Pipe-Parity Platform (Water-TAP 3 ) [187]. The goal of 57 lxv Water-TAP 3 is to identify high-impact early-stage research and development opportunities that promote pipe parity for nontraditional source waters. Water-TAP 3 is an excel-based, open-source decision support tool designed to facilitate consistent technoeconomic assessments of water treatment trains, regardless of source water or end use. The Excel model framework is capable of modeling up to five parallel water treatment trains with complete flexibility for in-plant by-passes and recycles. The tool can model the production of up to five target water uses (one per treatment train) from as many as 50 sources (ten per train). The model is also capable of assessing waste treatment and valorization opportunities in addition to co-located power and heat production options. 4.2.2.1 Evaporation pond area Evaporation pond area is a critical consideration in quantifying ZLD costs at NGCC power facilities. First, evaporation pond area was considered as a constant value, representing the industry-standard practice of maximizing use of existing evaporation pond area before turning to additional ZLD processes. From this, the capacity of required additional ZLD process was calculated. This “maximized” evaporation pond and “resulting” additional ZLD process capacity was considered the baseline. Then, a sensitivity analysis was conducted to determine the effect of increasing evaporation pond area on the levelized cost of water (LCOW) and the percent of the LCOW that is attributable to electricity. With more evaporation pond area available, can waste more water to an evaporation pond by operating at lower recovery rates. This would presumably decrease electricity costs but may increase other costs 58 lxvi such as disposal costs. Thus, the LCOW was chosen as a pipe-parity metric to quantify the value of increasing evaporation pond area. 4.2.2.2 Auxiliary power consumption, levelized cost of water, and energy intensity The Water-TAP 3 model was used to determine optimal evaporation pond area, auxiliary power consumption, LCOW, and energy intensity for the case studies. For all ZLD systems, the power required by the processes must be considered and will result in a reduction of the total house power (i.e., the power rating of the facility). Optimizing the auxiliary power consumption of a ZLD system is principally desired to improve heat rates and increase a facility’s net generating capacity [188]. For example, a brine concentrator’s specific energy consumption ranges from ~20-30 kWh/m 3 at salinities of ~250,000 mg/L while a brine crystallizer’s specific energy consumption ranges from ~50-65 kWh/m 3 at salinities of ~300,000 mg/L (values estimated from Tong and Elimelech [181]). This results in an efficiency penalty for the facility as a result of ZLD implementation. Facilities may choose an alternative to brine concentrators and crystallizers (e.g., CCRO) to minimize the reduction of house power. The levelized cost of water (LCOW) and energy intensity were also compared. LCOW is the sum of costs to treat the water divided by the total amount of water treated ($ per m 3 of water treated) and is calculated using: LCOW= )*+,-.+/ *12. 3 "($%") ' [($%") ' ]*$ 4 5 +667+/ 8&: *12.2 5 ;& *12.2 +<=>+?= +667+/ @-=/A -6 +*>=BC==. (1) 59 lxvii where the capital costs are amortized over the system’s life using a discount rate of r and useful life in years of n. O&M costs are annual operation and maintenance costs, and R&R costs are annual repair and replacement costs. Energy intensity, measured in kWh per m 3 of water treated, is similar to house power in that it accounts for the total work of a system but is not time-dependent and is normalized to m 3 of water treated. Energy intensity can be used to compare ZLD systems across facilities as it is not region- or facility-dependent [189], while LCOW is used to compare ZLD systems within a facility as cost of electricity varies by state (0.074 $/kWh in Colorado and 0.0628 $/kWh in Arizona [190]). 4.3 Results and Discussion 4.3.1 Existing Cooling Water Systems 4.3.1.1 Cherokee Generating Station The three main water uses at Cherokee Generating Station are: 1) make-up water for the main cooling tower; 2) process water for the steam-cycle (i.e., boiler/heat- recovery steam generator); and 3) make-up water for the service cooling tower. Many of the small systems covered by service water have small diameter cooling tubing, so the water must be low in particulate material. The wastewater from all of these systems is sent to treatment train A. Cherokee Generating Station (Figure 30) currently receives the majority of its main cooling water supply from Denver Metro’s municipal reuse water (2,059 acre-feet in 2019) and smaller amounts from Clear Creek and the Platte River (596 and 52 acre-feet in 2019, respectively). The reuse water is stored in a nearby reservoir and piped to the two cooling towers at the facility (Units 4 and Units 5-7). While in the evaporative cooling towers, the water is concentrated 5 to 10 times where 60 lxviii ~75% of the water is lost to evaporation for heat rejection and the rest (~0.9 MGD) is blown down to Treatment Train A. Along with this cooling tower blowdown, Treatment Train A treats the wastewater from Treatment Train B. Treatment Train B is used to treat the steam-cycle water that is supplied by Denver’s potable water supply and accounted for 75 of the 150 million gallons used at the facility in 2018 (the remaining 75 million gallons was used by the service cooling water). The steam-cycle water is first treated with anti-scalant to avoid precipitation in RO, then is filtered to remove particulates and is passed through activated carbon beds to remove chlorine prior to RO. The water is then piped to the RO unit where the RO permeate is sent through mixed bed ion exchangers and subsequently condensate storage tanks. The supply from the condensate storage tanks is then piped to the steam-cycle makeup as needed. The dissolved solids are concentrated in the drum boiler/heat-recovery steam generators and the boiler blowdown is used to control the solids concentration sent to the discharge clarifier in Treatment Train A. The remaining boiler blowdown that is not sent to the discharge clarifier is combined with the main cooling water treatment process for discharge. The service water is used to cool turbine lubricating oil, sample coolers, and a few other small systems. The water from Denver’s potable supply is not treated before it is directly used in the service water cooling tower. It is concentrated approximately twice before the blowdown is reused and combined with the main cooling water. There are plans to increase the cycles of concentration to four times with the addition of azole to inhibit corrosion. The facility’s combined wastewater (i.e., cooling tower blowdown, steam-cycle wastewater, and service cooling water blowdown) is treated in Treatment 61 lxix Train A that uses a clarifier followed by a settling pond before it is discharged to the river. Prior to entering the clarifier, the wastewater is dosed with coagulant, metal precipitator, and flocculant to remove metals and particulates. The proposed flow water flow diagram reconfigured Cherokee Generating Station after ZLD implementation is shown in Figure 30. A closed-circuit RO (CCRO) system that operates at 98% recovery was selected as the desalination process; by operating at such high recoveries, the flowrate of concentrate that is sent to the evaporation ponds is significantly reduced. The ZLD design (Figure 31) differs from the current design by the addition of a CCRO system, media filtration before the CCRO system, and 7 acres of evaporation pond. The on-site water reuse is expected increase in the pending ZLD scenario as a primary goal of the facility is to reduce city water use and limit the amount of water sent to the evaporation ponds. Thus, the facility will aim to achieve as many cycles as possible in the cooling towers. In the proposed ZLD system, all permeate will be reused and all RO concentrate will be sent to the evaporation ponds. The auxiliary tower blowdown and RO reject from Unit 4 will be reused for cooling tower makeup. The Boiler RO concentrate will also go the cooling tower if it is of higher quality than the cooler tower makeup, and the wastewater concentrate will be sent to the evaporation ponds. After wastewater treatment, all permeate will initially go to the new RO units in Treatment Train B before it is sent to the auxiliary tower. Permeate reuse will be prioritized in the following order: (i) boiler/heat-recovery steam generator makeup, (ii) cooling tower makeup, (iii) heat-recovery steam generator evaporative coolers, and (iv) main cooling tower makeup. 62 lxx Figure 30. Current Cherokee Generating Station cooling water flow diagram Figure 31. Proposed Cherokee Generating Station cooling water flow diagram with ZLD system Cherokee Pre-ZLD Raw Water Reservoir Platte River and Clear Creek (Summer) City Water Settling Pond Clarifier Overflow Treatment Train A Treatment Train B River (sampling required) Emergency Spill Pond West Stormwater Pond U567 Stormwater NW Area Stormwater Boiler ROs, U1-4 Service Water Denver Reuse Gray Water (fall/winter/ sping) U4 Cooling Tower U7 Cooling Tower U1-4 Stormwater Combined Plant Wastewater Cherokee Post-ZLD Combined Plant Wastewater Treatment Train A Treatment Train B River (sampling required) Raw Water Reservoir Sodium Bisulfite Injection Platte River and Clear Creek (Summer) Overflow Denver Reuse Gray Water (fall/winter/ sping) U4 Cooling Tower U7 Cooling Tower City Water Boiler ROs, U1-4 Service Water, CT Evaporative Coolers Emergency Spill Pond Emergency Spill Pond Evaporation Ponds Brine Landfill Filter Press Clarifier Solids Permeate Treatment (clarification filtration, TOC removal, CCRO) West Stormwater Pond Treated Water Storage Tank U567 Stormwater NW Area Stormwater U1-4 Stormwater 63 lxxi 4.3.1.2 Gila River Power Station Gila River Power Station receives its cooling water from a from a local aquifer and is stored onsite in a raw water tank [191]. To meet discharge regulations, the facility is equipped with a ZLD system that is composed of three thermal brine concentrators to process and recycle water as makeup for the facility. The wastewater from the brine concentrators is discharged to three 16-acre evaporation ponds that cannot be pumped for any purpose. In the event that the ZLD system needs to go offline, wastewater is discharged directly to the evaporation ponds. In 2013, the facility implemented a series of water conservation efforts to limit the amount of water sent to the evaporation ponds. Cooling tower blowdown was reduced by 33% as the result of a new water-chemistry program to increase cycles of concentration in the cooling tower, the service life of the brine concentrators was extended by 25% as the result of new anti-scalants and anti- foaming