Opportunities and Challenges for Direct Potable Water Reuse in Arid Inland Communities

A central question in planning and engineering today is: how can communities continue to develop and thrive while meeting the resource needs and providing a high quality of life for future generations? Critical among these needs is access to adequate water supplies for municipal, industrial, and agricultural uses. With increasing population and development pressures, many regions around the world face freshwater shortages (Arnold et al. 2012), which threaten quality of life and freshwater ecosystem environments (Grant et al. 2012). Drought conditions in the southwestern United States appear to be increasing in frequency and severity (Tinker 2014), climate change is expected to further reduce water supplies (Gutzler 2012), and a U.S. Department of the Interior (DoI) report identified hot spots for potential water crises that may result in water conflicts in the western United States by 2025 (U.S. Department of the Interior 2005). New sources of water must be identified to provide supplies for current and future demands.

Communities often have several demand- and supply-side options to meet future water needs (Grant et al. 2012; Hering et al. 2013; Hurlimann et al. 2009), and tools have been developed to help water planners and policymakers decide how best to develop their various options given future climate and demographic uncertainties (Ray et al. 2012). Reducing water demand through conservation measures is usually the first and most cost-effective step in water-scarce communities. For example, in Albuquerque, New Mexico, the community has reduced per capita water demand from 946 to $492\mathrm{L}/\mathrm{capita}/\mathrm{day}$ (250–130 gal. per capita per day) in less than 15 years strictly through incentive-based voluntary programs. However, it is important to recognize that water conservation programs may have environmental and water rights implications. These include issues such as reduced flows in receiving waters, which may impact the aquatic or riparian environments, and effects on water rights as a result of reduced return flow credits from decreased wastewater discharges.

Supply-side options consist of increasing the supply through obtaining new sources of water, such as transbasin diversions or seawater desalination. Some options are limited by sustainability and/or reliability considerations. For instance, while desalination of brackish and saline groundwater can be used in some areas to extend potable supplies, the water source is typically not replenished, and therefore may not constitute a sustainable supply (Thomson 2012). Desalination of brackish groundwater and seawater also comes with high-energy requirements and the challenge of brine disposal (Thomson 2012; Schroeder et al. 2012). Captured rainwater can help increase water supplies, but in arid climates rain is not only unreliable, but also predicted to decrease in coming years (Gutzler 2012; Thomson 2012). Nonpotable water reuse systems have been implemented in a number of communities, but it is important to consider that the consequent reduction in demand for potable water has hydraulic and flow implications for water and wastewater systems (Kandiah et al. 2016). Similarly, gray water reuse by individual households can reduce demand for potable water, but it has public health concerns, is expensive to implement in established communities due to the cost of replumbing residences (Arnold et al. 2012), and may lead to a detrimental decrease in sewer flow velocities (Ormerod and Scott 2012). Further, nonpotable and gray water reuse may not be considered water conservation in situations where return flow credits are required to meet streamflow requirements or downstream delivery obligations (Thomson and Shomaker 2009).

Historically, communities in the west have turned to large interbasin water transfers as a source of supply. Notable projects include diversions of Colorado River water to California, Arizona, and New Mexico. However, interbasin transfers are very costly, have enormous environmental consequences, and may not be reliable in times of need (Loaiciga 2015). More fundamentally, virtually all of the water in western watersheds is already fully allocated, making future interbasin transfers unlikely. Interbasin water transfers are also usually energy intensive, which incurs substantial environmental and societal costs (Grant et al. 2012; Schroeder et al. 2012).

