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.
4 DROUGHT MANAGEMENT OPTIONS Benedykt Dziegielewski Department of Geography Southern Illinois University Carbondale, Illinois Many municipal water systems face a risk of having major disruptions in water supply during droughts. The level of risk for any existing system can be assessed by preparing a probabilistic forecast of potential water supply deficits during a stated planning period. Such a forecast is obtained by comparing the probability distribution of the future availability of water in supply sources with the long-term forecasts of water requirements. The systems facing a high level of risk can reduce their vulnerability to shortages by expanding the capacity of supply sources or implementing nonemergency demand management programs. If the level of risk is low or moderate, the best strategy may involve formulation of drought contingency plans to cope with actual emergencies. This chapter summarizes the conceptual approaches to planning for droughts and presents a method for developing an optimal strategy for mitigation of water deficits caused by droughts. THE "DESIGN DROUGHT" APPROACH The traditional approaches to water supply planning treat the question of droughts as a part of the capacity expansion problem. Theoretically, the capacity expansion problem should be solved by balancing the cost of water supply augmentation projects against the expected damages that may result from recurrent shortages of water caused by droughts. However, the expected long-run drought damages are difficult to estimate, since both the hydrologic variability of -65-
-66- supply sources and the economic consequences of periodic disruptions of water supply are not adequately defined. Faced with inadequate data, water resource engineers devised a compromise solution that balances the cost of additions to supply capacity against the increments in the reliability of Atwater system. The term ''reliability" is related to the expectation of water supply sources to sustain a stated level of supply over time. Assuming that the hydrologic variability can be described by a reasonably well known probability distribution, reliability is defined as the probability that a desired outcome (or level of supply) will take place. In relation to droughts, the desired outcome is usually defined as the "safe yield" which determines the output of a supply project (or a combination of projects) that can be maintained during a severe drought, such as the worst drought in the historic record. In practice, the balancing of the cost of supply additions against the increments of reliability is reduced to the selection of a "design drought." In most applications, either the worst drought on record or a 100-year drought is chosen as the basis for deciding on the excess capacity of a water supply system. Since the design drought implicitly sets the magnitude of the economic losses that may be incurred, the selection of the probability of such a drought (or its recurrence interval) must assume that the incremental damages that may result from recurrent shortages of water (caused by droughts more severe than the design drought) during the planning horizon are balanced with the incremental costs of the additions to supply capacity. Typical capacity expansion projects usually involve the construction of new facilities for water storage, treatment, and transmission. In recent years, however, a number of unconventional alternatives have also been considered. These may include the following: - 1. more efficient utilization of existing water supplies (e.g., pumped storage or reduction of losses), 2. use of ground-water aquifers for storage of excess supply of surface water, 3. interbasin importation, 4. desalination of seawater or brackish ground water, 5. reclamation of waste water, and 6. implementation of cloud-seeding projects.
-67- Using the design drought approach, the optima1 combination of projects and the dates of their completion are determined by minimizing the cost of their implementation subject to the constraint that the safe yield is at least equal to the requirements for water at any time during a prescribed planning horizon. While structural solutions to water supply planning might have been efficient in the past, the cost of some unconventional supply augmentation projects have made this "fail-safe" approach to droughts prohibitively expensive for many existing systems. As a result, water resource planners have been forced to extend their perspective to include nonstructural alternatives. A combination of supply augmentation and long-term demand management projects has the potential for affecting the position of a water system, with respect to the risk of recurrent water deficits, by not only increasing supply but also by reducing future water use. The demand management projects that can substantially reduce future water use may include the following: 1. campaigns to educate the consumers on how to modify water use habits to reduce water consumption, 2. promotion or a mandatory requirement of use of water-saving devices and appliances, 3. promotion or a mandatory requirement of low-water-using landscaping, 4. adoption of efficient marginal cost pricing, and 5. adoption of zoning and land use policies to control the number of water users served by the system. Although the inclusion of demand management projects may reduce the overall cost of the long-term adjustments to droughts, it does not eliminate the problems associated with the arbitrary selection of the design drought. This is true in cases where the minimum-cost combination of the supply augmentation and nonemergency demand management projects is determined subject to the constraint that the safe yield is at least equal to the anticipated future water requirements with conservation at any time during the planning period. Also, the reliance on demand management projects introduces two additional sources of uncertainty that are associated with the effectiveness (water savings) and the cost of water conservation measures. Overall, the long-term adjustments to droughts can effectively reduce the magnitude and the frequency of