agents, and pond evaporation was enhanced as the result of a modified Title V air permit that allowed for the installation of spray evaporators [192]. In the facility’s water cycle (Figure 32), the cooling tower blowdown is sent to the RO reject tank that feeds the thermal brine concentrators and new treatment system. Before the water is processed through the brine concentrators, the pH is adjusted and anti-scalants and anti-foaming agents are added. The brine concentrator distillate is then sent through a mix-bed demineralizer that polishes the water to demineralizer quality. That demineralized quality water is then processed through the Block. In the Block, the water goes from the condenser through the heat-recovery steam generator to the steam turbines to generate electricity. The steam is then condensed, and the cycle is repeated. The water in this cycle needs to be of very high purity due to the high 64 lxxii pressure and temperature conditions. Thus, if the thermal brine concentrators are down, there is no way of feeding the system, and the facility would be reliant on whatever water is available in the demineralizer makeup and storage tanks. As a result, the facility uses the new treatment system to account for such emergencies. In 2021, the new treatment system was commissioned to reduce the volume of water sent to the evaporation ponds. Due to the unreliable operation of the brine concentrators, evaporation ponds were being filled to near-capacity. The new treatment system was made to primarily treat cooling tower blowdown and includes ultrafiltration and RO that can be run in a single or double pass. However, since the new system was commissioned, significant strides have been made to maintain the brine concentrators in service. As a result, the volume of water sent to the evaporation ponds was reduced and the new treatment system was then used for other emergency water needs. For example, the system could be used to process raw water that could help fulfill the high purity demineralizer water needs of the Block and it could also be used to recirculate the demineralizer feed tank if it needed remediation from contamination. 65 lxxiii Figure 32. Gila River Power Station process flow diagram 4.3.2 Water-TAP 3 Model Scenarios The Water-TAP 3 model was used to evaluate monetary and electrical costs, electrical intensity, auxiliary power consumption, and water recovery of several water treatment scenarios at Cherokee Generating Station and Gila River Power Station. A baseline and two “what-if” scenarios were modeled for each facility to evaluate current, future, and alternative water systems and practices. For Cherokee Generating Station, the facility’s baseline operation (Figure 30) was compared with the pending ZLD scenario using CCRO (ZLD-CCRO) (Figure 31) and a comparison scenario that utilizes a conventional brine concentrator technology (ZLD-BC). For Gila River Power Station (Figure 32), the facility’s baseline ZLD operation which utilized a BC system, and a new Gila ZLD2 Raw Water Tank New Treatment System: RO Cooling Tower RO Reject Tank Containment Condenser Demin Feed Tank Waste Water Surge Tank W-1 thru W-7 (10,316 gpm) RO Reject Tank Plant Drains Oily Water Separator BC Feed Tank Concentrated Waste Tank Evaporation Ponds Demin Feed Tank Demin Bottles HRSG’s ST GT’s BC-A BC-B BC-C ZLD Sump Seal/ Service Water System 66 lxxiv RO system (ZLD-BC+RO) was compared with the alternative scenarios of operation only with BC systems (ZLD-2BC) or operation only with RO systems (ZLD-2RO). 4.3.2.1 Summary of Water-TAP 3 Model Results The data in Tables 8 and 9 show monetary and electrical costs, electrical intensity, auxiliary power consumption, and water recovery at Cherokee Generating Station and Gila River Power Station, respectively. As can be seen by comparing the baseline non-ZLD scenario with the ZLD-CCRO scenario for Cherokee Generating Station in Table 8, the most significant increases in costs were the electricity supply costs, electricity intensity, and auxiliary power consumption that were 9.5X, 9.7X, 10X greater, respectively. This is primarily due to the addition of a CCRO system that has additional electricity supply costs and operation and maintenance (O&M) costs. This is further detailed in Section 4.3.2.2. There was less of a sizeable increase observed for the LCOW of the ZLD-CCRO scenario, which only increased by 2.5Xs from the baseline. This is due to the fact that the LCOW incorporates capital, O&M costs, and repair and replacement costs, which was substantial for the media filtration and holding tanks in the non-ZLD baseline scenario. This is also detailed in Section 4.3.2.2. Table 8. Summary of Water-TAP 3 model results for Cherokee Generating Station Facility Cherokee Generating Station Scenario Baseline What-If Variation from Baseline What-If Variation from ZLD-CCRO scenario Facility-Type Non-ZLD ZLD ZLD ZLD Technology N/A CCRO BC Status Decommissioning Pending N/A Total Capital Investment ($MM) 3.03 5.76 1.9X 17.03 3.0X Total Operating Cost ($MM/yr) 0.08 0.33 4.1X 1.99 6.0X Electricity Supply Costs ($MM/yr) 0.02 0.19 9.5X 1.75 9.2X 67 lxxv Electricity Intensity (kWh/m 3 ) 0.2 1.94 9.7X 18.39 9.4X Auxiliary Power Consumption (MW) 0.03 0.3 10.0X 2.7 9.0X Levelized Cost of Water ($/m 3 ) 0.25 0.63 2.5X 2.73 4.3X Water Recovery (%) 19.95 19.1 -- 18.25 0.96X Table 9. Summary of Water-TAP 3 model results for Gila River Power Station Facility Gila River Power Station Scenario Baseline What-If Variation from Baseline What-If Variation from Baseline Facility-Type ZLD ZLD ZLD ZLD Technology BC+RO 2BC 2RO Status In Commission N/A N/A Total Capital Investment ($MM) 33.57 36.89 1.10X 14.16 0.42X Total Operating Cost ($MM/yr) 3.45 3.75 1.09X 0.54 0.16X Electricity Supply Costs ($MM/yr) 2.99 3.25 1.09X 0.14 0.05X Electricity Intensity (kWh/m 3 ) 21.3 22.55 1.06X 1.3 0.06X Auxiliary Power Consumption (MW) 5.43 5.92 1.09X 0.26 0.05X Levelized Cost of Water ($/m 3 ) 2.88 3.05 1.06X 1.03 0.36X Water Recovery (%) 10.87 11.2 1.03X 8.5 0.78X The ZLD-BC scenario showed a similar increase with respect to the ZLD-CCRO scenario in electricity supply costs, electricity intensity, and auxiliary power consumption at 9.2X, 9,4X, and 9.0Xs, respectively. This is due to the replacement of the CCRO system with a BC system that had additional electricity supply costs and O&M costs. The ZLD-BC scenario resulted in a higher increase in the LCOW with respect to the ZLD-CCRO scenario (4.3Xs) as compared to the increase from the baseline to the ZLD- CCRO scenario (2.5Xs). This is because the BC had higher capital and repair and replacement costs in comparison to CCRO. Although the ZLD-CCRO scenario significantly increased overall costs in comparison to the non-ZLD scenario, employing CCRO instead of a BC as the primary ZLD technology resulted in significant savings. This is also reflected in the modeling results for Gila River Power Station (Table 9). 68 lxxvi Finally, water recovery, which is defined as the percent of useful water recovered from the system, did not vary as significantly between Cherokee’s baseline and ZLD scenarios. This is because WaterTAP3 considered the water that was discharged to the Platte River in the baseline scenario as useful, but the water that was discharged to the evaporation ponds in the ZLD scenarios as “not useful”. The water discharged to the evaporation ponds is considered wasted similar to the water that is wasted to evaporation in the cooling tower. Thus, since the useful water discharged to the Platte River in the baseline scenario cannot be directly recycled within the facility, water recovery can only be used to fairly compare the percent of water recovered and reused between the ZLD scenarios. Water recovery in the ZLD-BC scenario was 0.96X of the water recovery in the ZLD-CCRO scenario, which can be considered insignificant between systems. This suggests that the similar water savings can be achieved between the ZLD-CCRO and ZLD-BC scenario, but at a small fraction of ZLD-BC scenario costs. As can be seen in Table 9, all metrics slightly increased from Gila River Power Station’s baseline scenario to the ZLD-2BC scenario. All increases were within a smaller range (1.06-1.10Xs) because the BC system that replaced the RO system at Gila River required lower recoveries (59%) as compared to the original BC system (90% recovery) that dominates costs, electricity intensity, and auxiliary power consumption (detailed in Section 3.2.2). In the ZLD-2RO scenario, all costs, electricity intensity, and auxiliary power consumption decreased with respect to the baseline. The most significant decreases were the electricity supply cost (0.05X), electricity intensity (0.06X), and auxiliary power consumption (0.05X). This is due to the replacement of the 69 lxxvii energy-intensive BC system with an additional RO system. This is detailed in the subsequent section breaks down the electricity intensity, and monetary and electrical costs of treatment systems employed at Cherokee and Gila River by unit to provide further detail on the differences between the baselines and alternative ZLD scenarios. Gila River Power Station (Table 9). Gila River observed an increase in water recovery as more energy intensive ZLD technologies were employed (i.e., 2RO à RO+BC à 2BC). Gila River, however, had 6.9Xs greater evaporation pond area (48 acres) available as compared to Cherokee, so higher recoveries were not as necessary. Additionally, the capital and O&M costs of the larger evaporation pond area impacted the LCOW and likely required lower recoveries for Gila River as compared to Cherokee. 