Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 The Costs and Benefits of Desalination The promise of desalination to rid the world of water scarcity has been touted for nearly 50 years. During this period, public and private investment in developing and improving desalination technology has totaled more than a billion dollars (see Chapter 2). Although much pro- gress has been made and there have been successes in developing water supplies in very dry locales and regions, the promise remains largely un- fulfilled. The explanation lies with the fact that, although the process costs have been reduced, the total costs of desalination, including the costs of planning, permitting, and concentrate management, remain rela- tively high, both in absolute terms and in comparison with the costs of other alternatives. In assessing the future prospects and promise of desalination tech- nology, it is particularly important to examine the current and prospec- tive financial and economic circumstances that are likely to surround the technology as it develops. An examination of the structure of desalina- tion costs and of the determinants of those costs is important in identify- ing areas in which research might be pursued with the greatest effect. A consideration of the availability and costs of alternative supplies helps to place the future role of desalination in perspective. Finally, issues of reli- ability, water quality, and environmental impacts need to be understood if the costs and benefits of desalination technology are to be broadly un- derstood. All of these topics are considered in this chapter. GENERAL CONSIDERATIONS In assessing the economic desirability of any water supply project or facility, several general considerations and economic principles need to be understood. These principles, including the relative nature of cost- benefit comparisons, financial versus economic costs, and public versus private costs and benefits, are discussed below. 147
148 Desalination: A National Perspective Relative and Absolute Costs When evaluating water supply alternatives, it is important to under- stand that the benefits and costs should be evaluated in a relative rather than an absolute fashion. When the focus is on costs, for example, the absolute cost of the facility or project in question has little meaning unless it is compared with the costs of other alternatives for accomplish- ing the same purposes, if such alternatives exist. Thus, from an economic standpoint, employment of any desalination technology to augment water supplies will be attractive in circumstances where the alternatives are more costly or nonexistent, and unattractive in circumstances where less costly alternatives are available. Economic and Financial Costs In any assessment of economic viability, it is important to understand that economic costs and benefits will often differ from financial costs and benefits. Rarely, if ever, will a financial analysis be the same as an economic analysis. Economic analyses, including the analyses of costs and benefits, account for all of the costs to whomever they may accrue, irrespective of whether these costs are characterized by market-generated prices. Financial accounting and, consequently, financial feasibility are directly concerned with the availability and the costs of funds. Thus, for example, a utility or private-sector entity may be able to attract capital at a reasonable interest rate for the purposes of constructing a project if the returns will be sufficient to pay the interest charges, repay the principal of the loan, and allow appropriate return to management, labor, and other factors of production. In these circumstances, the project would be said to be financially feasible. There may, nevertheless, be costs or benefits that cannot (and need not, in this case) be captured because they are not internalized in the utilityâs cost stream or, in the case of benefits, not priced and subject to market-like exchanges. Consider a situation in which there are large environmental costs of a water supply project that do not have to be addressed by law or regula- tion. Because these costs do not have to be addressed, they do not change the financial attractiveness of the project, but the environmental impacts are costs nonetheless, which must be borne by some individual or group. Hence, they must be counted as economic costs even though they do not affect the financial feasibility. Similarly, if some of the benefits cannot be captured by the projectâs operator, those benefits are economic bene- fits and must be counted as such in an economic analysis even though they will have no effect on the financial feasibility of the project.
The Costs and Benefits of Desalination 149 Public and Private Costs It is also important to draw the distinction between private costs and public costs. Private costs are internalized within the operation of the project and must be borne by the utility or business operator. They in- clude costs such as wages, interest payments, energy, and equipment. Public costs, by contrast, are real costs that a business operator can es- cape but that will have to be borne by the public at large. In the absence of regulations, business operators usually have no incentive to mitigate or defray the costs of environmental damage that may accrue offsite as a consequence of their operations. In these circumstances, the costs will be borne by the public at large and are correctly accounted for as public costs. A similar distinction should be made on the benefit sideâthe dis- tinction between public benefits and private benefits. The benefits of a project are private benefits to the extent that they can be captured by the producer. When the producer sells the product, private benefits are cap- tured through the price of the sale and are fully appropriable by the pro- ducer. By contrast, when benefits are jointly conferred on consumers and cannot be fully captured by the producer, they are said to be public bene- fits. Stated differently, if a producer is unable to withhold the benefits of the good or service from a consumer because the consumer refuses to pay for it, the benefits are public in nature. Such benefits tend to be un- dersupplied by producers, from a strict efficiency perspective, because they cannot capture those benefits fully. Environmental services (e.g., the capacity to purify water and air, to provide environmental stability, to protect against disease) are examples of public benefits. Joint Costs Finally, it is important to understand the notion of joint costs and the problems they create for the accounting and allocation of costs. Fre- quently the financial and economic costs of desalination can be reduced by combining desalination operations with the production of some other water-related goods and services. Power is the primary example; these dual-purpose plants are termed cogeneration facilities. Some elements of the plant will jointly serve both the production of power and the produc- tion of water. The costs of these elements that contribute to both pur- poses are defined as joint costs. In theoretical terms, it is not possible to allocate joint costs back to individual purposes in any but an arbitrary way. That is, it is not possible to partition joint costs according to the contribution to the marginal product of each of the multiple purposes. There are principles and informal rules and practices for allocating joint
150 Desalination: A National Perspective costs, including the separable cost/remaining benefit method, methods based on energy production and consumption, and others, but these are all arbitrary from an economic standpoint (Friedman and Moulin, 1999). Thus, while there may be controversy associated with allocating a dis- proportionate share of the costs to one product (such as power) rather than another (such as water), there is no theoretical principle or analytical reason which bars it. The problem of allocating joint costs makes it particularly hard to analyze the costs of thermal desalination plants in a consistent and sys- tematic way. Many thermal desalination facilities are co-located with power plants so that the waste heat from the power plant can be utilized in the desalination process, and others are designed as cogeneration fa- cilities with large components of joint costs. The prevalence of joint costs in thermal facilities confounds efforts to allocate costs among the different inputs and to analyze the sensitivity of total costs to changes in various components of cost. For this reason, thermal technologies are largely omitted from consideration in the following section on the struc- ture of costs. A fully adequate economic assessment of a water supply project or strategic plan would be conditioned by and account for the considera- tions identified earlier. Benefit-cost analyses are relative, and the results will vary from situation to situation. Both economic analyses and finan- cial analyses have important roles in any project analyses, but one should not be confused with the other. Where public costs and benefits pre- dominate, there may be a case for governmental involvement. Where benefits and costs are largely private, the case for governmental in- volvement will be far weaker (Cornes and Sandler, 1996; Oakland, 1987). Although each project will have its own particular characteristics, it is possible to identify in some detail the elements and structure of costs and benefits. In the next section, the structure of costs for desalination facilities is considered and enumerated. THE STRUCTURE OF DESALINATION COSTS A summary of what is known about the current status and trends in desalination costs are presented in this section, followed by a detailed analysis of the determinants of those costs. An analysis of the structure of desalination costs is important to identifying research areas with the greatest potential for reducing costs.