Two supply-side options that hold promise for improving sustainability and reliability of potable water supplies are indirect potable reuse (IPR) and direct potable reuse (DPR). In both cases, the water supply is augmented by reusing wastewater, resulting in the potential to significantly increase water productivity (Grant et al. 2012). Though different configurations exist (Gerrity et al. 2013; Tchobanoglous et al. 2011), generally for IPR, wastewater treatment plant (WWTP) effluent is highly purified through advanced treatment and then directed to an environmental buffer, such as a stream, reservoir, or aquifer. The environmental buffer is intended to protect public health by providing an additional barrier and time for pollutant degradation processes as well as dilution with water from other sources (Raucher and Tchobanoglous 2014). With DPR, WWTP effluent is highly purified through advanced treatment and then either (1) combined with the raw water supply and directed to a water treatment plant, or (2) introduced directly into the distribution system, bypassing the water treatment plant. While the latter option is often mentioned in the literature, Raucher and Tchobanoglous (2014) suggest the former option as the preferred alternative until monitoring technology is available to detect low levels of contaminants of concern. Also, a DPR system may include an engineered storage buffer (ESB) in order to allow sufficient time for verification of specific water quality parameters before the water enters the distribution system (Raucher and Tchobanoglous 2014; Law 2008; Tchobanoglous et al. 2011; Leverenz et al. 2011; U.S. EPA 2012). Advanced treatment technologies and configurations are described below in more detail.

Numerous IPR systems exist around the world; however, implementation of this strategy requires access to a suitable environmental buffer (Raucher and Tchobanoglous 2014). The environmental buffer offers a benefit in that it allows for water storage during periods of decreased demand and increased withdrawal rates in times of higher demand (Trussell et al. 2013). Some researchers suggest that IPR reduces contamination risk by providing dilution and additional biological, chemical, and physical treatment (Rodriguez et al. 2009). Others have noted that IPR is inefficient in that highly treated water is degraded when mixed with lower-quality water in the environmental buffer, and therefore wastes energy and resources by treating the same water twice (Leverenz et al. 2011; Australian Academy of Technological Sciences and Engineering 2013). Whether or not the environmental buffer actually provides a benefit depends on many site-specific variables including the extent and nature of treatment provided, the residence time in the buffer, and the environmental chemistry of the buffer. Several scholars have found IPR to be more expensive than DPR because additional pumping is usually required (Law 2008; Tchobanoglous et al. 2011; Leverenz et al. 2011; Australian Academy of Technological Sciences and Engineering 2013; Venkatesan et al. 2011). Direct potable reuse’s smaller pumping requirements also result in a smaller carbon footprint—an important consideration because energy costs and greenhouse gas emissions are increasingly a consideration in water supply projects (Gutzler 2012; Law 2008; Australian Academy of Technological Sciences and Engineering 2013). Further, shorter water transmission lines for DPR are less vulnerable to natural and artificial accidents (U.S. EPA 2012). Despite the benefits of DPR, its implementation faces numerous challenges and requires many additional considerations, as discussed below.

The most cited DPR facilities in operation or under construction are in Windhoek, Namibia; Cloudcroft, New Mexico; and Big Spring, Texas; and other facilities will come online soon in Texas. In Windhoek, highly treated reclaimed water is blended directly into the potable water pipeline. The Cloudcroft plant, which is not yet operational, will blend highly treated reclaimed water with well water and springwater in a tank that provides a hydraulic detention time of 2 weeks; the combined flow will then receive additional water treatment before entering the distribution system. In Big Spring, filtered effluent from the WWTP receives advanced treatment and is blended with raw water in the transmission line; the combined flow is then treated in a water treatment plant (Tchobanoglous et al. 2011). The facilities all have monitoring schemes in place to measure specific water quality parameters prior to distribution. While DPR is relatively new to the United States, the Namibian plant has been operating successfully in various configurations since 1968, with no reports of significant adverse health effects (Crook 2010).

Communities adopt DPR because of severe water shortages and lack of alternative sources of supply (Tchobanoglous et al. 2011). As drought conditions worsen, DPR is increasingly being considered in communities facing water scarcity (Leverenz et al. 2011). Though many water-short communities are small to medium size and scattered throughout the inland southwestern United States—and the only DPR plants currently serving customers are in the inland Southwest—most DPR research has focused on large coastal communities with relatively high mean household incomes (U.S. Census Bureau 2012). Equally important, these larger coastal communities have very different laws governing water rights. Fig. 1 recreates the DoI’s hot spots map, highlighting the three most-studied communities in the United States and a sampling of inland ones within the DoI’s hot spots for high water conflict potential (Denver and Tucson, Arizona, are included as areas with DPR research, though far less has been performed there as compared with the large coastal areas shown).