4.3.2.2 Water-TAP3 results broken down by unit To understand the role of specific units (e.g., RO and discharge) in the treatment train on costs and other metrics, Figure 33 depicts TCI, O&M costs, electricity intensity, and LCOW for each unit in Cherokee’s three scenarios. In the baseline (non-ZLD) scenario, RO comprises the largest portion of (74% for both) the electricity supply cost and electricity intensity; this is followed by discharge, which comprises 22% of both. The high discharge costs are associated with the energy required to pump the discharge water to the Platte River. RO comprises only small portions (15 and 10%) of the LCOW and TCI, likely because RO is only used to treat makeup water for the service cooling tower, which is a relatively low flowrate. On the other hand, media filtration, which is used to treat the higher flowrate of Denver Municipal City Water, comprises greater portions (30 and 34%) of the LCOW and TCI. Holding Tank A, which is used to store 70 lxxviii cooling tower blowdown before it is sent to the clarifier, comprises 23 and 26% of the LCOW and TCI. This is likely due to the capital and O&M costs associated with the holding tank that do not contribute to electricity intensity and electricity supply cost. In the ZLD-CCRO scenario, energy and costs associated with CCRO comprise the majority of (89%) the electricity intensity. In comparison to the baseline scenario, values for electricity intensity and electricity supply cost are an order of magnitude higher for the ZLD-CCRO (note the scale of the y axis). RO costs for the ZLD-CCRO scenario did not change (RO is still used for the service cooling tower makeup water in this scenario) but now comprises very little of these graphs due to the change in scale of the y axis. Discharge also comprises a much smaller portion (1%) of the electricity intensity. Clearly, the high energy of the CCRO process significantly outweighs the cost of RO and discharge that were substantial in the baseline scenario. In Cherokee’s ZLD-CCRO scenario, discharge water that was previously sent to the Platte River is now sent to the on-site evaporation pond. Decreased discharge flowrate upon implementation of ZLD has nominal effects on other metrics; the discharge portion decreases from 4 to 2% of TCI, from 10 to 2% of O&M costs, and from 5 to 2% of LCOW. In the ZLD-CCRO scenario, there were two media filtration units, Media Filtration A processed clarifier effluent before it was sent to the CCRO system and Media Filtration B processed influent from Denver Municipal City Water. Compared to the baseline, the same media filtration unit that was used to process Denver Municipal City Water (Media Filtration B) decreased the LCOW and TCI by 66 and 67%, respectively, and increased O&M costs by 78%. However, combining both media filtration units in the CCRO process (Media Filtration A and B), the LCOW, TCI, 71 lxxix and O&M costs increased by 15, 10, and 497% as compared to the single media filtration unit used in the baseline. As can be seen by comparing the scale of the y-axis for the ZLD-CCRO scenario with the ZLD-BC scenario, a BC increases the overall costs of the system by an order of magnitude. The BC comprises the largest portion of TCI, O&M costs, electricity intensity, LCOW, TCI, and O&M costs at 77, 96, 99, and 87%. The remaining 1% of electricity intensity is attributed to the RO for the service cooling tower makeup water. For the ZLD-BC scenario, there is less distribution of costs as compared to the baseline or ZLD-CCRO scenarios; in other words, the BC occupies the majority of all costs and results in increased costs across the board. 72 lxxx Figure 33. Water-TAP 3 results broken down by unit for Cherokee Generating Station Figure 34 depicts the TCI, O&M costs, electricity intensity, and LCOW for Gila River’s baseline ZLD-BC+RO scenario, alternative ZLD-2BC scenario, and alternative ZLD-2RO scenario. In the ZLD-BC+RO baseline, the single BC system occupied the vast majority of TCI (85%), O&M costs (97%), electricity intensity (99%), and LCOW (91%). A combination of RO, UF, and the industrial unit used to process the BC and RO 73 lxxxi brine accounted for the remaining 1% of the electricity intensity. The evaporation ponds accounted for the next largest portion of the TCI, O&M costs, and LCOW at 9, 1, and 5% respectively. This is due to the capital and O&M costs required for land and liners. RO (1%), UF (~1%), the holding tank used to store cooling water blowdown (1-3%), and the industrial unit used to process the BC brine (0.5%) accounted for the remaining TCI, O&M costs, and LCOW. The replacement of the RO system with a BC system in the ZLD-2BC scenario did not increase overall costs as severely as it did for the Cherokee what-if scenario that replaced the CCRO system with a BC system. The overall costs of the what-if scenario increased TCI, O&M costs, electricity intensity, and LCOW by 10, 9, 6, and 6%, respectively. Similar to the baseline scenario with a single BC system, the what-if scenario with two BC systems was predominantly dominated by BC costs. The BC system that replaced the RO system displaced 8-9% of the TCI, O&M costs, electricity intensity, and LCOW from the original BC system. Additionally, it displaced the TCI, O&M costs, and LCOW for the evaporation ponds by ~27% each. Costs were significantly decreased by replacing the baseline BC system with an additional RO system in the ZLD-2RO scenario. Compared to the baseline, TCI, O&M costs, electricity intensity, and LCOW decreased by 58, 84, 94, and 64% in the what-if scenario with two RO systems, respectively. The RO system that replaced the BC system occupied the majority of O&M costs and electricity intensity at 70 and 86%, respectively. The original RO system, which only occupied <1% of costs in all categories in the baseline scenario, occupied a larger portion of costs (~4-8%) in the what-if scenario with two RO systems. Additionally, evaporation pond costs occupied a 74 lxxxii significantly larger portion of TCI (52%), O&M costs (18%), and LCOW (42%). Thus, when less energy intensive ZLD technologies like RO or CCRO are employed, the cost of evaporation ponds plays a more critical role. Figure 34. Water-TAP 3 results broken down by unit for Gila River Power Station 75 lxxxiii 4.3.2.3 Evaporation pond area The evaporation pond area available at each facility was used to determine the water recovery required from the ZLD technology (CCRO or BC or RO). Then, to consider the effect of having more or less available evaporation pond area, a sensitivity analysis was performed using the Water-TAP 3 model. The data in Figure 35 show the relationship between water recovery and evaporation pond area and the sensitivity of LCOW, % LCOW attributed to electricity, and electricity intensity to evaporation pond area. As can be seen in Figure 35a and e, water recovery increases as evaporation pond area decreases. As more water is recovered, less water is discharged to the evaporation pond. Similarly, as water recovery decreased, more water was wasted to the evaporation pond, requiring larger areas. Thus, final system costs were impacted by energetic costs (i.e., higher energetic costs required for higher recovery rates) and capital costs (i.e., higher capital costs for larger evaporation pond areas). The model assumed an evaporation pond liner thickness of 50 mil (equal to 1/1000 th of an inch), land costs of $5,000/acre, and land clearing costs of $1,000/acre. The LCOW for both Cherokee (35b) and Gila River (35f) increased with increasing evaporation pond area. Although there were CCRO and BC energetic savings as recovery rates decreased and evaporation pond area increased, the system treated less water which resulted in higher LCOWs. The LCOW is normalized per cubic meter of treated water, thus the LCOW increased as evaporation pond area increased (i.e., recovery rates decreased). Based on these results, facilities would be incentivized to limit their evaporation pond area as much as possible, however, based on interviews with facilities, this is known not to be the case. Instead, facilities aim to maximize 76 lxxxiv evaporation pond area to limit auxiliary power consumption caused by increased recovery rates. This perspective is reflected in the % LCOW attributed to electricity (35c and 35g), which did not consider capital costs, or the amount of water treated. In this way, the % LCOW attributed to electricity can be considered synonymous with auxiliary power consumption and would motivate facilities to increase evaporation pond areas as much as possible. Evaporation ponds comprised 7, 10, and 2% of the LCOW, TCI, and O&M costs of the ZLD-CCRO scenario, respectively. These were primarily due to capital and O&M costs of the evaporation ponds as they did not contribute to the electricity intensity (Figure 32). Figure 35. (a, e) Relationship between water recovery and evaporation pond area, and sensitivity analysis evaluating impact of evaporation pond area on (b, f) LCOW, (c, g) percent LCOW attributed to electricity, and (d, h) electricity intensity of Cherokee Generating Station and Gila River Power Station Similar to the LCOW, electricity intensity is normalized per volume of water treated and is calculated as: 77 lxxxv Electricity intensity= +667+/ *12. 1C =/=*.>-*-.@ =/=*.>-*-.@ ,>-*= 3 <1/7D= 1C E+.=> .>=+.=A (1) Thus, electricity intensity increased with increasing evaporation pond area due to the smaller volume of water treated. This was seen throughout Gila River (35h), however, there was an exception for Cherokee Generating Station (35d) at areas <35 acres and recoveries >15%. At recoveries >15%, the energetic savings of CCRO outweighed the volume of water treated. At recoveries <15%, not enough water was treated to decrease electricity intensity. Thus, this further highlights the importance of ZLD technology on electricity intensity and cost. As facilities aim to limit their monetary and electrical costs while still adhering to discharge regulations, employing less energy intensive ZLD technologies like CCRO can also help improve the land footprint (i.e., less electricity intensity and evaporation pond area needed at recoveries >15%). 4.4 Conclusion In this work, water and water reuse practices in the power industry were reviewed and a case study analysis of ZLD practices at two natural gas facilities was conducted. Power facilities are beginning to implement ZLD practices as the results of increasingly stringent discharge regulations. An addition to complying with regulations, facilities have the benefit of increasing water reuse at a facility, limiting their water footprint and environmental impact. To quantify the costs and electricity intensity associated with reuse based on ZLD implementation, two case studies were analyzed: (i) a non-ZLD facility that was retrofitted to ZLD with CRRO to increase reuse, and (ii) a ZLD facility with BC technology to increase reuse. 78 lxxxvi Similar to the Southern California Edison experience discussed in the Introduction, based on the LCOW and electricity supply costs, reducing water use does not result in substantial cost savings; this is in large part due to the low cost of water. However, in striving to build more sustainable power and water systems, power companies do see corporate-community value in reducing water withdrawals. On the other hand, reduced brine disposal costs can result in substantial cost savings, as in the case of Southern California Edison who replaced mobile demineralization trailers with high operating expenses with CCRO. In the case of Cherokee Generating Station, we don’t see these savings because the facility is not replacing an older technology with CCRO for cost savings, instead they are implementing CCRO in response to more stringent discharge regulations. However, that said, we do see some saving in terms of the LCOW – where we only see a relatively small increase in LCOW (small relative to the electricity intensity and auxiliary power consumption). Moreover, a sensitivity analysis on the effect of evaporation pond area on the LCOW, the percent of LCOW attributable to electricity, and electricity intensity was also conducted. Facilities often seek to maximize available evaporation pond area so that they can operate at lower recovery rates, minimizing auxiliary power consumption. The results from this study show, however, that depending on the ZLD technology used (i.e., CCRO vs BC), operating at higher recoveries can minimize LCOW and electricity intensity, resulting in water savings for the facility. 