The Costs and Benefits of Desalination 151 Difficulties of Estimating and Comparing Costs The cost to treat seawater or brackish waters to produce potable wa- ter is a function of numerous variables, and the components of these costs are frequently difficult to ascertain precisely from the literature. Although selling prices are reported for many international public-private projects, data on the components of the total cost and price are not re- ported and not available because they are regarded as confidential infor- mation by firms in the business and because of regulatory and public policies. The confidential nature of this information reflects the competi- tive nature of the international water business. Although water rates (or tariffs) are public information, these rates reflect the project-specific evaluation criteria, scope of work, and technical process impacts based on local conditions and requirements; therefore, they are not consistent from country to country or place to place. Consequently, tariffs do not provide a simple indicator for cost comparisons. Different project costs are also difficult to compare because virtually every desalination plant has its own unique design and site conditions and its own unique financing package. Table 6-1 provides an example of such comparative costs for three projects: the desalination facilities built and operated by the Inland Empire Water Agency in southern California for the purpose of desalting brackish water; the brackish water desalina- tion project in Texas developed by the El Paso Water Utilities in coop- eration with the U.S. Army (see Box 5-2); and the Tampa Bay seawater desalination plant in Florida. Although it is tempting to draw conclusions from comparisons such as these, particularly with respect to the sensitiv- ity of costs to source water salinity, great care must be exercised. For example, sometimes there are financing offsets that lower the apparent costs to the end users. Both Inland Empire and Tampa Bay will receive such offsets ($0.20/m3 or $250/acre-foot [a.f.] to Inland Empire from the Metropolitan Water District of Southern California; $0.09/m3 or $111/a.f. to Tampa Bay from the Southwest Florida Water Management District), although these offsets are not factored into the costs reported in Table 6- 1. Reviews of published data on costs can be confusing because costs are rarely reported consistently and some cost parameters are not re- ported at all. Additionally, the underlying assumptions (e.g., project life, project size) may differ and sometimes remain unstated (Almulla, 2002; Busch and Mickols, 2004; Dreizen, 2006; Frenkel, 2004; Hinkebein and Price, 2005; Miller, 2003). For example, some cost data include distribu- tion costs while others are for costs at the plant boundary. Miller (2003)
152 Desalination: A National Perspective TABLE 6-1. Financial Costs from Three Desalination Facilities Inland Empire El Paso/Ft. Bliss Tampa Bay Feedwater total 800-1,000 1,200-15,000 26,000 dissolved solids (TDS) (ppm) Average output 27,000 100,000 95,000 (m3/d) Operations and 0.31 0.45 maintenance 3 ($/m ) Admin and 0.029 0.02 3 general ($/m ) Capital 0.19 0.22 consumption ($/m3) Fixed costs 0.10 0.15 ($/m3) 3 Total ($/m ) 0.63 0.43 0.83 NOTE: The figure for the El Paso/Ft. Bliss project includes distribution costs whereas the figure for Tampa Bay does not. SOURCE: R. Atwater, Inland Empire Water Agency, personal communication, 2007; E. Archuleta, El Paso Water Utilities, personal communication, 2006; J. Maxwell, Tampa Bay Water, personal communication, 2007. summarizes costs reported in the published literature for a variety of de- salination projects and notes that the numbers can only be used as a rough guide because they are not calculated on a consistent basis. De- spite the limitations of proprietary data, there is a wealth of information available on the nature of desalination costs and on the ways in which those costs are determined. If this desalination cost information were compiled on a reasonably consistent basis, it would be particularly im- portant to water planners who are concerned with problems of meeting growing water demands in the future. However, there is so much varia- tion in the circumstances of individual projects, as well as in the bases upon which reported costs have been calculated, that the resulting num- bers must be interpreted with great care and strict comparisons are not usually possible. The information and analysis that follow result from efforts to pro- vide a detailed view of the various cost components that make up the costs of producing freshwater from seawater and brackish water sources using membrane technology. The conclusions that can be drawn from the analysis are limited by the lack of consistency and detail in the data. The conclusions are also limited by the committeeâs inability to analyze the
The Costs and Benefits of Desalination 153 costs of concentrate management, including environmental costs, in any- thing but a general way. As noted in Chapter 5, the environmental costs of concentrate discharge are not well understood, and available concen- trate management alternatives, and their associated components of costs, vary greatly from situation to situation. In this report, desalination costs that do not include the costs of concentrate management are specified as desalination production or process costs. It is also important to acknowl- edge that these challenges make it virtually impossible to generalize in any meaningful way about the structure of costs of inland facilities that utilize brackish feedwaters. In general, it is known that energy use and costs vary directly as a function of the concentration of total dissolved solids (TDS) of the feedwater (see Figure 4-10). Thus, where other things are equal, brackish waters should be less costly to desalinate than seawater. The problem is that brackish water desalination costs vary sig- nificantly, not only with the TDS of the feedwater but also with the costs of concentrate management, which can be very high in inland situations where brackish water desalination is otherwise attractive. For these rea- sons, brackish waters are not treated with the same level of detail or rigor in the cost analyses that follow. Reported Desalination Costs In general, the unit costs of producing freshwater from seawater have been reported in a range running upward from approximately $0.64/m3 ($800/a.f. or $2.46/thousand gallons [kgal]) (see, for example, Miller, 2003). Many of these estimates, particularly those at the lower end of the range, include subsidies or do not account fully for all costs (Miller, 2003). Some large seawater desalination plants appear to be operating in the range of approx $0.80/m3 ($3.06/kgal; $1000/a.f.) but new facilities are being proposed with substantially higher costs due to site-specific considerations (GWI, 2006a; Miller, 2003). The only operational exam- ples where substantially lower production costs have been achieved en- tail the use of membrane technologies and brackish feedwaters with TDS concentrations that are significantly lower than that of seawater. Treat- ment systems for such lower-salinity groundwaters can often produce water for less than half the costs of treating seawater. However, when low-cost concentrate management methods are not available, brackish groundwater desalination costs can reach or exceed seawater desalination costs. It is important to note that desalination costs found elsewhere in the world may be lower than those that can be realized in the United States because the permitting costs and the costs of meeting environ- mental regulations tend to be high in the United States.
154 Desalination: A National Perspective Membrane and thermal processes are both used widely in municipal- scale desalination plants worldwide and, as discussed in Chapter 4, each technology has strengths and weaknesses and differing operating condi- tions under which one or the other may be economically optimal. Ther- mal desalination systems consume more energy than reverse osmosis (RO) systems and are more capital intensive. Nevertheless, thermal sys- tems can use more diffuse or low-grade forms of energy (i.e., low- pressure steam) whereas membrane systems rely solely on electricity as an energy source. Global Water Intelligence (GWI, 2006a) reports the capital costs of seawater desalination by multi-effect distillation (MED) and multistage flash (MSF) distillation to be 1.5 to 2.0 times the capital costs of RO desalination systems, respectively. Additionally, GWI (2006a) approximates the costs for the seawater desalination process by RO to be $0.61/m3 as compared to $0.72/m3 for MED and $0.89/m3 for MSF. The breakdown for these costs is shown in Table 6-2. These costs are based on a system scale of 100,000 m3/day; a nominal interest rate of 6 percent; $450 element cost; $0.05/kWh energy cost; assumed electric- ity use of 4.5, 4.0, and 1.25 kWh/m3 for RO, MSF, and MED, respec- tively; and a 20-year capital-payback period. The costs for seawater de- salination by RO are slightly lower than costs reported from actual instal- lations. Most likely this is due to the favorable interest and energy costs used in the preceding analysis. Additionally, the calculated total costs for thermal technologies are likely exaggerated because offpeak electricity costs, cogeneration, or the use of low-grade or waste energy are not con- sidered in this analysis. If low-cost, dispersed sources of energy are available or energy can be jointly used with other purposes, seawater TABLE 6-2. Comparative Total Cost Data for the Desalination Process for 3 100,000 m of Seawater by Reverse Osmosis, Multistage Flash Distillation, and Multi-Effect Distillation SW RO SW MSF SW MED Annualized capital costs 0.15 0.29 0.22 Parts/maintenance 0.03 0.01 0.01 Chemicals 0.07 0.05 0.08 Labor 0.10 0.08 0.08 Membranes (life not 0.03 0.00 0.00 specified) Thermal energy 0.00 0.27a 0.27 a Electrical energy 0.23 0.19 0.06 ($0.05 k/Wh) Total ($/m3) 0.61 0.89 0.72 a The costs of thermal energy are likely exaggerated because offpeak electricity costs, cogeneration, or the use of waste energy are not con- sidered in this analysis. SOURCE: GWI (2006a).