Direct potable reuse options may be different for larger, wealthier coastal communities than for midsized or smaller inland ones. The principal challenges for small to medium size inland communities considering DPR include:

As a result, public utilities in inland communities are struggling with planning and selection of strategies to meet future water demand while minimizing constraints to sustainable community planning. A holistic assessment is needed to help decision makers in smaller inland areas weigh the opportunities and challenges of DPR.

This paper identifies the knowledge gaps and challenges that limit widespread implementation of DPR in arid inland communities. It summarizes the essential considerations for water managers in inland communities to aid in decision making related to DPR and long-term water supply reliability. The focus of this paper is on the following categories of DPR implementation challenges: water resources, regulatory uncertainty, technology, and costs. Also critical is public acceptance of potable reuse in arid inland communities, which is a major topic and the focus of other research.

One of the major distinctions between DPR for coastal communities and inland communities is that reduction of wastewater discharges by a reuse program will reduce flows in the receiving water. Although wastewater reuse is frequently proposed as a method of conserving water, if considered from a basinwide perspective it can be shown that in actuality it is not conservation. Rather, it is simply substitution of low-quality wastewater for high-quality surface or ground water (Thomson and Shomaker 2009; Fleming and Hall 2000). This reduces discharge of water that may be an important supply for in-stream environments or downstream users. The consequences of reduced discharge to streams in arid inland states can be manifested in three ways: (1) environmental effects resulting from reduced in-stream flows; (2) reduced flows to meet downstream obligations; and (3) impact of reduced return flows on a community’s water rights. These are discussed in the following paragraphs.

First, reduced discharges may have environmental effects associated with flow reduction in the receiving stream. This is particularly problematic in arid climates where WWTP effluent frequently comprises a major fraction of the flow of small streams, a phenomenon referred to as effluent-dominated streams. Flows needed to support aquatic and riparian environments are known as environmental flows. A growing body of literature describes the detriment to various ecological metrics as streamflows and hydrology are altered (Poff and Zimmerman 2010). An extreme manifestation of reduced environmental flows will occur if there are federally listed endangered species in the stream. For example, in New Mexico, the requirement to maintain adequate in-stream flows in the Rio Grande to protect two endangered species has been the subject of federal litigation (OSE 2016).

A second consequence of reduced wastewater discharge due to reuse programs is decreased flow to meet downstream obligations. Most interstate streams in the arid Southwest are subject to interstate stream compacts, federal legislation, and federal lawsuits and settlements. Interstate stream compacts are congressionally approved agreements between states that allocate water among the states sharing the stream. There are currently 26 interstate stream compacts in force; most are in western states and at least 50 years old (Muys et al. 2007). In addition to compacts, there is federal legislation regarding interstate waters as well as interstate lawsuits and settlements that relate to apportionment of interstate streams. The consequence of these laws, stream compacts, lawsuits, and settlements is that water managers of upstream states face federal requirements to assure that an adequate volume of water is delivered to downstream states. Thus, any activity that reduces the flow of water in a stream, such as wastewater reuse, is subject to careful scrutiny and possible denial, a challenge not faced by coastal communities with no downstream users.

The third constraint on wastewater reuse for inland communities has to do with state water law and water rights. The right to reuse wastewater varies by state and has been summarized by the National Research Council (NRC 2012). Water rights in many inland western states are based, at least in principle, on consumptive use, which is defined as the difference between the diversion amount and amount returned to the stream or aquifer, referred to as return flow credits. Therefore, a community that reuses its wastewater would reduce its return flow credits and would need additional water rights. This constraint, of course, depends on each state’s water law. In a recent review of Colorado water law, Hobbs (2013) notes that from a legal perspective wastewater return flows often “belong to the stream system as part of the public’s resource, subject to appropriation and administration.” Thus, while wastewater reuse in New Mexico may require acquisition of additional water rights to offset the reduction in return flows, wastewater from treatment plants in California that discharge to the ocean can be fully used by the utility (NRC 2012).