-68- potential water deficits; however j the criterion of the design drought with arbitrarily established levels of acceptable severity and unknown economic consequences is inadequate for developing optimal drought mitigation plans. Due to the rising costs of supply augmentation projects, the use of the design drought approach is very likely to lead to suboptimal plans in which the incremental cost of system reliability considerably exceeds the actual cost of shortages. DROUGHT EMERGENCY PROGRAMS The formulation of the canacitv expansion Droblem ~ r----a ---rid described in the previous section assumes that the costs of water shortage are prohibitively high and the situations in which the capacity falls below requirements are not permitted except for droughts more severe than the design drought. During an actual emergency, the knowledge of the system's safe yield is of limited value to the water manager. Since the severity (or recurrence interval) of an ongoing drought cannot be determined with a reasonable degree of certainty, to be on the safe side the manager is very likely to assume that the ongoing drought is more severe than the design drought. To keep the risk of running out of water reasonably low, the manager will always try to adjust the demand for water by imposing increasingly severe water use restrictions to forestall more severe cutbacks that may be required if the perceived shortage of water materializes at later stages of the drought. Past drought experiences show that the actions of water managers can greatly influence the magnitude of the monetary and nonmonetary losses from the drought. Although the manager apparently has little choice but to pass (or create) water shortage to the customers, the drought literature documents a great variety of drought emergency measures undertaken in response to anticipated shortages of water. Generally, these measures fall into three broad categories: ~ a, v demand reduction measures, 1. 2. improvements in efficiency in water supply and distribution systems, and 3. emergency water supplies.
-69- Table 4-1 classifies specific drought management options according to these three categories. Each urban area has considerable potential for temporary reduction of "normal" water consumption during drought emergency without significant costs or inconvenience to consumers. However, when the emergency actions are undertaken as uncoordinated ad hoc responses to changing storage conditions, the cost of these actions may be substantial. For example, high cutbacks in water delivery may very rapidly increase the losses suffered by local economies, especially when industrial output must be reduced. Therefore planning for water deficits in the long run should begin with the minimization of the cost of short-term deficit The central question is how management programs. potential water shortages can be averted at minimum cost to the supplier or, alternately, to the region served by the water utility. The information that would aid the most water managers in coping with water deficits during an actual drought is the following: 1. the level of deficit most likely to result from an ongoing drought if no action is taken, 2. the effectiveness of specific drought management measures (i.e., the quantity of water saved or obtained due to the implementation of the measure), and 3. the total cost of individual measures including the economic losses resulting from cutbacks in water delivery. Given this information, the manager would be able to devise a drought emergency program that would alleviate the expected deficit at the minimum cost. An important consideration in the formulation of minimumrcost drought emergency plans is that it is not likely that one emergency program can be optimal for all droughts that may occur during the planning period. The "best package" of various drought management measures may be different for different sizes of water deficits. Therefore in order to carry out a complete evaluation of drought management alternatives, it is necessary to develop a probabilistic forecast of future water supply deficits. Alternately, separate plans may be formulated for coping with deficits of increasing magnitudes, for
-70- TABLE 4-1 A Typology of Drought Management Options I. Demand Reduction Measures 1. Public education campaign coupled with appeals for voluntary conservation Free distribution and/or installation of particular water saving devices: 2.1 Low-flow showerheads 2.2 Shower flow restrictors 2.3 Toilet dams 2.4 Displacement devices 2.5 Pressure-reducing valves Restrictions on nonessential uses: 3.1 Filling of swimming pools 3.2 Car washing 3.3 Lawn sprinkling 3.4 Pavement hosing 3.5 Water-cooled air conditioning without recirculation 3.6 Street flushing 3.7 Public fountains 3.8 Park irrigation 3.9 Irrigation of golf courses 4. Prohibition of selected commercial and institutional uses: 4.1 Car washes 4.2 School showers Drought emergency pricing: 5.1 Drought surcharge on total water bill 5.2 Summer use charge - 5.3 Excess use charge 5.4 Drought rate (special design)