79 lxxxvii Chapter 5 5. Conclusion Global water stress has prompted recent interest in desalination systems for treatment of saline and/or otherwise impaired water sources for potable use. These desalination systems typically rely on reverse osmosis (RO) as the main treatment process. RO, however, can have high energy demands and produces a high-salinity waste brine that must be disposed of or further treated. Seawater RO (SWRO) facilities may be co-located with wastewater treatment facilities that also discharge a stream to the ocean. Currently, the only synergistic use of these waste streams is to use the treated wastewater to dilute the SWRO brine stream prior to discharge. However, discharge of treated wastewater to the ocean could be considered a waste of the water resource within this stream and has no impact on the high energy demands of SWRO. The main goal of this work was to provide solutions to increase potable water supply, reduce undesirable impacts associated with alternative water treatment, and improve water reuse practices at power facilities by evaluating the monetary, environmental, and public health costs associated with: (i) a membrane-based salinity gradient power process, (ii) a new concept for blending reclaimed wastewater and seawater upstream of disinfection, and (iii) zero- liquid discharge (ZLD) technologies employed at natural gas power facilities. For the first time, this work determined: (i) a power density phenomenon that limits osmotic power production for SWRO, (ii) an optimal wastewater-seawater blending and disinfection scenario to limit NDMA formation for potable reuse applications, and (iii) the levelized cost 80 lxxxviii of water for implementing ZLD systems at power facilities to adhere to discharge regulations and increase on-site water reuse. 81 lxxxix REFERENCES [1] “Water Scarcity,” 2015. [2] G. Tan, “Carlsbad desalination plant will be sold to investment firm for $1 billion,” Los Angeles Times, Los Angeles, CA, 2019. [3] B. Boxall, “L.A.’s ambitious goal: Recycle all of the city’s sewage into rinkable water,” Los Angeles Times, Los Angeles, CA, 2019. [4] N. Ghaffour, T. M. Missimer, and G. L. Amy, “Technical review and evaluation of the economics of water desalination: Current and future challenges for better water supply sustainability,” Desalination, vol. 309, pp. 197–207, 2013. [5] L. F. Greenlee, D. F. Lawler, B. D. Freeman, B. Marrot, and P. Moulin, “Reverse osmosis desalination: Water sources, technology, and today’s challenges,” Water Res., vol. 43, no. 9, pp. 2317–2348, 2009. [6] J. A. López-Ramírez, M. D. C. Oviedo, and J. M. Q. Alonso, “Comparative studies of reverse osmosis membranes for wastewater reclamation,” Desalination, vol. 191, no. 1, pp. 137–147, 2006. [7] J. E. Miller, “Review of Water Resources and Desalination Technologies,” Albuquerque, NM, 2009. [8] M. Elimelech and W. A. Phillip, “The Future of Seawater Desalination: Energy, Technology, and the Environment,” Science (80-. )., vol. 333, no. 6043, pp. 712 LP – 717, Aug. 2011. [9] D. R. Paul, “Reformulation of the solution-diffusion theory of reverse osmosis,” J. Memb. Sci., vol. 241, no. 2, pp. 371–386, 2004. [10] B. Van Der Bruggen, C. Vandecasteele, T. Van Gestel, W. Doyen, and R. Leysen, “A review of pressure-driven membrane processes in wastewater treatment and drinking water production,” Environ. Prog., vol. 22, no. 1, pp. 46–56, Apr. 2003. [11] J. G. Wijmans and R. W. Baker, “The solution-diffusion model: a review,” J. Memb. Sci., vol. 107, no. 1, pp. 1–21, 1995. [12] “Reverse Osmosis (RO) Membranes,” Sterlitech Corporation, 2020. [Online]. Available: https://www.sterlitech.com/reverse-osmosis-ro-membrane.html. [13] M. Ahmed, W. H. Shayya, D. Hoey, and J. Al-Handaly, “Brine disposal from reverse osmosis desalination plants in Oman and the United Arab Emirates,” Desalination, vol. 133, no. 2, pp. 135–147, 2001. [14] G. A. Tularam and M. Ilahee, “Environmental concerns of desalinating seawater using reverse osmosis,” J. Environ. Monit., vol. 9, no. 805–813, 2007. [15] D. Gerrity, B. Pecson, R. S. Trussell, and R. R. Trussell, “Potable reuse treatment trains throughout the world,” J. Water Supply Res. Technol., vol. 62, no. 6, pp. 321–338, Sep. 2013. [16] G. Tchobanoglous et al., “Framework for Direct Potable Reuse,” Alexandria, VA, 2015. [17] “Potable Reuse Compendium,” 2017. [18] S. L. Plata and A. E. Childress, “Limiting power density in pressure-retarded osmosis: Observation and implications,” Desalination, vol. 467, pp. 51–56, 2019. [19] J. E. Miller, “Review of Water Resources and Desalination Technologies,” 2003. [20] A. Achilli, T. Y. Cath, and A. E. Childress, “Power generation with pressure retarded osmosis: An experimental and theoretical investigation,” J. Memb. Sci., vol. 343, no. 1, pp. 42–52, 2009. 82 xc [21] J. Kim, M. Park, S. A. Snyder, and J. H. Kim, “Reverse osmosis (RO) and pressure retarded osmosis (PRO) hybrid processes: Model-based scenario study,” Desalination, vol. 322, pp. 121–130, 2013. [22] A. Achilli, J. L. Prante, N. T. Hancock, E. B. Maxwell, and A. E. Childress, “Experimental Results from RO-PRO: A Next Generation System for Low-Energy Desalination,” Environ. Sci. Technol., vol. 48, no. 11, pp. 6437–6443, Jun. 2014. [23] J. L. Prante, J. A. Ruskowitz, A. E. Childress, and A. Achilli, “RO-PRO desalination: An integrated low-energy approach to seawater desalination,” Appl. Energy, vol. 120, pp. 104–114, 2014. [24] A. Altaee, A. Sharif, G. Zaragoza, and A. F. Ismail, “Evaluation of FO-RO and PRO-RO designs for power generation and seawater desalination using impaired water feeds,” Desalination, vol. 368, pp. 27–35, 2015. [25] S. Loeb, “Production of energy from concentrated brines by pressure-retarded osmosis: I. Preliminary technical and economic correlations,” J. Memb. Sci., vol. 1, pp. 49–63, 1976. [26] S. Loeb, F. Van Hessen, and D. Shahaf, “Production of energy from concentrated brines by pressure-retarded osmosis: II. Experimental results and projected energy costs,” J. Memb. Sci., vol. 1, pp. 249–269, 1976. [27] K. L. Lee, R. W. Baker, and H. K. Lonsdale, “Membranes for power generation by pressure-retarded osmosis,” J. Memb. Sci., vol. 8, pp. 141–171, 1981. [28] T. Thorsen and T. Holt, “The potential for power production from salinity gradients by pressure retarded osmosis,” J. Memb. Sci., vol. 335, no. 1, pp. 103–110, 2009. [29] C. Klaysom, T. Y. Cath, T. Depuydt, and I. F. J. Vankelecom, “Forward and pressure retarded osmosis: potential solutions for global challenges in energy and water supply,” Chem. Soc. Rev., vol. 42, pp. 6959–6989, 2013. [30] K. L. Hickenbottom, J. Vanneste, M. Elimelech, and T. Y. Cath, “Assessing the current state of commercially available membranes and spacers for energy production with pressure retarded osmosis,” Desalination, vol. 389, pp. 108–118, 2016. [31] G. Han, S. Zhang, X. Li, and T.-S. Chung, “High performance thin film composite pressure retarded osmosis (PRO) membranes for renewable salinity-gradient energy generation,” J. Memb. Sci., vol. 440, pp. 108–121, 2013. [32] S. Lin, A. P. Straub, and M. Elimelech, “Thermodynamic limits of extractable energy by pressure retarded osmosis,” Energy Environ. Sci., vol. 7, pp. 2706– 2714, 2014. [33] G. Han, S. Zhang, X. Li, and T.-S. Chung, “Progress in pressure retarded osmosis (PRO) membranes for osmotic power generation,” Prog. Polym. Sci., vol. 51, pp. 1–27, 2015. [34] T. Y. Cath, A. E. Childress, and M. Elimelech, “Forward osmosis: Principles, applications, and recent developments,” J. Memb. Sci., vol. 281, no. 1, pp. 70–87, 2006. [35] N. Y. Yip, A. Tiraferri, W. A. Phillip, J. D. Schiffman, and M. Elimelech, “High Performance Thin-Film Composite Forward Osmosis Membrane,” Environ. Sci. Technol., vol. 44, no. 10, pp. 3812–3818, May 2010. [36] K. P. Lee, T. C. Arnot, and D. Mattia, “A review of reverse osmosis membrane materials for desalination—Development to date and future potential,” J. Memb. 83 xci Sci., vol. 370, no. 1, pp. 1–22, 2011. [37] B. S. Lalia, V. Kochkodan, R. Hashaikeh, and N. Hilal, “A review on membrane fabrication: Structure, properties and performance relationship,” Desalination, vol. 326, pp. 77–95, 2013. [38] S. Lin, N. Y. Yip, T. Y. Cath, C. O. Osuji, and M. Elimelech, “Hybrid Pressure Retarded Osmosis–Membrane Distillation System for Power Generation from Low-Grade Heat: Thermodynamic Analysis and Energy Efficiency,” Environ. Sci. Technol., vol. 48, no. 9, pp. 5306–5313, May 2014. [39] N. Y. Yip et al., “Thin-Film Composite Pressure Retarded Osmosis Membranes for Sustainable Power Generation from Salinity Gradients,” Environ. Sci. Technol., vol. 45, no. 10, pp. 4360–4369, May 2011. [40] Q. She, Y. K. W. Wong, S. Zhao, and C. Y. Tang, “Organic fouling in pressure retarded osmosis: Experiments, mechanisms and implications,” J. Memb. Sci., vol. 428, pp. 181–189, 2013. [41] N. Y. Yip and M. Elimelech, “Influence of Natural Organic Matter Fouling and Osmotic Backwash on Pressure Retarded Osmosis Energy Production from Natural Salinity Gradients,” Environ. Sci. Technol., vol. 47, no. 21, pp. 12607– 12616, Nov. 2013. [42] M. Zhang, D. Hou, Q. She, and C. Y. Tang, “Gypsum scaling in pressure retarded osmosis: Experiments, mechanisms and implications,” Water Res., vol. 48, pp. 387–395, 2014. [43] T.-T. Nguyen, S. Kook, C. Lee, R. W. Field, and I. S. Kim, “Critical flux-based membrane fouling control of forward osmosis: Behavior, sustainability, and reversibility,” J. Memb. Sci., vol. 570–571, pp. 380–393, 2019. [44] Q. She, L. Zhang, R. Wang, W. B. Krantz, and A. G. Fane, “Pressure-retarded osmosis with wastewater concentrate feed: Fouling process considerations,” J. Memb. Sci., vol. 542, pp. 233–244, 2017. [45] C. Y. Tang, Y.-N. Kwon, and J. O. Leckie, “The role of foulant–foulant electrostatic interaction on limiting flux for RO and NF membranes during humic acid fouling— Theoretical basis, experimental evidence, and AFM interaction force measurement,” J. Memb. Sci., vol. 326, no. 2, pp. 526–532, 2009. [46] C. Y. Tang and J. O. Leckie, “Membrane Independent Limiting Flux for RO and NF Membranes Fouled by Humic Acid,” Environ. Sci. Technol., vol. 41, no. 13, pp. 4767–4773, Jul. 2007. [47] C. P. Morrow and A. E. Childress, “Evidence, Determination, and Implications of Membrane-Independent Limiting Flux in Forward Osmosis Systems,” Environ. Sci. Technol., vol. 53, no. 8, pp. 4380–4388, Apr. 2019. [48] D. I. Kim, J. Kim, H. K. Shon, and S. Hong, “Pressure retarded osmosis (PRO) for integrating seawater desalination and wastewater reclamation: Energy consumption and fouling,” J. Memb. Sci., vol. 483, pp. 34–41, 2015. [49] N. T. Hancock and T. Y. Cath, “Solute Coupled Diffusion in Osmotically Driven Membrane Processes,” Environ. Sci. Technol., vol. 43, no. 17, pp. 6769–6775, Sep. 2009. [50] A. Achilli, T. Y. Cath, E. A. Marchand, and A. E. Childress, “The forward osmosis membrane bioreactor: A low fouling alternative to MBR processes,” Desalination, vol. 239, no. 1, pp. 10–21, 2009. 