Clearly the environmental, federal, and state constraints on reuse of wastewater are topics that must be addressed early in the planning stages for any inland reuse project. Bischel et al. (2012) reported that 37% of recycled water managers surveyed viewed these constraints as hindrances to water recycling. However, this study was done in northern California, and these topics are likely to be of much greater concern for arid inland communities.

One of the first technological challenges a community faces when considering a DPR project is determining what level of treatment is required. The federal Safe Drinking Water Act (SDWA) and subsequent regulations do not mention potable reuse. Although the effluent from most state-of-the-art WWTPs will meet the numeric criteria established under this act, it is clear that much additional treatment is needed before wastewater can be used for potable supply both from a health perspective and to satisfy public concerns about its use. However, there are no national criteria that establish the nature and degree of this additional treatment. The National Water Research Institute (NWRI), with funding from several national water research organizations, has recently published a framework for DPR that provides a good summary of the technical and regulatory challenges (NWRI 2015a). While the USEPA (2012) provides reuse guidelines, no federal water reuse regulations exist and regulation is left to the individual states. Several states have developed regulations, policies, or guidance for nonpotable reuse and a few have developed guidance on IPR, but no state yet has regulations for DPR (NRC 2012). In the absence of state or federal regulations, IPR and DPR projects are evaluated on a case-by-case basis.

Regulated and unregulated chemicals and pathogens are contaminants of concern in potable reuse. Pathogens are of greatest concern because they pose acute risks to health rather than the chronic risks associated with most chemical constituents (Raucher and Tchobanoglous 2014). This means that the risks to public health resulting from delivery of inadequately treated wastewater for a short period of time due to failure of a treatment process are primarily associated with the acute risk of disease from pathogenic organisms rather than ingestion of compounds that require years of exposure before posing a measurable risk.

Traditionally, the presence of pathogens in water has been inferred by measuring the concentration of indicator organisms, which are microorganisms present in water or wastewater that occur in human waste. The most common indicator organism, and that identified in the SDWA regulations, is E. coli. Enteric viruses are also increasingly used as an indicator of viral pathogens. Finally, Cryptospiridium oocysts and Giardia lamblia spores, two pathogenic protozoans that are particularly resistant to conventional disinfection processes, have recently been added to SDWA regulations. One of the challenges of establishing drinking water standards for pathogens is that they can be present in water at very high concentrations, often exceeding 1 million organisms per 100 mL, and yet present a health hazard at concentrations of less than 1 organism per 100 mL. The SDWA regulations deal with this wide variation by using the concept of log removal in addition to absolute concentrations of regulated pathogens.

For drinking water treatment, the log removal concept is used in identifying treatment processes that are appropriate for the quality of the influent feedwater. Each treatment process is assigned a log removal credit based on documented performance. If the feedwater is poor quality, a sequence of two or more treatment processes may be needed to achieve the desired log removal performance. A summary of existing and proposed log removal criteria for DPR applications is presented in Table 1.

The California Department of Public Health prohibits DPR until requirements for water recycling are established (Tchobanoglous et al. 2011). A task force is currently developing recommendations for these criteria (NWRI 2015a). The state is considering log removal requirements of 12, 10, and 10 for virus, Giardia, and Cryptosporidium, respectively, based on the pathogen levels in untreated wastewater; these levels are for IPR projects, and it is expected that the same levels will apply for DPR applications (Raucher and Tchobanoglous 2014).