-71- TABLE 4-1 (Continued) 6. Rationing programs: 6.1 Per capita allocation of residential use 6.2 Per household allocation of residential use 6.3 Prior use allocation of residential use 6.4 Percent reduction of commercial and institutional use 6.5 Percent reduction of industrial use 6.6 Complete closedown of industries and commercial establishments with heavy uses of water II. System Improvements Raw water sources: 1.1 Reservoir/lake evaporation suppression 1.2 Reduction of dam leaks 1.3 Transfer of surplus water between reservoirs 1.4 Pumped reservoir storage Water treatment plant: 2.1 Recirculation of washwater 2.2 Blending impaired quality water 3. Distribution system: 3.1 Reduction of system pressure to minimum possible levels 3.2 Implementation of a leak detection and repair program 3.3 Discontinuing hydrant and main flushing III. Emergency Water Supplies 1. Interdistrict transfers: 1.1 Emergency interconnections 1.2 Importation of water by trucks 1.3 Importation of water by railroad cars
-72- TABLE 4-1 (Continued) , Cross-purpose diversions: 2.1 Reduction of reservoir releases for hydropower production 2.2 Reduction of reservoir releases for flood control . 2.3 Diversion of water from recreation water bodies 2.4 Relaxation of minimum streamflow requirements . 3. Auxiliary emergency sources: 3.1 Utilization of untapped creeks, ponds, and quarries 3.2 Utilization of dead reservoir storage 3.3 Construction of a temporary pipeline to an abundant source of water (major river) 3.4 Reactivation of abandoned wells 3.5 Drilling of new wells 3.6 Cloud seeding example: 10, 20, 30, or 50 percent of unrestricted water requirements in any year of the planning period. The critical point in the formulation of drought emergency plans is the evaluation of specific emer~enev actions such as those listed in Table 4-1. ~ , ~ . _ The purpose of the evaluation of individual measures is to prepare an array of applicable, technically feasible, and socially acceptable conservation practices together with the information on quantities of water saved and the expenditures and monetary losses associated with their implementation. The specific steps in the evaluation of demand reduction measures must include the following: 1. determination of technical feasibility, i.e., the capability of a given measure to produce a reduction in water use upon implementation, 2. determination of social acceptability to predict the probable response to the measure of various sectors of the community,
-73- 3. analysis of implementation conditions to identify the agencies responsible for implementation and to define the temporal and sectoral coverage of each measure, 4 e determination of effectiveness, i.e., the reduction in water use that can be attributed to the implementation of each measure, 5. determination of the utility's expenditures on implementing the measures, and 6. dete~-~uination of the economic losses resulting from cutbacks in water delivery borne by the customers in various sectors of local economy. The identification and evaluation of emergency water supplies must follow similar steps; however, the information sought is of different character and it includes the following: 1. availability and quality of water in potential emergency sources during persisting dry weather conditions, 2. adequacy of existing treatment facilities to produce finished water of acceptable quality when emergency supplies make up some fraction of raw water supply, 3. lead time required to construct necessary water transmission and Pretreatment facilities (if required). 4. construction and operation-maintenance costs required to bring emergency sources on line, 5. forgone benefits associated with cross-purpose diversions of water from alternative uses, and 6. potential legal and institutional considerations involving permits, rights to water, or easements for transferral systems. The descriptive data on individual drought management options produced through these steps allow the manager to formulate optimal (minimum cost) drought emergency plans corresponding to deficits of varying magnitudes occurring in different points in time during the planning period. The selection of an emergency plan for implementation during an actual water crisis should be preceded by the determination of the expected magnitude of supply deficit. Preliminary actions of the water manager are contingent upon some indication of potential water shortage. When a shortage alert is in effect, the
-74- manager should initiate the analysis to assess the likelihood of water deficit. A probabilistic forecast of supply deficits can be developed by comparing the probable levels of supply in the short-term with short-term forecasts of unrestricted water demand. Provided with a cumulative probability distribution of water deficits that may be caused by the ongoing drought, the manager may either choose the expected deficit (expected value) or a volume of deficit with a "sufficiently" low probability of occurrence and match it with the corresponding minimum-cost shortage mitigation plan defined during the drought preparation stage. If after the implementation of the selected plan it becomes apparent that the volume of deficit will be different than predicted, the appropriate adjustments are made based on revised estimates of the supply deficit. An alternative approach to managing of actual drought emergencies involves a sequential implementation of increasingly ~ . . severe shortage mitigation measures as the supply conditions become more critical. Although this approach eliminates the need to assess the risk of shortage at the onset of a drought, the overall cost of coping with emergency may be higher, since the measures capable of providing substantial increases in supply (or water savings) at relatively low cost are not likely to be used during the critical period of the drought. When the drought emergency decisions are geared to the expected magnitude of supply deficits, the most effective measures are likely to be used at the earlier stares of the drought. and the decision as to when in the course of a drought to introduce and terminate - emergency measures becomes less critical. MINIMIZING THE LONG-RUN COST OF COPING WITH DROUGHTS The use of the minimum-cost plans during actual emergencies does not imply a simultaneous minimization of the long-term cost of coping with shortages. It is reasonable to expect that a system that has to resort to emergency measures every year or even every five years can deal with supply deficits more effectively by expanding the capacity of supply sources and/or implementing nonemergency water conservation programs. The need for expanding supply capacity of an existing system can be assessed by using the expected value of