84 xcii [51] G. T. Gray, J. R. McCutcheon, and M. Elimelech, “Internal concentration polarization in forward osmosis: role of membrane orientation,” Desalination, vol. 197, no. 1, pp. 1–8, 2006. [52] C. Boo, M. Elimelech, and S. Hong, “Fouling control in a forward osmosis process integrating seawater desalination and wastewater reclamation,” J. Memb. Sci., vol. 444, pp. 148–156, 2013. [53] R. W. Holloway, A. E. Childress, K. E. Dennett, and T. Y. Cath, “Forward osmosis for concentration of anaerobic digester centrate,” Water Res., vol. 41, no. 17, pp. 4005–4014, 2007. [54] M. M. Motsa, B. B. Mamba, J. M. Thwala, and A. R. D. Verliefde, “Osmotic backwash of fouled FO membranes: Cleaning mechanisms and membrane surface properties after cleaning,” Desalination, vol. 402, pp. 62–71, 2017. [55] A. Sagiv, N. Avraham, C. G. Dosoretz, and R. Semiat, “Osmotic backwash mechanism of reverse osmosis membranes,” J. Memb. Sci., vol. 322, no. 1, pp. 225–233, 2008. [56] A. Sagiv and R. Semiat, “Parameters affecting backwash variables of RO membranes,” Desalination, vol. 261, no. 3, pp. 347–353, 2010. [57] Q. She, R. Wang, A. G. Fane, and C. Y. Tang, “Membrane fouling in osmotically driven membrane processes: A review,” J. Memb. Sci., vol. 499, pp. 201–233, 2016. [58] J. Sim, J. Koo, S. Nam, E. Kim, and T.-M. Hwang, “Comparative evaluation of osmotically-driven cleaning methods for organic-inorganic fouling in pressure retarded osmosis (PRO),” Eng. Sci. Technol. an Int. J., vol. 21, no. 5, pp. 1018– 1023, 2018. [59] S. E. Skilhagen, “Osmotic power — a new, renewable energy source,” Desalin. Water Treat., vol. 15, no. 1–3, pp. 271–278, Mar. 2010. [60] P. H. Gleick and M. Palaniappan, “Peak water limits to freshwater withdrawal and use,” Proc. Natl. Acad. Sci., vol. 107, no. 25, pp. 11155 LP – 11162, Jun. 2010. [61] J. Eke, A. Yusuf, A. Giwa, and A. Sodiq, “The global status of desalination: An assessment of current desalination technologies, plants and capacity,” Desalination, vol. 495, p. 114633, 2020. [62] A. Covarrubias, “Santa Barbara working to reactivate mothballed desalination plant,” Los Angeles Times, Los Angeles, CA, 2015. [63] “Charles E. Meyer Desalination Plant,” City of Santa Barbara, 2020. [Online]. Available: https://www.santabarbaraca.gov/gov/depts/pw/resources/system/sources/desalin ation/default.asp. [64] H. Cooley and R. Phurisamban, “The Cost of Alternative Water Supply and Efficiency Options in California,” Oakland, CA, 2016. [65] “Green New Deal Plan Targets,” Los Angeles, CA, 2019. [66] J. Rice, A. Wutich, and P. Westerhoff, “Assessment of De Facto Wastewater Reuse across the U.S.: Trends between 1980 and 2008,” Environ. Sci. Technol., vol. 47, no. 19, pp. 11099–11105, Oct. 2013. [67] (National Research Council), Water Reuse: Potential for Expanding the Nation’s Water Supply Through Reuse of Municipal Wastewater. Washington, DC: National Academies Press, 2012. 85 xciii [68] D. Smith, D. Sedlak, R. Dower, E. Archuleta, and J. Mosher, “Mainstreaming Potable Water Reuse in the United States: Strategies for Leveling the Playing Field,” 2018. [69] A. Coate and D. Pedersen, “California WateReuse Action Plan,” 2019. [70] J. Arroyo, S. Shirazi, E. McDonald, K. Hood, and A. Plummer, “History of Water Reuse in Texas,” 2011. [71] “Potable Reuse,” 2020. [72] “Thinking Strategically about the Business of Water Reuse,” 2020. [73] “Groundwater Replenishment System,” Orange County, CA, 2020. [74] “Sustainable Groundwater: Seawater Intrusion,” U.S. Geological Survey. [Online]. Available: https://ca.water.usgs.gov/sustainable-groundwater- management/seawater-intrusion-california.html. [75] M. I. Michaels, A. S. Krayer, and T. P. Solomon, “Salinity Impacts on Coastal Wastewater Treatment Facilities,” Worcester Polytechnic Institute, 2017. [76] X. Wei, Z. M. Binger, A. Achilli, K. T. Sanders, and A. E. Childress, “A modeling framework to evaluate blending of seawater and treated wastewater streams for synergistic desalination and potable reuse,” Water Res., vol. 170, p. 115282, 2020. [77] R. Valladares Linares, Z. Li, S. Sarp, S. S. Bucs, G. Amy, and J. S. Vrouwenvelder, “Forward osmosis niches in seawater desalination and wastewater reuse,” Water Res., vol. 66, pp. 122–139, 2014. [78] A. Matin, Z. Khan, S. M. J. Zaidi, and M. C. Boyce, “Biofouling in reverse osmosis membranes for seawater desalination: Phenomena and prevention,” Desalination, vol. 281, pp. 1–16, 2011. [79] T. Younos and K. E. Tulou, “Overview of desalination techniques,” J. Contemp. Water Res. Ed., no. 132, pp. 3–10, 2005. [80] Y. Hong, S. Liu, H. Song, and T. Karanfil, “HAA formation during chloramination— significance of monochloramine’s direct reaction with DOM,” J. AWWA, vol. 99, no. 8, pp. 57–69, Aug. 2007. [81] C. J. Seidel, M. J. McGuire, R. S. Summers, and S. Via, “Have utilities switched to chloramines?,” J. AWWA, vol. 97, no. 10, pp. 87–97, Oct. 2005. [82] R. García-Pacheco et al., “Free chlorine exposure dose (ppm·h) and its impact on RO membranes ageing and recycling potential,” Desalination, vol. 457, pp. 133– 143, 2019. [83] J. M. Gohil and A. K. Suresh, “Chlorine attack on reverse osmosis membranes: Mechanisms and mitigation strategies,” J. Memb. Sci., vol. 541, pp. 108–126, 2017. [84] H. Weinberg, S. Krasner, S. Richardson, and A. Thruston, “The Occurrence of Disinfection By-Products (DBPs) of Health Concern in Drinking Water: Results of a Nationwide DBP Occurrence Study,” U.S. Environ. Prot. Agency Natl. Expo. Res. Lab. Athens GA, Jan. 2002. [85] S. D. Richardson et al., “Identification of new drinking water disinfection by- products from ozone, chlorine dioxide, chloramine, and chlorine,” Water. Air. Soil Pollut., vol. 123, no. 1–4, pp. 95–102, 2000. [86] S. W. Krasner et al., “Occurrence of a new generation of disinfection byproducts,” Environ. Sci. Technol., vol. 40, no. 23, pp. 7175–7185, 2006. 86 xciv [87] W. A. Mitch, J. O. Sharp, R. R. Trussell, R. L. Valentine, L. Alvarez-Cohen, and D. L. Sedlak, “N-Nitrosodimethylamine (NDMA) as a Drinking Water Contaminant: A Review,” Environ. Eng. Sci., vol. 20, no. 5, pp. 389–404, Sep. 2003. [88] I. M. Schreiber and W. A. Mitch, “Influence of the Order of Reagent Addition on NDMA Formation during Chloramination,” Environ. Sci. Technol., vol. 39, no. 10, pp. 3811–3818, May 2005. [89] J. Choi and R. L. Valentine, “A kinetic model of N-nitrosodimethylamine (NDMA) formation during water chlorination/chloramination,” Water Sci. Technol., vol. 46, no. 3, pp. 65–71, Aug. 2002. [90] Z. Chen and R. L. Valentine, “Modeling the Formation of N-Nitrosodimethylamine (NDMA) from the Reaction of Natural Organic Matter (NOM) with Monochloramine,” Environ. Sci. Technol., vol. 40, no. 23, pp. 7290–7297, Dec. 2006. [91] S. D. Richardson et al., “Occurrence and Mammalian Cell Toxicity of Iodinated Disinfection Byproducts in Drinking Water,” Environ. Sci. Technol., vol. 42, no. 22, pp. 8330–8338, Nov. 2008. [92] Y. Yang et al., “Toxic Impact of Bromide and Iodide on Drinking Water Disinfected with Chlorine or Chloramines,” Environ. Sci. Technol., vol. 48, no. 20, pp. 12362– 12369, Oct. 2014. [93] S. D. Richardson et al., “Identification of New Drinking Water Disinfection Byproducts Formed in the Presence of Bromide,” Environ. Sci. Technol., vol. 33, no. 19, pp. 3378–3383, Oct. 1999. [94] E. C. Ged and T. H. Boyer, “Effect of seawater intrusion on formation of bromine- containing trihalomethanes and haloacetic acids during chlorination,” Desalination, vol. 345, pp. 85–93, 2014. [95] N. Prihasto, Q.-F. Liu, and S.-H. Kim, “Pre-treatment strategies for seawater desalination by reverse osmosis system,” Desalination, vol. 249, no. 1, pp. 308– 316, 2009. [96] T. Asano and A. D. Levine, “Wastewater reclamation, recycling and reuse: past, present, and future,” Water Sci. Technol., vol. 33, no. 10, pp. 1–14, 1996. [97] G. P. Winward, L. M. Avery, T. Stephenson, and B. Jefferson, “Chlorine disinfection of grey water for reuse: Effect of organics and particles,” Water Res., vol. 42, no. 1, pp. 483–491, 2008. [98] B. A. Lyon, A. D. Dotson, K. G. Linden, and H. S. Weinberg, “The effect of inorganic precursors on disinfection byproduct formation during UV- chlorine/chloramine drinking water treatment,” Water Res., vol. 46, no. 15, pp. 4653–4664, 2012. [99] M. C. Dodd et al., “Kinetics and Mechanistic Aspects of As(III) Oxidation by Aqueous Chlorine, Chloramines, and Ozone: Relevance to Drinking Water Treatment,” Environ. Sci. Technol., vol. 40, no. 10, pp. 3285–3292, May 2006. [100] I. M. Schreiber and W. A. Mitch, “Nitrosamine Formation Pathway Revisited: The Importance of Chloramine Speciation and Dissolved Oxygen,” Environ. Sci. Technol., vol. 40, no. 19, pp. 6007–6014, Oct. 2006. [101] S. W. Krasner, W. A. Mitch, D. L. McCurry, D. Hanigan, and P. Westerhoff, “Formation, precursors, control, and occurrence of nitrosamines in drinking water: A review,” Water Res., vol. 47, no. 13, pp. 4433–4450, 2013. 87 xcv [102] W. A. M. Hijnen, E. F. Beerendonk, and G. J. Medema, “Inactivation credit of UV radiation for viruses, bacteria and protozoan (oo)cysts in water: A review,” Water Res., vol. 40, no. 1, pp. 3–22, 2006. [103] N. N. Mahamuni and Y. G. Adewuyi, “Advanced oxidation processes (AOPs) involving ultrasound for waste water treatment: A review with emphasis on cost estimation,” Ultrason. Sonochem., vol. 17, no. 6, pp. 990–1003, 2010. [104] W. R. Haag and J. Holgne, “Ozonation of Bromide-Containing Waters: Kinetics of Formation of Hypobromous Acid and Bromate,” Environ. Sci. Technol., vol. 17, no. 5, pp. 261–267, 1983. [105] T. P. Bonacquisti, “A drinking water utility’s perspective on bromide, bromate, and ozonation,” Toxicology, vol. 221, no. 2, pp. 145–148, 2006. [106] S. W. Krasner, W. H. Glaze, H. S. Weinberg, P. A. Daniel, and I. N. Najm, “Formation and Control of Bromate During Ozonation of Waters Containing Bromide,” J. AWWA, vol. 85, no. 1, pp. 73–81, Jan. 1993. [107] Z. Chen, L. Yang, X. Zhai, S. Zhao, A. Li, and J. Shen, “N-nitrosamine formation during chlorination/chloramination of bromide-containing water,” Water Supply, vol. 10, no. 3, pp. 462–471, Jul. 2010. [108] J. Le Roux, H. Gallard, and J.