The Costs and Benefits of Desalination 155 desalination using thermal technologies becomes more cost effective. Leaving aside situations in which energy can be obtained cheaply, the capital and operating costs of thermal systems appear significantly higher than the best-available membrane technology. As previously discussed, concentrate management costs can be widely ranging, based on the alternatives available, the volume and salin- ity of the concentrate, and other site-specific factors (see Chapter 4). There are a number of examples which illustrate the potential magnitude of concentrate management costs. One study, in which a zero liquid dis- charge (ZLD) option was evaluated as part of a desalination treatment train for an inland municipal application, estimated the cost of the ZLD steps as almost twice as large as the cost of the primary desalting step. The total desalination cost was estimated at $1.44/m3, of which the water production step accounted for $0.50/m3 while concentrate management costs amounted to $0.94/m3 (Sethi et al., 2007). For another project, cur- rently under construction and scheduled to be online in 2008, the bid construction cost for a 3,000 m3/day (0.8 million gallons per day [MGD]) RO facility was about $26 million, of which approximately $7 million was accounted for by the costs of a brine concentrator (Yallaly et al., 2007). Similarly, the brackish water desalting facility near El Paso, Texas, entailed significant costs for concentrate management. The pro- jected cost of the facility was $0.44/m3 ($1.64/kgal) of blended water including amortized capital, operation, and maintenance costs (assuming an energy costs of $0.07/kWh). Of the capital cost of $87 million, 26 percent was allocated for the concentrate disposal wells and lines. The estimated annual operating and maintenance costs for the concentrate management were lower, representing 0.04 percent of the estimated $4.8 million costs (E. Archuleta, El Paso Water Utilities, personal communi- cation, 2006). Comparing Desalination Costs against Other Alternatives When making water policy decisions, desalination costs need to be compared with the costs of other water supply or demand options avail- able in a given locale. While there is some prospect that the costs of pro- ducing freshwater from seawater may come down, the existing evidence suggests that they are still quite high when compared with the costs of alternatives in most locales (see Figure 6-1 and Box 6-1). Specifically, there are many instances in which demand management measures can make water available for new uses at costs that are signifi- cantly lower than the costs of desalination (See Box 6-1). Market-like transfers of water, in which water is reallocated away from relatively low-valued uses to relatively high-valued new uses, can also be less
156 Desalination: A National Perspective 1.2 1 0.8 Cost in dollars/m3 0.6 0.4 0.2 0 Seawater Desal. Reclamation Groundwater Core Transfers Conservation FIGURE 6-1. Financial cost ranges for a subset of available water alternatives for San Diego Water Authority in dollars per cubic meter, based on data from Robert Yamada, San Diego Water Authority, personal communication, 2006. These costs are not assumed to be representative of the nation as a whole but are pro- vided as an example of a cost comparison among alternatives for one commu- nity. The high range for reclamation accounts for the additional costs of con- structing more substantial conveyance facilities. expensive than desalting in many instances. Although these methods re- sult in the reallocation of some water among uses while desalination cre- ates ânewâ water, the end result of the reallocating alternatives is a more efficient pattern of water use in which water itself is more productive than it was prior to the reallocation (Colby, 1997; NRC, 1992). Comprehensive cost comparisons are difficult to make, because of uncertainty, rapidly changing technologies and prices, noncomparable variables, and other factors. Recent reviews (Chaudry, 2007; Cooley et al., 2006) suggest that, although energy costs have remained relatively stable over time (see Box 6-2), in the past 2 years energy price increases have begun to outpace desalination cost reductions due to improvements in technology. The problem is complicated not only by virtue of the fact that the total costs of alternatives vary with site-specific conditions but also by the fact that the total costs of desalination facilities differ from situation to situation. Factors such as the size of the plant, the salinity of the feedwater, the temperature of the feedwater, the prevailing costs of energy in the region, land costs, and the investment and operating costs
The Costs and Benefits of Desalination 157 BOX 6-1 Costs of Conservation Economists use a variety of methods to compare the relative costs of water supply and management alternatives, including cost-effectiveness, marginal costs, rates of return, and more. These different methods often produce different results depending on the point of view or perception of the target audience, the data used, and the assumptions made. For example, the cost of a conservation measure may differ substantially from the points of view of the water user as op- posed to the water supplier. In addition, the economic benefits may also differ substantially from different perspectives, even assuming that all benefits can be put in comparable terms. A typical approach is to use cost-effectiveness analysis to compare a unit cost of alternatives, for example in dollars per cubic meter of physical water sup- ply or conservation reduction. A water conservation alternative is considered cost-effective when the unit cost of conservation (sometimes called âthe cost of conserved waterâ) is less than the unit cost of the next unit of additional supply. A challenge in determining the costs of conserved water is the difficulty of identifying and quantifying many different cost factors. For example, the cost of conserved water depends on the capital cost and lifetime of the conservation technology (a low-flow toilet or front-loading washing machine, for example), but it also depends on the amount and cost of energy savings that might result, the amount and cost of changes in wastewater treatment costs resulting from a de- crease in overall water used, and the value of restored ecosystem water flows, if any, among other factors. Replacing lawn with xeriscaping may save water, and it may also save labor and energy costs associated with reduced frequency of mowing grass. Properly identifying and calculating all of the associated benefits of water conservation options is important in order to properly compare among alternative water choices. It is also important to distinguish between natural and accelerated replace- ment of water-using options. Natural replacement refers to the replacement of water devices due to age or failure; accelerated replacement refers to replace- ment of a device before the end of its natural lifetime specifically in order to re- duce water use. Perspective is important. For water conservation analyses, it has been ar- gued that the proper perspective is the viewpoint of the water consumer, as op- posed to the more traditional perspective of the water supplier (see, for example, Chapter 5 in Gleick et al., 2003). Analyzing the cost-effectiveness from the per- spective of the consumer requires calculating the cost of conserved water based on the investment required of the consumer and any changes in operation and maintenance costs that the consumer would experience from the investment. These costs must then be compared to the marginal costs of alternatives. This approach fairly compares the costs and benefits to the water supplier (which are passed on to the consumer) with the costs and benefits experienced by custom- ers apart from what they pay for water services alone, which often fail to account for associated benefits such as energy, labor, or other savings. When the cost of conserved water from a specific measure is less than the cost of water supply displaced by conservation, the customer and the water utility (collectively) will make money if the measure is implemented (Gleick et al., 2003). If water rates and utility rebates do not accurately and fully reflect the continued
158 Desalination: A National Perspective marginal costs of supply, the net benefits to consumers and the utility may not be apparent. This can be corrected by adjusting water rates. It is thus important to properly identify both the marginal variable costs and marginal capital costs of water supply displaced by efficiency improvements. A comprehensive analysis of water conservation alternatives was produced in 2003 by the Pacific Institute for California residential, commercial, and indus- trial efficiency options. The results of that study were presented as âsupply curves for conserved waterâ in a form similar to that used in the field of energy economics (CPUC, 2001; Koomey et al., 1995). They concluded that as much as 2.5 billion m3 of water could be conserved for less than $0.50/m3 of waterâa benchmark chosen because it is a commonly used measure for the marginal cost of new water supply and because no significant new water project in California has been proposed below that cost. They note, however, that at least 810 million m3 of conservation is cost-effective to conserve in the residential sector even if 3 new supplies cost only $0.05/m (Gleick et al., 2003). How is this low cost possible? It turns out that the estimated costs of some conserved waters are actually negative. This means that water could be free (have zero cost) and consumers would still save money by implementing the conservation option. For these options, nonwater benefits (such as savings in labor, energy, wastewater treatment, or fertilizer or pesticide use) are sufficient by themselves to pay for the efficiency improvements. For example, the cost of replacing inefficient toilets in restaurants and supermarkets was estimated to be on the order of $0.08/m3, but the cost of replacing low-flow showerheads in hotel rooms was estimated to be $-0.65/m3 because of the substantial energy savings associated with the reduction in hot water use. Similarly, improving housekeeping practices and using a dry-clean-up system in meat processing plants is not cost- effective based on the water savings alone ($1.10/m3), but becomes highly cost- effective ($-0.50/m3) when savings in wastewater treatment costs from biological oxygen demand fees are included. Water-efficient dishwashers in restaurants become tremendously cost-effective when the costs of energy and chemicals saved are properly included (Gleick et al., 2003; see, especially, Table 5-15). of particular environmental mitigation facilities all combine to make the total cost of each facility different. Examining the Determinants of Costs of Reverse Osmosis Desalination Given the difficulties in making broad conclusions based on reported desalination costs, the committee took an alternative approach and con- ducted an original analysis that illuminates the key determinants of de- salination cost and examines the sensitivity of cost to variations in these determinants. This analysis was conducted to illuminate the best oppor- tunities for additional cost reductions, and thereby to serve as a frame- work for focusing future research and development efforts aimed at low- ering the costs of desalination through technology innovations.