Texas uses log removal levels for virus, Giardia, and Cryptosporidium of 8, 6, and 5.5, respectively, based on effluent from secondary wastewater treatment; these levels are approximately equivalent to California’s proposed requirements if the log removal credits for wastewater treatment are counted (Raucher and Tchobanoglous 2014). However, new DPR projects in Texas must demonstrate more than compliance with these pathogen levels. In accepting the Big Spring DPR project, the Texas Commission on Environmental Quality set forth specific requirements “in design, operation, reporting, calibration, and record keeping” (LaCaille 2010). Additional recent DPR projects in Texas, such as the Wichita Falls facility, have been considered on a case-by-case basis by the Texas Commission on Environmental Quality.

In considering potential future regulations for DPR, Crook (2010) states that while several studies have indicated that potable reuse water does not pose an increased risk to public health, regulatory agencies would still need to resolve a number of issues in developing regulations, policies, and/or guidelines. Specifically, he notes that, “Assessment of the safety of using recycled water for direct potable reuse must consider several factors, such as microbial and chemical quality of the product water, treatment performance and reliability, multiple barriers, monitoring capability, and system operation and management” (Crook 2010). Others have suggested that it may be best to regulate treated wastewater as a third drinking water source, in addition to surface water and groundwater, with specific requirements for advanced treatment (Raucher and Tchobanoglous 2014).

With only a handful of municipal-scale DPR facilities in the world, there are few full-scale, operational examples from which to gain insights on the regulatory aspects of DPR. According to NWRI’s former executive director (J. Mosher, personal communication, 2014), regulators want to be responsive to communities’ needs and implementation of DPR, but many feel they do not have the tools needed to assist in moving DPR forward. Regulators’ priority is protection of public health, and there is little experience for them to build on in laying out regulations and reliability standards for DPR projects. Thus, many state regulators are presently approaching DPR with caution.

Direct potable reuse is predicated on the assumption that technologies exist to provide adequate wastewater treatment for potable use, while acknowledging the need for additional study (Arnold et al. 2012; Law 2008; Tchobanoglous et al. 2011; Leverenz et al. 2011; NRC 2012; Rogers and Laurer 1992). However, due to the complexities associated with advanced treatment technologies and the economies of scale for large treatment processes, DPR may not be appropriate for small communities.

While advanced treatment technologies can provide water of any desired quality for DPR, one must consider trade-offs of cost, complexity, reliability, and ease of operation. To ensure reliability, essential features that must be incorporated into the design of an advanced treatment system for DPR are robustness (i.e., safety factors), resiliency (i.e., ability to adjust to upsets), and redundancy (i.e., backup systems), along with adequate online monitoring and process control (Trussell et al. 2013).

Advanced treatment process configurations for potable reuse facilities usually include microfiltration (MF), RO, and advanced oxidation (Raucher and Tchobanoglous 2014). The state-of-the-art $2.65\times {10}^5{\mathrm{m}}^3/\mathrm{day}$ [$70\mathrm{million}\mathrm{gal.}/\mathrm{day}(\mathrm{mgd})$] IPR groundwater replenishment system (GWRS) at the Orange County Water District (OCWD) in California, which uses these processes, exceeds all drinking water quality standards and also removes unregulated contaminants, sometimes referred to as emerging contaminants or contaminants of concern (Raucher and Tchobanoglous 2014; Tchobanoglous et al. 2011). The approximately $680{\mathrm{m}}^3/\mathrm{day}$ (0.18 mgd) Cloudcroft treatment process is very similar to the OCWD configuration (Tchobanoglous et al. 2011). The OCWD and Cloudcroft treatment configurations are important examples for other communities examining the feasibility of DPR because they have already been “accepted by various regulatory authorities as being able to produce safe drinking water, and … implementation of these projects has been accepted by the public” (Tchobanoglous et al. 2011; Leverenz et al. 2011).

There are, however, four major drawbacks to including RO in advanced treatment: (1) the high energy requirements; (2) the water chemistry of inland supplies is much different than seawater and thus limits the fractional recovery of treated water; (3) the amount of water wasted from the process as concentrate; and (4) the challenge of concentrate disposal. Coastal communities have an unlimited supply of salt water with low scaling potential and can discharge concentrate into the sea (Leverenz et al. 2011) so that Drawbacks 2, 3, and 4 are of little consequence and low overall recovery is not a constraint. But inland communities have a limited supply of water. Assuming a RO feedwater recovery of 75%, a typical value for inland facilities, the RO plant must have a water supply that is 33% greater than the desired flow (i.e., 133 L of wastewater must be treated to produce 100 L of treated RO permeate). In contrast to seawater in which sodium and chloride make up 98% of the dissolved ions, inland water supplies are usually high in scale-forming constituents including calcium, magnesium, carbonate, sulfate, and silica. This chemistry limits feedwater recovery because high concentrations of scale-forming minerals in the concentrate are avoided by wasting more water. Furthermore, without a nearby ocean, concentrate disposal is more complicated at inland locations. The disposal options identified by NRC (2008) include surface water discharge, discharge to a wastewater collection system, deep well injection, evaporation ponds, and land application. However, concerns regarding surface water, groundwater, and soils at inland locations mean that deep well injection is often the only feasible option for RO concentrate disposal, and a suitable aquifer is not always close by. For instance, the City of El Paso, Texas, pipes concentrate from its desalination plant a distance of 35.4 km (22 mi) to deep injection wells (Thomson and Howe 2009). Numerous brine treatment options, including multistage RO, crystallizers, forward osmosis, membrane distillation, solar evaporators, and spray dryers, have been proposed to minimize the volume of concentrate requiring disposal; however, they are expensive, complicated, and unproven and have not been used in current inland DPR projects (Raucher and Tchobanoglous 2014). Despite these limitations, RO is perceived as a key technology for DPR projects because of its ability to provide good removal ($>90\%$) of most contaminants found in water, including unregulated contaminants (Lee et al. 2009). The limitations of RO for DPR applications, particularly with respect to removal of pathogens, are discussed by NWRI (2015a).

For inland communities, it is advantageous to consider advanced treatment options that do not include RO (Tchobanoglous et al. 2011) because they require less energy, recover nearly 100% of the feedwater supply, and do not produce a concentrate stream (Leverenz et al. 2011). Numerous IPR installations in the United States and around the world use advanced treatment trains that include a combination of MF, advanced oxidation processes, and biological activated carbon (BAC) (Trussell et al. 2013). Such treatment is capable of achieving comparable or better removal of chemical constituents and pathogens than RO (Trussell et al. 2013) with lower costs and operational requirements. Lee et al. (2012) demonstrated that treating MF effluent with ozone and biofiltration resulted in removal of micropollutants, including pharmaceuticals and personal care products (PPCPs), comparable to that achieved with RO (Lee et al. 2012). Use of activated carbon filter media would have further enhanced removal of these micropollutants. This alternative treatment scheme had advantages over the RO system in using less energy and generating no waste (Lee et al. 2012).

An important strategy for reducing a water system’s vulnerability to trace concentrations of micropollutants is to establish an effective program to prevent introduction of these constituents into the collection system. Such a program may consist of two components: an industrial pretreatment program that would regulate materials that could be introduced into the sewers, and an aggressive public education program to remind customers about the importance of assuring the quality of their drinking water through a source protection program.

Regardless of the treatment technology used, DPR system reliability and monitoring are of critical importance to control the process and protect public health (Arnold et al. 2012; Law 2008; Tchobanoglous et al. 2011; Australian Academy of Technological Sciences and Engineering 2013). Verification and process monitoring is needed to demonstrate that the DPR system can meet regulatory requirements and removes contaminants of concern (Law 2008). One scheme for monitoring water quality is to include flow equalization and monitoring at the influent to the DPR process train, and an ESB and monitoring at the effluent end to allow identification of off-specifications water before it reaches the distribution system; individual unit process performance could be monitored as well (Tchobanoglous et al. 2011; Leverenz et al. 2011). However, further consideration of the ESB calls into question its benefits because sampling and analysis of trace chemical constituents is very costly and analyses have lengthy time requirements. Depending on the parameters of interest, analytical costs range from $100 to greater than $1,000 per sample and analyses require days to weeks to complete. To implement this monitoring strategy would therefore require that the ESB have a hydraulic residence time longer than the laboratory turn-around time for critical constituents, an impractical requirement for all but the smallest treatment plants. Instead, online analyses of indicators of process performance and measurement of surrogates for chemical and biological parameters are more likely to be incorporated in DPR plants.

The challenges of assuring the reliability and performance of treatment processes incorporated in the wastewater and advanced water treatment trains are twofold. First, as discussed, the principal risk to the consumer of reuse water is the possible presence of pathogens. There are no continuous monitors capable of detecting pathogenic organisms in drinking water. This challenge is compounded by the fact that treatment processes capable of providing 10 log removal or greater of pathogens will produce a treated water with no pathogens present nearly all of the time. Thus, the problem of pathogen monitoring is exacerbated by the need to detect a nonzero constituent when it is absent nearly all of the time. The second challenge is in detecting very low concentrations of chemical constituents. While there are reliable online monitors for a few parameters such as chlorine, analyses of most constituents of interest can only be done in a sophisticated laboratory. As mentioned, laboratory analyses of trace constituents are time-consuming and expensive; the high costs are especially problematic for small communities. The NWRI (Mosher et al. 2016) has a lengthy discussion of quality assurance and quality control strategies required for DPR projects in which indicators of process performance are used instead of actual measurements of regulated constituents.

An additional concern faced by most small community utilities is the difficulty of operating the complicated processes involved in treating wastewater for a DPR project. Utilities serving less than perhaps 5,000 people likely only have one or two operators who have limited qualifications for operating simple water and wastewater treatment plants and the distribution system. An integrated treatment system for DPR will require operators with advanced certification for both the wastewater and reuse/advanced treatment plants. In addition, current training curricula such as that developed by the State of California (California EPA 2016) do not provide training in advanced water and wastewater treatment processes such as RO or advanced oxidation processes, nor do they include training in operation and maintenance (O&M) of advanced process monitoring and control equipment such as online analyzers for total organic carbon (TOC), ozone concentration, or absorbed ultraviolet (UV) dose. Recognizing this deficiency the NWRI (2015a) proposed the creation of a new advanced treatment category for operators of DPR facilities. Small communities will face a particular challenge of competing for operators with these qualifications because large utilities are able to provide better pay and benefits in addition to support in the form of backup operators. One possibility that small utilities may consider is contracting with commercial water and wastewater O&M firms to operate their plants. However, this adds costs that they may not be able to afford.

Little research has been completed regarding communities’ ability or willingness to pay for DPR projects. As a first step toward understanding the economic feasibility of DPR in a community, cost estimates are needed for DPR compared with other water supply alternatives. Raucher and Tchobanoglous (2014) performed an analysis of the capital and O&M costs for several water supply alternatives, including DPR, using data from some notable California projects. Their analysis is summarized subsequently—first for DPR and then for other water supply alternatives. The analysis was based on actual costs and operating expenses, bid prices, and estimated operating costs from operational and proposed facilities at the $3.79\times {10}^5{\mathrm{m}}^3/\mathrm{day}$ (100 mgd) OCWD GWRS; numbers from San Diego’s IPR project and the Santa Clara Valley Water District’s advanced treatment process were used for comparison and found to be similar (Raucher and Tchobanoglous 2014).

The costs for DPR were broken down into costs for advanced treatment (including MF, RO, and advanced oxidation), water conveyance, and concentrate management, as follows. Influent water characteristics, type and number of technologies included in the treatment train, and overall plant capacity may cause some variation in costs (Raucher and Tchobanoglous 2014):

The analysis by Raucher and Tchobanoglous (2014) of combined capital and operating costs for advanced treatment, water conveyance, and brine management found that the total annualized cost for DPR ranged between $\$0.67/{\mathrm{m}}^3$ and $\$1.63/{\mathrm{m}}^3$ ($820/AF and $2,000/AF, or $2.50 to $6.00 per thousand gallons). The advanced treatment costs were estimated based on facilities that include RO. As mentioned previously, Lee et al. (2012) suggested an advanced oxidation process that will provide equivalent treatment, but without generating a brine stream and using less energy, and therefore possibly resulting in significantly lower costs for advanced water treatment and waste management.

The costs for potable reuse and alternative new water supplies are difficult to compare because of site-specific characteristics and conditions, differences in facility sizes, regulatory requirements and uncertainties, and variability in reporting of existing cost data, among other issues. Acknowledging these difficulties, Raucher and Tchobanoglous (2014) have estimated ranges of probable combined capital and operating costs to facilitate comparison of various options. Table 2 shows a summary of likely ranges in total annualized cost per cubic meter for various new water supply options. All costs are based on facilities in California (Raucher and Tchobanoglous 2014). As another point of comparison, also using California data, Loaiciga (2015) estimated the cost of reclaimed water and the cost of seawater desalination to be 1.5 and 2.5 times the cost of water production at the water treatment plant, respectively.

Published cost estimates for DPR projects do not reflect the quality of the feedwater. It is likely that enhanced wastewater treatment would reduce the cost of advanced treatment processes. However, whether there would be an overall cost savings would depend on the relative costs of enhanced wastewater treatment compared with advanced treatment for DPR.

The capital and O&M costs for a very small DPR project have been summarized by NWRI (2015b). This study reported on the Cloudcroft, New Mexico, project in which wastewater was treated by RO, UV and oxidation, and granular activated carbon (GAC) adsorption, and then combined with groundwater and subjected to additional treatment by membrane filtration prior to introducing it to the community’s distribution system. The total capital costs were estimated at $5,000,000, which included $1,400,000 to upgrade an antiquated WWTP to the membrane biological filtration (MBR) process. The annual O&M costs were estimated to be $\$\mathrm{350,000}/\mathrm{year}$, which included salaries for one and a half operators; a half-time operator position was recommended to provide backup support when the primary operator is not available. Operators for the DPR plant must have advanced certification in both wastewater and water treatment with consequent increased salary and training costs. The O&M costs also include a sampling and analysis program and an equipment repair, rehabilitation, and replacement fund budgeted at 10% of the total equipment cost.

The capital and O&M costs of the Cloudcroft process illustrate the complex economic considerations of a DPR project. Cloudcroft is a small resort community with approximately 700 permanent residents that can swell to more than 3,000 people on holiday weekends. The initial cost of the project was feasible because more than 90% of the capital cost of the treatment plants was provided in the form of grants from federal and state agencies, justified in part by the fact that this was the first project of its kind in the country (it was originally designed and construction started in 2005). However, as the project has neared completion, the community has recognized that the very large O&M costs may not be acceptable because they would require almost doubling the community’s already high water and wastewater utility rates. And yet, because there is no alternative source of water, the community is divided as to whether the costs are justified.

The O&M costs alone for Cloudcroft are more than $\$2.44/{\mathrm{m}}^3$ ($3,000/AF)—this is significantly higher than the high end of Raucher and Tchobanoglous’s total annualized costs for DPR, which were estimated based on data from larger facilities. If the community were to pay for the capital costs using low-interest loans of 2% over 30 years, the capital costs would be just under $\$1.63/{\mathrm{m}}^3$ ($2,000/AF). This illustrates the economic disadvantages of very small treatment plants. In considering funding alternatives it is important to recognize that while financial assistance is sometimes available to communities in the form of federal or state grants or loans, this funding is only for capital costs. Almost without exception financial assistance is not available for O&M costs.

Bischel et al. (2012) found that 59% of recycled water managers surveyed viewed the “issue of who pays for program capital or operating costs” as a hindrance to water reuse. Due to federal and state funding declines, future water projects will likely be substantially funded by customers through increases in utility rates and fees (Thacher et al. 2010). However, a community likely will want to maximize its external financial assistance options prior to resorting only to rate increases and fees.