-P. Croué, “Formation of NDMA and Halogenated DBPs by Chloramination of Tertiary Amines: The Influence of Bromide Ion,” Environ. Sci. Technol., vol. 46, no. 3, pp. 1581–1589, Feb. 2012. [109] J. Luh and B. J. Mariñas, “Bromide Ion Effect on N-Nitrosodimethylamine Formation by Monochloramine,” Environ. Sci. Technol., vol. 46, no. 9, pp. 5085– 5092, May 2012. [110] R. Valentine, J. Choi, and Z. Chen, “Factors affecting the formation of NDMA in water and occurence,” in AWWA Research Foundation, 2006. [111] J. C. Ianni, “Kintecus,” 4.55, 2012. [112] J. Luh and B. J. Mariñas, “Kinetics of Bromochloramine Formation and Decomposition,” Environ. Sci. Technol., vol. 48, no. 5, pp. 2843–2852, Mar. 2014. [113] C. T. Jafvert and R. L. Valentine, “Reaction scheme for the chlorination of ammoniacal water,” Environ. Sci. Technol., vol. 26, no. 3, pp. 577–586, Mar. 1992. [114] M. Gazda and D. W. Margerum, “Reactions of Monochloramine with Br2, Br3-, HOBr, and OBr-: Formation of Bromochloramines,” Inorg. Chem., vol. 33, pp. 118–123, 1994. [115] M. Deborde and U. von Gunten, “Reactions of chlorine with inorganic and organic compounds during water treatment—Kinetics and mechanisms: A critical review,” Water Res., vol. 42, no. 1, pp. 13–51, 2008. [116] W. Beita-Sandí, C. U. Erdem, and T. Karanfil, “Effect of bromide on NDMA formation during chloramination of model precursor compounds and natural waters,” Water Res., vol. 170, p. 115323, 2020. [117] H. K. Hall Jr., “Correlation of the Base Strengths of Amines,” J. Am. Chem. Soc., vol. 79, pp. 5441–5444, 1957. [118] M. E. Huang, S. Huang, and D. L. McCurry, “Re-Examining the Role of Dichloramine in High-Yield N-Nitrosodimethylamine Formation from N,N- Dimethyl-α-arylamines,” Environ. Sci. Technol. Lett., vol. 5, no. 3, pp. 154–159, Mar. 2018. 88 xcvi [119] M. Selbes, W. Beita-Sandí, D. Kim, and T. Karanfil, “The role of chloramine species in NDMA formation,” Water Res., vol. 140, pp. 100–109, 2018. [120] A. Kwon, “Natural gas generators make up the largest share of overall U.S. generation capacity,” U.S. Energy Information Administration, 2017. [Online]. Available: https://www.eia.gov/todayinenergy/detail.php?id=34172#tab1. [121] A. (Duke U. Vengosh and D. (Duke U. Patino-Echeverri, “Replacing coal with gas or renewables saves billions of gallons of water,” Science Daily, Durham, NC, 2019. [122] B. Park, “Water withdrawals by U.S. power plants have been declining,” Washington, DC, 2018. [123] “What is U.S. electricty generation by energy source?,” U.S. Energy Information Administration, 2021. [Online]. Available: https://www.eia.gov/tools/faqs/faq.php?id=427&t=3. [124] T. Hodge, “EIA forecasts less power generation from natural gas as a result of rising fuel costs,” U.S. Energy Information Administration, 2021. . [125] “Annual Energy Outlook 2021 with Projections to 2050,” Washington, DC, 2021. [126] “Thermoelectic Water Use: Consumption,” Washington, DC, 2017. [127] C. A. Dieter et al., “Estimated Use of Water in the United States 2015,” Reston, Virginia, 2017. [128] “Thermoelectric cooling water data.” Oct-2019. [129] “Thermoelectric Power Water Use,” U.S. Geological Survey, 2015. [Online]. Available: √. [130] V. Dorjets, “Many newer power plants have cooling systems that reuse water,” U.S. Energy Information Administration, 2014. [Online]. Available: https://www.eia.gov/todayinenergy/detail.php?id=14971. [131] J. Macknick, R. Newmark, G. Heath, and K. C. Hallett, “Operational water consumption and withdrawal factors for electricity generating technologies: a review of existing literature,” Environ. Res. Lett., vol. 7, no. 4, p. 45802, 2012. [132] “Steam Electric Power Generating Effluent Guidelines - 2015 Final Rule,” U.S. Environmental Protection Agency, 2015. . [133] S. R. Saxer, “The fluid nature of property rights in water,” Duke Envtl. L. Pol’y F., vol. 21, p. 49, 2010. [134] J. Smyth, “Coal and water conflicts in the American West,” 2020. [135] B. I. Cook, T. R. Ault, and J. E. Smerdon, “Unprecedented 21st century drought risk in the American Southwest and Central Plains,” Sci. Adv., vol. 1, no. 1, p. e1400082, 2015. [136] Muhammad Yaqub and W. Lee, “Zero-liquid discharge (ZLD) technology for resource recovery from wastewater: A review,” Sci. Total Environ., vol. 681, pp. 551–563, 2019. [137] M. Whited, F. Ackerman, and S. Jackson, “Water Constraints on Energy Production: Altering our Current Collision Course,” Cambridge, MA, 2013. [138] “Water Rights,” California Water Boards, 2021. [Online]. Available: https://www.waterboards.ca.gov/waterrights/board_info/faqs.html#toc178761079. [139] “The Water Rights Process,” California Water Boards, 2020. [Online]. Available: https://www.waterboards.ca.gov/waterrights/board_info/water_rights_process.html . 89 xcvii [140] M. A. Burch, “Water rights: Supply issues for local agency formation commission,” Monterey, California, 2005. [141] G. Andersen, M. Cleveland, and D. Shea, “Water for Energy: Addressing the Nexus between Electricity Generation and Water Resources,” 2019. [142] C. V. Stern and P. A. Sheikh, “Management of the Colorado River: Water Allocations, Drought, and the Federal Role,” 2020. [143] M. Berkman, “The electricity–water nexus: is a crisis imminent?,” Water policy, vol. 17, no. 6, pp. 1163–1175, 2015. [144] M. Buzec, “What Are Apartment Utilities and How Much Will They Cost Me?,” RENTCafe Blog, 2019. [Online]. Available: https://www.rentcafe.com/blog/apartment-search-2/money/apartment-utilities- breakdown/. [145] S. Ray, “Some U.S. electricity generating plants use dry cooling,” U.S. Energy Information Administration, 2018. . [146] J. Macknick, R. Newmark, G. Heath, and K. C. Hallett, “Operational water consumption and withdrawal factors for electricity generating technologies: a review of existing literature,” Environ. Res. Lett., vol. 7, no. 4, p. 45802, 2012. [147] “Reducing water consumption for cooling of thermal generation plants,” European Environment Agency European Commission, 2021. . [148] “Cooling Power Plants,” World Nuclear Association, 2020. . [149] Effluent Limitations Guidelines and Standards for the Steam Electric Power Generatig Point Source Category. U.S.A., 2015. [150] “Technical Development Document for the Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category,” Washington, DC, 2015. [151] “EPA Finalizes Power Plant Effluent Limitation Guidelines that Save Money and Reduce Pollution,” U.S. EPA Press, Washington, DC, Aug-2020. [152] “2020 Steam Electric Reconsideration Rule,” U.S. Environmental Protection Agency, 2020. . [153] M. Bornhoft, “Considerations for implementing a zero liquid discharge program,” Water Technology, 2018. [Online]. Available: https://www.watertechonline.com/wastewater/article/15550690/considerations-for- implementing-a-zero-liquid-discharge-program. [154] G. P. Richer, “4 Ways to Maximize Your Steam Boiler’s Cycles of Concentration,” Process Heating, 2017. [155] I. Khamis and K. C. Kavvadias, “Trends and challenges toward efficient water management in nuclear power plants,” Nucl. Eng. Des., vol. 248, pp. 48–54, 2012. [156] G. R. Nelson, “Water Recycle/Reuse Possibilities: Power Plant Boiler and Cooling Systems,” Corvallis, Oregon, 1974. [157] R. H. Borgwardt, “Combined flue gas desulfurization and water treatment in coal- fired power plants,” Environ. Sci. Technol., vol. 14, no. 3, pp. 294–298, 1980. [158] M. S. Mohsen, “Treatment and reuse of industrial effluents: case study of a thermal power plant,” Desalination, vol. 167, pp. 75–86, 2004. [159] “Coal Ash Pond Remediation,” WesTech Engineering. [Online]. Available: https://www.westech-inc.com/en-usa/industry-solutions/industrial-overview/coal- 90 xcviii ash-pond. [160] C. Charisiadis, “Brine Zero Liquid Discharge (ZLD) Fundamentals and Design,” Netherlands, 2018. [161] T. K. Poulson, “Strategic Alternatives for Brine Management in the Valley of the Sun,” 2010. [162] “Brine-Concentrate Treatment and Disposal Options Report,” 2009. [163] “Treatment and Disposal of Brine Concentrate Report,” 2009. [164] M. Goodin, “Deep Well Injections,” 2017. [165] C. Charisiadis, “Brine Zero Liquid Discharge (ZLD) Fundamentals and Design,” South Miami, FL, 2018. [166] “Brine-Concentrate Treatment and Disposal Options Report.” [167] E. D. Mackey and T. Seacord, “Regional Solutions for Concentrate Management,” Alexandria, VA, 2008. [168] D. M. Warsinger, E. W. Tow, K. G. Nayar, L. A. Maswadeh, and J. H. Lienhard V, “Energy efficiency of batch and semi-batch (CCRO) reverse osmosis desalination,” Water Res., vol. 106, pp. 272–282, 2016. [169] V. A. Como, T. Cannon, and M. Mueller, “Operating a ZLD, What Does it Take?,” 2013. [170] D. Mukhopadhyay, “Method for high efficiency reverse osmosis operation,” US9073763B2, 2015. [171] “Closed Circuit Reverse Osmosis System Squeezes Money Savings Out of Water Management,” Power Magazine, 2017. [172] “RO systems make their case for brine concentration applications,” Global Water Intelligence Magazine, Oxford, UK, 2019. [173] E. B. Casey, “U.S. Power Profile Shift Sparks Water Opportunity,” Water Online, 2016. [174] “Duke Energy selects HERO Membrane Process for ZLD Application,” Canonsburg, PA. [175] “Zero Liquid Discharge (ZLD) Systems Market - Growth, Trends, and Forecast (2020-2025),” 2020. [176] W. A. Shaw, “Fundamentals of Zero Liquid Discharge System Design,” Power Magazine, Rockville, MD, 2011. [177] C. Charisiadis, “Brine Zero Liquid Discharge (ZLD) Fundamentals and Design,” South Miami, FL, 2018. [178] “Wastewater Treatment for Power Plants: Considering Zero Liquid Discharge,” Salt Lake City, UT, 2017. [179] “How much will a zero liquid discharge system cost your facility?,” Samco Technologies, Inc., 2017. . [180] A. Panagopoulos and K.-J. Haralambous, “Minimal Liquid Discharge (MLD) and Zero Liquid Discharge (ZLD) strategies for wastewater management and resource recovery – Analysis, challenges and prospects,” J. Environ. Chem. Eng., vol. 8, no. 5, p. 104418, 2020. [181] T. Tong and M. Elimelech, “The Global Rise of Zero Liquid Discharge for Wastewater Management: Drivers, Technologies, and Future Directions,” Environ. Sci. Technol., vol. 50, no. 13, pp. 6846–6855, Jul. 2016. [182] M. Bornhoft, “Consideration for implementing a zero liquid discharge program,” 92 xcix Water Technology, 2018. [183] “Wastewater treatment for power plants: Considerting zero liquid discharge,” WesTech Engineering, Salt Lake City, UT, 2017. [184] M. Klembara, “Energy-Water Desalination Hub,” U.S. Department of Energy, 2019. . [185] “Climate,” Metro Denver EDC, 2021. [Online]. Available: http://www.metrodenver.org/lifestyle/climate/. [186] A. Meadow et al., “Climate Profile of Gila River Indian Community,” 2017. [187] “Water-TAP3,” National Alliance for Water Innovation, 2021. [Online]. Available: https://www.nawihub.org/watertap. [188] S. Korellis, “A systematic approach to reduce power plant auxiliary power,” Palo Alto, CA, 2019. [189] H. Cooley and R. Wilkinson, “Implications of Future Water Supply Sources for Energy Demands. WateReuse Research Foundation,” Alexandria, VA, 2012. [190] “One Energy Electricity Rate Data Tool,” Findlay, OH, 2019. [191] “Gila River Power Station,” Energy Justice Network. [Online]. Available: http://www.energyjustice.net/map/displayfacility-68801.htm. [192] “Rethinking water cycle yields big savings at Gila River,” Combined Cycle Journal, 2014. [Online]. Available: https://www.ccj-online.com/1q-2014/best-practices-gila- river/. 93