The Costs and Benefits of Desalination 159 Sensitivity Analysis Methods The following analyses are based on a set of baseline capital and op- erating cost data that were obtained for actual operating RO desalination systems of three different scales (38,000 m3/d, 189,000 m3/d, and 380,000 m3/d) (M. Shah, GE Global Research, personal communication, 2007). Energy costs and membrane life in these baseline data were $0.07/kWh and 5 years, respectively. Process recovery was set at 40 per- cent, and the operating pressure was set at an average of 5,516 kPa (800 BOX 6-2 The Importance of the Price of Energy The costs of desalination have always been sensitive to the price of energy because desalination processes, whether thermal or membrane, are energy in- tensive. Although it is possible that the energy intensity of membrane processes can be reduced, as discussed in this report, the likelihood is that the price of en- ergy will continue to play a major role in the costs of desalination for the foresee- able future. The costs of electricity for the 46-year period 1960-2006 are shown in Figure 6-2. While it is true that electricity costs are not always the pertinent costs for thermal technologies, the trends in these costs are generally represen- tative of energy costs for thermal technologies and reasonably exact costs for membrane technologies. The data in Figure 6-2 reveal several important conclu- sions. First, and somewhat surprisingly, the costs of electricity in the period 2004- 2005 were not higher than the costs in 1960 in constant dollar terms. This means that while there may have been ups and downs, the general electricity price pic- ture for the period was relatively favorable for electricity users. Second, although there was a trough in the early 1970s and a peak in the early 1980s, electricity costs have been trending down since about 1985. During the intervening period of about 20 years, the costs of electricity became increasingly favorable in terms of their impact on desalination technology. Third, recently the costs of electricity have begun to trend upward. The price of oil on international markets has risen, and this suggests that there is reason to expect that electricity costs will continue to rise for the indefinite future. Although electricity costs are not directly tied to the costs of oil, because considerable quantities are generated from hydroelectric and coal-fired power plants, oil prices will provide a rough approximation of what is likely to happen to the cost of electricity. A rising trend in electricity prices is likely to affect the costs of desalination adversely unless there are parallel offset- ting reductions in other costs. As noted in this chapter such reductions are quite possible. The costs of energy, adjusted for inflation, are presented in Figures 6-2 and 6-3. These data show that real oil prices peaked in the late 1970s and early 1980s and then were relatively stable until 2002, when they began to rise mark- edly. Similarly, Figure 6-2 shows that electricity prices, which are typically less volatile than world oil prices, began to rise around 2002. If these price increases continue, they will slow the rate at which desalination costs can decline due to technological improvements. continued
160 Desalination: A National Perspective 12 10 Total Cents per Kilowatt Hour 8 6 4 Industrial 2 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Year FIGURE 6-2. Inflation-adjusted electricity costs in the United States from 1960 to 2006 for all sectors and for the industrial sector alone. Prices are in cents per kilowatt-hours in constant 2000 dollars and include relevant taxes. SOURCE: Data from U.S. Energy Information Administration (2007a). $100 $90 $80 $70 2006 Dollars per Barrel $60 $50 $40 $30 $20 $10 $0 $- Jan-74 Jan-78 Jan-82 Jan-86 Jan-90 Jan-94 Jan-98 Jan-02 Jan-06 Year FIGURE 6-3. Inflation-adjusted crude oil price in the United States from 1974 to 2007. Prices are in dollars per barrel constant 2006 dollars. SOURCE: Data from U.S. Energy Information Administration (2007b).
The Costs and Benefits of Desalination 161 psi). All cases reflect the use of the best commercially available mem- brane and energy recovery technologies. Subsidies, land costs, and dis- tribution costs have not been included, nor have the costs of concentrate management. This analysis did not use a baseline cost model of the de- salination processâthere are other tools available to estimate desalina- tion costs developed by various companies and agencies in the industry, but there are limitations to each of them. Instead, the committee con- cluded that the transparency of actual operating data from existing de- salination plants would serve as a better baseline for this sensitivity analysis. Brackish water desalination analyses were conducted where feasible and where trends were expected to differ from the seawater de- salination analysis. In this analysis, it is important to be precise about terminology. Total process costs include all of the costs incurred (or estimates of those costs) during construction and the productive lifetime of the project (ex- clusive of the costs of concentrate management). Capital costs include all of the costs of capital, including principal and interest. Capital costs typi- cally do not vary during the life of the project and are treated as fixed costs. Operating costs include the variable costs of operation over time and are expressed as annual operating costs. For this analysis, it was nec- essary to put capital costs and operating costs on a common footing even though the former are fixed and the latter are variable. Annualized capital costs were determined by calculating the yearly capital consumption costs and interest charges as a function of different assumptions about equipment lifetime (i.e., depreciation schedule) and the cost of money (i.e., interest rate). Annualized capital costs were added to the annual operating cost to obtain the annual cost. Annual costs can be multiplied by the life of the project to obtain total project costs (exclusive of con- centrate management). The annual costs per cubic meter are derived by dividing annual costs by the annual production of the plant. Most of the analysis is presented in terms of total annual costs, although a few key factors affecting capital costs are discussed separately. The sensitivity analysis was performed on these empirical data by linearizing the cost data associated with each respective variable. For example, in the analy- sis of the effect of membrane life on desalination cost, as membrane life was increased to 10 years from the baseline life of 5 years, costs per cu- bic meter contribution from membrane life were multiplied by 5/10 (i.e., 0.5). When membrane life was decreased to 1 year from the 5-year base- line, the costs per cubic meter contribution from membrane life were multiplied by 5/1 (i.e., 5). All other costs that were not affected by mem- brane life were held constant. All of the cost data were then normalized to the actual baseline cost data for the conventional pretreatment sea- water desalination case supplied for the analysis so that the results re- ported are relative costs per cubic meter. The results are expressed in
162 Desalination: A National Perspective relative terms, so that the cost sensitivities could be more easily com- pared (as percent change) across the different variables and to eliminate the effects of different absolute magnitudes. All of the data from the sen- sitivity analysis are supplied in tabular form in Appendix C. Impact of Water Source, Scale, and Process Design on Capital Costs Initially, it is important to acknowledge that membrane desalination facilities exhibit very real economics of scale in both seawater and brack- ish water applications. Data on capital costs of seawater desalination as a function of plant size are presented in Figure 6-4. As the scale of the sys- tem is increased from 38,000 to 380,000 m3/day (10 to 100 MGD), capi- tal costs, expressed as cost per cubic meter of product water, decline by 20 percent. Other published data also illustrate the impact of scale on capital costs, especially as the scale is decreased below 38,000 m3/day where the costs rise even more dramatically (USBR, 2003). When feed- waters are brackish, cost also declines as plant scale increases, albeit to a smaller degree (<10 percent). Data also are presented to illustrate how capital costs vary with dif- ferent pretreatment operations included in the total treatment system. The marginal increase in the capital costs of a system with an ultrafiltration (UF) or microfiltration (MF) pretreatment process compared to a conven- tional pretreatment process should be noted (Figure 6-4). The significant benefit of the UF/MF-based pretreatment is realized through reduced operating costs. As shown in Figure 6-5, the annual costs for a seawater RO system with UF/MF pretreatment are projected to be approximately 5 percent lower than one with a conventional pretreatment system. Care should be taken to account for these sorts of trade-offs between capital and operating costs. Impact of Critical Operating Variables on Annual Costs Annual costs, which are made up of annual operating costs and an- nualized capital costs, are a function of many variables. The key vari- ables for RO include the quality of the water source (measured in terms of TDS or salinity concentration, total suspended solids, and organic fou- lants), the scale of the plant, the process, membrane life, operating pres- sures, and the cost of funds. The two largest components of annual costs in RO desalination of seawater are the cost of energy to operate the plant and the annualized capital costs, which include the annual repayment of principal and the interest payment (see Table 6-2 and Figure 6-6). The
The Costs and Benefits of Desalination 163 140% 2-Stage Seawater RO with Conventional Pretreatment 120% 2-Stage Seawater RO with UF/MF Pretreatment Brackish Water RO with Conventional Pretreatment 100% 3 Relative Capital Cost per m 80% 60% 40% 20% 0% 0 50 100 150 200 250 300 350 400 Plant Output Size (1000 m3/day) FIGURE 6-4. Relative capital costs per cubic meter for seawater and brackish water RO desalination according to facility size. Brackish water is 1,000 mg/L TDS. 140% 2-Stage Seawater RO with Conventional Pretreatment 120% 2-Stage Seawater RO with UF/MF Pretreatment Brackish Water RO with Conventional Pretreatment 3 Relative Annual Cost per m 100% 80% 60% 40% 20% 0% 0 50 100 150 200 250 300 350 400 3 Plant Output Size (1000 m /day) FIGURE 6-5. Effect of facility size, water source, and pretreatment process on relative annual costs per cubic meter for RO plants. The baseline assumptions for this scenario are as follows: energy costs are constant at $0.07/kWh; mem- brane life is assumed to be 5 years; nominal interest rate is 5 percent; deprecia- tion period is 25 years.
164 Desalination: A National Perspective Total Maintenance 6% Annualized Capital Costs Total Labor 37% 4% Energy ($0.07/kWh) 36% Chemicals Membranes 12% (5 yr life) 4% Filters 1% FIGURE 6-6. Annual cost breakdown in a 189,000 m3/day seawater RO plant with conventional pretreatment. The baseline assumptions for this scenario are as follows: energy costs are constant at $0.07/kWh; membrane life is assumed to be 5 years; nominal interest rate is 5 percent; depreciation period is 25 years. specific impacts of these variables are described in the following para- graphs. Water Source. The annual costs of membrane desalination plants are very sensitive to the salinity content of the source water (Figure 6-5). In this analysis, energy costs, membrane life, and the cost of money are held constant. As a general rule, it will cost 50 percent more per cubic meter to produce freshwater from seawater as opposed to brackish water with a salinity of approximately 2,000 ppm. The sensitivity of operating costs to the salinity content of the source water supports the general proposition that, when concentrate management costs are ignored and all else is equal, it will virtually always be cheaper to desalinate brackish water than seawater when using membrane technologies. As discussed previously, concentrate management costs can be very high, where deep- well injection and ZLD increase the total costs of brackish water desali- nation by 50 to 200 percent above the desalination process costs alone (Mickley, 2007). With thermal techniques, operating costs are not sensi- tive to the salinity of the water. Therefore, thermal technologies are rarely used to desalinate brackish waters. Cost of Money. It is important to recognize that, as with any capital investment, interest costs (or the costs of money) are invariably one of the larger components of total project cost. For example, the total repay-
The Costs and Benefits of Desalination 165 ment of a $1 million loan over 30 years at a nominal interest rate of 5 percent is $1.93 million. More pertinently, as shown in Figure 6-7, inter- est costs rise significantly as interest rates rise. An increase in the interest rate from 3 to 7 percent leads to a 15 percent increase in annual costs. Thus, the ability to secure relatively favorable rates of interest has a strong bearing on both the financial and the economic feasibility of any project. The longer the life of the project, the lower the annualized capi- tal costs, other things equal, because the payment of principal is spread over a longer period of time. Nevertheless, the longer the project life, the higher the total costs since the magnitude of interest charges grows ex- ponentially with time. As a matter of current practice, for this analysis the payback period is set at 25 years; however, some owner operators set more aggressive payback periods of 10 to 20 years for seawater desalina- tion projects. Energy Pricing. Despite recent improvements in energy efficiency (see Chapter 4), RO desalination processes are energy intensive and de- pend on prime energy sources. The sensitivity of annual costs (and total project costs) to the cost of energy is illustrated in Figure 6-8 for both brackish and seawater desalination. An increase in energy costs from 140% 2-Stage Seawater RO with Conventional Pretreatment 2-Stage Seawater RO with UF/MF Pretreatment 120% 3 Relative Annual Cost per m 100% 80% 60% 40% 20% 0% 2% 3% 4% 5% 6% 7% 8% Interest Rate FIGURE 6-7. Effect of cost of money on relative annual costs per cubic meter for a 189,000 m3/day seawater RO plant with conventional pretreatment or mem- brane pretreatment. The baseline assumptions for this scenario are as follows: energy costs are constant at $0.07/kWh; membrane life is assumed to be 5 years; nominal interest rate is 5 percent; depreciation period is 25 years.
166 Desalination: A National Perspective 180% 2-Stage Seawater RO with Conventional Pretreatment 160% 2-Stage Seawater RO with UF/MF Pretreatment Brackish Water RO with Conventional Pretreatment 140% 3 Relative Annual Cost per m 120% 100% 80% 60% 40% 20% 0% 0 0.05 0.1 0.15 0.2 0.25 Energy Costs ($/kWh) FIGURE 6-8. Effect of energy costs on relative annual costs per cubic meter in a 3 189,000 m /day seawater RO plant with conventional pretreatment or membrane pretreatment and a 189,000 m3/day brackish water RO plant with conventional pretreatment. The baseline assumptions for this scenario are as follows: mem- brane life is assumed to be 5 years; nominal interest rate is 5 percent; deprecia- tion period is 25 years. Please note the different scale for this figure. $0.04 to $0.10/kWh results in an increase in total costs of over 35 per- cent for seawater desalination. Because brackish water RO desalination uses less energy than seawater desalination (see Figure 4-10), smaller increases in annual costs are observed. Efforts to reduce energy costs, as well as reductions in the total capital costs of the system, offer the great- est prospect of significant reduction in the total costs of seawater desali- nation systems. Operating Pressure. Figure 6-9 shows the results of a sensitivity analysis of the effects of changes in operating pressure from a typical baseline operating pressure on relative annual costs per cubic meter, when other system variables are held constant. Although todayâs best available seawater RO membranes are operating at pressures that are only 40 percent greater than the osmotic pressure of seawater, it is possi- ble that further improvements in membrane permeability and salt rejec- tion could lead to additional reductions in operating pressures in the range 10 to 15 percent with corresponding reductions in energy usage for the seawater desalination unit operation by as much as 15 percent (see Chapter 4 and Appendix B). For a fixed energy cost of $0.07/kWh, a
The Costs and Benefits of Desalination 167 140% 120% 3 Relative Annual Cost per m 100% 80% 60% 2-Stage Seawater RO with Conventional Pretreatment 40% 20% 0% -30 -20 -10 0 10 20 30 40 50 60 % Change in Baseline Operating Pressure FIGURE 6-9. Effect of operating pressure on relative costs per cubic meter in a 3 189,000 m /day seawater RO plant with conventional pretreatment. Energy costs held constant at $0.07/kWh; membrane life held constant at 5 years; nominal interest rate is 5 percent; depreciation period is 25 years. reduction in operating pressure of 15 percent could result in approxi- mately 5 percent reduction in annual costs (Figure 6-9). At higher energy costs, such improvements in membrane technology could lead to total annual cost reductions approaching 10 percent. In addition to permeability and rejection characteristics, the potential of membranes to foul has an important impact on treatment costs. Foul- ing requires increases in operating pressures if the membrane is to re- main effective, and increases of 25 percent are not uncommon. An in- crease in operating pressure of this magnitude would increase annual costs by over 8 percent per cubic meter (Figure 6-9). At higher energy costs, the impact of fouling on annualized operating costs is even more severe and could increase costs in excess of 15 percent. Membrane Life. The cost of membranes has fallen in recent years, and this fact is widely cited as one explanation for the increasing attrac- tiveness of desalination. Membrane costs are now quite modest, ranging from only 3 to 5 percent of annual costs (Figure 6-6). One of the major operating issues in membrane seawater desalination plants is the issue of shortened membrane life that can result from system imbalances that lead to fouling and the need for accelerated cleaning cycles. Thus, for exam- ple, a decrease in membrane life from 5 to 3 years will increase annual
168 Desalination: A National Perspective 140% 120% 3 100% Relative Annual Cost per m 80% 60% 2-Stage Seawater RO with Conventional Pretreatment 40% 20% 0% 0 1 2 3 4 5 6 7 8 9 10 Membrane Life (Years) FIGURE 6-10. Effect of membrane life on relative annual costs per cubic meter in 3 a 189,000 m /day seawater RO plant with conventional pretreatment. Energy cost is assumed to be constant at $0.07/kWh; the nominal interest rate is 5 per- cent; and the depreciation period is 25 years. costs by over 3 percent, as shown in Figure 6-10. Catastrophic, irreversi- ble membrane fouling leading to membrane life of less than 1 year will increase annual costs by over 25 percent. On the other hand, for conventional systems, total project costs can be reduced by 3 percent if membrane life can be extended to 10 years over the 5-year baseline period that is common today (see Figure 6-10). Such a change would have a somewhat greater positive impact on costs for membrane-based pretreatment systems. These data and analyses, to- gether with the fact that membrane life accounts for such a small propor- tion of total project costs, suggest that further research focused on ex- tending membrane life is unlikely to have significant payoffs. Key Determinants of Desalination Costs An examination of the available cost data and a sensitivity analysis of the determinants of water production costs permit some general con- clusions to be drawn. Five conclusions have emerged: 1. The two largest components of annual costs are annualized capi- tal costs and energy costs. This is true of both thermal and nonthermal
The Costs and Benefits of Desalination 169 technologies as illustrated in Table 6-2. Such costs are very significant, each representing over one-third of annual cost for a 189,000 m3/day seawater RO facility (Figure 6-6). Significant reductions in the water production component of desalination costs will require significant re- ductions in the costs of one or both of these components. Frequently, interest charges are the largest or nearly the largest single component of cost for desalination plants. It is important to note that reductions in scale or in the capital cost of a facility will invariably reduce interest costs. 2. Development of membranes that operate effectively at lower pressures could lead to 5 to 10 percent reductions in annual costs (ex- cluding concentrate management costs) for seawater desalination, pri- marily through reduced energy use. 3. Plant size is an important variable for seawater desalination pro- jects as there are significant economies of scale in the production of freshwater. This has significant implications for small communities who are not able to take advantage of economies of scale and are therefore confronted with the higher costs of small projects. Economies of scale also exist for the production of freshwater in brackish desalination plants although they are smaller. 4. Membrane desalination costs are sensitive to the salinity of fe- edwater. Therefore, when concentrate management costs are ignored and other things are equal, RO desalination of brackish source water will nearly always be cheaper when compared with seawater. However, de- pending on the available alternatives, concentrate management costs, especially at inland locations, can increase the total costs of brackish wa- ter desalination significantly, up to 200 percent above the desalination process costs alone. 5. Lengthening membrane life from 5 years to greater than 5 years is unlikely to have a significant impact on overall costs. Economic Costs The costs described in the preceding section are all appropriately characterized as financial costs. However, as explained earlier, economic costs of desalination and other water supply alternatives, including tradi- tional sources, also need to be considered. For the most part, economic costs include the external costs that are borne by the public at large that the water provider can avoid. In the case of desalination facilities, the most significant category of external costs are environmental costs. These could take several forms (see Chapter 5). The environmental costs of surface water concentrate discharges are virtually never monetized and in many instances are not even well understood from a biological perspective. Examples of other external costs that might be associated
170 Desalination: A National Perspective with desalination facilities include the loss of environmental amenity values because the facilities may be unsightly or interfere with lines of sight. The cost of air pollution stemming from energy generation neces- sary for desalination would be yet another example. Although such costs are rarely monetized, there are numerous techniques which allow them to be estimated, either directly or indirectly, with considerable accuracy. These techniques were recently summarized and analyzed by the Na- tional Research Council (NRC, 2004c). Many external costs can be attenuated or mitigated if they are ac- counted for in project design and operations. Federal and state water quality and marine protection legislation normally require that the water providers consider environmental costs and take appropriate steps to minimize or mitigate them. This can require either additional capital fa- cilities or changes in operating routines or both. The costs of these facili- ties and operations will appear as financial costs. External costs are, however, rarely treated in an economically opti- mal fashion in the planning of desalination facilities. Most environmental quality legislation does not require a precise balancing of the costs and benefits of facilities and operations that protect environmental quality; thus, the financial costs may be equal to, less than, or greater than the true environmental costs. Often external costs are completely ignored and, where they must be considered because of laws and regulations, there is little attention to whether the external costs that are ultimately borne are optimal. Pricing Historically, production of desalinated water has not been a particu- larly attractive water supply option because the costs have been high. Ultimately, costs need to be recovered through some combination of government support and the pricing of water. While government subsi- dies have been a part of the picture, the pricing of desalinated water is set to defray the majority of the costs, and prices have not been competitive with those from other sources. There is evidence that this may be chang- ing in some specific situations. Nevertheless, the pricing of water is complex and there are complicated problems that must be solved in es- tablishing pricing structures and deciding what level of price to charge consumers. In almost every instance, trade-offs are involved. In this sec- tion, the principles and purposes of pricing are elucidated and some im- portant practical applications are discussed. Goods or services are said to be scarce when the wants for them ex- ceed the supply. In these circumstances, the good or service in question is said to have value, and a method must be found for rationing the lim-
The Costs and Benefits of Desalination 171 ited supply among the more or less unlimited wants. Markets are one commonly used institution whose function is to ration scarce assets, and they accomplish their tasks by generating prices and rationing the limited number of goods or services according to who is willing to pay for them. Where markets function reasonably well, the prices that emerge reflect the value of the commodity or service in question with higher prices, in- dicating higher value, or lower prices, indicating lower value. Prices are very efficient forms of information and tell consumers, at a glance, whether a commodity is scarce or plentiful and what its value is. Water is not usually traded in well-functioning markets. Instead, the price is usually regulated and covers the costs of capturing, treating, and distributing the water; the water itself is frequently assigned a scarcity value or price of zero. A price of zero sends a number of important addi- tional signals. First, it suggests that the commodity is freely available or available in limitless amounts, when in fact it is not. Inasmuch as con- sumers tend to make decisions about how much of a commodity to con- sume by consuming up to the point where the cost of the last unit con- sumed is equal to the benefit from the last unit consumed, pricing water at zero encourages consumers to use the water up to the point where the last unit used has a marginal value to the consumer of zero. If water is assigned a price of zero (or if water is underpriced) consumers are in- duced to use more water than they would if it were accurately priced, and it encourages uneconomical uses in a time of scarcity. One means of re- sponding to water scarcity would be to charge consumers rates that in- clude a scarcity value or a reasonable approximation thereof. Some pur- veyors have used price effectively as a tool of demand management. Faced with a need to reduce per capita rates of consumption, the El Paso Water Utility increased rates on a number of occasions, which led to re- ductions in water uses each time. The last of these involved a price in- crease of 35 percent, which caused a 5 percent reduction in total domes- tic water consumption (E. Archuleta, El Paso Water Utility, personal communication, 2006). Historically, in most locations the cost of desalinated seawater has been very high, both in absolute terms and in relative terms compared with the cost of other alternatives. Where desalination capacity has been built in recent years, water scarcity has been high (as in Trinidad, Israel, and Perth, Australia), the value of reliability and local control has been important (as in Singapore), or certain kinds of subsidies have artificially lowered the apparent cost (such as long-term energy contracts, reduced land costs, or low-interest loans). Nevertheless, process cost reductions attributable to improved technology combined with the rising cost of most water supply alternatives suggests that the gap between the costs of desalinated seawater and the costs of alternative sources is narrowing. It is not unusual, however, for the cost of desalinated seawater to be twice
172 Desalination: A National Perspective the cost of existing water supplies. In spite of this fact, the addition of desalinated water is often paid by incremental increases in the price of water paid by users. The San Diego Water Authority is a case in point. Existing rates are approximately $0.24/m3 ($300/ a.f.). The costs of sea- water desalination are estimated to range somewhere between $0.64 and $1.04/m3 ($800 and $1300/a.f.). Yet, the Water Authority estimates that if seawater desalination were used to augment the existing supply (ap- proximately 800,000 a.f.) by a little over 10 percent (89,600 a.f.), the rate increase would be approximately $40 annually. How can the rate in- crease be so modest and what does it mean? Public water providers typically price water according to the average cost of acquiring, treating, and delivering it. In some states and locales, this is required by law. Historically, this was necessary because capital- intensive industries such as water purveyors have very high fixed costs relative to operational or marginal costs. Pricing structures that recovered only marginal or operational costs would leave the operator unable to cover fully the costs of debt service and repayment of the capital costs. Today, the average cost pricing rule continues to be followed even though marginal or operating costs are now much higher than they once were. One contemporary consequence of this is that the costs of rela- tively high-priced increments of additional water supply are averaged in with the (much) lower costs of the existing supply. Consider, for exam- ple, a new supply that costs $0.8/m3 ($1,000/a.f.) and an existing supply that costs $0.24/m3 ($300/a.f.). Suppose the existing supply is augmented with an additional 10 percent of the high-cost supply. The resulting aver- age costâthe cost of the existing and augmented supply averaged to- getherâwould be $0.29/m3 ($363/a.f.) even though the costs of the new supply was almost three times that high. The implications of using such average cost pricing rules are mixed. On the one hand, it keeps the water rates paid by consumers from rising sharply. On the other hand, by buff- ering the consumer from the higher costs of the new supply, the practice sends erroneous signals about the true cost of the additional water, sug- gesting to consumers that water is cheaper and more plentiful at the pre- vailing price than is true in fact. Consumers respond by using more of the higher-cost (e.g., desalinated) water than they would if faced with the true price or costs (although total water use would be expected to decline in the face of higher prices). The result is that the high-cost supply is in- efficiently used. More of that supply is consumed than would be the case if consumers were charged the true price. In contemporary circum- stances, the practice of average cost pricing, then, stimulates water use artificially beyond the point where it is efficient. Stated differently, the value of water in the artificially stimulated uses is less than what it costs to make the marginal increments of supply available.
The Costs and Benefits of Desalination 173 In an era when water scarcity is both pervasive and intensifying, policies and institutions that create rates and costs that are lower than the true or âmarginal costâ appear perverse.1 While it may be difficult to separate high-cost water from low-cost water in distribution systems, it is not difficult to identify the consumers who require the additional, high- cost supplies to be developed. There is no rule that requires that new- comers and long-time residents be charged the same price. The use of marginal cost pricing would raise the cost of water to users of the mar- ginal or most recent addition to supply very sharply. However, there are concerns about the fairness of charging sharply higher prices for the new water. Most of these concerns are directed at the question of how the poor can afford to pay higher prices for water, which is, at some level, an essential commodity. In general, the benefits of marginal cost pricing can be gleaned without unduly penalizing the poor by devising rate schedules that mimic marginal cost pricing. Such schedules entail very low or zero prices for the first block or âlife-lineâ quantity of water. This is the amount of water necessary for drinking, cooking, and basic sanitation. Thereafter, the price of ensuing increments of water rises so that the more water a consumer uses, the higher the price paid at the margin. This type of rate structure is widely used with electricity and has been adopted by many water purveyors in recent years. Fundamentally, such rate struc- tures mimic marginal costs by pricing successive blocks of use at pro- gressively higher rates. The intention is to discourage progressively lar- ger quantities of consumption beyond some lifeline amount, and the evi- dence suggests that such rate structures are associated with lower levels of consumption when compared with the declining block rate structures (Haneman, 2006; Renwick and Green, 2000). The true cost of water, including desalinated water, is likely to re- main unstated or understated. This means that, in general, consumers will behave as if water scarcity is less intense than it is in actuality. Although it is desirable to reform rate structures so that consumers are confronted with a realistic approximation of the cost of the water they use, political and institutional inertia sometimes makes this difficult. The result will be overinvestment in desalination facilities and the continued application of desalinated water to uses whose value is less than the cost of making the 1 This discussion of marginal cost pricing refers to short-run marginal costs and not long-run marginal costs. The difference is that in the short run (i.e., over the life span of a desalination plant) certain costs such as capital costs are consid- ered fixed and invariant. Thus, in the short run, the marginal costs of these fac- tors of production are counted as zero because they cannot be altered. In the long run (i.e., considering replacement plants) all costs are variable, including capital costs, so long-run marginal costs would include capital cost and other costs which are fixed in the short run.
174 Desalination: A National Perspective water available. A first step toward correcting this situation would be for water purveyors to disclose fully and accurately the costs of desalinated water (as well as the costs of water from all other sources). Such disclo- sures should include a full accounting of all costs and report subsidies and as well as distortions introduced through regulation or market imper- fections. Ideally, the local price of water would be established by the in- teraction of supply and demand such that the resulting price covers fi- nancial, environmental, and social costs. Other Cost Considerations The Role of Subsidies The development of most desalination facilities entails public subsi- dies of some sort. These are usually found in research and development efforts and occasionally in construction and operation. Although subsidi- zation has been practiced extensively in the development of water sup- plies, it is reasonable to ask whether such subsidies are both necessary and justified. The economic justification of subsidies relies on the notion that subsidies are justified when the benefits of the subsidized activity are widespread in nature and cannot be fully captured by the firm or util- ity that generates them. As a rule there will be underinvestment in such activities in the absence of the subsidy. Thus, for example, subsidies may well be justified to promote environmental protection and enhancement activities and to underwrite the costs of water quality maintenance and improvement when the benefits from such improvements are widespread. There will be many instances where research and development activities related to the development and implementation of desalination technolo- gies can justifiably be subsidized because the developing agents cannot capture all of the returns to the scientific information that is developed when that information is freely available to others. Subsidies do not meet the standard economic tests of justification where the research and de- velopment information is held as proprietary. In some instances, subsi- dies are promoted in order to keep water rates down or in an affordable range. There is no scientific basis which justifies this kind of use of sub- sidies. Rather, they are the result of political processes and the value and equity preferences inherent in them. Public and Private Roles Earlier studies have concluded that, because the demographic, eco- nomic, political, and physical circumstances vary so widely, there is no
The Costs and Benefits of Desalination 175 single model of public or private water service delivery that is preferred in all situations (Gleick et al., 2002; NRC, 2002c). A canvas of desalina- tion projects around the world reveals that some facilities are entirely publicly constructed and operated, some are private, and some are mixed. The presence of such a diversity of public, private, and mixed arrange- ments for the design, construction, and operation of desalination facilities would appear to confirm these earlier conclusions about the desirability, or lack thereof, of privatizing the provision of water and wastewater ser- vices. Indeed, it now seems to be recognized that privatization is but one of an array of approaches to acquire sufficient capital as well as opera- tional experience with water and wastewater services of all types (Raftelis, 1989). It is important to recognize that there is a range of privatization op- tions. They include (1) private provision of services (e.g., laboratory work) and supplies (e.g., chemicals); (2) private contracting for the op- eration and maintenance of a capital facility; (3) negotiated contracts for the design, construction, and operation of new facilities; or (4) sale of water utility assets to a private firm. The first three options are found throughout the global desalination industry and while there are currently no examples of the fourth without the government retaining some owner- ship, as in Aruba and Malta, for example, it does remain an option. Each of these options is known to work well in some circumstances. It is, however, difficult to generalize about such circumstances (NRC, 2002c). A more detailed discussion of various common public-private partner- ship models is provided in Chapter 7. One circumstance in which privatization may prove attractive is the provision of water supply and wastewater treatment services for smaller communities, which tend not to have the resources to take advantage of economies of scale and which may lack the resources to acquire the nec- essary scientific, technical, and financial expertise. Opportunities for âregionalizingâ utility services across a number of communities may not always favor private provision of those services, however. In general, in selecting arrangements along the public-private continuum, it will be important to recognize that private and public agencies face different kinds of incentives and have different strengths. For example, public agencies tend to be more responsive to political currents and tend to do a better job of providing public goods such as environmental protection (Baumol and Oates, 1988; Oakland, 1987). On the other hand, private entities are thought to operate more efficiently, although they are fre- quently subject to regulation (NRC, 2002c). There is reason to believe that these conclusions about public and private provision of water and wastewater services will apply to desalination. Indeed the existing diver- sity of public and private arrangements found around the world lends strong support to the proposition.
176 Desalination: A National Perspective THE STRUCTURE OF BENEFITS The benefits of desalination accrue in terms of the value of the water produced. Where markets function well, market clearing prices tend to be good indicators of value and can be used to compute the value of a fresh, reliable water supply to consumers in the various water-use sectors. The difficulty in computing the benefits of water supply in this conventional fashion lies with the fact that, as discussed previously in this chapter (see Pricing), water is rarely priced in markets. As a practical matter, the value of water produced from desalination facilities is frequently estab- lished with reference to the costs of acquiring water from the least costly alternative source. At its simplest, if the cost of desalinated water is $0.8/m3 ($1000/a.f.) and the next most attractive option costs $1.2/m3 ($1500/a.f.), then the value of the desalinated water would be counted as $0.4/m3 ($500/a.f.). This illustrates why the value of the water cannot be established in the absence of reference to the costs of acquiring water from other sources. Unfortunately, the problem of valuation is often not usually so simple. The qualities of water from different sources differ; the reliability of water from different sources differs; and the environ- mental costs associated with different sources differ. In some instances desalinated water can be substituted for other water supply sources that are not being exploited in a sustainable fashion (e.g., persistently over- drafted groundwater; see Box 3-3), and benefits accrue to the provision of desalinated water in the form of reduction in unsustainable uses of alternative groundwater and surface water sources. An accurate valuation of desalinated water would include a premium on quality because such waters are likely to be of the highest possible quality and have a pre- mium on reliability since desalination processes (particularly those that use seawater as a source) are uncoupled from the hydrologic cycle and, thus, are among the most reliable of water sources. Value of Reliability Water supply reliability can be defined as the consistent availability of water in response to demand. Reliability of supply, for example, is most obviously manifested by the fact that water flows from taps in most U.S. homes when they are turned on. A reliable water supply is more valuable than one that is susceptible to interruption. Indeed, water pur- veyors are willing to pay a premium for water that is a reasonably accu- rate indicator of the value of reliability. This value and the premium de- pend on a number of factors, including the use to which the water is put, the availability of alternative sources, production costs, and the costs of a supply disruption. The importance and implications of these factors will
The Costs and Benefits of Desalination 177 differ from region to region and, thus, the value of reliability will differ from case to case (Cooley et al., 2006). One of the most desirable attributes of desalination is the fact that the availability of brackish estuarine waters and seawater is independent of the hydrologic cycle. This means that the capacity of coastal desalination projects to produce water is not affected by severe droughts, an attribute that is particularly valuable in circumstances where climate is highly variable. Brackish groundwater desalination can also provide reliable water supplies during short-term droughts, although longer periods of drought can affect regional groundwater availability. This reliability will also be significant in arid and semi-arid areas where rainfall and runoff are often inadequate to serve existing demands, where water is overallo- cated, and where water allocations, water rights, and the ability to access and use other sources of water are in dispute. Purveyors are willing to make substantial investments to avoid sup- ply interruptions, often because the cost of such interruptions may be larger than the cost of improving the water supply reliability. Thus, for example, the East Bay Municipal Utility District (EBMUD) invested more than $200 million to protect reservoirs, pipes, pumps, and treatment plants against a large earthquake. EBMUD serves over one million con- sumers on the eastern margins of San Francisco Bay, and it is estimated that in the absence of such strengthening, water-related losses to con- sumers from a large earthquake on the EBMUD system could be as high as $2 billion (EBMUD, 2005). Earthquakes are not the only potential threats to reliability. Other threats to reliability include water quality con- tamination, climate change and its impacts on quality and water avail- ability, changes in the regulatory environment, and price volatility in the various factors needed to operate the system, such as energy. Unantici- pated threats such as the renewed potential for terrorist attacks can also arise at any time. Water supply reliability is measured in different ways. The most common measure is to portray the risk of a projected supply falling be- low a projected demand over some specified time period. When a system is described as having a reliability of 95 percent, that implies that supply will be equal to or exceed demand in 19 instances (e.g., days, months, or years) out of 20. Other approaches depict the severity of water shortfalls, and this should always be taken into account. A system with a reliability of 90 percent might be more attractive than one with a reliability of 95 percent if the shortfalls in the former supply are very small and those associated with the latter are less frequent but much larger (Cooley et al., 2006). There are a number of alternatives available for improving supply re- liability. For example, dams and reservoirs and the associated convey- ance facilities tend to insulate consumers from short-term variations in
178 Desalination: A National Perspective precipitation and runoff by storing water during wet periods for use in dry times. Multiple sources of supply also augment system reliability, particularly when the different sources are hydrologically independent of each other or provide a means of buffering against nominal hydrologic variability. No source is completely invulnerable to supply disruption. For all its reliability attractiveness, desalination technology is especially vulnerable to adverse changes in the energy supply picture, for example. Where reliability is valued it should be counted as a benefit of de- salination technologies. There is frequently an issue of whether, in fact, a more reliable supply justifies higher values and how reliability ought to be valued. If water were traded in well-functioning markets, it would be relatively easy to answer this question by examining the difference in prices as between a more reliable supply and a less reliable supply. The absence of market-generated prices means that techniques for imputing the price must be employed in order to value reliability. One means that is frequently used by economists is to evaluate the willingness to pay (WTP) of consumers to reduce the probability of a water shortage. Sup- pose, for example, that the risk of a water shortfall that would require rationing is 1 chance in 50. What would consumers be willing to pay to reduce that chance to 1 in 75? Economic studies have shown that the WTP to avoid restrictions on water use due to drought or other factors ranges from $32 to $421 per household per year in constant 2003 dollars (Carson and Mitchell, 1987; Griffin and Mjelde, 2000). When the esti- mated reduction in water use due to drought is multiplied by the prob- ability of the drought actually occurring, these estimates imply that the reliability value of extra water in severe drought circumstances could be as high as $3.12/m3 ($3900/a.f.) (Raucher et al., 2005). Although there may be controversy surrounding the magnitude of this figure, the study results do indicate that greater water supply reliability has value. There are other techniques for estimating the value of reliability. One, developed by the Pacific Institute, is described in Appendix D. The method borrows from and adapts tools from financial portfolio theory and allows a comparison of water supply alternatives that have differing degrees of reliability. CONCLUSIONS AND RECOMMENDATIONS Historically, the relatively high financial costs of desalination con- strained the use of desalination technologies in all but a few very specific circumstances, but the cost picture has changed in a number of important ways. There have been significant reductions in membrane costs and in other components of cost in the production of desalinated water. Perhaps, more significantly, the costs of other alternatives for augmenting water
The Costs and Benefits of Desalination 179 supplies have continued to rise along with the degree of treatment re- quired of existing supplies, making desalination costs more attractive in a relative sense. A continuation of these trends would likely make desali- nation costs more attractive and less of a constraint in the future. The trend of desalination process cost reduction may be abetted through a program of strategically directed research aimed at achieving potentially large cost reductions. The following recommendations are based on the detailed discussion and analyses of this chapter. Substantial reductions in the financial cost of desalination will require substantial reductions in either energy costs or capital costs. Energy and capital costs are the two largest components of financial cost for both thermal and membrane seawater desalination processes. It is important to recognize that reductions in scale or in the capital costs of a facility will have associated reductions in interest costs. In most in- stances, interest costs will be a large component of total costs. Future trends in energy costs will also be important inasmuch as significant in- creases in energy prices could offset or more than offset cost reductions in other areas and make desalination technologies less attractive. For brackish water desalination, the costs of concentrate man- agement can vary enormously from project to project and may rival energy and interest costs as the largest single component of cost. The high cost of concentrate management at some inland locations ultimately offsets the cost advantage that can be obtained from utilizing feedwaters with lower salinity. There are small but significant efficiencies that can be made in current membrane technologies that will reduce the energy needed to desalinate water and therefore offer potentially important process cost reductions. Todayâs best available seawater RO membranes are operating at pressures that are only 40 percent greater than the osmotic pressure of seawater and therefore are approaching the theoretical limits of energy efficiency for membrane desalination. However, development of membranes that operate effectively at lower pressures could lead to 5 to 10 percent reductions in annual costs of desalinating seawater associ- ated with a 15 percent decrease in energy use. Extending membrane life is likely to have a very small impact on desalination costs. Todayâs best-available seawater RO membranes rou- tinely operate for 5 or more years before needing to be replaced. The ability to extend membrane life past 5 years to 10 years will have a minimal impact on total costs given the small contribution of membrane replacement costs to total costs over a 5-year lifetime. However, the pre-
180 Desalination: A National Perspective vention of catastrophic failure is especially important because membrane failure within the first year of operation can cause an annual cost increase of over 25 percent. Future research efforts should be focused on mistake- proof, robust prefiltration to ensure against premature failure of the RO membranes. The costs of producing desalinated water have fallen in recent years but may rise in the future if the price or cost of energy rises faster than cost decreases from technological improvements. In- creases in energy costs lead disproportionately to increases in desalina- tion costs and in the costs of transporting water long distances. The ulti- mate size of these increases, however, may be limited when the costs of fossil fuels reach the costs of other energy technologies, especially re- newable energy technologies that can substitute for fossil fuels. Conse- quently, energy costs will not rise indefinitely even if possible fuel prices do rise more or less indefinitely. In considering the implications of in- creasing energy costs, it is important to recognize that alternative supply measures that also have high energy demands will be sensitive to future energy prices. Conservation and transfers from low- to high-valued uses will usually be less costly than supply augmentation schemes, including desalination. In many circumstances, low-cost methods of demand man- agement could provide significant water savings. Low-cost demand- management techniques have not been exhausted and, so long as poten- tial remains, demand management will offer the possibility of freeing up water to serve new uses at lower cost than desalination. Similarly, mar- ket-like transfers of water can also offer relatively low-cost ways of ac- quiring additional supplies of water. This is particularly true where addi- tional water supplies are needed to support urban growth and where agri- cultural water is available for reallocation. Conservation and efficiency improvements that reduce the total demand for water often come with associated benefits (such as reduced energy costs), require little capital investment, and can be implemented relatively quickly. Ultimate costs will vary depending on the local details of water use, water available for transfer, previous efforts to improve efficiency, financial perspectives, and institutional factors that encourage or discourage different water pol- icy choices. To make the true costs transparent, the economic costs of desali- nation should be accounted for and reported accurately. Failure to price water accurately can lead to inefficient use and overuse. Melded pricing or average cost pricing is frequently used pursuant to law or to address equity consideration. This practice understates the cost of desali-
The Costs and Benefits of Desalination 181 nated water to the consumer, and the supplier should take care in pub- licly reporting the true and accurate economic costs.