-75- the cost of coping with drought emergencies during the planning period. The expected value of the coping cost is a common metric for comparing the alternative long-term adjustments to drought. For any given water supply system, the expected value of the cost of coping with emergencies can be determined on the basis of the probabilistic forecast of supply deficits and the cost of drought emergency plans. For each future year, minimums drought emergency plans are determined for a range of possible supply deficits. The probability of occurrence of these deficits is assigned to the cost of the corresponding emergency plans so that higher costs associated with large volumes of deficit have lower probability. The expected value of the coping cost in each year is found by summing the products of the costs and their respective probabilities. For a specified planning period, the expected values of annual costs are reduced to a single number by finding the sum of the present worths of the expected values of coping costs in each future year. This number represents the expected value of the cost of coping with water deficits in the long run. The expected value of the long-term costs to cope with emergencies in the supply of water allows water planners to examine the trade-offs between the short-term and the long-term adjustments to droughts. Any combination of the long-term supply augmentation and demand management projects will affect the probability distribution of supply deficits in each future year thus resulting in the new expected value of the long-term cost of coping with emergencies. Theoretically, the optimal strategy for dealing with droughts would be determined by balancing the incremental cost of the long-term adjustments with the decrements of the cost to cope with emergencies. However, the optimal solution selected in this manner would be based on the comparison of the relatively certain costs of system expansion with uncertain expenditures and economic losses during droughts. A more appropriate approach would be to compensate for the differences in uncertainty by assigning subjective weights to each of the two cost categories.
76 AVAILABILITY OF ANALYTICAL TOOLS A more complete elaboration of the drought planning approach outlined in this discussion is given in Dziegielewski et al. (1983a,b). This approach follows the concepts and decision criteria formulated by Russell et al. (1970), Young et al. (1972), and Russell (1979~. The practical application of this method largely depends on the availability of analytical tools for forecasting the water supply and demand. There exists a large body of literature on forecasting water supply. The methods most relevant to drought planning are described by Stall and Neill (1963), Lampe and Smith (1982), and Sheer (1980~. The methods for forecasting water demands are also available. A computerized water use forecasting system known as IWR-MAIN (Crews and Miller, 1983) is most useful for the evaluation of drought management alternatives, since it estimates water use at a highly disaggregate level based on demographic and socioeconomic characteristics of the water service area. Although further refinement of forecasting methods can considerably improve the planning for water deficits, the most critical research needs are related to the measurement of the effectiveness and costs of various emergency actionse A complete evaluation of short-term drought management measures will allow water utility managers to make more informed decisions during crisis situations and will also result in more efficient long-term water supply planning. REFERENCES Crews, J. E., and M. A. Miller. 1983. Forecasting Municipal and Industrial Water Use: IWR-MAIN System User's Guide for Interactive Processing and User's Manual. Contract Rep. 83-R-3. U.S. Army Corps of Engineers, Institute for Water Resources. Dziegielewski, B., D. D. Baumann, and J. J. Boland. 1983a. Evaluation of Drought Management Measures for Municipal and Industrial Water Supply. Contract Rep. 83-C-3. U.S. Army Corps of Engineers, Institute for Water Resources. Dziegielewski, B., D. D. Baumann, and J. J. Boland. 1983b. Prototypal Application of a Drought Management Optimization Procedure to an Urban Water Supply
-77- System. Contract Rep. 83-C-4. U.S. Army Corps of Engineers, Institute for Water Resources. Lampe, L. K., and R. L. Smith. 1982. A Drought Contingency Manual for Kansas Public Water Supplies. Prepared for the Kansas Department of Health and Environment by the Center for Research, Inc., University of Kansas. Russell, C. S. 1979. Water deficit planning, Proceedings Paper _ Southeast Regional Conference on Water Conservation and Alternative Water Supplies, Georgia Institute of Technology. Pp. 209-219. Russell, C. S., D. G. Arey, and R. W. Kates. 1970. Drought and Water Supply. The Johns Hopkins University Press, Baltimore, Md. Sheer, D. P. 1980. Analyzing the Risk of Drought: Occoquan Experience. Journal of American Water Works Association, 72:246-253. Stall, J. B., and J. C. Neill. 1963. Calculated risks of impounding reservoir yield. Journal of Hydraulics Division, ASCE, 89(HY1~. Young, G. K., R. S. Taylor, and J. J. Hanks. 1972. A Methodology for Assessing Economic Risk of Water Supply Shortages. Contract Rep. 72-4. U.S. Army Corps of Engineers, Institute for Water Resources