Abstract (if available)

Abstract Global water stress has prompted recent interest in desalination systems for treatment of saline and/or otherwise impaired water sources for potable use. These desalination systems typically rely on reverse osmosis (RO) as the main treatment process. RO, however, can have high energy demands and produces a high-salinity waste brine that must be disposed of or further treated. Seawater RO (SWRO) facilities may be co-located with wastewater treatment facilities that also discharge a stream to the ocean. Currently, the only synergistic use of these waste streams is to use the treated wastewater to dilute the SWRO brine stream prior to discharge. However, discharge of treated wastewater to the ocean could be considered a waste of the water resource within this stream and has no impact on the high energy demands of SWRO. The main goal of this dissertation is to provide solutions to increase potable water supply, reduce undesirable impacts associated with alternative water treatment, and improve water reuse practices at power facilities by evaluating the monetary, environmental, and public health costs associated with: (i) a membrane-based salinity gradient power process, (ii) a new concept for blending reclaimed wastewater and seawater upstream of disinfection, and (iii) zero-liquid discharge (ZLD) technologies employed at natural gas power facilities. For the first time, this work determined: (i) a power density phenomenon that limits osmotic power production for SWRO, (ii) an optimal wastewater-seawater blending and disinfection scheme to limit disinfection by-product formation, and (iii) the levelized cost of water for employing ZLD technologies at power facilities used to increase on-site water reuse.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Integrating systems of desalination and potable reuse: reduced energy consumption for increased water supply

PDF

The formation, fate and transformation of nitromethane in potable reuse processes

PDF

Wastewater reclamation and potable reuse with novel processes: membrane performance and system integration

PDF

The roles of surface and pore properties in wetting resistance for membrane distillation membranes

PDF

Thermally driven water treatment with membrane distillation: membrane performance, waste heat integration, and cooling analysis

PDF

Catalytic oxidation and precursor identification of disinfection byproducts in recycled water

PDF

Water desalination: real-time membrane characterization for performance prediction and system analysis for energetic enhancement

PDF

Wastewater-based epidemiology for emerging biological contaminants

PDF

Optimizing biomembrane reactor systems for water reclamation and reuse applications

PDF

Developing high-resolution spatiotemporal methods to model and quantify water use for energy

PDF

Advancing energy recovery from food waste using anaerobic biotechnologies: performance and microbial ecology

PDF

Interdependency of the water-oil-nuclear industries in the Persian/Arabian Gulf: understanding risk and improving prevention and preparation of disasters

PDF

Fate of antibiotic resistance in anaerobic membrane bioreactors

PDF

Developing frameworks to quantify the operational and environmental performance of energy systems within the context of climate change

PDF

Electrokinetic transport of Cr(VI) and integration with zero-valent iron nanoparticle and microbial fuel cell technologies for aquifer remediation

PDF

Groundwater contaminant transport predictions in a sustainable water management scenario

PDF

Impacts of heat mitigation strategies and pollutant transport on climate and air quality from urban to global scales

PDF

Bioelectrochemical treatment of anaerobic process effluents: mitigation of dissolved methane and sulfide

PDF

Reproducible and rapid computational approaches for assessing contamination in natural aquifers

PDF

Development of novel techniques for evaluating physical, chemical and toxicological properties of particulate matter in ambient air

Asset Metadata

Creator Plata, Sophia Lauren (author)

Core Title Integrated technologies, blending schemes, and reuse practices to address contaminant and energy challenges in water reclamation

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Engineering (Environmental Engineering)

Degree Conferral Date 2021-12

Publication Date 12/07/2023

Defense Date 07/14/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag desalination,OAI-PMH Harvest,water reuse,zero-liquid discharge

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Childress, Amy (committee chair)

Creator Email sophia.plata1793@gmail.com,splata@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC18010016

Unique identifier UC18010016

Legacy Identifier etd-PlataSophi-10283

Document Type Dissertation

Format application/pdf (imt)

Rights Plata, Sophia Lauren

Type texts

Source 20211210-wayne-usctheses-batch-903-nissen (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu