IDENTIFICATION OF ROBUST WATER RESOURCES PLANNING STRATEGIES by FRANCISCO JAVIER MARTIN CARRASCO B.S. Civil Engineering Polytechnic University of Madrid, Spain (1984) Submitted to the departments of URBAN STUDIES AND PLANNING and CIVIL ENGINEERING in partial fulfillment of the requirements for the degrees of MASTER CITY PLANNING and MASTER OF SCIENCE IN CIVIL ENGINEERING at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY June 1987 Francisco Javier Martin Carrasco The author hereby grants to M .I .T. permission to reproduce and to distribute copies of this thesis documents in whole or in part. Signature of the Author Department htJrban,8tudies and Planning and 7 Department of Civil Engineering . June, 1987 Certified by Thesis S'gupervi'sor Ralph A. Gakenheimer Departrent of Urban Studies and Planning Certified by Theps Supervisor, Dennis B. McLaughlin Department of Civil Engineering Chairman, Phillip L. Clay Master City Planning Committee Certified by_ Certified by Ole S. Madsen Civil Engineering Master Committee IBARI IDENTIFICATION OF ROBUST WATER RESOURCES PLANNING STRATEGIES by FRANCISCO JAVIER MARTIN CARRASCO the departments of Urban Studies Submitted to on June 25, and Planning a nd of Civil Engineering 1987, in parti a 1 fulfillment of the requirements for of Master City Planning and Master of the degrees Science in Civ i I Engineering. ABSTRACT This thesis presents a two-objective method for of the objectives is water resources planning. One and the other is expected net benefits, to maximize this thesis, robustness to maximize robustness. In a measure of the sensitivity of the is considered conditions. An to uncertain performance projects index of robustness is developed. The index is based on the variation of net benefits from a project as consequence of the uncertainty in determining the parameters of water resources systems. by problem is solved The two-objective obtaining the Pareto curve, which represents the set method uses screening of non-inferior projects. The models as the optimization technique. To prove the practical viability of the method, a case study. The case study applied to it is select from al l possible projects to be consists in that represent built in a hypothetical bas in, those the best development altern at ive. Thesis supervisors: Dr. McLaughlin, Dennis Engineering Dr. Gakenheimer, Ralph Studies and Planning professor of Civil professor of Urban and Civil Engineering 2 ACKNOWLEDGMENTS This thesis is the final step of two Masters that have been possible thanks to the Fundacion Juan March of Spain, which granted me the scholarship that allowed me to come to MIT. I also want to mention here some people without whom this thesis would have not been possible. In first place Aurelio Menendez, who always provided the support and advice I needed, not only for this thesis, but also for other academic and personal situations. Secondly, professor Dennis McLaughlin, with whom I have worked closely during all the thesis period. He made of this thesis the most important part of my education at MIT. Besides, he provided the final economic support. Finally, I also want to thank professor Ralph Gakenheimer who, although this thesis did not fit in his area of main interest, was interested and supportive throughout all the period. 3 I want to dedicate this thesis to my friend Aurelio. 4 IDENTIFICATION OF ROBUST WATER RESOURCES PLANNING STRATEGIES TABLE OF CONTENTS 1.- INTRODUCTION 1.1.- 1.2. 2.- - OVERVIEW OF A MULTIOBJECTIVE METHOD FOR CONSIDERING ROBUSTNESS 14 ORGANIZATION OF THE THESIS 21 LITERATURE REVIEW 2.1.- CURRENT METHODS IN WATER RESOURCES PLANNING 2.1.1.2.1.2.- 2.2.- 2.3.- Optimization models and Simulation techniques 25 27 27 28 ADAPTATIONS OF SYSTEM ANALYSIS TECHNIQUES TO DECISION MAKING NEEDS 31 2.3.2.- 2.3.3.2.3.4.- A METHOD FOR DECISIONS 3.1.- techniques CURRENT USE OF PLANNING METHODS 1. Institutional resistance to use system analysis techniques. 2. Lack of communication between decision makers and analysts 3. Institution's conditions to use system analysis 4. Institutional situations LDC's 2.3.1.- 3.- 12 Multiobjective analysis of nearly optimal Identification alternatives Stochastic planning Indices 1.Reliability index 2.Vulnerability index 3.-- Resiliency Index 4.- Robustness index INCORPORATING ROBUSTNESS OF THE METHOD Screening models 3.1,1.Step 1: Derive the ideal curve - Decision Variables - Uncertain Parameters IN PLANNING 30 30 32 36 37 43 44 44 45 47 49 50 DESCRIPTION - net benefits 52 5 3.1.2.- 3.1.3.- Ideal Step net benefits 2: Select Step 3: Assess the of the alternatives Curve of net - - 4.- Index of alternatives performance of Net Benefits of an robustness Step 4: Compare Final remarks among alternatives 75 DESCRIPTION OF THE CASE 76 STUDY 81 82 83 87 95 SCREENING MODEL 4.2.- 4.2.1.4.2.2.- Decision variables Objective function 4.2.3.- Constraints 4.2.4.- Uncertain parameters APPLICATION OF THE METHOD TO THE CASE STUDY 4.3.1.- Step 1: Derive the ideal net benefits curve 4.3.- - Ideal net benefits - Curves of - Expected value of net benefits for th e alternatives Indices of robustness for the alternatives Step 4: Compare among alternatives Conclusions about robustness and expected net benefits 4.3.4.4.3.5.- net 113 114 115 117 121 Summary of Characteristics water 103 111 benefits CONCLUSIONS - 97 98 curve - Optimal sets of design decision variables 4.3.2.- Step 2: Select candidate alternatives 4.3.3.- Step 3: Assess the performance of the alternatives 5.- 68 70 Summary CASE STUDY 4.1.- 57 61 benefits - Expected Value Alternative 3.1.4.3.1.5.- curve candidate the method of the resources method planning within the 121 123 framework APPENDIX C 126 137 154 BIBLIOGRAPHY 169 APPENDIX A APPENDIX B 6 INDEX OF TABLES AND FIGURES CHAPTER 1: Figures FIGURE 1-1: FIGURE 1-2: FIGURE 1-3: Two-objective tradeoff curve for alternatives A, B, and C 16 Potential of a basin used for irrigation under uncertain crop prices 17 Net benefits uncertain 19 of an Alternative crop prices FIGURE 1-4: Delta curve of an Alternative FIGURE 1-5: Comparison Alternatives of A, A under A three delta curves, for B, and C 20 21 CHAPTER 2: Figures FIGURE 2-1: Theoretical multiobjective optimal solution 33 FIGURE 2-2: Typical form of utility 39 FIGURE 2-3: Continuous and discrete probability functions for uncertain parameter Ck curves 41 CHAPTER 3: Figures FIGURE 3-1: Discrete probability functions for the uncertain parameter 01 z discount rate. 54 FIGURE 3-2: Typical ideal net benefits curve for a basin under uncertain discount rate. 57 FIGURE 3-3: Typical bar graph representation of an hypothetical probability distribution of values of a design decision variable Xu obtained from the M1 M2 Mn optimization runs performed in Step I 59 FIGURE 3-4: Net benefits curve of an hypothetical 63 7 alternative under uncertain discount rate FIGURE 3-5: Delta Curve of an hypothetical 65 alternative, by subtracting the net benefits curve (Figure 3-4) from the ideal net benefits curve of the basin (Figure 3-2) FIGURE 3-6: Robustness related to the shape of the delta curve. Comparison between Alternative A, a non-robust alternative, and Alternative B, a robust alternative. 66 FIGURE 3-7: Typical curves 69 FIGURE 3-8: Derivation form of the Pareto frontier of the Pareto frontier curve 71 CHAPTER 4: Tables TABLE 4-1: Fixed and variable possible projects basin TABLE 4-2: Parameters needed for the screening model 91 TABLE 4-3: Intervals in which the continuous probability function of the uncertain parameters have been divided 99 TABLE 4-4: costs for to be build Ideal net benefits the in the curve 86 102 TABLE 4-5: Probability distributions of the values of the design decision variables 104 TABLE 4-6: Probability reference variables 106 TABLE 4-7: Projects TABLE 4-8: Compatibility table 109 TABLE 4-9: Reference sizes for the project tII TABLE 4-10: Candidate alternatives 112 TABLE 4-11: Expected net benefits and index of robustness of each candidate alternative 115 distribution of the Values of the design decision after custering more likely to be built 107 8 TABLE 4-12: Decrease in expected net benefits and in index of robustness of the non-inferior alteranti ves with respect to the maximum values 117 Figures FIGURE 4-1: Scheme of FIGURE 4-2: Total river year FIGURE 4-3: Seasonal distribution of total inflows for the typical year FIGURE 4-4: Water demands for typical year FIGURE 4-5: Return of the irrigation non-comsumed by crops FIGURE 4-6: Assumed probability crop prices distribution of (9: 84 FIGURE 4-7: Assumed probability electricity prices distribution of e2: 84 FIGURE 4-8: Assumed probability distribution discount rate of 93: 85 FIGURE 4-9: Assumed probability distribution or decrease in general increase costs construction of 94: 88 FIGURE 4-10: Assumed probability distributi on of e65: general increase or decrease in i rr i ga ti on water demands FIGURE 4-11: Hierarchical relationships FIGURE 4-12: Pareto the basin 77 inflows for typical irrigation graph two-objective 78 river 79 the 79 in water representation among projects of candidate alternatives APPENDIX the 81 110 of for 93 the 116 the case study A: Figures FIGURE A-1: Construction costs FIGURE A-2: Costruction costs of of the dams 135 the irrigations 135 9 FIGURE A-3: Construction costs of the power plants 136 FIGURE A-4: Construction costs of the 136 transfer APPENDIX B: Tables TABLE B-1: Probabilities of the 243 elements of 138 the uncertain parameter vector & TABLE B-2: Optimal sizes of Dam A 139 TABLE B-3: Optimal sizes of Dam B 140 TABLE B-4: Optimal sizes of Dam C 141 TABLE B-5: Optimal sizes of Dam D 142 TABLE B-6: Optimal sizes of Irrigation 143 TABLE B-7: Optimal sizes of Irrigation 144 145 TABLE B-8: Optimal sizes of Irrigation TABLE B-9: Optimal sizes of hydropower Plant A 146 TABLE B-O: Optimal Plant C 147 TABLE B-1I: Optimal sizes of Transfer sizes of hydropower 148 Figures FIGURE B-1: Bar graph of the probability of the sizes of Dam A distribution 149 FIGURE B-2: Bar graph of the probability of the sizes of Dam B distribution 149 FIGURE B-3: Bar graph of the probability distribut ion of the sizes of Dam C 150 FIGURE B-4: Bar graph of the probability distribution of the sizes of Dam D 150 FIGURE B-5: Bar graph of distribution the probability of the sizes of 151 Irrigation B Bar graph of distribution the probability of the sizes of Irrigation FIGURE B-6: 151 C 10 FIGURE B-7: FIGURE B-8: Bar graph of the probabil ity distributi on of the sizes of 152 Irrigation D Bar graph of the probabil ity distributi on of the sizes of hydropower 152 Plant A FIGURE B-9: Bar graph of the probability distribUti on of the sizes of hydropower Plant FIGURE B-10 APPENDIX 153 C Bar graph of the probability distributi on of the sizes of 153 Transfer C: Tables TABLE C-i: Ne t benef i ts of A 155 TABLE C-2: Ne t benef i ts of Al ternative B 156 TABLE C-3: Ne t benefi ts of Al ternative C 157 TABLE C-4: He t benef.zts of Al t erna t i ve D 158 TABLE C-5: Ne t benef j ts of Al ternative E 159 TABLE C-6: Ne t benef i ts of Al ternative F 160 TABLE C-7: Ne t benef i ts of Al ternative G 161 TABLE C-8: Ne t benefi ts of Al ternati H 162 TABLE C-9: Ne t benef i ts of Al ternative I 163 TABLE C-10: Ne t benef i ts of Al terna t i ye 164 TABLE C-11: Ne t benef i ts of Al t erna t i ve K 165 TABLE C-12: Ne t benef i ts of Al ternati ve L 166 TABLE C- 13: Ne t benef i ts of Al t erna ti ve M 167 TABLE C-14: Net 168 benefits of Alternative Alternative ve N 11 IDENTIFICATION OF ROBUST WATER RESOURCES PLANNING STRATEGIES CHAPTER INTRODUCTION 1: This thesis particular, water resources planning and, deals with with a strategies. Through for method developing in robust planning a litera ture review, I have realized that water resources planning has a different meaning for engineers than for planners. Engineers resources planning as a series of optimization models. Engineers under the generic name of system planning methods mean generally consider water mathematical techniques and often group these techniques analysis. For improvement of the engineers new mathematical techniques, perhaps at the sacrifice of decision-making needs. On the other institutions, social and hand, planners consider water objectives, decision making processes, economical resources and other issues. But planners do not usually use analytical tools to introduce their concerns into a practical analysis. This thesis should as narrowing gap the between be considered theory and an effort for practice, between 12 engineers and planners. The main planning is objective to maximize in traditional expected net water resources benefits. With this single criterion, the project or system of projects chosen for implementation is generate benefits. most But the one water that will more net resources planning situations are subject to some degree of uncertainty. The actual construction of the that project begins time years after the plan was completed. In interval, planning forecasts. some variables Hence, the may change actual net from the benefits from the project may differ from the predicted net benefits. As a consequence of uncertainty, there of possible is a net benefits to obtain from a project. Robustness measure possible net of the dispersion benefits. If net benefits depend heavily other words, non robust the value Robust projects, to maintain of that distribution of the distribution of net benefits widely spread, the project depend on is a distribution is considered on the is non robust, because uncertain conditions. In projects are those whose performance that uncertain variables happen to take. on the other hand, relatively constant are those which are able net benefits under a range of conditions. Since resources of uncertainty planning, a project, because is almost robustness it is indicates always present in water a desirable characteristic relative guarantee of net benefits from the project. Another desirable characteristic of 13 a project is to produce as much net benefits as possible. In most water resources systems, however, the most robust project is not the one which produces the greatest net benefits. There exists a tradeoff between projects with robustness and net benefits: the greatest robustness are those those with lowest expected net benefits and vice versa. The central decision making method does modification topic method this for not represent of existing decision makers' robustness of thesis water a is to resources mathematical planning information needs. introduce a new planning. The advance, practices to but the provide The method evaluates the and expected net benefits of the projects and ends up with a two-objective problem. 1.1.- OVERVIEW OF A MULTIOBJECTIVE METHOD FOR CONSIDERING ROBUSTNESS The multiobjective this thesis is screening model from all intended is to be projects an objective problems comes from identified maximization that would the traditional programming. with of respect benefits generate the use of A to propose in applied to screening models. A which that could be built function. The linear method an optimization technique possible projects the set of mostly decision-making an identifies in the basin, optimal value of solution to screening optimization algorithm, single optimal an unique minus costs. alternative is objective function: The incorporation of 14 robustness into screening the mode Is, considered and decision because provides expected net be nefits) thes is This benefits to that we projects which have great candidate are even or, able no single robustness both. To Such alternative an decision optim and or expected net great These projects are ca lIed alternat ives candidate al alternative to exist, exists. In fact, it should have the the greatest expected net benefits. is problems. hig hl y Therefor e, unlikely we need in to two-objective compare candidate alternatives, one of w hich will be the final Comparisons are plotting the points alternatives. Figure Alternative A has 1-i c or responding shows three the lowest choice. the of is obtained candidate this curve. the greatest expected net benef its, expected net neither the greate st robustness benefits. However, makers, the to points lowest robustness. Alternative C has the but among made by using a two-objective tradeoff curve. The tradeoff curve, also calle d the Pareto curve, by and final project cho ice. consider for a single optimal alternati ve robustness (robustness be identify several alterna tive to better, greatest to both robustness and expected net criteria for alte rnatives. implies that uncertainty two-objective considers Suppose benefits allows traditional tradeoff graph. important be it a improves process probab le if greatest benefits. nor the but the robustness, Altern ative B has greatest expected net this graph were presented t o a decision chosen alternative would be B , because it 15 has almost as great robustness as C, and almost as great expected net benefits as A. FLISIILITY ALTERATIV C FLICI ~- A-, T M(3 (A) Two-objective Alternatives A, The Pareto i I IM(C FIGURE 1-1: I ALTERMI or tradeoff I n tradeoff B, and C curve for curve is the final result of the two-objective analysis. But before obtaining the Pareto curve, two values for every candidate alternative must be calculated. The first value is the alternative. expected Cost-benefit net techniques benefits are widely calculate this value. The second value is the the I found given alternative. formulation of robustness in prior stage to obtain have the literature. the Pareto method to measure robustness not curve, I of the given used to robustness of any practical Therefore, as a had to develop a in water resources projects. Robustness may be evaluated by assessing the potential of 16 the basin to produce net benefits under uncertain The potential of the basin is the obtained from an ideal project which all hydraulic resources that are favorable (for example, more projects benefits if crop be has. When conditions prices are greater that of the obtained basin from is greater the irrigation in the basin). However, if conditions are unfavorable if crop prices drop) the potential of the basin (for example, decreases. Figure benefits can net benefits is always able to exploit the basin what was forecasted) the potential (i.e. expected conditions. of 1-2 basin a shows the assuming potential or maximum net only irrigation projects and uncertainty in crop prices. The upper curve shows the potential form of the distribution of expected of the net basin in the benefits as a function of crop price. The lower curve shows the uncertainty EETED EV NWITS FOTENIALY ITS f MfTMI CM0F PFICES (:CJ DISTIJTI(3 OF CROP PFRCES Frob (C) FIGURE 1-2: Potential of a basin used for irrigation under uncertain crop prices iT in it because small, this high. In and Curve, will crop prices of the potential thesis, basin the of be that is called alternatives the is as reference importance its robustness evaluating for curve these high potentials has unlikely that is Net Benefits Ideal basin that the Therefore, small. values is that crop prices take these high the probability see that the probability we also but increases, crop prices basin increases as of the the potential see that curve. We its probability distribution represented by prices, crop is discussed below. a measure of Robustness of a project has been defined as possible net benefits as in the distribution of the variation net benefits consequence of uncertainty. Variation of project has measured be to appropriate reference net other reference, benefits net variation of The most something. is the potential of the basin to produce any With benefits. to respect with is misleading the measure of the measure of a as robustness of projects. If we do not take the potential of the which give basin as reference, robust projects would be those the net same certain may conditions performance should the most robust yields benefits The net that basin in fact would be same net benefits benefits of depends also a given be ignoring that condition, any be expected from the project exactly the under favorable and project. As a result, nothing, which to build for any conditions: alternative to on the uncertainty of better zero. be built in the variables. 18 For example, net benefits of an prices, as shown in Alternative A Figure 1-3. When crop depend on crop prices are lower than Pi, Alternative A yields negative net benefits. words, the agricultural costs are greater In other than the benefits from selling the agricultural products at price P 1 . EIVECTE ET EEFITS KT RMITS & ALTEATIVE A CmF ICM FIGURE 1-3: Comparison Figure 1-3 Alternative Net benefits of an Alternative A under uncertain crop prices of Figure 1-2 (potential of the basin) and (net benefits of Alternative A is from realizing curves are is directly Curve" for Alternative A the two the shape of two delta curves, for Alternative A curves to be (see Figure i-4). Delta important because the robustness related to how far the full potential of the basin. We define the difference between the "Delta A) indicates of an alternative its delta curve. Compare and for Alternative C 19 FOTEIIIL KT DEEFITS OFN NET HIUlTS EIPECTED .. ,/ f' 811I . ....... . ET DEFUITS OFETERiTE I 011TI CUIE OF ETEltTITE I CIPlCES /icto Delta curve of an Alternative A FIGURE 1-4: (Figure 1-5). The horizontal. This curve for Alternative crop benefits depend Alternative A strongly Alternative potential but Alternative A the is that Alternatives basin, for increasingly less A any the net on the crop achieves the full other crop price, attractive. We may conclude B and C are more robust than Alternative A. Although the assessment and original almost price. However, prices. For crop price P 2 , of is shows that the net benefits from Alternative C are relatively independent of from C contribution of this criterion for decision making. delta curves were shown thesis, it of robustness is an is not a sufficient Consider two in Figure robust than Alternative A. study alternatives whose 1-5. Alternative But Alternative A C is more is much closer 20 ELTA C" rucis It FIGURE 1-5: Comparison of Alternatives A, three delta curves, for B, and C than Alternative C to reaching the potential of any crop price. In other words, although A is C, A always produces more net benefits the basin for less robust than than C. There is, therefore, the need to consider robustness along with expected net benefits as 1.2.- in the Pareto tradeoff curve in Figure 1-1. ORGANIZATION OF THE THESIS Chapter 2 serves as an introduction to the resources The planning methods, first part techniques and of the called system analysis chapter discusses their reviews the is that in real problems system analysis techniques. techniques. existing planning current use situations. There are some institutional the utility of current water Part of planning that reduce the problem these techniques do not completely respond to decision 21 making information needs. Four improvements to traditional system analysis techniques are analyzed in the second the chapter: (1) multiobjective analysis, of nearly optimal alternatives, (4) the use of (2) identification (3) stochastic planning, and performance indices. Chapter 3 resources introduces the planning. consequence of My concept of concern robustness with is a the presence of uncertainty in water resources in the distribution of of uncertainty. Also step in water robustness variables and parameters. An index of robustness based part of method for in projects' outcomes as consequence this Chapter, identifying the is formulated I describe a step-by- robustness-net benefits tradeoff curve based in a screening analysis. In Chapter 4, in a the general specific hypothetical application. This case is appropriate Possible set of projects are and solved, and tables and three intrabasin 5 used resources planning deciding the most with be implemented in a basin. irrigation transfer. The computer summarizes water resource systems, method limitations are to 3 is outputs areas, hydropower case study are is fully accompanied in appendixes. Chapter the dams, of Chapter water concerned projects plants, an approach represents of the importance and comments on to traditional the robustness method the of robustness in improvements that screening models. Some and possible improvements also discussed. 22 LITERATURE REVIEW CHAPTER 2: Variables describing the system is called a model. Models forming what equations, mathematical of response the can be used to predict through represented be can relationships their and system is of mathematical development the the analysis. for techniques planning resources water require to complex enough a of design The system the and to evaluate the benefits derived from water resource project. of formal and objective evaluation method is Some degree always present in use planning water some mathematical management science, many names: analysis. This actual operations planning is system system analysis techniques section summarizes used in water resources planning modeling, engineering, etc. The system water resources most standardized name in for techniques analysis, or solution. There are given research, Therefore most planning. resources and comments on their use in planning situations. not all Of course, issues reducible to mathematical form, systems fully model. In political, understood, or water resources economical, and of interest to the planner are easy to there all are nor water resource identify, describe, and are many institutional other social, factors which can 23 only partially be in found introduced provide designs, optimal information making process. Since is planning will on as multiobjective such optimal nearly the is than search alternatives, planning, do not end up with the optimal stochastic provide of identification analysis, Concern to decision makers. New help planning, methods in water resources models. which, rather new ones techniques and to devise for formal adapt existing system analysis to literature the in and system but for the decision- alternatives the decision-making oriented approach to issue a central in this in this be extensively considered later these techniques thesis, chapter. CURRENT METHODS IN WATER RESOURCES PLANNING 2.i.- et Friedman al. (1984] in water currently used (1986] Rogers and Fiering Rogers related problems. [1979) and is study which a describe and techniques models review similar but limited to problems which involve some optimization. Following Rogers groups of system optimization models and Fiering [1986], techniques: analysis and techniques, with search and sampling techniques, and techniques, (4) related techniques analysis, and system analysis discussed (3) (cost-benefit techniques The first and (2) (1) the Analytical Simulation combined Probabilistic models techniques, Statistical game theory). there are five main and analysis, two only (5) Other input-output are the most used ones that are here. 24 2.1.1.- Optimization models and techniques. planning technique variables, define the configuration Variables Variables. is system which the solution provide and constraints. objective function, and operation of the value They are formed by decision 1979]. [Rogers, parameters, Decision widely spread water resources are a Optimization models being optimized. Their to the problem. For example, when trying to obtain the best reservoir size, the main variable the volume is the reservoir. of Parameters. Parameters fixed properties of describe the the system to be modeled. Parameters are independent and their values do However parameters which are likely to which are or Known, of the life be frequently demands) can prices and well not the change during of the model. particular run during one not vary project (i.e. water varied in independent runs creating sensitivity analysis. function. Objective measure quantitative The relationship of main objective the function objective common most of decision variables Constraints parameters and variables that and characteristics. equations in form of are the equalities, and of the the is a projects. mathematical that parameters project. relationships among the system operation constraints are mathematical describe Normally is from the describes the benefits minus costs Constraints. function objective The inequalities, integral, and 25 differential constraints season is equations. A typical for a reservoir: equal example are the continuity the water stored at the end of a to the water that was stored at the beginning of the season, plus all the inflows received, and minus all the releases and diversions during that season. Optimization techniques require a formal search procedure for the set of decision variables that optimize function while satisfying all the constraints. When objective function and constraints can be expressed equations, set the of the objective as linear algebraic which maximize decision variables objective function can be found with a technique called programming. Several are available in commercial software packages. algorithms the linear to solve linear programming When some of the variables can only take an integer value (zero one), or the integer programming. way to for optimization problem may be solved with The use of constraints introduce more example, programming fixed integer programming provides a costs is also available for in the facilities. of solution for for special the decision non-linear cases, variables. problems, but such as quadratic constraints remain linear, but as, Integer from linear programming objective function and constraints functions quadratic linear problems, software packages. Non-linear programming differs that the into the in may be non-linear There techniques programming objective is not general are available (in which the function takes form). 26 Dynamic programming linear problems is a method to solve linear and non- which have a sequential problems can be divided in stages decisions are required at given stage general affects to software (i.e. character. years or seasons), and each stage. A decision taken in a the next stage. Although there available computational procedures Such for dynamic are relatively is not programming, simple for a limited number of stages and decisions. 2.1.2.- Simulation techniques Simulation performance techniques of the system parameters. Simulation function. In produce under techniques that case, for each information different can on sets include the of input an objective simulation run, a value of the objective function is obtained. By performing many runs, a response surface formed by the values of the objective function can be created. Some sampling or search procedure can examine the response surface and obtain nearly optimal solutions. 2.2.- CURRENT USE OF PLANNING METHODS Assessment on in actual authors. water the use of system analysis planning methods resources planning situations differ among Two recent surveys show very distinct results. Rogers and Fiering conclude that (1986], using agencies and in part results from Rogers major consultants [1979], only appear to 27 use system analysis techniques in few cases. This pess imistic view differs from the conclusions of Friedman et al. [1984], using results of a study performed by the U.S. Congress Off ice of Technology Assessment (OTA) in 1982. Friedman et al. [1984) conclude that water agencies and organizations extensively use mathematical problems models to find solutions conclusions part from the different meaning of the authors. Rogers used to solve potential techniques system analysis techniques [1986] [1982] only researched surveyed techniques any kind of water related problem. Despite their differences, there is a that the in these studies result in and Fiering optimization techniques, while OTA can be institutional currently water resources efficiently and effectively. The contradictory for to of common system analysis improved issues conclusion in with the water limit the application both studies: and other mathematical understanding resources of in the of some agencies techniques. which In what follows, four issues are discussed. 1. Institutional techniques. require resistance Sophisticated great not easily planners understood and techniques were mathematical specialization. their study conclude that to Rogers complex water by engineers, use system models of and Fiering analysis analysis [1986] resources models in are many decision makers. Even senior trained before system used, have problem to understand analysis the methods. 28 Then, those who departments, supposedly have head difficulties agencies divisions to accept system analysis new planning methods. In addition, the use of for non specialized people Rogers and Fiering [1986] confidence 2. of the of system analysis produce wrong results. For may communication Part of the newness and difficulties of is also lack of undermine communication the is and of the them. But there communication due to the use of optimization the given fact economical more the a consequence techniques tools that "guarantee" the best problem. Single best solutions do not leave any room for negotiation, and that come in even decision makers and decision makers to understand lack of solution for between complexity of techniques as mathematical that as in the new techniques. Lack analysts. may this and ignore environment part of in what the social, decisions from techniques political, are and taken. When analysts use system analysis as substitute for decision making judgement, decision makers imposition and However, are likely to perceive it as an a threat to their authority. if techniques are used with the perspective of providing decision making needs for information, communication between analysts makers Miller, 1985). under the and decision Techniques are different decision makers need able is to improved perform (Meyer and the analysis optimization criteria and policies that to evaluate their decisions. In this 29 sense, system analysis can become very useful 3. Institution's Development resources of support system agencies specialized requires at and within facilities, institutions do some of finally planning method situation, planning agencies A priory developed (DC's). no have overall in their evaluation decide [Friedman et al., systems countries in already in place and analysis do not result the exists 1984], among DC's current water planning in developed resources is systems, in resources systems to which system analysis effective (Rogers, 1979). LDC's there decision making process are lost. in less developed countries. than major water and the possibility should be more effective (LDC's) to use was expected. To complicate coordination Institutional situations While of conditions could not be techniques that not of sharing resources and experiences among agencies is 4.- water training agencies these four available and system analysis this analysis. least four main conditions: introducing system analysis system analysis, more system techniques computer methods. Consequently, when efficient use and availability of data. For Friedman et al. agencies strategies for to analysis personnel, people, [1984], conditions and accepted. still only in less countries systems are made on small large undeveloped water Other advantage application in LDC's is is simpler and easier. Simpler is more that the in the 30 sense that decisions number of parties are more centralized, what reduce the involved in the process. Easier because objectives are fewer (normally national or regional economical growth) and technical more clearly analysis. Institutional similar to those on DC's, arise from resources, the and lack (1979) the trained of application reports of in main the LDC's are problems could personal, detailed system simplifies problems although of what lack or non reliability of Rogers cases defined, computational data. characteristics analysis techniques of 22 in the developing world. Most of them were successful. 2.3.- ADAPTATIONS OF SYSTEM ANALYSIS TECHNIQUES TO DECISION MAKING NEEDS There are optimal differences between best in the real world and in the mathematical is unlikely 1982]. To is enter more improve system variables, complex equations. Large-scale for research models, significant because not increase they understand, analysis more the solution first tendency relationships, and more models represent a challenge and devise of solution techniques. But complicate models may be simple mathematical the best solution for the planning problem (Chang et al., to world. The may less only effective because in the real they in the efficiency of be too complicate large and world than may not represent a the model, for other but also people to apply, and solve. 31 System analysis techniques are supposed to making aim decision- information needs. The analyst should realize that what causes a project to be implemented is a decision, and models, and techniques. that the decision system analysis alternative is These, however, increase the chance correct. Therefore, to identify projects not the and analysts should use performances and consequences of be able to clearly present the information, recognizing that decision makers are normally not familiar with the technical analysis. Four improvements to system analysis are literature. These analysis, identification of nearly optimal stochastic (2) improvements planning, and properties of are: (4) definition of (1) found in the multiobjective solutions, indices to (3) indicate the alternative designs. 2.3.1.- Multiobjective analysis River basin problems are concerned with the allocation of water among several uses and development Traditionally planners have used a single maximization efficiency for of national income, (benefits minus costs). river objectives reduction, basin are to is environmental, regional etc. Solutions forward when development development, water is economic also objective: called economic However, public investment multiobjective. Different recreation, national optimization a single objective alternatives. unemployment self-sufficiency, problems are straight considered and the rest are 32 ignored. The solution is not that easy when the problem has conflicting objectives. However, theoretically, single optimal solution exists. The procedure steps: (I) find the to find it consists possibility frontier curve projects which represent the best possible objectives); (2) find the in three (formed by the tradeoff among the social indifference utility curves (that show the social tradeoff among the objectives); find the curve with tangent of possibility (3) indifference the maximum possible social objective the and frontier. Figure 2-1 displays the method for the case of two objectives. cum SOCIAL lilti22 WIERI I POSnLIY MML t SOLUION OBJECTIVE I FIGURE 2-1: Theoretical solution Although theoretically simple has almost determination unsolvable of optimal multiobjective and practical logical, problems. the measure of merit for each the process First the objective. How 33 to measure the benefits according to the markets for conventional was a number many of of ecological p e i preservation?: wild animals? There these objectives, and would be with monetary measures of the problem problem arises possible to merit for are not then there are not measure of benefits and costs. market, it convert from In fact, if there fi nd a way to end up any and to objective, in single-objective. The multiobjective are not because merits for different objectives comparable. Second indifference practical analysts when problem curves. the select the curves, authors do not because preferences indifference curves, determina tion They are very difficult applications. Many quantifying have is of of of to formulate for recommend that the prac tical the social difficulties society. Without social the multiobjective problem do not likely single optimal solution. The third problem is curve. For many analysts, step. Almost centered finding efforts new the frontier curve. optimization, constraints: of the analysis multiobjective and more The since effective basic is f rontier ju st this are planni ng meth ods for problem to s olve the objective functi on is is a a scalar: objective function: and in studying instead of generation multiobjective in called vector vector all the Max Z Z ZR CRi Aij Xi [Zi, Z2 '.' Zp] Xi - Bj 0 34 Xi A vector cannot be directly maximi zed or minimized. a set of non-inferior defines 1 0 the called solutions be can obtained. the optimization the problem. techniques of more yet to be developed up off trade resources to study, many available, for cases worth In their are method objectives, the the vector conclude that of not have practical an efficient method only the is called subrogate be appli ed when great computer For Cohon available). four can [1975) res ources problems. Also, than four objectives (presently, Marks solve they do to multiobjective water application to developed techniques This set possibilit ies frontier curve, or Pareto simply frontier curve or Pareto curve. C ohon and summarize Only and best Marks, for me thod cases with is the constraint method. The constraint method consists in the following optimization: objective function: Max Zm m with the new constraint: ZI 1 LC R old constraints: ZR = Cki and the Aij Xi Then, Zm is maximized other objectives. the advantage If this Xi that sensitivity - Bj 1 0 1 0 subject to problem is XI lower limits Ll in the LP, it has solved using analysis (included in most of LP packages) can rapidly obtain solutions for different values 35 of LI's. There is small chance Therefore available. alternatives could no that social utility curves further be taken, choice are among non-inferior and the frontier curve is to be presented to the decis ion makers. Identificatio n 2.3.2.The only global optim identification same the object ive problem but with agreement for terms of are significantly dif the in not the maker; with in acceptable function represent involved is decision alternatives sol utions to the fo r ne gotiation and the dec ision very clos e process. These other alternatives, solution is model more possibilities par ties among mathematical d ifferent of ranges of um of a interest of solution of nearly optimal solutions making t o the optimal the objective function val ue, but which ferent decisions in terms of the decision variables values, are called nearly optimal solut ions. The solutions is Optimization optimal pr oblem major the of computer algorithms are designed nearly optimal generate ne arly softwar e to little reac h optimal available. the single info rmation about solutions. Harrington and interesting procedure which has identifying lack solution and they when using for a LP Gidley [1985] to analyze package. The a single optimal suggest a simple and nearly optimal alternatives procedure assumes a LP problem solution: 36 objective function: Max and constraints: Z Z = CI Xi Xi - Aij Xi and the single optimal Bj S 0 1 0 solution is: =Z Z To generate alternative optima the software package is repeatedly applied but introducing a new constraint: Ci In case 1 a ZN Xi = i the global optimal solution ZN that a again. For other nearly any values inferior to nearly optimal" Harrington and nearly optimal the optimal alternatives, is obtained a should take i. As closer is the value to 1, "more solution Gidley [1985) solutions is. The prove exists studies made by that a great number of within 0.5% of the global optimal objective function value. Other attempts functions in which variables are varied, new optimal the been made maximum to value generate objective of some decision randomly constrained. When these constraint are objective solutions The have functions are maximized, and nearly are obtained. consideration potential not only of producing better of nearly optimal solutions has the producing better decisions, but understanding of the nature of also of the decision problem itself. 2.3.3.- Stochastic Planning 37 Many of resources system the factors that define the systems is cannot be known Most of the planned. performance with certainty when the p1 ann ing is uncertain circumstances. The simplest a pproach simple probabilistic measure uncertain variables, and then of water (mean, under is to take some median, proceed as made for the mode) on a deterministic problem. There is a second approach to deal with is to evaluate the consequences system design. Loucks et al. this approach. considers The maximum (X W)] rev iew some expected net methods under benefits (X W ) are i given design p the net is described benefits of probability of occurrence expected net benefits for the system whose by the decision variab the uncertain parameters taR e the of value given alternat ive In Other of the max-min project the methods method, uses given projects, are the calculated, a most the the one wit h better which provide s olut ion. chosen benefi al. ts projec t whose smallest net benefits are the greatest. favor alternatives thi s method Loucks et net the with greater [1981], pessimistic cri teri on: smalI le st and when is Pj essence, pr ojects, in dicated by les Xi, and The project Wj chosen. the Wj, average performance. It can be call ed compromise all the method j i selects, of uncertainty in a given the following problem: E [HB where NB (1981) uncertainty that It is from of every is the one a a minimum of benefits way to every 38 year over other alternatives more profitable but that may produce very low net benefits In between both methods, outcomes and other ignoring them, Its is basic decision point maker's risk premiums, Figure 2-2). include analysis, there define preferences, the in same conditions. overreacting is the utility to poor theory. utility function. From indifference situations, and a continuous utility curve might be drawn (see Keeney decision analysis how to to one on the average, and Raiffa [1976], the traditional book, show how to obtain utility curves and in them other attributes than money. In utility the new objective function value is utility, and the problem is to maximize it. VTILIYT I UflLITT CUM ET BIEMlTS FIGURE 2-2: Typical form of utility Utility theory has many curves favorable points. It represents directly social or decision maker's preferences and tradeoffs. 39 For example, bad outcomes ones. Sensitivity utility curves, participation. are more heavily weighted analysis and Difficulties likely makers, preferences. Consensus and then the in applied for different stated utility assess representing different have be conclusions quantifying preferences to decision could for social debate theory utility perspectives arise rejected and when curves. Different and political groups, of social values in a utility curve might be conclusions than good for those and impossible, with different utility curves. The second approach to deal identify optimal alternatives under main method of suggested original this approach in part by Loucks NB Aij Xi where Ci, Aij, In stochastic and (1981]. programming, We assume the Bj First, the Xi - Bj 1 0 1 0 represent the parameters Assume probability distribution There = Ci Xi linear programming deterministic. to be al. linear NB, Subject to: has is to uncertain conditions. The stochastic et uncertainty deterministic formulation: Max no is with that some of Ck is is represented are certain steps to be taken these of the model. parameters are uncertain, and its in Figure 2-3. to solve the problem. continuous probability distribution function for Ck approximated by a discrete distribution function, as 40 }I DISCET= FMCi0 v/i C0TIPM FEm MEITI FIGURE 2-3: p(CkI), its correspondent P(CR .. , two linear kind programming, define operation rules of The model of as a finite second dependent step it is is on Cki, important to design decision variables. Design decision Operational the projects extensively discussed discrete (q is of decision variables: variables define project sizes. is Cj probability associated p(Ckl) variables and operational decision This .. , )- In stochastic distinguish Ct Continuous and discrete probability functions for uncertain parameter Ck done in graph. Each interval Ct, CR2, number) has PAiAwnTU decision variables once they are built. in Chapter 3. to write the with i original = 1, 2,. ., q (q intervals of the uncertain consider the decision variables Xj, with parameter j optimization is the number Ck). We = 1,2, .. , k-1, also to 41 with j and the decision variables Xj, decision variables, be design = R,+1, ..n ,I decision variables. We to be operational can write: NB = C 1 Xi + C 2 X 2 + . NB 2 = C1 NBq Xi + C 2 X 2 + . = CI X, + C 2 X 2 + depending on condition, there a design + Cn Xn 2 + .. + Cn Xn 2 Xk-I + Ckq Xkq + .. + Cn Xnq Xk-i + CR + CK-i + Ck 2 conditions: set of operational possible benefit decision XK take variables uncertain are .. + CR-i the m aximum that obtains However, the + Xk- decision Operational Xk1 + C-i variables because the size o f the project can can different values for future decision variables from the projects. take only a value, be not any adapted to the possible future co nditions. The Max new optim ization problem is now: [NB1 p(C gi) + NB 2 p(CR 2 ) +.+ NB p(Ckql E NBM p(Cgm) Max m = t, 2,..q Subject to: Aij XiM - Bj 1 0 Xim ) 0 Note that the value of the superscript m Xi" depends on the type of decision variable: design decision variable: operational m does decision variable: m constraints of and the variables problem is is in not have any value =1, 2,..q. The new problem continues being linear, size in decision variables but the number of significantly increased. The big fact the main limitation of 42 stochastic As with linear programming. a conclusion, uncertainty: uncertainty, yield there (1) and to (2) uncertain are two forecast to choose benefits. evaluating methods main approaches to deal consequences et al. solutions, but (1981), for both approaches, that uncertainty models should not be used to best to of between alternatives which Loucks and models the eliminate when concludes identify single clearly inferior alternatives. 2.3.4.- Indices Indices performance are of criteria. All the quantitative that measures different alternatives indices discussed here under compare some given are consequence of uncertainty in water resources planning. Ind ices rank projects in regarding their performance under the fol lowing criteria: 1. Probability that the project fails 2. How bad are the consequences of the f ailure 3. Probability that the well i and 2 are reliability and vulnerability slightly reasonably under different demand conditions. Criteria indices p erform project will are appropriate, different, resiliency, and and robustness. respectively indices. although so For the 3, two definitions are criterion their their by measured proposed In what follows, formu Iae: the four ind ices are proposed and analyzed. 43 1.- Reliability Index Reliability index has no failure within [1982a] proposes measures probability the planning that the project period. Hashimoto et al. the fol lowing formulation: = Prob [ Xt E S ] Irel where: Irel Xt : S : Reliability Index of Reliability Variable that describes the system's output at time t (t takes the discrete values 1, 2,......) Set of all satisfactory outputs is opposite probability of failure. as 2.- 1 to risk, or what In this sense, is the same, risk index the is defined -Irel' Vulnerability Index Vulnerability consequences of not consider number of measures the a failure, given that the length failures, Vulnerability only Hashimoto index of nor time how magnitude it has until long the the of the occurred. It does failure, nor the failure lasts. refers to how severe the consequences are. et al.[1982a) proposes the following formula: Ivul = E Sj Ej, E F where: Ivul S Ej F Index of Vulnerability Severity of the consequences Probability that the consequences be with severity Sj Set of unsatisfactory outputs (failures) 44 3.- Resiliency Index Resilience is a term known ecology, materials, economy, and common definition generalized resiliency is associated to situations, and to in other structures. There for resiliency. adaptation to recovery sciences from like is not a In general, new (no projected) surprises and adverse situations. In water resiliency. resources First, for describes how quickly the failure. Their of there system will mathematical is the two Hashimoto et al. recovery. They propose the If TF are time that (1982a], formulation is [T F] E is possible 1 / (Prob to resilience likely recover from a based on the time following measure of resiliency: a system remains unsatisfactory after a failure, the index of resiliency is expected values, approaches i/TF; considering to define TF: [ X E S I X E F ) where: Xt+ i = Variable which describes the system's output at time t+i Set of satisfactory outputs Variable which describes the system's output at time t Set of unsatisfactory outputs (failures) S Xt F and by def inition of I re s : Prob Second, for the [ X Fiering probability that unlike events index of resiliency Ires: t+i E S [1982a, I X t E F ) b, c, the system will and d] resiliency is operate well occur. This concept of resiliency enough when is similar to 45 of notion the a when the partial for these robust to system is Fiering, to Fiering's distinction is and this to is small systems response the system if between robustness and resiliency, even robust According in certain variables changes derivative of the variables. But, below. described robustness is not it may be resilient as a whole. A certain variables resilient system accommodates the surprise produced in several of its by variables in changes relations among derivatives: dz/dxi A remaining variables. measured as Resiliency, therefore, should be total the : E (dz/dxj) (dxj/dxi) of the combination linear all total derivatives dz/dxi measures the resilience of a given system as a whole. Other criteria and alternative measures of resiliency are In fact he proposes (982b]. Fiering proposed by different alternative residence time in of indices non-failure up to eleven based resiliency, state, and in combinations of passage time between failure and non-failure states. total derivatives The method of the [1982a] used was resiliency of They that resilient when the the surprises" expected a system expected in and Marks systems agricultural concluded "unpleasant Allan by the degradation design performance in the [1984] by Fiering to measure the developing countries. in planning proposed can be expected to be degradation parameters planning is due to less than parameters themselves. 46 For practical considered to applications, be those which highly resilient systems contain many are redundances of design. For these systems, proper operation rules minimize the unpleasant effects of surprises. Large systems with many connections have proved to be the most resilient. 4.- Robustness Index For Fiering much the [1982a], same. robustness has measure of the robustness and resilience mean very However for Hashimoto other meaning. et They consider possibility and expenses of al. [1982b], robustness as a adapting a system to future conditions different from those for which the was calculated. information It is the cost about the future. The Prob I [ C(qJD) - of not having index they propose L(q) 1 3 L (q) system perfect is: ], where: Irob = Index of Robustness q = Future conditions D = A particular alternative (Project or Design) C(qjD) = Cost of accommodating the alternative D to the future demand conditions q L(q) = Minimum accommodating cost among all the alternatives 13 = Level of robustness Under the al. [1982a), which affect "demand conditions" there are the project which are uncertain and This thesis already introduced grouped the uses term used by Hashimoto et set of (demand, future conditions costs, prices, ... ) and likely to vary. a different in Chapter concept i. Robustness of of robustness, a project is a 47 measure of the dispersion project Accordin g as consequence to this of of possible net benefits uncertainty criterion , an index from the in some parameters. of robustness is develope d in Chapter 3. 48 CHAPTER 3: A METHOD FOR INCORPORATING PLANNING DECISIONS This chapter describes a method to resources planning ROBUSTNESS ident ify robust water strategies. The method is a multiobjective analysis of the water resources system. One of is to maximize expected net robustness and the is a measure of how sensit ive constant rob ustness 1, the project In exists. under is measure of project. The Chapter to maximize analyzing robustness the project parameters. A project is ca lied robust when thesis, is of uncertain conditions in the bas in. the presence relatively the objectives other benefits. The reason for is based project to uncerta in is its perf ormance In is th is a desirable property of a robustness, on the distr is Robustne ss conditi ons. different cons idere d IN ibution as of net indicated in benefits from under uncertain condit ions. mul tiobjective an alysis, The refore outcome the several n on- inferi or develo pment no of single best this method is alternatives for project to define the water resources system. The metho d begin s by displaying all possible projects to be bui It method with is in applied, the great er identified. One robustness of t he bas in. projects or and them will be From these, combinations expected the final net after the of projects benefits are development choice. 49 3.1.- DESCRIPTION OF THE METHOD applying the Previously to the screening model for the basin. projects the valu e greatest basin). maximize a partic ular, In uses for the water. Projects variables. the optimal most functio n and all consists expressions of define the The model contains Both, mathematical are parameters and decision variables. etc. maximize economic the mathematical If, for efficiency, economic the criterion the example, function objective for a typical to be repeated during the water system Constraints are mathematical the minimize efficiency, expression of benefits minus are considered measure of a quantitative maximize criterion: unemployment, in obtaining constraints. function and objective function is policy that possible decision their by constraints, and water system, the projects to build. se t of objective and costs group of defined are two main el ements: objective main a is value o f every decision variable, profit able The from the to be built solv e the screening model To (normally the benefits model represent poss ible projects including all net screening relati onships that mathematical those which provide the objective function to is function objective of the basin in built be that could possible all among is an model The screening selects that techn ique optimization itself, we must write method costs. year, that is is could to be Benefits assumed lifetime. expressions which show the 50 physical conditions institutional etc.), and (water continuity, conditions (water social-economical prices and costs, etc.) maximum sizes, priorities, etc.), operation rules, conditions (demands for water, of the basin or water resources a screening model system. The looks mathematical representation of like: (for the typical - year) objective function Maximize: for {XjJ - constraints Gi where Xj (XI, F (Xi, X2, .. Xm) , (i equations): X2, . . , Xm) are the decision variables. Once the technique screening problem is formulated, an optimization may be used to decision variables Xj, operating These rules. variables define the (irrigation areas, their optimal The models. (i equation): which normally are optimal most the optimum values of values profitable hydropower plants, project sizes of projects etc.), the and the decision to be built their sizes, and operation rules. most used Linear (LP), technique. The screening model screening screening programming - obtain that is models the mathematical is: models can most linear screening be solved using available representation (for the typical objective function are linear optimization of a linear year) (i equation): 51 Maximize: for (XjI - constraints (XI, Gi F (XI, X2 - , Xm) F (X) = Cj Xj (I equations): X2 9 . . = Aij Xm) , Xj - Bi 1 0 where: Xj are the decision variables, parameters 3.1.1.- and Cj, Aij, Bi are input (some of them may be uncertain) Step 1: Derive the ideal net benefits curve Decision Variables Decision variables rules. There screen ing are two model examples and 0f be of variables decision in the variabl es and operational dec ision variable s define the sizes bu ilt. capa city de sign project design and operating dec ision Des ign projects to reservoir, types des ign dec is I on variabl es. of the define of Area of hydropower irr igation, volume of p lant are typical decision variables. Op erational decision variables def Ine the operation rules of the projects once they are built. Rel eases irrigation are Design from examp les dec is ion of var lables r eservoirs and water diverted for operational a re of primar y analysis, s ince they defi ne what p roj ects what projec ts are not to be built. Al so, be built, design decision variables The a priori. values of In fact, d ecision variables. indicate interest are to be in our bui It and f or those proj e cts to their the design decision variables when we search for candidate sizes. are not Known alternatives 52 in Step 2, what we decision actually do variables alternative. In which is to find define general, design the the set candidate decision of design development variables form a vector X, such that: = (Xi, X 2 , X where m , . XmJ is the number of design decision variables, and where X, may represent volume of reservoir, X 2 Irrigation area, and so on. Uncertain Parameters The reason why predicted and actual project may differ is the uncertainty the particular case of linear parameters are a subset of the Bi. performance of the in some parameters. In screening models, uncertain input parameters Cj, Aij, and We group this subset formed by the uncertain parameters in a vector 0: e = (( , 1 4 2, ' n I where n is the number of uncertain parameters where ej basin, and agricultural We need uncertain and so on. the parameter. distributions example: may represent discount rate, e 2 price of products, to have used gaussian, existing in the The in distribution probability most common models engineering lognormal, and Gumbel, risk or of into a finite number of discrete every probability analysis log Pearson) continuous distributions. For present purposes, these divided of (for are must be intervals with their 53 associated probabilities If we assume that, is rate" as shown for example, and uncertain, in Figure it is 2-3 in Chapter 2. the parameter "discount approached with a gaussian distribution, Figure 3-1 shows the.division of the continuous distribution in discrete intervals. Prob (1) : VROAIllLITY 1 : DIs functions for the probability uncertain parameter 94 : discount rate. FIGURE 3-1: In general, T TE () Discrete after dividing In discrete continuous uncertain parameter vector e1 intervals, the is converted to: e = Oj where: I = t, 2,..,n j = 1,2, .. where Mi 1. We is Mi the number of also obtain intervals of the corresponding the uncertain parameter discrete probability of each interval: 54 p(E 1 ) To compute this uncertain probability parameters correlation exists some cases (for are we have independent; among them. example, to assume in other This assumption no correlation irrigation water demands and discount rate), cases some correlation may be present words, no obvious in exists between although in other (for example, between irrigation water demands and agricultural Ideal is that the prices). net benefits curve The ideal net benefits net benefits of ideal net robustness, the basin under benefits curve is as shown curve, we pick up parameters vector E, the maximum and, exploit the full conditions. The reference to obtain one point of that the We also for the optimal set potential calculate this element eiJ of the uncertain using the screening model, net benefits that uncertain the a particular decision variables defined by is formed by the potential in Chapter 1. To conditions defined by ei. design curve of basin may yield under the obtain the conditions of we obtain design optimal e5J. The set of projects decision variables the basin for that situation eij. For the element E 1 J of the uncertain let X*(e variables NB 1 J) represent the optimal parameters set of obtained from the screening model. [X*(E9J )] represent the net benefits vector e, design decision And let obtainable from the 55 projects defined by the design decision variables X*(e1J). The value NB [X*(e 1 5J)] eij, and, the situation the ideal net the ideal net benefits curve for: I j It is not = that we process is the uncertain parameter is fully defined. representation correspond project. to a single indicates, may expect. for be useful of the ideal typical zero when t hat def ines the Oi3 the of net benefits c urve optimum projects set to be of decision built. This net benefits curve of r ate is shown in Figure 3-2. n ever takes the discount rate equals discount rates element ideal net step s of the idea that point, to the project of there dis count the system. At The th e maximum possible an form net benefits curve Final Iy, a lso note that, ass ociated with th e next in every is 943, example, the final costs same ideal under uncertain equal a point of n .,Mi variables X*(e 1ij) of the important to note that the element A of curve mathematical uncertain parameters vector will When basin for is: , 2,. 1,2, curve benefits every of the [X* (8 J.- I)], NB does curve. the other elements Therefore, the net benefits the potential consequently, represents benefits repeated for all vector G, represents costs, the building any project internal rate of benefits from the project and there higher than negative values. the a basin In this It is return are are no net benefits. For internal rate of are greater than the return, the benefits 56 NET BENEFITS (dol lars) ,IDEAL, I I I ~~ I I I NET BENEFITS CUI I ! I - . . DISCOUNT RATE (11 Typical ideal net benefits curve for a basin under uncertain discount rate. FIGURE 3-2: which yields zero benefits and zero costs. curve remains internal rate of return of the system. net candidate We should identify alternatives benefits or both. that have decision of Select Step 2: that design There have Alternatives variables. Then the ideal net zero for discount rates higher than the benefits 3.1.2.- is none, build project to obtained. Consequently the optimal alternatives that have high robustness or, are Selection of defined by high expected even better, their design alternatives means selection identifying candidate decision variables. is no formal alternatives. However, procedure from Step for i, while obtaining the ideal 57 net benefits curve, we projects suggest the to procedures. Mi M In fact, number of optimal each element Gij. using the use of enough information some very effective in Step i it was optimization independent Mn 2 developed procedures analysis proposed informal necessary to perform of gave equal which runs, sets of decision variables X*(e From the on the these 3i), 1 one for optimal sets, below, we can obtain candidate alternatives. Several of decision procedures for the variables X*(O 1 analysis of j) are suggested procedure is to calculate the probability values the design decision var iables. of procedure that produces design. C onsider (XI, -. X2, the set , Xm, and in great of one into of variable volume of a reservoir). When the problem was uncertain Xu Mn 2 values optimization runs, of Xu"(E91 d is tr i but ion for Xu* (Gij) for: of ), one with j 1, 2, have the we E5j, the have M, = optimized for the value M 2 is therefore: p(e 1 3) . Mi a discrete probability design X them, the design for eac h combination of probability 1,2, .. , n system's of Xu was p(e i3 ). Since we performed the values of Xu i Therefore we values of the ma y, for example, represent the element with probability (Eij ), MI M parameters the is a very simple decision variables: particu lar (that here. The first This insight des ign sets distribution of decision Xu the optimal decision variable Mn different i and j. The distribution of the Xu. The analysis of 58 that likely values optimiza tion Step distribution probability I for is LP, de sign the algorithm us ed to the vertices for decision variable Xu. solve the thre e values. wh Ich the most reduce to If the screening model in tends to be highly This is because LP polygon formed by a few the otherwise possible values o f the design decision variable Xu- The second in shown identify of the n-dimension vertices the cons traints, infinite to the probabi li ty d istribution concentr ated around two or searches helps Figure is procedure 3-3 probabil ity distribution fo r a graphical representation, hypothetical an of Xu* (eiJ). Using easier to visualize the most probable values variable Xu. These values seem clustered example the like of the graph, it is of the decision in three groups. PROBABILITT Prob (I) II1I~, DECISIIARIABLE FIGURE 3-3: Typical bar graph representation of an of probability distribution hypothetical values of a design decision variable Xu, obtained from the MI M2 Mn optimization runs performed in Step I 59 The purpose we expect that values of Xu- likely three more this of variable, is procedure do not to use that producti on more that 1 eft when other and l es s Pro fitable pro, jects are built projects do not require the relationships among of existence of optimization runs project: (1) projects that prof itabl e more other projects. These by insp ection resu Iting from the M, performed relationships among projects are built. only after more identified be projects can the decision variables vectors different H owever built. projects are some there projects Therefore, are are are that is projects effectively are built sti 1l projects of in water the effective capac ity the table problems resources of a third among design decision res ourrces more projects. Il is a create that are for med by se veral 0 ther hav e profitable among combinations char acteristic of optimizat ion proje cts before than is incompati bil ities and resources systems those With the va riables. the decision variables, procedure compatibilities variables. A combining the most sugg ested. third The by possible all of values likely profitable. These decision but bef ore examining more the and the likely values of every decision identify may we graphs design alternatives identify are con structed the of robust both be to is to procedure candidate alterna tives likely values identify the could of each group, we the averag e value Taking M2 in Step I. Three suggested; a lways built for Mn kinds a given when that 60 particular project is built given projects one); (2) (more profitable projects that that particular project is built given one); with that to and (incompatible particular project. they especially are alternatives, very built when projects among These tables they the in reduce any relationship are not difficult more developing the likely that with the for small water resources helpful because combinations never (3) projects that do not have construct, and are the number values systems, candidate of of possible the design decision variables. 3.1.3.- Step 3: From Step Assess the performance of 2, we alternatives. Step alternative this in have a 3 reduced but assesses group. performance of an value net and calculate them, the curve of Curve of it performance is necessary to net benefits of the use every two characteristics alternative: (2) of of (1) the robustness an expected index. To intermediate step: every alternative. net benefits For benefits benefits, the promising group In this thesis, define the of the alternatives a given candidate indicates the net conditions. We alternative, benefits assume Alternative A its curve of net obtained under uncertain defined by the decision variables: XAg with k 1, 2,..m. 61 This Alternative A, of the yields some net benefits vector G, parameters uncertain element Gij the particular under (or, if negative, net costs). Mathematically these net benefits are represented by: NB e [XA j) that reads: XA given Eaj. variables To alternative defined by the deci sion of the net benefits e j] [XA NB calculate use we an optimiza tion 1 technique before to (like LF), cal cul a te the an optimi zation calculate NB decision [XA element ideal net E j] is benefits curve. require d, Alternativ e Alternative A procure the maximum possible net benefits curve benefits of 1 j. In words, A (while are held th e when e1 j, design con stant) in is obtained computed for every the uncertain p arameters the performance take the value indicated by Alternat ive A indicates the net benefits the curve will op erational net benefits. are A p rocess to because the for Alternative A Alternative other T he use of op timization of involves for variabl es The full when net could be the same technique used technique decision variable s of order to that curv e of that Alternati ve A yie ld. Fig ure 3-4 shows of net uncertain benefits of an discount rate. The hypothet ical alternative under internal rate of return of the alternative is for what the net is zero. For discount rates higher than the benefits curve 62 WT REFITI m UTjW hM=am(I Net benefits curve of an hypothetical FIGURE 3-4: alte rnative rate internal rate of return, that costs of building benefits of performance are of concern in and robustness) For of our are study based Net Benefits of on of candidate (expected situation probabilities of e 3, e1 j. The alternatives value of net the net benefits curves. an Alternative a given candidate alternative, the benefits can be calculated by uncertain curve takes negative it two measures Expected Value discount from the alternative are smaller than The benefits net under uncertain the net values, because benefits the I 1 adding net weighted net benefits by expected value of benefits for each the respective are given by the net 63 benefits curve just calculated. before: been calculated the 1iJ). [NB [NB (A)] (A)] is E NB [XA e i = i, 2,.., n j = 1, 2, .. Mi probabilities Therefore, expected value of net benefi ts E E p( The have also for Alternative A, is given by: p(e j) J evidently a scalar value. The same process has to be done for the rest of candidate alternatives. Index of robustness For a given candidate alternative, to the possible variation uncertainty. This potential the of of potential benefits curve represented given We calculated by its net the beginning benefits curve of that in as shown is represented in Step benefits curve, net benefits respect by the are be related to their shape as The mathematical related to the And the ideal net other hand, for under uncertainty are already calculated curve "Delta 1. the the Curve" important because ideal at net how far the shows is from reaching the potential difference is result from Chapter On the i. benefits net alternative. Delta curves Alternative A with measured Step 3. The difference between and the alternative call basin alternative, given the the basin, benef its which in net variation robustness of the basin. of the given robustness may proved below. representation of the delta curve of is : 64 [XA Figure 3-5 alterna tive) curve s shown 3-4 alterna of I = NB [X*(E j] shows tive Figure (dll M44ili I e the in (net 3)] - delta Figure 3-4, e [XA curve benefits from Fi gure 3-2 NB of j the obtained curve of (the assumed hypothetical by subtracting the hypothetical ideal net benefits the basin). *C *0 IEAL E MITS CMR e En MUITS W to1 *N FIGURE 3-5: Delta Curve of an hypothetical alternative, by subtracting the net benefits curve (Figure 3-4) from the ideal net benefits curve of the basin (Figure 32) It 3-5. is interesting to note two characteristics First, any point Second, the by these corresponding The delta ideal the net benefits net benefits two curves discount curve is meet rate, in curve of is not the exceeded for curve our is in alternative. a point. Therefore, the delta the basis curve in Figure at that zero. study of robustness. 65 Its measure of differences in evaluate the variations Consider the two this alternative A takes benefits, which indicates on ei. On the other but constant even losses of for very there are big sensitive e1 (its net potential different values alternatives. The e 1 . When ei is to be built. losses of But potential performance depending is never the best hand, Alternative B alternative for any value of zero), alternative to performance of sensitive uncertain variable the best net of be used in Figure 3-6. depicted indicates other values, optimum can performance curves under the eil, Alternative A is when el the delta curve for Alternative from the delta curve are benefits of el, is never like almost e 2 and e9 3 . This indicates insensitive performance of Alternative B 411 FIGURE 3-6: Cit il Robustness related to the shape of the curve. Comparison between delta alternative, a non robust A, Alternative alternative. B, a robust and Alternative 66 under the e5. Alternative B is more robust uncertain variable than alternative A with respect As said before, robustness to the shape measure this robustness curves. We the of its delta were the robustness. However applied all to piecewise curve. this process cases, and coefficient of variation delta curve. If the delta is very hor izontal of the simpler basin ), if curve ve ry is of net the Similarly, summary of has be the and can not be curves curvat ure to may be can not be calculate g reat the that form the robustness, the low dispersion in with v ariation low means coefficient <---------> Low CV robustness curvature: would the values has the delta is very robustness, the delta h igh dispersion respect to the of variation low. in the potential is very high. of A is: High CV The CV of delta is of (which bene fits and the complex how to benefits with res pect to the potential ti Ited the basin), is (which me ans a Iter native of net brief (CV) the coefficient distribution higher of is related shap e of the radius of procedure al ternative and question now is then the radius A the distribut ion The because calculated. curve alternat ive the ei. parameter an this with curvature, the linear, of associated with the could measure higher to the uncertain the index > ----- values of of the Low Robustness High Robustness the delta alternative. performance of every candidate alternative curve is After this called the step, the is defined with two 67 values: an expected value of net benefits and a robustness index. 3.1.4.- Step 4: The Compare among alternatives performance evaluated in of Step 3. the candidate Two performance measures with every alternative: expected net benefits The objective all candidate alternatives and to present an easily of comparisons candidate and and robustness. collect these values for benefits is an is and superior but objective another. The case both choice. Thus, a Iternative robustness decision robustness and final expected net are associated this information in that facilitate comparison and a two-objective benefit s not have is are was among alternatives. We have net the present step understandable form s e I ec t ion alternatives is matter we superior in expected alternative say that a another are expected argument to where could only to robustness not problem when greater. If net benefits, prefer one both only we do alternative to similar when only expected net benefits superior but not robustness. To easi ly visualize robustness for all the projects, benefits are in the horizontal axis vert ic al. Each graph. Figure These curves, shows called net benefits we consider a graph where expected net alternative is 3-7 and expected the Pareto and robustness represented t ypical form as of a is point in the in the these graphs. curves or Pareto frontier, are 68 two-objective evaluation methods. in common The vertical robustness Increase to N - robustness by or like 4 axis 5). If CV as we move (where instead of robustness directly CV, vertical axis. in is the representing as this kind of axis, we represent fixed number, N - CV, we represent we go down to an Alternative A, when curves. formed by robustness from the final B has than A. evaluation in the al though valid, candidate alternatives is inferior inferior to others. An Alternative B which are not order for In a "large" increases The Pareto frontier is and N up This creates an unconventional, representation of benef its robustness. represents smaller not expected net B could Alternative and does both enter in be omitted the Pareto front ier. wS$ ( - CI UKCtD FIGURE 3-7: Typical form of curves the Pareto K ITS IE (4ellan) frontier 69 inferior alternatives, Elimination of direct inspection which fall should inside the of of robustness a l so may requirements Mi nimum be yield expected be may i Ind icates to to the me an value. equal an twice the value of c an be if CV= 2, the the mean. This CV=2 is minimum inde x of robustness CV=i robustness. low d ecide cases, very which smaller than delta in t he values of the delta political the delta curve is 1. In of dispersion minimum rnatives the acceptable v al ue. Howev er, is To is h arder. An I n some and than z ero are viable. imposed. that th e CV of of a (greater o r th e standard devi at ion other words, benefits i of view, alt greater politically fo r robustness requirements equal benefits inferior and Developme t mi nimum require me nts However, other zero) net ar e benefits decision. From the analysts I poi nt done from alternatives net expected ne t be those curve requi red. expecte d for curve: f ron tier the elimin a ted. be Pareto can values is cons idered st andard dev iation indicates curve, great and consequently almost always an unacceptable value. Figure 3-8 case, with illustrate th is minimum required concepts net benefits for of an hypothetical NBI and minimum requi red robustness of CV 1 3.1.5.The this Final remarks two-objective stage. The outcome analysis is not of one the single water best system ends at alternative, 70 LW IOMTEUS (I ALTTITD ~- CY ~aLmygT Iffia IIU ALWTEMI ......- 3-cl1 LM wMn KT 102M hLTUMYI1 IEECTD ET 1ITS (sliRs) Derivation FIGURE 3-8: although it could formed for one There not it exists, there benefits: to is a choose necessary to trade-off a give up a final limitation of global more the making process, purpose of some net at the method is such an same time. robustness it But and net may be and vice versa. The could that limited it to be regarded as is an help number alternatives with enough data for comparison to general includes the greatest alternative benefits, I think a the between best alternative revealing The method applies curve always exists robust the method. But because frontier it should have and expected net benefits normally lack of If the Pareto if the non-inferior set only alternative. optimum alternative robustness be of a advantage, the decision of candidate and decision. and complex water resources 71 systems. The imposed by method does the optimization screening models A not of practical Step have more limitations technique used than the solve the to 1. summary of the method is indicated below. Objective: To identify a limited a two-objective benefits), set analysis and display the of non-inferi or (robustness information al ternatives and in in expected net a Pareto graph. Procedure: STEP I. Derive The the ideal ideal net net benefits benefits evaluating the performance of is derived as curve Formulate the b.) Perform the candidate M screening model M, Mn one for c. ) Obtain the net d.) Obtain an curve, ideal optimal Select candidate candidate and one reference for alternatives. set It runs by using the every vector benefits of basin curve: decision e0j NB [X* (G8j )) variables: X* (ei ) alternatives alternatives of for the optimization screening model, The serves as follows: a.) STEP 2. curve them will will be chosen later for define the Pareto implementation. 72 a.) Calculate probability distributions values of each design Depict b ar graphs the two identify of or a compatibility - d.) Generate candidate 3. Assess of the most likely by design decision variables of table the candidate alternatives Formulate a model for every alt ernative b.) For each rules and values of combining the a.) alternative , optimization runs and incompatibility table compatibility the Performance candidate Mn obtained distributions, alternatives sat isfying the while 2 variables Construct like ly values M Step 1 these c.) most STEP of three every des ign decision the M, decision variables, from the optimization runs b.) of in calculate perform optimize to order M2 Mn operation pos sible maximum the MI net benefits c.) Obtain the net benefits d.) Calculate E [NB] curves: values e) [XA NB for each candidate alternative: [NB E = E (A)) NB e [XA j e.) Obtain [[XA f.) the I Calculate robustness delta = j] CV of 3)] p(E 1 1 curves [X* (E NB (5) each )] NB - values. candidate [XA e j These measure the alternative 73 STEP 4. Compare among alternatives a. ) Plot b.) Eliminate c.) Set minimum an Expected Net Benefits versus N - CV (b) graph and d.) inferior alternatives requirements for expected net benefits robustness Present the Pareto non-inferior set of alternatives CASE STUDY CHAPTER 4: The 4. Chapter case This chapter correspond to the Rio the enough to to not coefficie nts and parameters in A rgentina, exposed case study of the the correspond to a real applicability easily demonstrate method described in of the Colorado basin ge neral show the also small of is does [1979] . The size and Lenton enough to study most However, basin. this viability and usefulness practical Major of purpose of illustrate is in large the method, but each step of the method. The models optimization and alternatives to optimize is linear simplification function and expressed as limitation of of 1979]. linear case of the equations. the as but real the computer is is that the candidate the screening model objective have to be is not a simplification Imposed by the LP optimization planning optimization run of (LP). Therefore, the first This world a common used to rules study approximation of by linear equations The programming the method, programming Hence, the screening used to solve the operation constraints technique. Most of linear algorithm situations technique relationships use [Rogers, and constraints practice. the LP package was a personal 75 computer IBM AT. Computer time was extensive because, the method does facilities, 243 runs not requires it requires for the many times runs about 231 4.1.- 2 minutes. of about (14 candidate in 2402 total runs). 14 minutes; the candidate Therefore the to optimize each alternatives computer time was hours. DESCRIPTION OF THE CASE STUDY The object of projects assume of alternative took optimize operating rules took about runs for each, results Each run for the screening model run to optimization runs: and 243 candidate 243 sophisticated computer independent screening model, operating rules of each alternatives very although are that to this case built be study is in a hypothetical the development of a regional development to decide what hydraulic the basin plan, and river is an we are basin. We important part interested in getting as much benefit as possible from the river. But, since there of are several benefit to uncertain be obtained concerned with obtaining a is insensitive identify the trade-off The four variables to the from the basin dams, basin, we projects three is amount are also strategy which existing uncertainty. Our task curve between robustness possible the affect the robust development non-inferior set of scheme of that is to that represents the and expected net benefits. shown possible in Figure 4-1. irrigation There are areas, two possible hydropower plants, and a possible transfer. Any of 76 MI NY tOPUEl PlaIT TUNS MN C IniTIcE C II1TIIA I' MMD IllQTIMN 0 FIGURE 4-1: these projects Scheme of could be the basin built in the basin. Most of the data for these projects has been taken from the real projects in Chapters in the 9 and Rio Colorado characteristics of in Argentina, as described 10 of Major and Lenton [1979]. Before formulating a screening model we have to define the considers that benefits repeated throughout and the assumed at the heads of main river. enter Inflows typical year. costs for project lifetime. in the model as the basin, The screening model this the three upstream for typical River year are inflows tributaries average of are the Inflows for 77 of each season the typical year. There are no tributary inflows nor outflows in the river course. A downstream minimum flow requirement exi sts total provide water to downstream users ecological reasons. The minimum downstream flow is and for X of to rive r upstream the total yearly 4-2 shows the inflows. Figure inflows of the river in 40 its upstream branches. INFL0 f3 : 2toll/& INFLOW fi : a6 3/s INFLOW ft : 4 3/1 FTAM FLOW : 0 1 if1 f4 jf4 FIGURE 4-2: river Total for Inflows 3 the typI ca I year The Season II, I, typical seasonal for irrigation and zero III, (high flow season). Figure 4-3 shows ye arly are medium in Season Season and Season of distribution (medium flow season); flow season); (low from May to August December seasons. The three seasons are to April from January September to t hree year has III, inflows. in Season as shown The demands I, maximum in Figure from the for water in Season II, 4-4. 78 PEICET Of THETOTAL IULV SWSO 3 SM 2 SLSON i Im ASMAL DISimT (il uTLW or TOTAL Seasonal distribution of total inflows for the typical year FIGURE 4-3: river 0356 4150 I sag i EASM I SASa3I IflIGTIM VATER DVuADS (i3/Ha) FIGURE 4-4: Water demands typicai year For simplicity, the have fixed two Irrigation hydropower plants heads. Otherwise, we would obtain in the screening model (power is of and head for and flow), to the are assumed to a non-linear term proportional linearize in to the product it would be very time 79 consuming. without since In this case study, hydropower plants requiring the dam irrigation the is construction not areas do needed the irrigation case non-consumed losses the area without water to to groundwater are returns the rest of water month later). irrigation After two stages: (before in two seasons, all refers non-consumed water. to the return in Season during I, and irrigation because there The of the bottom in Season II. connects river with the right upstream only one Water is pumps nor graph is no irrigation transfer direction: from assumed to move some water returns some water (between 8 and 12 the seasonal figure, the top graph water during irrigation is for non-consumed water for Season III left upstream branch of the in Season branch. the and 4-5 shows No graph exits the later); non-consumed water has In this non-consumed inflows or consumed by 4 months); seasons already returned to the river. Figure return of areas return (between 4 and 8 months returns their storage. included. The water not in the next season of divert water to river. No groundwater in the same season However, any water, or could also that crops returns to the river in three immediately only to interseasonal the head. construction storing study considers be built their associated dam, create could serve serve for regulation and This to require associated dam. The dam of can III. The transfer acts in left branch to right branch. by gravity, without needing elevation operation costs. 80 PROPORTIOS SEASU 3 SusO 2 ETM oF Tr VATEUK1 0015 U 11 SUSON I SusM I PR7OPorl(1 SEASON i SE0N 3 SEASO 2 t URN OFTHEVATER 90 CONSMED 11 SEASON FIGURE 4-5: 4.2.- Return of the Irrigation consumed by crops that describe in the to be particular case mixed of this (10 integer variables, objective below. Appendix A summary of model is mathematical during the project programming and 24 (whose function, contains the variables. equations and economic relationships existing year The model with In the model 34 is is 0 or 1). an decision operational variables) value and that is lifetime. case study, the screening project sizes variables formed by identified for a typical repeated linear-integer variables is the physical basin. These are assumed the non- SCREEHING MODEL The screening model 10 water and Decision are discussed constraints screening model formulation and for this screening framework taken from Chapter 5 of Major and Lenton (1979]. 81 4.2.1.- DECISION VARIABLES There are screening two model. variables. types The Design of first decision projects of the water variables decision type "des ign" is are variables resources system. in the There this decision s ize of the are 10 design decision variables, one for each project: VA VB VC VD AB AC AD CA CC T 3 Volume of Reservoir A (Hm ) Volume of Reservoir B (Hm 3 ) Volume of Reservoir C (Hm 3 ) Volume of Reservoir D (Hm 3 ) Area of Irrigation B (Ha) Area of Irrigation C (Ha) Area of Irrigation D (Ha) Capacity of hydropower Plant A Capacity of hydropower Plant C : Size of the Transfer (m 3 /s ) : : : : : : : : : The second is type (MW) (MW) "o perational" de cision variables. Releases from reservoirs, wate r diverted to and water diverted to the decision variables. There are 12 corresponding dams times to three diverted for seasons), and operation of from the reservoirs. operational 9 correspondi ng (th ree irrigation the water are not any o perational If depends If the operational 24 operation de cision variables, 3 corresponding to the hydropower plants. operation are re leases from the reservoirs seasons), irrigation transfer. There built, the the transfer irrigation areas, to (four the water areas times three diverted to the deci sion variable for the dam associated with the plant the plant the dam directly decision variables on i s given is the not f low Is by the releases built, in the the river. plant The are: RA, t : Releases from Reservoir A (t = season i,2, or 3) 82 RB, t : Releases from Reservoir B (t = RCt : Releases from Reservoir C (t = RD,t : Releases from Reservoir D (t = IB,t : Water diverted to Irrigation B ICt : Water diverted to Irrigation B ID,t : Water diverted to Irrigation B Tt : Water diverted to the Transfer In this case design decision study, we variables. are season 1, 2, or 3) season 1,2,or 3) season i,2,or 3) (t = season 1,2, (t = season 1,2, (t = season 1,2, (t = season 1,2, only concerned Design decision variables what projects are to be built and what projects are built. Also, for those variables 4.2.2.- projects indicate their optimal 3) 3) 3) 3) about the indicate not to be to be built, design decision sizes. OBJECTIVE FUNCTION objective The efficiency, or typical in function Is other words, of benef its from the year. Benefits at evaluated come selling electricity evaluated at mathematical equation for benefits = E [E 1 LS, t] s = B,C,D t = 1,2 + E s t minus market of economic costs for the agricultural products and prices, its [e2 measure a their B or or or or from the produced price. The general is: Es, t = A,C = 1,2,3 where: Annual benefits ($) Price of agricultural products ($/Ha) Land irrigated at Irrigation s in season t (Ha) Price of electricity ($/MWh) Power produced at Plant s in season t (MWh) B P1S Ls, t 02 Ps, t The c oefficients prices) are assume a ei (crop prices) and 0 2 (electricity considered uncertain. Instead of a fixed value, we probability distribut ion of their possible values 83 (Figure 4-6 and Figure the reasons 4-7). 4.2.4. briefly Section explain to consider e1 and e 2 as uncertain parameters. Probability dC N% 3,1 N 3/ K I0 30 AGIICLTUAL PlCES ($/R ) FIGURE 4-6: Assumed probability distribution of el: crop prices ribability I 3' is rEmil"T FiCEs (w1&h FIGURE 4-7: Costs projects, for the are because typical Assumed probability electricity prices only incurred from no operating costs year result are from distribution of e2:. construction of considered. The the the costs amortization of construction costs over discount rate (3. C = CC the project lifetime with a given The formula is: (I + 93)n E 3 ) / [(i + e 3 )n - Ij where: C : Annual costs (s) CC : Total construction costs e3 : Discount rate (. ) ($) : Projects lifetime (years) n lifetime the discount if For example, coefficient 03 uncertain. Figure 4-8 shows Section 4.2.4. sts for the typical year are 0.101 construction costs. times the total The the co 50 years, is 10x and the projects is 9 3 rate (discount considered distribution, and its probability briefly justi fies is rate) the uncertainty in (3. 0W-0.4 \ 501 / Probability / / II a costs fixed cost variable distribution Assumed probability discount rate FIGURE 4-8: The 12 P DISC09ET lATE (I'J III of construction of term any project have independent cost term dependent on the on the size of (3: two terms: project size, and of the a project. For 85 example, the cost of construction of (fixed machinery, personnel, that depend volume of on how concrete, includes four graphs every facility. dams, Figure the amount that A-2 for the etc.) dam of is has fixed costs and variable costs (volume of labor, excavation, etc.). Appendix shows construction irrigation areas, costs for the Figure A-3 for the and Figure A-4 for the transfer. the numerical TABLE 4-1: B indicate the construction costs for Figure A-i hydropower plants, summarizes big offices, a dam Table 4-1 coefficients. Fixed and variable possible projects to costs for the be built in the basin FIXED (10 PROJECT DamA Dam B Dam C Dam D Irrigation B Irrigation C Irrigation D Plant A Plant C Transfer /(m 3 /s) FVA: FVB: FVC: FVD: FAB: FAC: FAD: FCA: FCC: FT : $) I.50 i -50 VARIABLE aA: aB: aC: aD: QB: SC: PD: TA: TC: 3 2 0 0 .00 .00 .25 .50 1 .00 1 .00 1 .50 1 .50 mathematical The general 6 P : 0.0044 0. 0176 0. 0056 0. 0104 0. 000082 0. 000247 0. 000536 0. 050 0. 077 0. 140 ((106 $/Hm3) (106 (106 (106 (106 (106 (106 (106 (106 (16 $/Hm3 ) $/Hm 3 ) $/Hm 3 ) $/Ha) $/Ha) $/Ha) $/MW) $/MW) equation for construction costs (CC) is: CC = E4 + s5 [ E (Vs as + FV 5 S = A, B, C, D E (Cs = A,C TS + FCs YVS) YCS) + + E (As fs + FAs YAs) + s = B, C, D (T p + FT YT) I where: e4 : General increase or decrease in construction costs (X) 86 Vo Iume of Reservoir s (s = A,B,C,D) (Hm3) Variabi e costs of Reservoir s ($/Hm3 ) Fixed c osts of Reservoir s ($) Integer variabl e for Reservoir s: Vs as FVS YVs = i YVs = 0 As Area of Irrigat ion s (s : BC,D) (Ha ) s ($/Ha ) :Variable costs of Irrigation 13s FAs : Fixed costs of Irrigation s ($) YAs : Integer variabl e for Irrigation s: If Irrigation s is built: YAs = i If Irrigation s is not built: YAs = 0 C5s Capacity of hyd ropower Plant s (s = A, C) Ts :Variable costs of Plant s ($/MW) FCs : Fixed costs of Plant s ($) YCs : Integer variable for Plant s: If Plant s is built: YCs = i If Plant s is not built: YCs = 0 T : Size of the Transfer (m 3 /s) p : Variable costs of Transfer [$/(m 3 /s)] FT : Fixed costs of Transfer ($) YT : Integer variable for Transfer: If Transfer is built: YTs = I If Transfer is not built: YT, = 0 If Reservoir s is built: YVs If Reservoir s is not built: The variable FAS, r s IFCs, and fixed and P FT (MW) costs coefficients are found Table in coefficient 94 (general increase or decrease costs) Figure 4-9 shows is considered uncertain. probability distribution, After can be and Section defining benefits and costs, the FVs, 4-1. QSs The in construction its assumed discusses it. objective function expressed as: Max (decision variables] where B costs 4.2.4. as, are the for 4.2.3.The B - C benefits for a typical a typical year and C are the year. CONSTRAINTS screening model of this case study includes seven 87 Probability -25 FIGURE 4-9: 415 1 YAIATIO II CMSTlICilm 0STS (1) I Assumed probability distribution increase general construction costs sets of constraints (1979), Chapter Instead of or 5, this section briefly discusses the special characteristics of the constraints refers to Major and Lenton [19793 for for this case study, and details. (1) Continuity constraints. Continuity constraints conservation of mass in the reservoirs: in a reservoir must lost through three be stored evaporation of in for continuity equations are necessary requirements. The between to it, insure that enters all water released, diverted, subsurface for every dam, one equations equations - in and Lenton Major repeating of e 4 : decrease There are leakage. per season. There dams. Three or are also additional indicate the minimum downstream explicit mathematical notation is: for Dam A: 88 SA,t SA,t+i - RAt + fit - eAt SA,t where: SA, t RAt fit eAt - : : : : Storage in Reservoir A in season t (t = 1, 2, 3) Releases from Reservoir A in season t Inflows of the left branch of the river in season t evaporation coefficient of Reservoir A in season t continuity between Dam A and Dam B: IBt = RAt - TRt where: IBt TRt - : Inflows in Reservoir B in season t : Flow of the Transfer in season t for Dam B: SB,t + SB,t+i IBt - RBt - DBt - eBt SB,t where: SB,t RBt DBt eBt - : : : : Storage in Reservoir B in season t (t = 1,2,3) Releases from Reservoir B in season t Water diverted to Irrigation B in season t evaporation coefficient of Reservoir B in season t for Dam C: SC,t+1 SC,t + f3t + TRt - RCt - DCt - eCt SC,t where: SC,t f3t RCt DCt eCt - : : : : : Storage in Reservoir C in season t (t = 1,2,3) Inflows of the right branch of the river in season t Releases from Reservoir C in season t Water diverted to Irrigation C in season t evaporation coefficient of Reservoir C in season t continuity between dams B and C and Dam D: IDt = RBt + RCt + f2t + RIBt + RICt where: IDt : Inflows in Reservoir D f2t : of RIBt Inflows the center in season t branch of : Water that return to the river Irrigation B RICt : Water that return to the river the river in in season t from in season t from season t 89 C Irrigation - for Dam D: SD,t + SD, t+ I IDt - RDt - DDt - eDt SD,t where: SD, t : Storage in Reservoir D in season t (t = 1,2,3) : Releases from Reservoir D in season t RDt : Water diverted to Irrigation D in season t DDt : evaporation coefficient of Reservoir D in season t eDt - for downstream requirements RDt + RIDt 0.4 (40 X of (fit + f2t + all river inflows): f3t) where: Water that return to Irrigation D : RIDt The values of = 1,2,3) values (t= can of 1, 2, 3) (2) reservoir reservoir. the inflow parameters fit, f2t, be calculated from Figure 4-2 the evaporation are indicated Reservoir in are mathematic formulation SA, t ! VA SB, t SCt SD, t ! VB VC ! VD coefficients in Table maximum any season There in season t from the river can three and Figure 4-3. eAt, eBt, eCt, (t The and eDt 4-2. storage. not and f3t exceed The the storage volume equations per dam. The of of the the explicit is: VD are where VA, VB, VC, of represent the volume above) decision variables that the design the reservoirs (see Section 4.2.1. 90 (3) Water requirements for show the irrigation. These relationships between constraints water divested for irrigation and TABLE 4-2: Project Parameters needed for the screening model lifetime: n 50 years reservoirs: Evaporat ion from the SEASON i est SEASON 2 SEASON 3 DAM A IX DAM B 2X 5% 3% 5X 9% 3% 7% 1o% 3% DAM C DAM D Irrigation channel losses: Consumptive use of the Es 4% 6% toy IRRIGATION B IRRIGATION C IRRIGATION D 65%Xof 1% water in irrigation: irrigation water is consumed Coeff ic ients of hydropower plants Head of Plant A: Head of Plant C: Efficiency: Load factor: Factor of utilization: the land are also irrigated. Water irrigated losses considered. There are irrigation area. surfaces 100 m 50 m 0.65 0.86 0.68 six in the equations, Six more equations in each period can irrigation project areas. The mathematical - water requirements for (I - EB) DBt = 5 "t irrigation two indicate not channels for each that the exceed the formulation is: irrigation LBt 91 (1 - EC) (1 - ED) DCt = DDt = E 5 Ft LCt e 5 Ft LDt where: EB EC ED (5 : water losses in Irrigation B channels :water losses in Irrigation C channels : water losses in Irrigation D channels : general increase or decrease in irrigation water demands Ft : irrigation water demands in season t (t 1,2) LBt : land irrigated in Irrigation B in season t LCt : land irrigated in Irrigation C in season t LDt : land irrigated in Irrigation D in season t - maximum irrigation per season: LBt LCt LDt 1 AB 1 AC S AD where AB, AC, represent the 4.2.1. above) AD are surface The values of Table 4-2. indicated decrease (4) the parameters in irrigation water demands) its Ft (t (general indicated 1,2,3) = in are increase or is considered uncertain. assumed probability distribution, that in Section 4.2.4. Irrigation water diverted for river of of in Figure 4-4. The parameter e 5 discussed are the parameters EB, EC, ED The values Figure 4-10 shows is the design decision variables that or the irrigation areas (see Section the non water return. irrigation with This the constraint relates return consumed water. The mathematical to the main formulation is: RIBt lt [(i1 RICt RIDt lt E(1 Ilt [(i - - EB) EC) ED) DBt + EB DCt + EC DBtI DCtl DDt + ED DDt] 92 Probability 33 1 33 1 33 ' I ~- N / / / a~~ -501 FIGURE 4-10: I eve 4 VAIIATlS II IRIQTiO§ VATER DAMiDS (1) Assumed probability distribution of 95: general increase or decrease in irrigation water demands where: Olt : water non consumed in season 1 (1 = 1,2) to the river in season t water consumed by crops The values of 4-5. The the value (t parameters Qlt can of the parameter 0 be flow with equation for indicates not each is indicated that exceed - for the electricity the representation electricity season capacity of return calculated from Figure (5) Hydropower constraints. The first relates that 1, 2, 3) set production. each plant. production the plants. in Table of constraints There The 4-2. is one second set in each season can The mathematical is: electricity production: PAt PCt = (2.73 i0-6) RAt HA SA Rt = (2.73 10-6) RCt HC sC kt 93 where: power produced at Plant A power produced at Plant C PA, t PC, t HA HC head season t season t (t 1,2,3) Plant A head of Plant C efficiency of Plant A efficiency of Plant C number of seconds in season t sA sC Rt - of in in maximum electricity ( PAt PBt ht ht If per season: u CA If U CC where: ht : number of If u : load factor : factor of utilization hours in season t = 1, 2, 3) (t and CA and CC are the design decision var iables the capacity of the hydropower plants (see above) The values are indicated of the in Table HC, parameters HA, transferred transfer. TRt in sA, sC, If, and u 4-2. (6) Transfer size . These constraints water that represent Section 4.2.1. each season In a general mathematical to limit the the amount of capacity of the form: 1 T where: TRt : Flow of the Transfer in season t and T is the design decision variable that of the Transfer (see Section 4.2.1. above) (7) Conditionality constraints indicate that corresponding dam has and to also maximum build to be (t represent the size sizes. an = 1,2,3) Conditionality irrigation built. area the The maximum size 94 if a project constraints ensure that are included. ten are There project. The mathematical built is of these the fixed costs equations, one for each is: representation irrigation-dam conditionality: - AB - AMAXB YVB 0 AC - AMAXC YVC 1 0 AD - AMAXD YVD 0 - maximum size constraints: VA - VMAXA YVA 1 0 0 VB - VMAXB YVB VC - VMAXC YVC 5 0 VD - VMAXD YVD 1 0 AB - AMAXB YAB 5 0 - AMAXC YAC AD - AMAXD YAD AC 0 0 CA - CMAXA YCA CC - CMAXC YCC 5 0 T - TMAX YT ! 0 1 0 The coefficients AMAXC, AMAXD, for the VMAXA, VMAXB, VMAXC, VMAXD, AMAXBI and TMAX represent maximum sizes CMAXA, CMAXC, projects. These values can be obtained from Figures B- i through Figure B-4. 4.2.4.In UNCERTAIN PARAMETERS this study case uncertain parameters: (3) discount water demands. constraints. rate, we consider the existence of (1) crop prices, (2) electricity prices, (5) irrigation the irrigation four uncertain parameters affect the (4) construction costs, and Irrigation water demands affect The other objective function. This section briefly discusses for the uncertainty five the reasons in these parameters. 95 Crop from the prices and projects. Market agricultural products crop productions, prices 4-6 electricity prices are and shows the common for depending on climatological conditions, international estimated prices are normally tend to rise. subsidized when incentive used for electricity cost are very uncertain and affected Electricity an fluctuations affect the benefits In exports. Crop for complex factors. distribution more as imports and stable rural of than areas, crop crop agricultural an uncertain prices and may be irrigation development. variable prices. electricity energy for pumping and Figure This as makes difficult to estimate (Figure 4-7). The meaning and stressed in other parts discount rate is construction, years. the money this thesis. We interest rate that has variable. Figure objective function. those of to benefits in have We have costs, all are reduced. real of a assumed Projects costs could assume that money borrowed for returned yearly during 50 also projects situations. construction been is strongly affected distribution study. construction. Therefore, construction factor of the has 4-8 shows the assumed discount rates for the case Construction costs the of of the discount rate Therefore, the objective function for this of importance if that there are more costs Most of a direct are the effect on the only costs are is an increase in expensive and net a very uncertain the elements that define project (labor costs, row 96 materials, fuel, etc) are subject to uncertainty. The probability distribution for construction figure water demands are uncertainty. First, unexpected water irrigation greater channels amount requirements. and of water Second, differ from the be with planted factors create demanded for forecasts. great irrigation, an This will the in of soil great in happen require a irrigation requirements and are not plants may Third, different crops may water requirements. uncertainty as to could satisfy in the can be seen here that the important aspect People of that area of want, benefit as they can. effect uncertainty. of "guarantee" that they can projects. projects built shown All these amount of water in Figure 4-10. APPLICATION OF THE METHOD TO THE CASE STUDY We have supposed bad, to conditio ns different a losses irrigation initial subject installat ions. perfectly defined. Physical is is 4-9. Irrigation 4.3.- costs assumed Our job projects to what course, They if to as want be built, a rural obtain as area. much net robust projects that uncertain conditions happen to analysts Figure projects the river equally concerned with the expect satisfactory indicated in and of development of But they are even still the development of is 4-1, should we have performance from the to decide, from all what be projects the should be not be built. Also for the to decide their sizes. These 97 have projects to provide acceptable net benefits and also the robustness that people want. The method described their robustness evaluates in the the basin, expected net 4.3.1.- Step e = in deciding how to obtain more follows describes is solved using the method. The description i: Derive the the vector how is indicated in Chapter 4. net benefits ideal There are five uncertain Therefore, compared with give up the method. What the step by step process in involved the parties benefits. The subsequent decision making process study the case and benefits. After net consists people willing to beyond the scope of based among process The comparison others. suited to solve this each candidate alternative may be much robustness are is expected and a decision making process in that, is a few candidate alternatives method identifies The case. in Chapter 4 parameters in curve this case study. e is formed by five elements: e 1 , 0 2, 0 3 , E 4 , E51 where: agricultural products prices ($/Ha) electricity prices ($/MWh) discount rate (X.) E3 )4 = construction costs (Z increase) e 5 = irrigation water demands (X increase) e = E2 Figures 4-6 of the method, through 4-10 uncertain we divide each have showed parameters. to divide probability curve the probability distribution To them in use these in discrete three curves in the intervals. We intervals. Table 4-3 98 summarizes the values for of the intervals and their probabilities every uncertain parameter. TABLE 4-3: Intervals in which the continuous probability distribution of the uncertain parameters have been divided VARIABLE ist INTERVAL Value Prob, Discount Rate Constr. Costs Irrig. Demands 82 -252 -50Z 301 502 331 fox even even 50Z 351 332 122 +25Z +502 Agric, Prices Electr. Prices 30 10 301 251 40 20 401 401 50 30 2nd INTERVAL Value Prob. Now, for example, rate, e33 has = 12%. three values The same parameters. The 243 elements. To associated: be e is clarify e, 2 3 said UNITS 202 152 332 1 301 351 $/a $/Mh 2 Increase I Increase parameter 03 1 G 3 , discount =6, 8 for the 3 2 meaning one of ,9 4 1, 3 5 ]. its = ox, and other uncertain therefore formed by: the consider [e 1 1 ,e2 2 ,0 element uncertain can vector parameters vector the the 3rd INTERVAL Value Prob. of 3-3-3-3-3 the = uncertain elements, for example This element indicates uncertain conditions defined by (according to Table 4-3): G) 02 )3 E4 E5 From table element the = agricultural products prices = 30 $/Ha = electricity prices = 20 $/MWh = discount rate = 10% = construction costs = 25X decrease = irrigation water demands = 50% increase 4-3 we [9 1 1 ,e 2 2,0 uncertain reasonable 3 can also 2, 4 1 , e 5 3 ] . We parameters, assumption. The what obtain the assume in independence between this probability probability of the of case seems [011,(2 2 a very 32941, 0 5 3 ) is: 99 p [e 1 1 ,e2 2 e3 2 E = 0.50 0.50 3 1 4 ,0 5 0.33 ] 0.30 After doing this for each p(e 0.40 0.010 the shows B Appendix p(E ) = 2 3 ) p(e 4 1 3 ) p(e 5 ) 1.000 % 243 of the elements ha ve defined all the uncertain of occurrence. probability their with situations 2 the of one parameters vector, we uncertain 2 p(e 1 i) probabili ty Table B-1 each of occurrence of in uncertain element. Ideal net benefits curve We have defined 243 possible correspondent possibilities benefits curve indicates obtained from the future of occurrence. basin for each we curve screening model. [E1 1 , E 2 2, E 3 2 1 4 For water demands are multiplied build. possible the net and situation, The are ideal by t hat possi ble the ideal net of to screening products 1.50. solve the model for prices discount rate by can be 0.75, and are 30 is 10%, irrigation The computer gives us that cot Ild be obtained for this the optimal si zes of process has al so situations. screening model The I he multiplied benefits same future in net the 2 43 package LP prices are 20 $/MWh, cos ts maximum point agricultural construction the the example, , 953], electricity $/Ha, utilize one of e ir th ideal The the maxi mum net benefits future situations. To calculate one benefits situations with net have to In total, to be benefits De the projects done for the 243 to other 242 optimization runs of performed. curve can not be graphically 100 represented, Table because 4-4 shows curve for the E4 1 ,0 5 3 ], of 1.03 Optimal is defined study. We ideal that indicates curve net benefits [ for space. 1, 22,E32, net benefits million $. sets In of design decision variables this study case there variables. Then, the vector X = {XI, X of the can see net benefits ideal a 5-dimensions in the numerical values case the it X2 , X3 , X4 , X 5 , are design ten decision is formed by: X 6, X7, X 8, X9, x 10 1 where: = Volume of Reservoir A (Hm 3 ) 3 X2 = Volume of Reservoir B (Hm ) 3 X3 =Volume of Reservoir C (Hm ) 3 X4 =Volume of Reservoir D (Hm ) Area of Irrigation B (Ha) X5 X6 =Area of Irrigation C (Ha) Area of Irrigation D (Ha) X= Capacity of Plant A (MWh) X8 X9 = Capacity of Plant C (MW) X0 =Size of the Transfer (m 3 /s) X, From the ideal net benefits to obtain runs performed 243 optimization 243 curve, we also obtained the set s of optimal generically represented them by: decision variables. We X* [(] To the illustrate particular value X*[e optimal set run optimization particular for we obtained of meaning 1 1 of X* [E], ,e 2 2 ,E 3 2 ,E 4 1,E decision performed for [ei 1 , ( obtained Xj 2 2 ,( the represents the ]. This variables decision variable the optimal value: 3 5 consider lets 2 3 from e4 1 ,E 3 5 ]. the In (volume of Reservoir A), Xj*[G1t,922,932,E i E531 101 TABLE 4-4: Ideal net benefits curve WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even DISCOUNT RATE: (million $) i WATER REQUIREMENTS = +50 % |EI.Pr.=10 EI.Pr.=20 EI.Pr.=30!E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 1 !Agr.Pr.=50 3.67 5.19 6.71 4.01 5.51 7.01 4.49 5.99 7.50 1.50 2.20 2.91 1.80 2.52 3.26 2.26 3.00 3.75 0.80 1.27 1.73 1.14 1.57 2.03 1.63 2.01 2.51 CONSTR.:Agr.Pr.=30! COSTS= !Agr.Pr.=40 even |Agr.Pr.=50 3.30 4.79 6.31 3.60 5.09 6.61 4.02 5.51 7.00 1.20 1.90 2.60 1.50 2.20 2.90 1.80 2.53 3.27 0.50 0.97 1.43 0.98 1.27 1.73 1.47 1.69 2.03 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: 425 1 |Agr.Pr.=50 3.00 4.40 5.92 3.30 4.70 6.21 3.60 5.03 1 6.52 . 0.90 1.60 2.30 1.20 1.90 2.60 1.64 | 2.20 2.90 0.34 0.67 1.13 0.82 1.04 1.43 1.30 1.53 1.75 ------- :---------- ----------------------------- :-----------------------------|----------------------------- REQUIREMENTS = even WATER REQUIREMENTS = -50 1 | WATER DISCOUNT RATE: : WATER REQUIREMENTS = +50 1 10 1 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30|E.Pr.=10 EI.Pr.=20 EI.Pr.=30|E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.|Agr.Pr.=30 COSTS= Agr.Pr.=40 -25 X Agr.Pr.=50 3.39 4.91 6.43 3.69 5.20 6.72 4.16 5.64: 7.15 | 1.29 1.99 2.69 1.59 2.29 2.99 1.92 | 2.67 3.41 | 0.59 1.05 1.52 1.03 1.35 1.82 1.51 1.74 2.17 CONSTR.!Agr.Pr.=30 COSTS= !Agr.Pr.=40: even Agr.Pr.=50! 3.02 4.42 5.94 3.32 4.72 6.23 3.62 : 5.06 : 6.54 | 0.92 1.62 2.32 1.22 1.92 2.62 1.64 1 2.21 : 2.91 0.35 0.68 1.15 0.83 1.05 1.45 1.26 1.53 1.76 ------- :---------- !-----------------------------|-----------------------------|----------------------------- CONSTR.|Agr.Pr.=30: COSTS= !Agr.Pr.=40 +25 1 |Aqr.Pr.=50 2.65 4.05 5.45 2.95 4.35 5.75 3.24 4.65 ! 6.05: WATER REQUIREMENTS = -50 DISCOUNT RATE: 12 1 0.55 1.25 1.95 0.96 1.55 2.25 1.44 | 1.84 1 2.54: 0.20 0.42 0.78 0.63 0.85 1.08 1.11 1.33 1.55 z : WATER REQUIREMENTS = even I WATER REQUIREMENTS = +50 1 IE1.Pr.=10 EI.Pr.=20 E1.Pr.=30!E1.Pr.=10 EI.Pr.=20 E1.Pr.=30|E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= !Agr.Pr.=40! -25 1 Agr.Pr.=50 3.18 4.63 6.15 3.48 4.93 6.45 3.82 1 5.31 ' 6.81 | 1.08 1.78 2.48 1.38 2.08 2.78 1.73 2.38 3.08 ' 0.43 0.85 1.31 0.92 1.14 1.61 1.40 1.62 1.91 CONSTR.|Agr.Pr.=30| COSTS= |Agr.Pr.=40! even 1Agr.Pr.=50! 2.74 4.14 5.57 3.04 4.44 5.86 3.34 4.74 6.16 1 0.64 1.34 2.04 1.01 1.64 2.34 1.49 1.94 | 2.64 I 0.24 0.46 0.87 0.68 0.90 1.17 1.16 1.38 1.60 CONSTR.IAgr.Pr.=30| COSTS= |Aqr.Pr.=40I 2.30 3.70 2.60 4.00 2.90 | 4.30 ! 0.39 0.90 0.77 1.20 1.26 | 1.59 0.08 0.28 0.47 0.66 0.95 1.14 +25 % |Agr.Pr.=50I 5.10 5.40 5.70 | 1.60 1.90 2.20 0.50 0.88 1.37 102 0 Hm 3 . Besides this optimal value E 5 3 ), there are other 242 optimal to the other 242 possible Appendix B shows the The same variables. B-3 243 [0 1 1 , for values of values of 2 2 ,e3 2 641 X1 , corresponding future situations. optimal Table B-2 in X 1. could be said for the other nine design decision There are 243 through B-il show the variables X 2 of X 1 optimal values optimal for each one. values through X 1 0 . These tables for Tables design decision are the basis for next step. 4.3.2.- Step 2: We have Select candidate alternatives 243 values for each design When w e consider one decision vari ab le, get th e probability looR in g at Hm 3 table B-2, an d 0 Hm 3 of t o calculate X1 . And the dec is i on distri the of the these h av e the With thi s, be Xj, value S. can we For we probabil ity have all probability d istribution same can variables. butions B-1 we to occur. 0f for example Xi, o nl y two different values: there are . In table optima I value need distribution decision variable. said Table 4- 5 values of for the shows the ten of 84 e ach the data we the val ues of other nine des ign the probabil ity desig n decis ion variables. The probability distributions also be plotted. Figures B-I representation of provide a in Table 4-5 can through B-10 show the bar graphs the probability quick means indicated distributions. These graphs for identifying the most likely values 103 TABLE 4-5: Probability distribution of the of the design decision variables VOLUME DAN A 0 81 I 88.501 0 0.001 11.501 9 100.001 PROB. N 0 8 100.00! VOLUME DAN B 99.281 0.121 100.00' AREA IRRIGATION B FROD. I --------- 0 81 3/t FROe. MV 100.001 Jo3 TRANSFER SIZE CAPACITY PLANT A PROB. Hm3 values Ha - 12.411 87.591 FROB. 0 16550 16150 27880 100.001 9 ----------- 0.911 83.111 11.501 3.881 100.001 VOLUM DAM C 3 AREA IRRIGATION C FROB. O 105 I to. 581 59.421 ---100.001 Ha FROB. 0 18410 19660 10.581 48.161 10.661 --------- CAPACITY PLANT A 0 6 to 100.001 ------ ---- - VOLUE DAN D 0 21 122 123 100.001 - ----- 61.141 38.111 0.T21 --------- -- - - AREA IRRIGATION D PROB. I 1a3 ----- PROB. mw 68.45 10.661 3.501 11.0 ------ 00.00 Ha PROB. I 0 1550 1650 0 206TO 20110 20818 68.15: 1.431 9oo o 8.851 0.38 14.441 2.651: 3.501 104 of the B-7, design decision variables. Consider for example Figure for the design decision variable D. The values of (2) around is formed X 7 are 1600 Ha, and cluster is formed the second 1550, the of 1650, of 20670, The weighted instead of by the the is 1640 considering cluster. cluster obtaining The three 20710 Ha clustering more likely shows the reference values of built. Table likely to be built, Analyzing 4-7 B-2 to this cluster, values, we only done for t he third value. identify the t (with their respe variable. decision zero means su mmarizes each value). t he refere n ce value as decision their s izes, Tables in close values clustering procedure. A siz e of not Then, 1640 the zero. For included serves probabilities) for each des ign The the three values very refere nce 20840 Ha. obtained the reference proced ure third can be be analogously as The is obviously Ha. t hree can It and probability of consider the representative value for Ha. representative value (weighted average (1) 0 Ha, second cluster 1720 20770, first cluster cluster, the cluster groups: 0 Ha. The and by obtaining the weighted average of in the irrigation (3) around 20700 Ha. The first cluster by values reference value of area of clustered in three exclusively by the value is formed by values of = X7 ive ct afte r the the proje ct is the projects that are m ore and their probability. through B-I, we can infer s ome re Iationships among projects. Consider, for example, Dam A. is easy to see, by comparing or Tabl e 4-6 variables that NO Table B-2 with Table B-3, It that 105 TABLE 4-6: Probability distribution of the reference values of the design decision variables after clustering VOLUME DAM A Hs3 CAPACITY PLANT A PROD. I mV --------- ----------- 84 0 0.001 100.001 9 100.001 100.001 ------------ PROD.I 3/1 ---- 0 99.281 8 0.721 100.00' ---------------------------------------------------------------------------------------- YOLUME DAM B AREA IRRIGATION I PROD. I 033 --------- 0 84 Ha - PROD. I ----------------------- 1?.41 0 87.591 16510 27880 VOLUME DAM C ------------ : ---------------------a-- 88.501 11.50 0 PROD. 1 TRANSFER SIZE 0.9 95.211 3.881 AREA IRRIGATION C 3 PROB 0 105 40.581 59.421 ------- PROD. I Ha 0 40.581 8680 59.121 ------- VOLUME DAN D "3 PROD. I - a 0 6 61.141 38.141 0!2 - - - - .0 100.001 --------------------------------------- AREA IRRIGATION D Ha PROD. P a 21 123 MV PROa.I ---------------------------------------------------------------------------------------- 100.0010 --------- CAPACITY PLANT A 10.661 20.60 2010 100.001 :0o.00 160 10.66 20.60 a a a ata enii10g6 a TABLE PROBABILITY 100.00 95.21 % 87.59 x built. is Also we project with Table X % % % % % X Dam can Dam A: f or Dam B-2 when Dam A Dam D, % (size different that when Therefore, X built is B that they are mutually facil Study We Irrigation C 7) that built. built. between see are ities now the We times Dam A and to MW Ha Hm 3 Hm 3 1640 Ha 3 21 Hm 27880 Ha MW 10 8 m 3 /s zero), A Dam B Dam B is is is never never built. an incompat ibl e e x c Ius i ve. the t ables , Compa ring we see that are always bu i lt: and P I ant A. profi t able This m eans than Da m A, a l ready built be fore be relationship betwe en Dam A and from the ir respective Table s (B-2 an d B- Dam concl ude c an D, more ha ve some times Dam A is Othe r 6 20710 123 84 oth er facilities Irr igat ion MW Ha Hm 3 Ha 3 105 Hm Dam say that these f our facil.ities building Dam A. built, then I our built, the se is A with the rest of Irri gation B, because 9 16570 84 18680 Plant A Irrigation B Dam B Irrigation C Dam C Plant C Irrigation D Dam D Dam A Irrigation D Dam D Irrigation B Plant C Transfer 59.42 x 59.42 X when Dam A SIZE PROJECT % 38.14 20.60 20.60 11.50 10.66 10.66 3.88 0.72 0.72 to be built more likely Projects 4-7: bu ilt and is built A that Irrigation there C: Irrigation C is and Irrigation C is is no no one unique also not relation is more or less 107 than profitable other. the called the Compati bilities table. table shows Irrigat at D, and that are above that have to be Dam D) Irri gation D can and Dam D projects (for example, built together or not built without building not be not can it be built without also building D. for de velopment create candidate alternatives To plants, strategies. But, combinations among (Table 4-8) number of when we candidate have the we have decided alternative, we in this projects must decide reference sizes for the Figure is very reduc ed. what their projects. that the the compatibility graph or the hierarchical possible combinations planning con straint projects have to respect 14 possible combinations are allowed Once the impose the can be made result in d ifferent will that of irrigation (dams, Many combinat ions and transfer). projects these projects possible the combine we basin, table (Figure form "hierarchical" are There built. all: Irr igation with also ion D areas, information same The a project, those projects to build be have to Dam incompatible Top project s are more profitable than bottom ones. When deci de built in project, this the projects, related projects. graphicall y represented be 4-11). we the non given For any profitable more the and project s, can we obtain Table 4-8, other des ign decision variables, the all for comparison process made with Dam A Repeating the same It the 4-11), In fact, only case study. are part of a sizes. We already is reasonable to 108 ------ table Compatibilities TABLE 4-8: facilities that )AY or Facilities that Additional facilities that : KATNOT be built are NEVER built are ALWAYS built: ...................................... Initial facility: PLANT A TIT IRRIG B DAN A DAN a DAM C 111I C PLANT C DAN D IRRIG D DAN A PLANT A TRI 1 1IG B DANE DADC IRIGC PANT C DAM D IRRIG D ------------- ---------- ------------ ---------- --------------- TRI DAN A DAM B PLANT A DAN C IRRIG B IRRIG C PLANT C DAN D 11IG D ------------- ---------- ------------ ------------------------------ TIRFDAN A DAN B DAM C PLANT A D DAM IRIG B IIIG C PLANT C IRRIG D ---------------- ------------ ------------------------------ TEI DAN A DAM B PLANT A I1IG B DAN C IRIG C PLANT C DAN D I116 D ---------------------- -------------------------------------------- IIG B 1IG C . . . TRI DAN A PLANT A DAM I PANT C DAN C DAMD I1IG D ------------------------- ---------- ------------------------------ TRI DAN A DAN B DAN C IRRIG C PLANT C PLANT A IRIG B IRRIG D DAN D - - - - - - -- - --- ----- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - DAN A - - TRF DAN E IRIG B PLANT A DAN C IRIG C PLANT C DAN D 11IG D PLANT A DAM A DAM I IUIGB PLANT C TRY DAN C IRIs C DAN D 1IG D ------------------------------------------------------------- PLANT A TEF PLANT C DAM A DAN IRIG DAN C IHIG C D IMIG D DAM t019 FIGURE 4-11: Hierarchical representation relationships among projects think that the sizes alternative project, reference for a most likely sizes has with similar project. a main In included of in the candidate the projects. For each reference probability. In some cases si ngle project the projects are the ref erence sizes the greatest of size two or is reference case size study, that probabilities of of the main of summarize s the reference size for Now projects, we have three most identified 14 and for each project we reference identi f ied however, stands have probability with even Secondary reference sizes the the one probabilities can be this of clea rly. at least size. each Table half 4-9 every project. possible have combinations identified of its 110 Reference sizes TABLE 4-9: for the projects SIZE PROJECT Dam A 105 123 16570 18680 20710 A Plant 9 6 Plant C Transfer reference size. alternatives, alternatives of are 4.3.3.- this have described satisfy the identified in Compatibility Assess the performance of Step 3: former step step we have of net we to asses conditions have the are possible future situations. To under alternative independently evaluate under each one of Lets % 20.60 100.00 38.14 0.72 X X % % X X % 4-10. These by combinations table, and each 14 candidate performance alternatives. of every one of in the basin. benefits Uncertain candidate 95.21 59.42 X the alternatives them under the uncertain conditions existing Curves % 14 candidate Table are formed 11.50 87.59 59.42 20.60 its reference size. From the In are we candidate because projects that project has 8 Therefore that Hm 3 Hm 3 Hm 3 Hm 3 Ha Ha Ha MW MW m 3 /s 84 84 Dam B Dam C Dam D Irrigation B C Irrigation D Irrigation PROBABILITY study the the the approximated assess by defining 243 performance the of a uncertain conditions we have to performance of the alternative 243 possible future situations. performance of Alternative A. We consider 111 TABLE 4-10: ALT ERNOI1VE A: Plant A: Candidate alternatives 9 MW ALTERNATIVE 1: Plant A: 9 MW Irrigation B: 16,570 Ha ALTERNATIVE F: Plant 1: irrigation B: Dan F: ALTERNAI VE C: Plant A: Plant C: 9 MW Dam A: Plant C: 84 Hmr3 6 MW 16,5/0 Ha 84 Hm31 9 MW 6 MW ALTERNiiVE D: 9 1W Irrigation B: 16,570 Ha Dam A: 84 Hm ALTERNATIVE J: 9 Plant A: Irrigation B: 16,570 Dam A: 84 Irrigation C: 18,6 Dam C: 105 MW Ha Hm3 Ha HM3 Plant A: ALTERNATIVE E: 91MW Plant A: Irrigation P: 16,570 Ha Dam B: Plant C: 84 Hm 6! MW ALTERiJAIVE F: Plant A: 9 MW w Irrigation B: 16,570 Ha Dan F: 84 Hm3 Irrigation C: 18 ,680 Ha De C: 105 ALTERNAT IVE G: Plant A: 9 irrigatian B: 16, 57') DaR : 84 Irrigation D: 20,710 Dam D: 123 MW Ha Hm 3 Ha ALTERNATIVE H: Plant A: MW w Irrigation P: 16,570 Dam B: 84 irrigation C: 18,680 Ha DamC: 105 Irrigation D: 20,710 Ha Hm3 Dan D: 123 ALIERNATIVE K*: 9 MW Plant A: Irrigation B: 16,570 Ha 84 HW3 Datq A: Irrigation D: 20,710 Ha 123 H 3 Dam D: ALTERNATIVE L: 9 MW Plant A: Irrigation B: 16,570 Ha 84 H 3 Dam A: Irrigation C: 18,680 Ha Dan C: 105 Hm3 Irrigation D: 20,710 Ha 123 HW3 Darn0: ALTERNATIVE M: Plant A: 9 MW Irrigation B: 16,570 Ha Dam B: 84 HI3 Plant C: 6 f1W Irrigation D: 20,710 Ha 123 Hm3 Dam D: ALTERNATIVE N: Plant A: 9 MW Irrigation B: 16,570 Ha 84 Hm3 Dam A: 6 MW Plant C: Irrigation D: 20,710 Ha Dan, D: 123 HW3 i 12 the future calculate benefits [0 1 1 ,022,032 E)41 we obtain the optimal yield maximum net Alternative A would the evaluate 2 use We ,E3 2 ,4IE and the LP 3 5 ]. To of 9E4 1 and policies that A. Alternative benefits package, irrigation are not We but only about the operating rules, net [e1 1 t62 2 ,3 conditions ]. for benefits maximum of 3 5 releases about the really concern To that 2 [ 1 1 ,e 2 2 ,E 3 2 , e 4 1 ,e 5 3 ] does situation defined by conditions value 2 optimize operating rules for Alternative A under the we come, [E 1 1 ,0 by maximum net the if the produce defined situation Alternative A under 9(5 3 1. of Alternative A for all performance the possible future situations, we must run the LP program a total of 243 future situation E9i. maximum net benefits A. for Alternative Alternatives B same M. through the Alternatives optimization runs the C-2 Tables under was repeated for to be to C-14 show their uncertain situations. of net benefits of to C-14 are called curves C-i it has process benefits These Tables In total, indicates the Appendix C in resulting from the 243 optimization runs The respective maximum net candidate C-1 Table for each possible rules optimize operating to times, A to N. necessary to perform to calculate the curves 14'243 of net = 3402 benefits for 14 candidate alternatives. Expected value The of expected net benefits value of for net the alternatives benefits for a candidate 113 are the probabilities The weights given is also required is to net benefits of first stage of the to calculate both we need again conditions uncertain compute The expected va lue shown alternative in is of use Chapter equal 4, to f ar how potential probabilities the B-1. the the last The of the net the of the to Index values (note To of the is to stage of the stan dard is just the value ex pected alway s has delta ideal the delta curve of values var iation the by The delta c urv e of in dica t es coefficient of variation divid ed deviation As Ta bI e in the coef ficient of delta curve is between the curve reaching the t he deviation of standard it 4-4). (Table cal Ied the But to c alculate the expected value and is second stage for al ternative alternative and the is fro m is alternative basin. The curve curve delta The alternative. candidate the candid a te The differen ce benefits curve. candidate calculate the difference the benefits net benefits. net benefits ideal the benefits. in Table 4-11. curve of its calculated from also net net expected robustness for a given index of The of curve only the alternatives robustness for the of is of net the uncertain conditions g iven every candidate alternative are Indices of curve benefits the of The values Table B-1. in of its from net values the of weighted average of value expected The benefits. calcul ated directly is alternative that the be greater than ze ro). of coefficient Robustness of an of variation of the 114 values of of its robustness delta of curve. Table the TABLE 4-11: candidate 4-11 alternatives. EXPECTED NET 4.3.4.- BENEFITS NB Curve 2.818 0.000 2.000 0.769 1.231 Alternative Alternative Alternative Alternative B C D E 1.698 0.710 1.657 1.913 0.858 0.846 0.825 1.057 1.142 1.154 1.175 0.943 Alternative F 2.718 1.343 0.657 Alternative G Alternative H Alternative I Alternative J Alternative K Alternative L Alternative M Alternative N 2.331 1.722 1.871 2.693 2.292 1.706 2.545 2.507 0.455 0.374 1.008 1.079 0.437 0.381 0.972 0.872 1.545 1.626 0.992 0.921 1.563 1.619 1.028 1.128 Step 4: Compare information where horizontal indicated - CV, among alternatives expected value in where study. in and value of robustness a "large" 4-12 shows Each number. the Pareto alternative net benefits is net is benefits in is the and the to organize The best method is Chapter 4, robustness N is of alternative, we have a clear form expected axis 2. Figure case CV (N=2) 0.496 index of robustness for every graph N - INDEX OF ROBUSTNESS Alternative A Once we have the this indices Expected net benefits and index of robustness of each candidate alternati ve ALTERNATIVE Ideal summarizes the to use a in the vertical. As should be represented by N In this case, two-objective represented as N is equal to graph for our a point in this graph. The non-inferior set is formed by Alternatives H, K, G, 115 ALYT1YE I 1.2 I hLTT ALTIMYJI E 1.8 ALTED&YIVE 1.4 1.2 TETIVE J 1 ALTETETIET A LTIEATIVE I LTENTIVI F 0.8 ILEYWAIVE D 8 0.4 0.2 I I 1 2 PECTD IET FIGURE 4-12: N, M, J, and Alternative provided by n E H, 3 EFNUITS (9llin$J Fareto two-objective graph for the candidate alternatives of the case study F. The and the Alternative maximum robustness maximum F. expected With respect values, we can calculate the percent and expected net benefits of of is provided by net benefits is to these extreme losses in robustness the other alternatives (Table 4- 12). Alternatives K and G seem to be These alternatives (only about 15X in the "compromise" have relatively high expected net benefits less expected net are relatively robust benefits than (only about 20X less Any other of the non-inferior alternatives has significant decrease benefits. The final outside of zone. choice the scope of F), and they robustness than H). (H, M, N, J, or F) in either robustness or expected net of this an alternative thesis, to because implement is it depends on 116 TABLE and in Decrease in expected net benefits of the non-Inferior index of robustness to the maximum with respect alternatives va ILies 4-12: DECREASE IN EXPECTED NET BENEFITS NON-INFERIOR ALTERNATIVE process. conclusion as However, that Alternatives K and G. may obtain some conclusions benefits we shows study the the most expected (those Therefore, Alternative (Plant A, which we expected net net benefits. to likely be the greater contains Irrigation B with its Table The built. projects of more than 50% have F, which of value most the with probability are the alternatives projects in Figure 4-12, and their project likely alternatives which contain built being chosen are and robustness. First, 4-7 the study, about the relationship between the alternatives the of characteristics case benefits and robustness the Pareto curve analysis of From the the decision-making the of expected net Conclusions about 4.3.5.- in possibility of have more alternatives 16.87% 21.75% 133. 50% 160. 29% 188.92X 259. 58X involved parties the among IN OF ROBUSTNESS -------- 36.64% 15.67%/ 14.26% 7.77%Y 6.36%. 0.93% Alternative H Alternative K Alternative G Alternative N Alternative M Alternative J Alternative F- agreement DECREASE INDEX to be in Table 4-7) net benefits. the most profitable associated Dam B, and 117 C Irrigation with the formed g reatest less by For benefits. projects t han bui It expected Plant C than the the E w ith C, the in condition S formed by more by seven c as e, of some expected projects than net benefits comparing Alternative should expec t that those projects. are the most m itigate the p rojects. In fact, alternative of effect (Alternat ives H in an discussed at exception. possibilit ies mor e the most for this reason We are that the robust. But, and of varying unfavorable two alternatives and L, cons istent relationship included built Irrigatio n D. However interrelated have project s is no the robustness. projects) are there be alternati ve s containing greater Irrigation may be to to D, they rule s operation interesting An of study becaus e smaller I with Altern ative K. The alternatives formed by more robust, has G o r Alternative chapter, Se cond, we same pr obability of F. likely ry happens, a s can be seen by of the I rr igation C by lower altern atives containing G, by substitut in g Plant C by T he r ef ore, have to formed Al ternative is less 4 -7 ). Table expected net A Iterna tive than on D circumstances end a happ en s when Irrigati D Alternative s pe c ial the Irrigation rule smaller subs tituting than s ame contra is D has ar e supposed Plant G Irrigation (s ee C bu t h ave Since benefits Other alternatives benefits. Alternati ve F, is the alternative C), Dam proj ects profitable this exception to net Alternative ne t Irrigation expected example, Irrigation D. being associated its with eac h one formed except in this between the number its robustness. For 118 hiie rarchical From the Dam A and Dam B are the s ame in present is with Dam D, I, N, Table This 4-7). smaller have explains that expected net include upst ream the of conf igurat ion the Dam B . of for This for used regulati on poss ibi I it y conf ers grea t e r There is composition of that include a of of and A Plant Irrigation B, can to the robustness consistent the alternatives Irrigation D on sever al alternat relat of uses G (see al ternatives look i nrg at by Dam A is addition of of also be used t he Transfer. of the water ives. ionship and robu stness: (with its and used for reg ulati on in regulati Dam B 4-1.). A, for F of alternat ives which Dam Dam A for using second be However, regulation of than Figure (see can only Dam B Irrigation B. water of be in g system the Al ternatives M, Dam B may be explained of instead A Dam benefits) contai ning Dam A alternatives benefits be a Iternative than built greater robustness containing Dam B. The of E, B, to be likely less is respectively. Dam A an case the is This that exp ected net less Alternatives over K and J, having with Dam B. than A found We alternative. that not c an and proj ect s see we 4-11), (Figure incompatibl e ( although robust more gra p h Dam B by replace which alternatives first The robustness. and alternatives to the refers relationship Dam A. the of composition the between consistent relationships however, two are, There projects. six by each formed and M, N than Alternatives robust more is project, one only by formed A, Alternative example, between the alternatives associated Dam D) are more 119 robust than example, sa me the Alternative B Alternative D etc. project built the an to be increase alternative using the extra downstream river and the this C uriously, that water because flow is is no t D), exist ence of clear conc lusion relationship always and of characteristi cs net of very explanation excess otherwise will this be use lost section of Then, alternatives in over the from Irr igations efficient D last chance of in benefits b etween of B and C although the water with Irrigation than without Irrigation provides wit h the the possibility of individual as a whole. that, while there configuration net benefits, respect of is the expected seem to be more syster th e in likely any condition. Irrigation D) the a is a G, than Irrigation tributary. this section their exits D present ext ra water relat ionship alternatives that middle Irrigat ion D under The including F not poss ible the water the reas on why some small er expected using is make D have the when is D For D. Alternative Alte rnative The return flows the does 4-7), I rrigation than K, Irrigation (that th e of robust Irrigation (see Table ro bustness is without Alternative Extra wat er Irrigation D (and in less requirem ents) downstream. the is tha n Alternative H, of alternative to their of is a the no such clear robustness. The projects important than the (as Dam A and configuration 120 CONCLUSION CHAPTER 5: the summarizes in characteristics of The first part thesis. this The the method within genera I framework of water resources planning. the S umar! A the method of This reason for, The uncertain p rocess to maximize is water resources the is existence of planning water systems. of th e concept problems, determi nistic have the in for water to maximize robustness. is robustness considering parameters ional tradit obj ective other the method obj ectives the (th e benefits and objective) of One planning. net expected two-objec tive thesis presented a resources In the parts. described planning method second par t discusses two in divided is chapter This robustness does of not any meaning. The thesis. of consideration Robustness is of performance correspond maintain a measure a a proj e ct performance relia ble Ch apter 2, of to robust performance, I t he orig ina I is sensit iv ity to uncer ta in conditions . that the un certain parameters In robustness projects, t hat independently in of this the Inse ns itive are able to of the value happen to take. reviewed some indices to measure 121 robust ness exposed and d) do in to use robustness may be benefits a correspond this thesis. in the index. the of consequence b] Fiering to index value of net obtain the is and the of the of possible net depending the hand, on the robust project a project but we of parameters. indicates input of the as can not project, the is benefits If net this non-robust is a wide range of dispersion indicates of it c, robustness distribution uncertain variation on small from a possible b, robustness uncertainty, we benefits there indices I developed a new robustness disperse, project: of quantify analysis, of The [1982a, concept to of the very performance performance and two-objective a function distribution systems. Therefore, This to performance of other water uncertainty. Because of able distribution of distribution single as [1982a, not it to related predict resiliency by Hashimoto indicated and and project parameters. of the because the values performance, On similar even under different is conditions. The planning in this process to obtain alternatives thesis requires to the is solve use technique. The method facilities that could the part the of indices of screening begins be by built the method is a set of non-inferior robustness general two-objective performing the four steps described of the of method problem. models as displaying of the proposed The method optimization all basin. possible in the Then, after in Chapter 3, the result development strategies 122 for the basin, to prove to a n, hypotheti cal in th e of interesting inf the the net benefits the bas in Cu rve. wo lI d al tern at i ve. benefits of projec ts of Character istics planning as also mo St was applied that, method a whole method shows in the produces and on the evaluates yield can for never goal the to the candidate the ideal profitable incompatible after on the net exceeded for information least insight maximum each uncertain be of provides summary, In the possible and met hod the time section any within planning on net the projects combinations method the potential the of water the and amount expenses discusses the of in is method Different decision makers. characterized by This the of purpose information to by it 4, is the resource framework The and Chapter projects. and basin the The a good can have In is the as This cI ose detected. are we applied, method Also co mpa tible and identified. are as The proj ects. individual the benefits The met hod curve. of ideally any to be method, e xample, net is the basin the These alternatives of curve, future conditio n. candidate graph. t o p roduce net benefits, represented by of benefits on For basin th e form Pareto ormation potential ideal ity main ad vantages proj ects. individual Pareto study. case fin ding of a iabil practical the One process presented relevant to provide planning methods are information provided generating improvements the to information. the planning 123 process that about, the method described in particularly multiobjective There maximizes is However, deal solution et But the since limited very some methods still linear (stochastic of practical subject of are basin, of the an and there hydrological in theory, that may, optimal unique described by Loucks these methods utility of require they if we situation of the obtain programming, objective our estimation of in the randomness and uncertainty 1981). al., conditions are There itself. with uncertainty economical the procedure, resources systems is is also uncertainty due to process and optimization the real represents water great uncertainty. There and wrong with this nothing the physical bring uncertainty solution that that the model the with resources were certain basin. will thesis water Traditional only provide one measure. dealing this analysis. Uncertainty. models in extraordinary is computer facilities. The method to developed with the deal uncertainty this task by performing Any available problems method fact the case Another to study thesis issue. is optimization runs. independent solve specially suited The method accomplishes deterministic (screening models, not requires does XT personal many algorithm be used can this in optimization in particular). sophisticated computer equipment; in Chapter 4 was fully solved using an The in IBM computer. advantage of the method is that there is no 124 the limitatoion to want the analysis, number, time is not method The in the when future only it quantifies is the ability and between the basin the in computer uncertainty of our by cond itions in a measure of thesis This calls projects to in the area of n is of sensiti vity the predict surely not projects. the because about uncertainty, can the of uncertainty of effect the under planning for serves we present, performance robustness stud y relations hi ps concerned are projects. We uncertainty actual not also but uncertainty, th e a nd mo re . form mathematical but it se If, exist ing the modeling variables t o the the . Of course, the reased, inc we in parameters applied De limi Lation represent to knowledge is runs the method by given still If consider. can we uncertain ot her necessary. The is uncertainty can method optimization of of or more include to level uncertainty. The in uncertainty projects. this projects outcomes of the However, the project example, pollution) practical to problem of measure criterion use them. may that can have be considering Methods alone are is the t he most merits correct ed quantifi ot her could c riterio evalua te to criterion evaluation To Economic analysis. Multiobjective utilized planning. resources water' re search the advances further me thod when of the all the in m onetary terms. non-economic effects (for interest decision maRers. non-economic found in effects the The is how literature to 125 quantify ecological, these methods recreational are general, other and and are some effects. Some of for specific just cases. The this thesis said before, this of method As techniques. net benefits as part alone. under the countries, issue of may be critical. In subsistence net some produces that produce nothing. The method is to be Pareto projects given curve is Pareto or in other words, in curve to is easily the i s the dec is ion that the visuali2 resources projec t result makers. First, when regions. that to than provide of The a but may also the finding method, advantage about a set of and it of t he the alternati ve and under stood. is much decision always implement non-inferior the the decision-making process curve. projects to problem by information ed developing underdevelo ped preferable two-objective Pareto The reasons why the is be performance implemented reliable solves the or, curve projects. benefits to be to element may produce muc h more net benefits, project Pareto a of and consider criteria project of water people the implemen t to cases, those the areas rural robustness to robustness for decision making process projects may be These are important in In conditions. of an reliability indicates future a mean is Robustness it the multiobjective studies net benefits in the decision making considered because thesis of a two-objective problem. We of that neither robustness nor used part is There are two improved with making involves a 12 6 collective the probaba I it y solu o s sI this object ive wate r the of proc eS decision guaranteed of agreement is not the the one of thi s not exist. the tradeoffs and the only makers to be t he case planning resource best nego t iat ion. roum for negotiati on, objectives, the method the This in concept compromise s in g Ie traditional that present alternative does to that is provide much two- objective methods like best alternative does of is in increases not process A dec is ion-mak ing among in curve finding techn iques solution. the of optima I one However, thesis, Pareto thi necessary to define s case robustness and net benefits. Secondly, between analysts decision-making and group it in the objectively compared case of non-infer'ior being inferior, improves strongly supports the Pareto with other it The candidate should be communication a if example, a particular project, of performance graph. the For makers. decision the analyst may evaluate locate also the project and can project projects, disregarded in and favor now be in the of a project, 127 APPENDIX A 128 FORMULATION OF THE SCREENING MODEL MAX BENEFITS - DISCOST Alter parameters DISCOST - 0.0817 COST =C COST - u.75 CONSTRUC = 0 IRRIGAT ).f IRRAREA =( IRRAREA = 0 IRRBEN - 0. (3 ELECBEN - 0.1 FOWER = -) Benefits and costs BENEFITS - IRRBEN - ELECBEN = 0 CONSTRUC - DAMCOST - IRRCOST - ELECCOST - TRFCOST = 0 DAMCOST - 1.5 YDAMA 1.5 YDAMB - 3r YDAMC - 2 YDAMD - 0.44 VOLDAMA - 1.76 VOLDAMB - 0.56 VOLDAMC - 1.04 VOLDAMD = ( IRRCOST - 0.25 YIRRB - 0.5 YIRRC - YIRRD - 0.082 AREAB - 0.247 AREAD = C) .56 AREAC ELECCOST - YFLANTA 1.5 YPLANTC - 5 CAFACA - 7.7 CAPACC = 0 TRFCOST - 1.5 YTRF - 0.14 TRFSIZE = 0 Continuity 9.513 SA2 - 9.418 SA1 SA3 + + RELA3 RELB1 + RELB2 + SA1 - 9.228 9.418 SEC SB3 SB1 SC2 SC C1 SD = - 9.037 SB 9.418 SB3 - 9.037 SCi1 - 8.657 SC2 - 9.228 SC3 - 8.847 SD1 .3:2 9.51 SD- + + + Reservoir 3B1 - 8.562 SD 2 68 9.228 SD + + + + + = = - VOLDAMEL- = 0I =~ 0 = 0 = = = ( - S3 C1 - OLDAMhB VOLDAMB - ~VOLDAMC IRRB1 IRRB2 RELB. - RELA3 + RELC1 + IRRC1 RELC2 + IRRC2 RELC: - TRF3 = RELDI + IRRDI - + RELD2 + + + RELD3 - IRRD2 RELB3 - RELA1 + TRF1 = 0 RELA 2 + TRF 2 = TRF3 = 0 TRF1 = 6.6 TRF2 = 3.4 10D IRRTB1 RELBI - RELC1 RELB2 - RELC2 RELC3 IRRTB3 - - IRRTB2 - - - IRRTC3 = ) = U VOLDOMC (9 VOLDA4MC 0 SD1 - VULDAMD '= SD2 - VOLDAMD 1= = SD7 - ViLDAMD Irrigaton loses and - 5.28 2.72 8 volumesE VLDAMA VOLDAMA OVOLDA 5812 SC2 8C3 =A = IF:RTD1 IRRTD2 >= 2.72 7 j>= IFRTD3. - A2- = 9.323 SB1 - FELL RELD2 FE L D. RELA1 RELA2 + SA3 IRRTC2 =. 9.513 SDI1 + S2 9.513 9.513 9.513 9.513 9.513 9.513 9.513 9.513 9.51 IRRTC1 water requirements for irrIga.tion 129 129 IRRB1 - LANDB11 2.125 IRRB2 - LANDB2 =) 2. 08 IRRC1 - LANDC1 =( 2. 8 IRRC2 - L-ANDC2 =0 1.912 RFRD1 - LANDD1 = 1.9c92 IRRDZ - LANDD2 = IRRIGAT - LANDB1 LANDB2 - LANDC1 - LANDC2 LAND1 - AREAB= LANDB2 - AREA1B = 0) LANDC1 - AREAC= LANDC2 - AREAC 0 LANDD1 - AREAD = ( LANDD2 - AREAD <=0 Irrigation return flow IRRTB1 162 IRRB1 ).135 IRRB2 = C) IRRT2 -IRRB1 - 0.03) 0.026 IRRB2 = 0 IRRTB3 ). 184 IRRB1 - 0.214 IRRB2 = 0 IRRTC1 - D.167 IRRC1 - D.140 IRRC2 = 0 2. 12 IRRTC - C.C31 IRRC1 - LANDD1 - - LANDD2 C) = C.027 IRRC2 = 0 = 0 IRRTC3 - C. 191 IRRC1 - 0.222 IRRC2 IRRTD1 -1 . 178 IRRD1 - 0.149 IRRD2 IRRTD2 - 0.033 IRRD1 - 0.029 IRRD2 IRRTD3 - 0. 203 IRRD1 - 0.237 IRRD2 Hydropower constraints 5.37'61 POWERA1 - RELA1 := 0 5.361 FOWERA2 - RELA2 = C 5.361 F0WERA3 - RELA3 <= C 1(.722 FOWERC1 - RELC1 <= 0 10.'722 FOWERC2 - RELC2 <= C 10. 722 FWERC3 - RELC3 <= C) 0. 059 POWERA1 - CAFACA 0 0. 059 FOWERA2 - CAFACA := C) C. 059 F0WERA3 - CAFACA (= 0 0.059 FOWERCI - CAFACC 0 .j059 POWERC2 - CAPACC =0 C). 059 FOWERC3 - CAFACC <= C POWER - FOWERA1 - FOWERA2 - FOWERA3 = 0j 1 r-ansfers size TRFi - TRFSIZE := 0 TRF2 - TRFBIZE <= U TRFSIZE <= 0 Condi ionalitv and maxiLLfn ize AREAB - SQ YDAMB AREAC - 75 YDAMC Y YDAMD < AREAD 'LDAMA YDAMY= (9 VOLDAMB - 7 YDAMB = VOLDAMC - 10 YDAMC .= VOLDAMD () YDAMD <= REA - 50 YIFB <:.= 0 AREAC - 75 YIRRC i= (9 = C) = 0 = 0 - FOWERC1 - POWERC2 - FOWERC3 130 AFEAD CAPACA - 9) YI RRD YF'LANT CAPCC- 3YP'LANTC TRFEZE - END INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER INTEGER YDAMA YDAMB YDAMC YDAMD YIRRP YIRRC 'YIFRD YFLAN'TA YFLANTC YTRF 30 ( = (9 .=0 YTRF 131 SUMMARY OF VARIABLES AND PARAMETERS OF THE SCREENING MODEL Design decision variables Volume of Reservoir A (Hm 3 ) Volume of Reservoir B (Hm 3 ) Volume of Reservoir C (Hm 3 ) Volume of Reservoir D (Hm 3 ) Area of Irrigation B (Ha) Area of Irrigation C (Ha) Area of Irrigation D (Ha) Capacity of hydropower Plant A Capacity of hydropower Plant C Size of the Transfer (m3 /s) VA VB VC VD AB AC AD CA CC T Operational RA, t RB, t RC, t RD, t B, t IC, t D, t Tt (MW) (MW) decision variables Releases from Reservoir A (t = Releases from Reservoir B (t = Releases from Reservoir C (t = Releases from Reservoir D (t = Water diverted to Irrigation B Water diverted to Irrigation C Water diverted to Irrigation D Water diverted to the Transfer season season season season (t s (t s (t s (t = s 1, 2, or 3) 3) 3) 3) 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, or 1, 2, or 1, 2, or eason eason eason eason or 3 or 3) or 3) or 3) Other variables and parameters B C CC e E2 E)3 E)4 E)5 Lst LBt LCt LDt Ps, t PA, t PC, t n as Ts Annual benefits ($) Annual costs ($) Total construction costs ($) Price of agricultural products ($/Ha) Price of electricity ($/MWh) Discount rate (X) General increase or decrease in constr. costs (%. general increase or decrease in irrigation water demands Land irrigated at Irrigation s in season t (Ha) land irrigated in Irrigation B in season t land irrigated in Irrigation C in season t land irrigated in Irrigation D in season t Power produced at Plant s in season t (MWh) power produced at Plant A in season t (t = 1,2,3) power produced at Plant C in season t Projects lifet ime (years) Variable costs of Reservoir s ($/Hm3 ) Variable costs of Irrigation s ($/Ha) Variable costs of Plant s ($/MW) 132 Transfer [$/( M3 /s R eservoir s ($) I rrigation s ($) P lant s ($) T ransfer ($) Integer variable for Reservoir s: If Reservoir s is built: YV, = i p :Variable FVS FAs FCS FT : : : : YVS : Fixed Fixed Fixed Fixed costs o f costs costs costs costs )] of of of of EB If Reservoir s is not built: YV S= 0 : Integer variable for Irrigation s: If Irrigation s is built: YAs = I If Irrigation s is not built: YAs =0 : Integer variable for Plant s: If Plant s is bu ilt: YCS = 1 If Plant s is no t built: YC, = 0 : Integer variable for Transfer: If Transfer is b Ui It: YTS = 1 If Transfer is n ot bu i It: YTS = 0 : Storage in Reser vo ir A in season t (t = i 2, 3 Storage in Reservoir B in season t (t = 1 2, 3 Storage in Reservoir C in season t (t = 1 2, 3 Storage in Reservoir D in season t (t = 1 2, 3 Re l ease s from Reservo ir A in season t Re I ease s from Reservo i r B in season t Release s from Reservo ir C in season t Release s from Reservo ir D in season t Inflows in Reservoir B in season t Inflows in Reservoir D in season t Water d iverted to Irr ig at ion B in season t Water diverted to I rrigati on C in season t Water diverted to I rrigati on D in season t evaporation coeffic ient of Reservoir A in season evaporation coeffic ient of Reservoir B in season evaporation coeffic ient of Reservoir C in season evaporation coeffic ient of Reservoir D in season : Inflows of the left branch of the river in season t : Inflows of the center branch of the river in season t : Inflows of the right branch of the river in season t : Water that return to the river in season t from Irrigation B : Water that return to the river in season t from Irrigation C : Water that return to the river in season t from Irrigation D :water losses in Irrigation B channels EC :water YAS YCS YT SA t SB, t SC, t S D, t. RAt RBt RCt RDt IBt IDt DBt DCt DDt eAt eBt eCt eDt fit f2t f3t RIBt RICt RIDt ED t Qlt return to losses in Irrigation C t t t t channels : water losses in Irrigation D channels :irrigation water demands in season t (t = 1, 2) : water non consumed in season I (1 = 1, 2) that the river in season t (t = 1, 2, 3) 133 HA HC water consumed by head of Plant A head of Plant C SA SC kt ht efficiency of Plant A efficiency of Plant C number of seconds in season t number of hours in season t (t if u load factor factor of utilization TRt Flow of crops the Transfer in season 1,2,3) t (t = 1,2,3) 134 (1FUR Fixed (14 20 $) larlile (1.6 $A3) iA 1.5 3 1.5 S.0044 6.6if6 ~C 3.S II 2. *.w5% 6.61W4 w I w I C low 15W EEil FIGURE A-i: Construction costs of SIZE (O) the dams COST (1364$) Tariable Fired (186 j(1g6$/83) IflIGTIM 0 3, IUIQTI 31 0.25 IBIGTIO C 0.50 Ifl6QTIm D i.0 8.00002 O.000247 l.0536 20 IuI&ATIcm C 10 250M SO0 (Ba) InIGTIOU SIZE FIGURE A-2: Costruction costs of the irrigations 135 Fired (l'$ tilable (16g/fgU PAIaT A . PLAWi C .58 M.g50 8.11 (186 s) 26 MIT C to FnAl' A IN mno CWCTf Construction costs FIGURE A-3: of the power plants (16 u le Filed Tism 150 Turble .14 36 1 TnAmTE SilE (*3/i) FIGURE A-4: Construct i on costs of the transfer 136 APPENDIX B 137 TABLE B-1: Frobabilities of uncertain WATER REQUIREMENTS = -50 DISCOUNT RATE: 8 % the 243 elements parameter vector | MATER REQUIREMENTS = even | of the & WATER REQUIREMENTS = +50 % ======================================================== E1.Pr.=10 EI.Pr.=20 EI.Pr.=30|E1.Pr.=10 E1.Pr.=20 El.Pr.=30:El.Pr.=10 EI.Pr.=20 E1.Pr.=30 CONSTR. Agr.Pr.=30: COSTS= Agr.Pr.=40: -25 1 :Agr.Pr.=50 0.3751 0.5001 0.3751 0.6001 0.8001 0.600% 0.5251: 0.7001 0.5251: 0.3751 0.5001 0.3751 0.6001 0.8001 0.600% 0.5251: 0.7001: 0.5251: 0.3751 0.5001 0.3751 0.600% 0.8001 0.600% 0.5251 0.7001 0.5251 CONSTR.Agr.Pr.=30: COSTS= Agr.Pr.=40: even :Aqr.Pr.=50: 0.2621 0.3501 0.2621 0.4201 0.5601 0.420% 0.367%: 0.4901: 0.3671: 0.262! 0.350% 0.2621 0.420% 0.560% 0.4201 0.3671: 0.4901: 0.3671: 0.2621 0.3501 0.2621 0.4201 0.560% 0.4201 0.3671 0.4901 0.3671 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 1 :Agr.Pr.=50: 0.1121 0.150% 0.1121 0.1801 0.2401 0.1801 0.1571: 0.210% 0.1571: 0.1121 0.1501 0.112! 0.180% 0.240% 0.1801 0.157%| 0.2101: 0.157%: 0.1121 0.1502 0.1121 0.1801 0.2401 0.180% 0.1571 0.2101 0.1571 WATER REQUIREMENTS = -50 1% DISCOUNT RATE: = 10 = = = :WATER REQUIREMENTS = +50 % WATER REQUIREMENTS = even = = = = = = = = :El.Pr.=10 E1.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.220 E1.Pr.=30:E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 % :Agr.Pr.=50: 0.625% 0.8331 0.6251 1.0001 1.3331 1.0001 0.8751: 1.1671: 0.8751| 0.6251 0.8331 0.6251 1.000% 1.3331 1.000% 0.875%| 1.1671: 0.875%: 0.6251 0.8331 0.6251 1.0001 1.333% 1.000% 0.8751 1.167% 0.875% CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: even :Agr.Pr.=50: 0.4371 0.5831 0.4371 0.7001 0.933% 0.7001 0.612%: 0.8171: 0.6122: 0.4371 0.583% 0.4371 0.700% 0.9331 0.7001 0.612%: 0.8171: 0.6121: 0.4371 0.583 0.4371 0.700! 0.933% 0.7001 0.6121 0.8171 0.6121 0.3001-. 0.2621: 0.4001 0.350%: 0.300% 0.2621: 0.1871 0.250% 0.1871 0.3001 0.4001 0.3001 0.2621 0.3501 0.2621 |----- -- ------------ - CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 % :Agr.Pr.=50: 0.187% 0.2501 0.1671 0.300% 0.4001 0.3001 ---------------------------------- 0.262%: 0.3501: 0.2621: - ------------------------------------ WATER REQUIREMENTS = -50 1 DISCOUNT RATE: 0.1871 0.250% 0.1871 - --------------------------- :WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 % :EI.Pr.=10 El.Pr.=20 E1.Pr.=30:EI.Pr.=10 El.Pr.=20 EI.Pr.=30 E.Pr.=10 E1.Pr.=20 EI.Pr.=30 CONSTR. :Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 % :Agr.Pr.=50: CONSTR. :Agr.Pr.=30: COSTS= :Agr.Pr.=40: even :Agr.Pr.=50: 0.2501 0.3331 0.2501 0.400% 0.3501: 0.5331 0.4671: 0.400% 0.3501: 0.2501 0.3331 0.250% 0.4001 0.5331 0.400% 0.350%: 0.4671: 0.350%: 0.2501 0.3331 0.2501 0.4001 0.533% 0.4001 0.350% 0.4671 0.3501 0.1751 0.2331 0.1751 0.280% 0.245%: 0.373% 0.3271: 0.280% 0.2451: 0.1751 0.2331 0.1751 0.2801 0.373% 0.2801 0.2451: 0.3271: 0.245%: 0.1751 0.2331 0.1751 0.2801 0.3731 0.2801 0.2451 0.3271 0.2451 0.1051 CONSTR. :Agr.Pr.=30: COSTS= :Agr.Pr.=40: 0.0751 0.120% 0.1051 0.0751 0.1201 0.1051 0.0751 0.120% 0.100% 0.1601 0.1401: 0.100% 0.160% 0.140%: 0.100% 0.160% 0.1401 +25 % :Agr.Pr.=50: 0.0751 0.120% 0.1051 0.075% 0.120% 0.1051 0.0751 0.120% 0.1051 138 TABLE Optimal sizes B-2: of Dam A REQUIREMENTS = even REQUIREMENTS 2 -50 1 : MATER WATER DISCOUNT RATE: WATER REQUIREMENTS = +50 1 E1.Pr.=10 E1.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 El.Pr.=30 E1.Pr.=10 EI.Pr.=20 E1.Pr.=30 0.00 0.00 0.00 -- 0.00 0.00 0.84 CONSTR. Agr.Pr.:300.00 0.00 0.00 0.00 0.00 0.84 COSTS= Aqr.Pr.=40 0.00 0.00 0.00 0.00 0.00 0.00 -25C IAgr.Pr.=50 |--------------------------:------------------------------------------0.00 0.00 0.00 0.00 0.00 0.00 25NSTRAgr.Pr.:30: 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR.:Agr.Pr.=30 0.00 0.00 0.00 0.84 0.00 0.84 COSTS= :Agr.Pr.=40: |Agr.Pr.=50| 0.84 0.84 0.84 CNSTR. Agr.Pr.=30 COSTS= :Agr.Pr.=40: +251 !Agr.Pr.=50 0.00 0.00 0.84 0.00 0.00 0.84 0.00 0.00 0.84 even ------ -------- | 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ------ ------------------- ---------------- ------------- REQUIREMENTS = even REQUIREMENTS = -50 1 | WATER WATER DISCOUNT RATE: 10 1 - 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ------------------0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 : WATER REQUIREMENTS = +50 1 |E1.Pr.=10 El.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 EL.Pr.=30:E1.Pr.=10 EI.Pr.=20 El.Pr.=30 CONSTR. Agr.Pr.:30COSTS= Agr.Pr.=40: -25C :Agr.Pr.=50| 0.00 0.84 0.84 0.00 0.84 0.84 0.00 -- 0.00 0.00 0.84 0.00 0.00 0.00 0.00 0.00 0.00 | 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 COSTR.Ar.Pr.=30! COSTS= :Agr.Pr.=40: even1Aqr.Pr.=50: 0.00 0.00 0.84 0.00 0.00 0.84 0.00 0.00:! 0.84 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -- ----CONSTR.:Aqr.Pr.=30: +25n :Agr.Pr.=50: ----------------- ------0.00 0.00 ! 0.00 0.00 :----------------------- :Agr.Pr.=40: 0.00 0.00 COSTS= DISCOUNT RATE: : : : ------------ | --------- 0.00 0.00 0.00 0.00 |: --------------------------0.00 0.00 0.00 0.00:: 0.00 0.00 0.00: ---------------------------- MATER REQUIREMENTS = -50 I | 0.00 REQUIREMENTS = even WATER 2------------------------------------------------0.000 0.00 0.00 0.00 0.00 CONSTR.EAgr.Pr.=30| 0.0 ----0.0 0.00------------0.00-- 0.00 0 COSTS:- :Ag 0.00 0.00 0.00 0.00 0.00 COSTS= Agr.Pr.=40| 0.00 0.00 0.00 0.00 0.84 :Agr.Pr.=50: -25 : WATER REQUIREMENTS = +50 1 0.00 .0 0.00----------------0.00 0.00 : 0.00 0.00 0.00 0.00 0.00---0.000.00 0.00 0.00 0.00 0.00 0.00 0.00:1 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00 0.00 0.00 0.00 0.00 0.00 0.00: 0.00 0.00 0.00 0.00 0.00: 0.00 0.00 0.00 : 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00: 0.00 0.00 0.00 0.00 0.00 | 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 Agr.Pr.=30: CONSTR 0.84 0.84 0.00 CONSTR= Agr.Pr.=430: 0.00 0.00 0.00 even Agr.Pr.:50 0.84 0.00 CONSTR.:Aqr.Pr.=30: 0.00 COSTS= :Agr.Pr.=40 +251 :Agr.Pr.=50 0.00 0.00 : 139 TABLE B-3: Optimal sizes of Dam B MATER REQUIREMENTS x -50 1 WATER REQUIREMNTS = even : WATER REQUIREMENTS = +50 1 DISCOUNT RATE: EI.Pr.=10 El.Pr.=20 EI.Pr.=30 E1.Pr.=10 E1.Pr.=20 El.Pr.=30;E.Pr.=10 EI.Pr.=20 EI.Pr.30 CONSTR. Agr.Pr.:30 COSTS= Agr.Pr.=40: -25C :Agr.Pr.=50: 0.00 0.00 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 -0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 CONSTR. :Agr.Pr.=30: COSTS :Agr.Pr.=40: even !Agr.Pr.=50: 0.84 0.00 0.00 0.84 0.00 0.00 0.84 0.84 0.00 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 CONSTR.:Agr.Pr.=30: COSTS= :Aqr.Pr.=40: 4251 :Agr.Pr.=50: 0.84 0.84 0.00 0.84 0.84 0.00 0.84 0.84 0.00 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 WATER REQUIREMENTS = -50 1 DISCOUNT RATE: 10 I :WATER REQUIREMENTS = even ; WATER REQUIREMENTS = +50 Z -========================================================================================= :E1.Pr.=10 EI.Pr.=20 El.Pr.=30:E.Pr.=10 El.Pr.z20 E1.Pr.=30|E.Pr.=10 EI.Pr.=20 E1.Pr.=30 C-SR-grP.3 CONSTR.:Agr.Pr.=30: COSTS :Agr.Pr.=40: -251 :Aqr.Pr.=50: C-. 0 0.84-0.-4--4------------------------------------------ 0.84 0.00 0.00 0.84 0.00 0.00 0.84 : 0.00 : 0.84 1 84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 | 0.84 0.84 . 0.84 |.4 CONSTR Agr.Pr.=30| COSTS :Agr.Pr.=40: riven :Agr.Pr.:5o: 0.84 0.84 0.00 0.84 0.84 0.00 0.84 0.84 | 0.00 : 0.84 0.84 0.84 CONSTR.:Agr.Pr.=30: COSTS: :Agr.Pr.=40: 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84---------0.84 -0.84 - ----- ----25-1 .-. :---------------i-----|-------------- Ar.Pr.-5- : 0.84 0.84 0.84 WATER REQUIREMENTS = -50 1 DISCOUNT RATE: 12 1 0.84 0.84 0.84 0.84 0.84 ------- 484 0.84 0.84 0.84 0.84 : 0.84 0.84 0.84 0.84 0.84 0.84 .4 8 0.84 0.84 0.84 0.84 0.84 0.84 0.00 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 ---------------------0.84 0.84 0.84 : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 :EI.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:El.Pr.=10 EI.Pr.=20 EI.Pr.=30 -- ------- : CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: -5 1 :Agr.Pr.=50: - ------ 0.84-0.8--------------|- 0.84 0.00 0.00 0.84 0.84 0.00 0.84 0.84 0.00 ------------------------ --------- 0------------------ 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 ----------------- 0.84 0.84 0.84 0.84 0.84 0.84 ---------------- 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 ;Agr.Pr.=50: 0.00 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 1 :Agr.Pr.=50: 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.84 0.00 0.84 0.84 0.00 0.84 0.84 0.00 0.84 0.84 CONSTR. :Agr.Pr.=30: COSTS= |Agr.Pr.=40: even 140 TABLE B-4: DISCOUNT RATE: Optimal sizes of Dam C WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS =evn MATER REQUIREMNTS x +50 2 :EI.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 EL.Pr.=20 EI.Pr.=30 C-SP-gr R. 3 1 0.00 0.00 05--.05--.0--.05--.00 ---. --. 0 COSTR Agr.Pr.=30: 1.05 0.00 0.00| 1.05 1.05 0.00 1.05 0.00 0.00 COSTS: Agr.Pr.=40 1.05 0.00 0.00: 1.05 0.00 0.00 1.05 1.05 0.00 25 :Aqr.Pr.=50: 1.05 0.00 0.00| 1.05 0.00 0.00 1.05 1.05 0.00 CA- -STR 1.05 . 1.05 05-1.05--.00--.00--.00 -. --. 00 CONSTR.:Agr.Pr.=30: 1.05 1.05 0.00 1.05 1.05 0.00 1.05 0.00 0.00 COSTS: :Agr.Pr.=50: 1.05 1.05 0.00 1.05 1.05 0.00 1.05 1.05 0.00 CONSTR.:Aqr.Pr.=30: COSTS= :Agr.Pr.=40: -25 1 :Agr.Pr.=50: -- 10U============= - - - - -- ---------- 1.05 1.05 1.05 - - - - 1.05 0.00 0.00 - - - - 1.05 1.05 1.05 - - - - - 1.05 1.05 1.05 - - - - 0.00 1.05 1.05 - - - 0.00 1.05 1.05 - - - - - - 0.00 0.00 1.05 - - - - 0.00 0.00 0.00 - - - - - - ---------------------------------------------------------- DISCOUNT RATE: 10 1.05 1.05 1.05 ATER REQUIREMENTS : ATER REQUIREMENTS -50 : MATER REQUIREMENTS even 1- -- 500 - CEI.Pr.10 EI.Pr.20 EI.Pr.30:E1.r.10 EI.Pr.20 EI.Pr.30:E1.Pr.10 EI.Pr.20 E 0.Pr.0 -- -- --- --- - -- - CONSTR.:Agr.Pr.=50: 1.05 COSTS: -- - - -- 1.05 0.00 1.05 1.05 0.00 1.05 1.05 0.00 Cg-r---. -R A = ----- .0 COSTR.:Agr.Pr.=30: 1.05 1.05 . 1.05 COSTS: 1.05 0.00: :Ar.Pr.=40: -25% 1Agr.Pr.:50: :Ar.Pr.=40: 1.05 - --- - - -- - ---------RATE: - --- : 1.05 1.05 0.00 : 1.05 1.05 0.00 -5 15 1.05 1.05 :---------------------- - -- : - - - ----- - - - - - - 1.05 0.00 0.00 1.05 1.05 0.00 1.05 1.05 0.00 ---------------------00 -0 . 1.05 0.00 0.00 0.00 0.00 1.05 1.05: i.0s 0.00 0.00 1.05 1.05 1.05 0.00 0.00 1.05 1.05 0.00 El 1.05 1.05 0.00 0.00 1.05 1.05 ---0.00 0.00 1.05 ------------------------------- --------W ATER - 0.00 !Agr.Pr.=50: 1.05 1.05 0.00 1.05 12I ============================ ======= CONSTR. 1.05 EAgr.PPr..30: 1.05 1.05 E.1.05 COSTSR :Agr.Pr.=40: 1.05 1.05 1.05 1.05 +25C :Agr.Pr.=50: 1.05 1.05 1.05 1.05 DISCOUNT -- 1.05 even ----- - 1.05 0.00 0.00 0.00 ---------- ------------------------------------------------ REQUIREMENTS = -50 MATER 21 REQUIREMENTS = even : MATER REQUIREMENTS = +50 1 12 1 :E2.Pr.1A E1.Pr.=20 EI.Pr.=30E.Pr.=10 El.Pr.=20 E0.Pr.=30E.Pr.10 EI.Pr.:20 EI.Pr.=30 --------------------------------------- CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25e :Agr.Pr.=50: 1.05 1.05 1.05 ---------------------------- 1.05 1.05 1.05 0.00 0.00 | 0.00 1.05 1.05 1.05 1.05 1.05 1.05 0.00 1.05 0.00 0.00 1.05 1.05 ------------------. 0.00 1.05 1.05 . . 0.00 0.00 1.05 . . CONSIR. :Agr.Pr.=3o: 1.05 1.05 1.05 : 1.05 0.00 0.00 0.00 0.00 COSTS= :Aqr.Pr.=4o: 1.05 1.05 1.05: 1.05 1.05 1.05: 0.00 0.00 0.00 even :Agr.Pr.:so: 1.05 1.05 1.05: 1.05 1.05 1.05: 1.05 1.05 0.00 CONSTR. :Agr.Pr.=30: 1.05 1.05 1.05 0.00 0.00 COSTS= :Agr.Pr.=40: +25 2 :Agr.Pr.=50 0.00 0.00 0.00 1.05 1.0 0.00 1.05 1.05 1.05 1.05 : 1.05 1.05 1.05 1.05 0.00 1.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 141 TABLE B-5: Optimal sizes of Dam D WATER REQUIREMENTS = -50 1 : MATER REQUIREMENTS x even DISCOUNT RATE: 81 WWATER REQUIREMENTS z +50 1 ;EI.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 EI.Pr.=20 E1.Pr.=30!E.Pr.=10 EI.Pr.=20 E1.Pr.=30 -------------- :----------------------------- :-----------------------------|:-- ------------------- ------- CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40! -25 1 :Agr.Pr.=50 0.21 0.21 0.21 1.23 1.22 1.22 1.23 1.22 1.22 0.00 0.00 0.21 0.00 1.23 1.23 1.23 1.23 1.23 ------- |:---------- :----------------------------- ----------------------------- CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40! even 1Agr.Pr.=50 0.00 0.21 0.21 0.00 0.21 0.21 1.23 1.23 1.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.23 1.23 ------- |:---------- :----------------------------- ------ ------------------- --- CONSTR.|Agr.Pr.=30 COSTS= 1Agr.Pr.=401 +25 1 1Agr.Pr.=50 0.00 0.00 0.21 0.00 0.00 0.21 0.00 1.23 1.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even DISCOUNT RATE: 10 0.00 0.00 0.00 0.00 0.00 1.23 0.00 0.00 1.23 --------------------------0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.23 ------------------------ ---0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 WATER REQUIREMENTS = +50 1 :EI.Pr.=10 E1.Pr.=20 E1.Pr.=301E1.Pr.=10 EI.Pr.=20 El.Pr.=30 E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.1Agr.Pr.=30t CDSTS= lAgr.Pr.=40: -25 1 1Agr.Pr.=50 0.00 0.21 0.21 0.00 0.21 0.21 1.23 1 1.23 1 1.22 1 0.00 0.00 0.00 0.00 0.00 0.00 1.23 1.23 1.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.23 CONSTR.:Agr.Pr.=30 COSTS= 1Aqr.Pr.=40! even :Agr.Pr.=50: 0.00 0.00 0.21 0.00 0.00 0.21 0.00 | 1.23 1.23 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 | 0.00 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 C -NSTR. -Agr.Pr.=300.00 COSTS= Agr.Pr.=401 0.00 +25 Z lAgr.Pr.=50: 0.00 WATER REQUIREMENTS = -50 2 | WATER REQUIREMENTS = even DISCOUNT RATE: 12 1 WATER REQUIREMENTS = +50 Z :EI.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 El.Pr.=20 E1.Pr.=30 CONSTR. -Agr.Pr.=300.00 COSTS: Agr.Pr.=40 0.21 0.21 -25 C Agr.Pr.=50 0.00 0.21 0.21 1.23 -- 0.00 1.23 0.00 1.23 1 0.00 0.00 0.00 0.00 0.00 0.00 1.23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. Agr.Pr.=30: COSTS: 1Agr.Pr.=40~ even Agr.Pr.=50 0.00 0.00 0.21 0.00 0.00 0.21 0.00 0.00 0.21 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. Agr.Pr.=30 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 COSTS: |Agr.Pr.=40 +25 |Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 , 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 14 2 TABLE B-6: Optimal sizes of Irrigation B I WATER REQUIREMENTS = -50 DISCOUNT RATE: 1 : WATER REQUIREMENTS = even | WATER REQUIREMENTS = +50 1 :EI.Pr.=10 E1.Pr.=20 EI.Pr.=30|E1.Pr.=10 E1.Pr.=20 E1.Pr.=30:E1.Pr.=10 E1.Pr.=20 E1.Pr.=30 16.75 16.75 27.88 16.55 27.88 27.88 16.55 27.88 27.88 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 CONSTR. Agr.Pr.=30 -- 16.55 16.75 COSTS= Agr.Pr.=40| 16.75 even :Agr.Pr.=50: 16.55 16.75 16.75 16.55 16.55 16.75 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.75 16.55 16.55 16.75 16.55 16.55 16.75 16.55 16.55 16.55 16.55 16.55 16.55 16.55 : 16.55 1 16.55 1 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 CONSTR. Agr.Pr.=30; COSTS= Agr.Pr.=40 -25 Z Agr.Pr.=50 CONSTR. Agr.Pr.=30: COSTS= Agr.Pr.=40: +25 1 :Agr.Pr.=50| WATER REQUIREMENTS = -50 1 IWATER REQUIREMENTS = even DISCOUNT RATE: ! WATER REQUIREMENTS = +50 1 ==================================== 10 1 |E1.Pr.=10 EI.Pr.=20 E1.Pr.=30:E1.Pr.=10 E1.Pr.=20 E1.Pr.=30:E1.Pr.=10 EI.Pr.=20 E1.Pr.=30 CONSTR|Ar.Pr.=30: 2OSTS= :Agr.Pr.=40: -25 i|Agr.Pr.=50 16.55 16.75 16.75 16.55 16.75 16.75 16.55 | 16.75 27.88 16.55 16.55 16.55 16.55 16.55 16.55 16.55 -- 16.55 16.55 1 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 O0MSTR.:Agr.Pr.=30 O0STS= |Agr.Pr.=:40 even |Agr.Pr.=50 16.55 16.55 16.75 16.55 16.55 16.75 16.55 1 16.55 16.75 16.55 16.55 16.55 16.55 16.55 16.55 16.55 1 16.55 | 16.55 | 16.55 16.55 16.55 16.55 16.55 16.55 0.00 16.55 16.55 -NSTR.Agr.Pr.=30 -- 16.55 16.55 16.55 +25 Z :Agr.Pr.=50: :-----:--------------12--- 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 | 16.55 | 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 OSTS= |Agr.Pr.=40 ISCOUNT RATE: ----------------------------------------- WATER REQUIREMENTS -50 I WATER REQUIREMENTS = even WATER REQUIREMENTS = +501 IE1.Pr.=10 E1.Pr.=20 E1.Pr.=30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30|E1.Pr.=10 E1.Pr.=20 EI.Pr.=30 ONSTR.1Agr.Pr.=30 COSTS= :Aqr.Pr.=40 -25 Z Agr.Pr.=50 16.55 16.75 16.75 16.55 16.55 16.75 16.55 16.55 16.75 16.55 16.55 16.55 16.55 16.55 16.55 16.55 : 16.55 | 16.55 | 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 ONSTR. IAgr.Pr.=30 Agr.Pr.:40W 16.55 16.55 16.55 16.55 16.55 1 16.55 16.55 16.55 16.55 16.55 16.55 1 16.55 16.55 16.55 16.55 16.55 16.55 16.55 Agr.Pr.=50 16.75 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 | 16.55 | 16.55 | 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 16.55 10.00 16.55 16.55 0.00 16.55 16.55 0.00 16.55 16.55 COSTS= even CONSTR. :Aqr.Pr.=30: COSTS= :Agr.Pr.=40: :Agr.Pr.=50: +25 143 TABLE Optimal B-7: sizes of Irrigation C T WATER REQUIREMENTS z -50 1 1 WATER REQUIREMENTS = even DISCOUNT RATE: 8 WATER REQUIREMENTS = 450 Z === EI.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 EI.Pr.=20 EI.Pr.=30 -- -- --- -- - CONSTR. Aqr.Pr.=30 COSTS= Agr.Pr.=40 -25 1 Agr.Pr.=50 - - -- - 19.66 19.66 19.66 - -- - - 0.00 0.00 0.00 - - - : - - - - -- -- - - -: - - - - - - - 0.00 0.00 0.00 - - - - - - ----- -- 18.47 18.47 19.66 - - - - - - - - - - - 18.47 0.00 0.00 - - - - - - .-- 0.00 0.00 0.00 - - - - -- -- - -- -- 18.47 18.47 18.47 - - - - - -- -- 0.00 18.47 18.47 - - --- -- -- - 0.00 0.00 0.00 --- -- -- CONSTR.:Agr.Pr.=30: COSTS= |Aqr.Pr.=40 even tAqr.Pr.=50 18.47 19.66 19.66 18.47 19.66 19.66 0.00 0.00 0.00 18.47 18.47 18.47 18.47 18.47 18.47 0.00 0.00 0.00 18.47 18.47 18.47 0.00 18.47 18.47 0.00 0.00 0.00 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40 +25 1 !Agr.Pr.=50 18.47 18.47 19.66 18.47 18.47 19.66 18.47 0.00 0.00 18.47 18.47 18.47 18.47 18.47 18.47 0.00 18.47 18.47 0.00 18.47 18.47 0.00 0.00 18.47 0.00 0.00 0.00 ------------------------------- ---------------- -- - - --- - - - - - - - - - - - - - - - - - - - - - ------- - - - - - - WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even DISCOUNT RATE: 10 1 - -- - - - -------- -- -- -- -- - | WATER REQUIREMENTS = +50 1 !EI.Pr.=10 EI.Pr.=20 EI.Pr.=30|E.Pr.=10 EI.Pr.=20 El.Pr.=30 E.Pr.=10 EI.Pr.=20 E1.Pr.=30 -- - -- - :I-- - - - - - - - - - - - - - - - - - - - - - - - - --- - - -- - -- -- -- ---- -- - CONSTR.:Aqr.Pr.=30: 18.47 18.47 0.00 COSTS= Agr.Pr.=40 19.66 19.66 0.00 -25 1 :Agr.Pr.=50 19.66 19.66 0.00 ---- :-----:-----------------------CONSTR.2Agr.Pr.=30 18.47 18.47 18.47 COSTS= Agr.Pr.=40 18.47 18.47 0.00 even |Agr.Pr.=50 19.66 19.66 0.00 18.47 18.47 18.47 18.47 18.47 18.47 0.00 0.00 0.00 18.47 18.47 18.47 0.00 18.47 18.47 0.00 0.00 0.00 18.47 18.47 18.47 18.47 18.47 18.47 0.00 18.47 18.47 0.00 18.47 18.47 0.00 0.00 18.47 0.00 0.00 0.00 CONSTR.:Agr.Pr.=30 COSTS= :Agr.Pr.=40: +25 1 :Agr.Pr.=50 18.47 18.47 18.47 0.00 18.47 18.47 0.00 18.47 18.47 0.00 0.00 18.47 0.00 0.00 18.47 0.00 0.00 0.00 DISCOUNT RATE: 12 1 18.47 18.47 18.47 18.47 18.47 18.47 18.47 18.47 18.47 WATER REQUIREMENTS = -50 Z WATER REQUIREMENTS = even | WATER REQUIREMENTS = +50 1 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 E1.Pr.=20 El.Pr.=30 E1.Pr.=10 El.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.=30 18.47 18.47 0.00 18.47 18.47 0.00 0.00 0.00 0.00 COSTS= Aqr.Pr.=40 19.66 19.66 0.00 18.47 18.47 18.47 18.47 18.47 0.00 -25 1 Agr.Pr.=50 19.66 19.66 0.00 18.47 18.47 0.00 18.47 18.47 18.47 ---------------------------------------------------- ----------------------------CONSTR. Aqr.Pr.=30 18.47 18.47 18.47 18.47 0.00 0.00 0.00 0.00 COSTS= Agr.Pr.=40 18.47 18.47 18.47 18.47 18.47 18.47 0.00 0.00 0.00 Agr.Pr.=50 19.66 19.66 19.66 18.47 18.47 18.47 18.47 18.47 0.00 CONSTR. Agr.Pr.=30 18.47 18.47 18.47 0.00 0.00 0.00 0.00 0.00 0.00 COSTS= Agr.Pr.=40: +25 1 Agr.Pr.=50 18.47 18.47 18.47 18.47 18.47 18.47 18.47 18.47 18.47 18.47 0.00 18.47 0.00 0.00 0.00 0.00 0.00 0.00 even 0.00 144 TABLE B-8: Optimal sizes WATER REQUIREMENTS = -50 DISCOUNT RATE: of Irrigation D WATER REQUIREMENTS = even WATER REQUIREMENTS = +50 Z W El.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 E1.Pr.=20 E1.Pr.=30 CONSTR. Aqr.Pr.=30 1.65 20.67 20.67 0.00 0.00 20.67 0.00 0.00 0.00 COSTS= |Agr.Pr.=40 -25 Z :Agr.Pr.=50 1.65 1.72 20.84 20.84 20.84 20.84 0.00 1.55 20.67 20.67 20.67 20.67 0.00 0.00 0.00 0.00 20.67 20.67 CONSTR.:Agr.Pr.=30 COSTS= |Aqr.Pr.=40 even :Aqr.Pr.=50 0.00 1.65 1.65 0.00 1.65 1.65 20.67 20.67 20.77 0.00 0.00 0.00 0.00 0.00 Q.00 0.00 20.67 20.67 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 20.67 CONSTR. Agr.Pr.=30COSTS= Agr.Pr.=40 +25 Z :Agr.Pr.=50 0.00 0.00 1.65 0.00 0.00 1.65 0.00 20.61 20.77 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 DISCOUNT RATE: 10 : WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even WATER W REQUIREMENTS = +50 Z :-================================================ :EI.Pr.=10 EI.Pr.=20 E1.Pr.=30 E1.Pr.=10 El.Pr.=20 EI.Pr.=30|E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.JAgr.Pr.=30 COSTS= tAgr.Pr.=40: -25 Z Agr.Pr.=50 0.00 1.65 1.65 0.00 1.65 1.65 20.67 20.77 20.84 CONSTR. Agr.Pr.=30- -COSTS= Agr.Pr.=40 even :Agr.Pr.=50: 0.00 0.00 1.65 0.00 0.00 1.65 0.00 20.67 20.77 -ONSTR. Agr.Pr.=30 COSTS :Agr.Pr.=40: +25 C :Agr.Pr.=501 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00: 0.00 0.00 20.67 : 20.67 t 20.67 | 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 WATER REQUIREMENTS = -50 tI WATER REQUIREMENTS = even DISCOUNT RATE: - 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 20.67 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 :WATER REQUIREMENTS = +50 Z == 12 I :El.Pr.=10 EI.Pr.=20 El.Pr.=30 E1.Pr.=10 El.Pr.=20 EI.Pr.=30:El.Pr.=10 E1.Pr.=20 EI.Pr.=30 -------------- ------------------------------ ----------------------------- ----------------------------- CONSTR.|Agr.Pr.=30: COSTS= |Agr.Pr.=40 -25 Z Agr.Pr.=50 0.00 1.65 1.65 0.00 1.55 1.65 20.67 20.67 20.77 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 20.67 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. Agr.Pr.=30 -COSTS= Agr.Pr.=40: even :Agr.Pr.=50: 0.00 0.00 1.65 0.00 0.00 1.55 0.00 0.00 1.55 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. :Agr.Pr.=30: COSTS= Agr.Pr.=40; 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 +25 1 Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 145 TABLE Optimal sizes B-9: DISCOUNT RATE: WATER REQUIREMENTS = -50 1 of hydropower Plant WATER W REQUIREMENTS = even A | WATER REQUIREMENTS = +50 Z E1.Pr.=10 EL.Pr.=20 El.Pr.=30:EI.Pr.=10 El.Pr.=20 EI.Pr.=30:EI.Pr.=10 EL.Pr.=20 EI.Pr.=30 CONSTR. -Ar.Pr.=-300.09 0.09 0.09 | 0.09 0.09 0.09 0.09 0.09 0.09 COSTS: Agr.Pr.=40 -25C Agr.Pr.=50 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 even Aqr.Pr.=50 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 CONSTR. Agr.Pr.=30 COSTS= Aqr.Pr.=40 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 +25 I Agr.Pr.=50 :WATER REQUIREMENTS = -50 Z | WATER REQUIREMENTS = even DISCOUNT RATE: 101 WATER REQUIREMENTS = +50 1 ::=========================================================================================... CEI.Pr.=0 E1.Pr.=20 E0.Pr.=30E0.Pr.=10 E0.Pr.=20 E0.Pr.=301.Pr.=10 E0.Pr.=20 El.Pr.=30 CO2STR. Agr.Pr.=30 COSTS=: Agr.Pr.=40-25 I :Agr.Pr.:50 W 0.09 0.09 0.09 0.09 0.09 0.09 0.09 | 0.09: 0.09 0.09 0.09 0.09 CONSTR. :Agr.Pr.=30: COSTS= :Agr.Pr.=40: even 1 '.Agr.Pr.=50: Agr.Pr.= 50 +25 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 : 0.09 0.09 I ---- ----- ----------------0 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: 0.09 0.09 DISCOUNT RATE: 0.09 0.09 0.09 0.09 0.09 0.09 : 0.09 0.09 : 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 | 0.09 0.09 0.09 0.09 0.09 0.09 |:I 0.09 0.09 0.09 0.09 0.09 0.09 ---------------- 0.09 0.09 WATER REQUIREMENTS = -50 0.09 0.09 0.09 0.09 WATER REQUIREMENTS ------------------ 0.09 : 0.09 even 0.09 0.09 0.09 0.09 0.09 0.09 WATER REQUIREMENTS = +50 - :EI.Pr.=10 EI.Pr.=20 El.Pr.=30 E1.Pr.=l0 E1.Pr.=20 EI.Pr.=30!E1.Pr.=10 EI.Pr.=20 E1.Pr.=30 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40: 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 IAgr.Pr.=50: 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 CONSTR.Agr.Pr.=30: COSTS :Agr.Pr.=40: even :Agr.Pr.=50: 0.09 0.09 0.09 0.09 0.09 0.09 0.09: 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 -- 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.40 +25 :Agr.Pr.=50: 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 : 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 -25 146 TABLE B-O: Optimal sizes of hydropower Plant C I WATER REQUIREMENTS = -50 Z DISCOUNT RATE: 8------- E-----------------------------: ' WATER REQUIREMENTS = +50 1 ----------------------------- ----------------------------0.00 0.00 0.00 0.00 0.06 0.06 0.06 0.06 0.06 0.00 0.00 0.00 0.06 0.00 0.00 0.06 0.06 0.06 0.06 0.06 0.06 0.00 0.00 0.00 0.00 0.06 0.06 0.00 0.00 0.06 0.00 0.00 0.00 0.06 0.00 0.00 0.06 0.06 0.06 0.00 0.06 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.06 0.06 0.06 0.06 0.00 0.00 0.06 CONSTR. :Agr.Pr.=30: COSTS= Agr.Pr.=40 -25 1 :Aqr.Pr.=50: 0.00 0.00 0.06 0.06 0.06 0.06 0.00 0.06 0.06 CONSTR. :Agr.Pr.=30: COSTS= !Agr.Pr.=40: even !Agr.Pr.=50: 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. :Agr.Pr.=30: COSTS= Agr.Pr.:40+25 1 Agr.Pr.=50! 0.00 0.00 0.00 0.00 0.00 0.00 DISCOUNT RATE: WATER REQUIREMENTS = even M t WATER REQUIREMENTS = -50 Z : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 :EI.Pr.=10 EI.Pr.=20 E.Pr.=30;E.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 ------ :---------------i----------------------CONSTR.:Agr.Pr.=30: 0.00 0.00 0.06 0.00 0.00 0.06 COSTS= Aqr.Pr.=40: 0.00 0.00 0.00 0.00 -25 1 :Agr.Pr.=50: 0.00 0.06 !----------------------------0.00 0.00 0.00 0.06 0.06 0.06 0.00 0.00 0.00 0.06 0.00 0.00 0.06 0.06 0.06 0.00 0.00 0.00 0.00 0.06 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.00 0.00 0.06 0.10 Agr.Pr.=50: 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.06 0.06 rCONSTR. :Agr.Pr.=30 COSTS= Agr.Pr.=40: +25 1 :Aqr.Pr.=50! 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.06 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.06 0.06 0.06 0.06 0.00 0.00 0.06 : WATER REQUIREMENTS = +50 1 CONSTR. :Agr.Pr.=30 COSTS= :Agr.Pr.=40: even WATER REQUIREMENTS = -50 1 DISCOUNT RATE: WATER REQUIREMENTS = even :E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 El.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 EI.Pr.=30 0.06 0.06 0.00 0.00 0.06 0.00 0.00 0.06 0.06 0.00 0.06 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.06 0.06 0.00 0.06 0.06 0.06 0.06 0.00 0.00 0.06 0.06 0.00 ' 0.00 0.00 0.00 0.06 0.06 0.06 0.10 0.06 0.06 CONSTR. :Agr.Pr.=30: 0.00 0.00 0.06 0.00 0.00 0.06 COSTS= :Agr.Pr.=40: 0.00 0.00 0.06 0.00 0.00 0.00 -25 Z :Agr.Pr.=50: 0.00 0.00 0.06 0.00 0.00 CONSTR. Aqr.Pr.=30 COSTS= !Agr.Pr.=40: even :Aqr.Pr.=50: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR. Agr.Pr.=30 COSTS: Agr.Pr.=40: +25 1 'Aqr.Pr.=50: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ' 0.00 0.00 0.00 1 47 T ABL E B-1 1 : DISCOUNT RATE: Optimal sizes of Transfer WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even | WATER REQUIREMENTS = +50 1 E1.Pr.=10 El.Pr.=20 EI.Pr.=30|E.Pr.=10 El.Pr.=20 El.Pr.=30:E.Pr.=I0 EI.Pr.=20 EI.Pr.:30 CONSTR. :Ar.Pr.=300.00 0.00 COSTS :Agr.Pr.=40: 0.00 0.00 -25 I |Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 CONSTR2 Agr.Pr.=30 COSTS= Agr.Pr.=40 even Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR.AAgr.Pr.=30 COSTS= Agr.Pr.=40 +25 1 :Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 - 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00: 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 - : : : 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ------- |----------- ----------------------------- ------------------------------------------------------------ DISCOUNT RATE: 10 | WATER REQUIREMENTS = -50 %| WATER REQUIREMENTS = even WATER REQUIREMENTS = +50 1 W |= :El.Pr.=10 El.Pr.=20 EI.Pr.=30 E.Pr.=I0 El.Pr.=20 EI.Pr.=30:E.Pr.=10 El.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.:30 COSTS :Agr.Pr.=40: -25C Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 -- 0.00 0.00 0.00 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CONSTR Agr.Pr.=30 COSTS=: Agr.Pr.=40even Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 6.00 0.00 0.00 CONSTR. Agr.Pr.=30 OSTS= Agr.Pr.=40 +25 1 Agr.Pr.=50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00 | 0.00 : 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 DISCOUNT WATER REQUIREMENTS =-50 1I WATER REQUIREMENTS = even iWATER REQUIREMENTS = +50 1 RATE: :E1.Pr.=l0 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 E1.Pr.=20 EI.Pr.=30!E1.Pr.=10 El.Pr.=20 E1.Pr.=30 ONSTR. !Agr.Pr.=30 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 OSTS= :Aqr.Pr.=40 0.00 0.00 0.00 0.00 0.00 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00: 0.00 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 : 0.00 : 0.00: 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 : : 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 8.00 0.00 0.00 0.00 0.00 : 0.00 .00 0.00 0.00 0.00 0.00 0.00 -2 :Agr.Pr.=50: ONSTR.:Agr.Pr.=30: OSTS= Agr.Pr.=40 even Agr.Pr.=50 CONSTR. :Agr.Pr.=:30 COSTS= :Agr.Pr.=40: +25 Agr.Pr.=50 : 148 70%~ -.. enx 5D% D 401 301 20% 1Xx 20 40 no 100 d \CKLUME I I I I I I I I I I I I 0 120 ~1 140 m3 Bar graph of the probability distribution of the sizes of Dam A FIGURE B-1: tont -J E0t - 1C I 5 20 ; I II on I II S0 I I II 1np II iI i2n I I t 14n '--U ME Hm 3 FIGURE B-2: Bar graph of the probability distribution of the sizes of Dam B 149 a901 -T 5ax - sax IL 20% a% 0 20 40o I Do so I I 101 I I 120 I v 140 '.OL1ME Hm FIGURE B-3: 100% Bar graph of distribution the probability of the sizes of Dam C - all- IT jdlt 20t rat -a I Ii 21 40 I I I on in0 1211 14') '.COi3ME rAm3 FIGURE B-4: Bar graph of the probabili ty distribution of the sizes of Dam D 150 00% 4- E- - 50X - 40%2a% - 413% - 20%3m - 1 ,4, 0 FYGURE F-5: 1001 II """ a 12 1. (Thumamd} AREA Na 20 24 25 Bar graph of the probability distribution of the sizes of Irrigation B - 01- T 4 r r e -y ___________ --- -T 12 -r11 *(Thounmdv'e FTGUwR F-6: lID liii -- ---a1) 24 28 ME Ho Bar graph of the probability distribution of the sizes of Irrigation C 151 aos 50% E- 40% 20x 10% Wx I ." 0 +4 12 . 1 (Thaummdu} 20 24 25 MU!L.4Na FYGUWE B-T: Bar graph of the probability distribution of the sizes of Irrigation D 1nOT- sa( 803 - Jolt +4 FIGURE U-6: 10I 12 14 Bar graph of the probability distribution of the sizes of hydropower Plant A 152 5D' CL LL 4B.1 70% 0% am. 10% 0 a 4 a 10 12 14 [CZP.ZCW MAI FYGURE V-9: Bar graph of the probability distribution of the sizes of hydropower Plant C E I.J a 10% - i 2 4 I i I I 8 a I - T 1? -I i 14 ':!ZE m.3/s FIGURE B-O: Bar graph of the probability distribution of the sizes of Transfer 153 APPENDIX C t54 TABLE C-i: Net benefits of Alternative WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even DISCOUNT RATE: A : WATER REQUIREMENTS = +50 1 :E1.Pr.=10 El.Pr.=20 El.Pr.=30|El.Pr.=10 El.Pr.=20 El.Pr.=30:El.Pr.=10 El.Pr.=20 EI.Pr.=30 0.21 0.21 0.21 0.51 0.81 : 0.51 0.51 0.81 0.81 0.21 0.21 0.21 0.51 0.51 0.51 0.81 0.81 0.81 -- 0.21 0.21 0.21 0.51 0.51 0.51 0.81 0.81 0.81 0.18 CONSTR. :Agr.Pr.=30: COSTS= |Agr.Pr.40 -- 0.18 0.18 even |Agr.Pr.=50: 0.48 0.48 0.48 0.78 0.78 0.78 0.18 0.18 0.18 0.48 0.48 0.48 0.18 0.78 1 0.78 -- 0.18 0.18 0.78 1 0.48 0.48 0.48 0.78 0.78 0.78 0.15 0.15 0.15 0.45 0.45 0.45 0.75 0.75 0.75 0.15 0.15 0.15 0.45 0.45 0.45 0.75 0.75 0.75 0.15 0.15 0.15 0.45 0.45 0.45 0.75 0.75 0.75 CONSTR.-Ar.Pr.=30 COSTS= Agr.Pr.=40 -25C :Agr.Pr.=50 CONSTR.|Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 1 |Agr.Pr.=50! 1 REQUIREMENTS = even WATER REQUIREMENTS = -50 1 : WATER DISCOUNT RATE: 10 1 WATER REQUIREMENTS = +50 1 W :El.Pr.=10 El.Pr.=20 EI.Pr.=30|E1.Pr.=10 E1.Pr.=20 E1.Pr.=30:El.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.Agr.Pr.=30: COSTS= Agr.Pr.=40! -25 C |Agr.Pr.=50: 0.19 0.19 0.19 0.49 0.49 0.49 0.79 0.79 0.79 0.19 0.19 0.19 0.49 0.49 0.49 0.79 0.79 0.79 0.19 0.19 0.19 0.49 0.49 0.49 0.79 0.79 0.79 CONSTR.|Agr.Pr.=30: 0.15 0.45 0.75 0.15 0.45 0.75 0.15 0.45 0.75 COSTS: -Agr.Pr.=40--- 0.15 0.15 even Agr.Pr.=50: 0.45 0.45 0.75 0.75 0.15 0.15 0.45 0.45 0.75 0.75 0.15 0.15 0.45 0.45 0.75 0.75 0.12 0.12 0.12 0.41 0.41 0.41 0.71 ! 0.71 0.71 0.12 0.12 0.12 0.41 0.41 0.41 0.71 0.71 0.71 0.12 0.12 0.12 0.41 0.41 0.41 0.71 0.71 0.71 CONSTR.:Agr.Pr.=30; COSTS= :Agr.Pr.=40| +25 1 !Agr.Pr.=50: REQUIREMENTS = even WATER REQUIREMENTS = -50 1 : WATER DISCOUNT RATE: -- : WATER REQUIREMENTS = +50 1 EI.Pr.=10 EI.Pr.=20 E1.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:EI.Pr.=10 E1.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.=301 COSTS= Agr.Pr.=40| -25 X Agr.Pr.=50| 0.!7 0.17 0.17 0.47 0.47 0.47 0.76 0.76 0.76 0.17 0.17 0.17 0.47 0.47 0.47 0.76 0.76 0.76 : ONSTR. Agr.Pr.=30| COSTS: Agr.Pr.=40 0.12 0.12 0.42 0.42 0.72 0.72 0.12 0.12 0.42 0.42 0.72 0.72 |0.12 Agr.Pr.=50: 0.12 0.42 0.72 0.12 0.42 0.72 | even - 0.08 0.02 0.08 0.38 0.38 0.38 0.68 0.68 0.68 0.08 0.08 0.08 0.47 0.47 0.47 0.76 0.76 0.76 0.12 0.42 0.42 0.72 0.72 0.12 0.42 0.72 ----------------------------- -------- -------------------------- CONSTR. Agr.Pr.=30t COSTS= Ar.Pr.=40: +25 1 Agr.Pr.=50: 0.17 0.17 0.17 0.38 0.38 0.38 0.68 0.68 0.68 0.08 0.08 0.08 0.38 0.38 0.30 0.68 0.68 0.68 155 TABLE C-2: DISCOUNT RATE: Net benefits of Alternative WATER REQUIREMENTS = -50 1 WATER W REQUIREMENTS = even B WATER REQUIREMENTS = 450 1 EI.Pr.=10 EI.Pr.=20 El.Pr.=30:El.Pr.=10 El.Pr.=20 El.Pr.=30:EI.Pr.=10 El.Pr.=20 El.Pr.=30 I ---------------------------------------------------------------CONSTR. Agr.Pr.=30 1.^1 2.21 2.51 0.92 1.22 1.52 0.59 0.89 1.19 COSTS= :Agr.Pr.=40: 2.57 2.87 3.17 1.25 1.55 1.85 0.81 1.11 1.41 -25 1 lAgr.Pr.=50; 3.24 3.53 3.83 1.58 1.88 2.18 1.03 1.33 1.63 CONSTR.:Agr.Pr.=30: COSTS= |Agr.Pr.=40 even |Agr.Pr.=50 1.79 2.45 3.11 2.09 2.75 3.41 2.39 3.05 3.71 ! 0.80 1.13 1.46 1.10 1.43 1.76 1.39 1.72 2.06 0.47 0.69 0.91 0.76 0.99 1.21 1.06 1.28 1.50 CONSTR. Agr.Pr.=30 COSTS= |Agr.Pr.=40 1.67 2.33 1.96 2.63 2.26 2.92 0.67 1.00 0.97 1.30 1.27 1.60 0.34 0.56 0.64 0.86 0.94 1.16 2.99 3.29 3.59 1.34 1.63 1.93 0.78 1.08 1.38 +251 Agr.Pr.=50! --------------------------------------------------------------------------------------DISCOUNT RATE: WATER REQUIREMENTS = -50 1 | WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 10 1 E1.Pr.=10 El.Pr.=20 EI.Pr.=30|E1.Pr.=10 EI.Pr.=20 El.Pr.=30:E.Pr.=10 EI.Pr.=20 E1.Pr.=30 CONSTR.2Agr.Pr.=30! COSTS= Agr.Pr.=40: -25 1 Agr.Pr.=50: 1.83 2.49 3.15 2.12 2.79 3.45 2.42 3.08 3.75 0.83 1.16 1.49 1.13 1.46 1.79 1.43 1.76 2.09 0.50 0.72 0.94 0.80 1.02 1.24 1.10 1.32 1.54 CONSTR.2Agr.Pr.=30| COSTS= |Agr.Pr.=40: even |Agr.Pr.=50! 1.67 2.33 3.00 1.97 2.63 3.29 2.27 2.93 3.59 0.68 1.01 1.34 0.98 1.31 1.64 1.28 1.61 1.94 0.35 0.57 0.79 0.65 0.87 1.09 0.95 1.17 1.39 CONSTR.2Agr.Pr.=30| COSTS= :Agr.Pr.=40: 1.52 2.18 1.82 2.48 2.12 2.78 0.53 0.86 0.83 1.16 1.13 1.46 0.20 0.42 0.50 0.72 0.79 1.01 +25 1 Agr.Pr.=50| 2.84 3.14 3.44 1.19 1.49 1.79 0.64 0.94 1.24 : -------------------- ----------------------------- ----------------------------DISCOUNT RATE: WATER REQUIREMENTS = -50 i WATER REQJIREMENTS = even WATER REQUIREMENTS = t50 1 12 1 El.Pr.=10 El.Pr.=20 El.Pr.=30 E .Pr.=10 El.Pr.=20 E1.Pr.=30;EI.Fr.=10 EI.Pr.=210 Ei.Pr.z 30 CONSTR. Agr.Pr.=301 COSTS= Agr.Pr.=40 -25 1 Agr.Pr.=50 1.74 2.40 3.06 2.04 2.70 3.36 2.34 3.00 3.66 0.75 1.08 1.41 1.05 1.38 1.71 1.34 1.67 2.0 0.42 0.64 0.86 0.71 0.94 1.16 1.01 1.23 1.45 CONSTR.Agr.Pr.=30 COSTS= lAgr.Pr.=40 even Agr.Pr.=50: 1.56 2.22 2.88 1.86 2.52 3.18 2.16 2.82 3.48 0.57 0.90 1.23 0.86 1.20 1.53 1.16 1.49 1.82 0.24 0.46 0.68 0.53 0.75 0.97 0.83 1.05 1.27 CONSTR. :Agr.Pr.=30: COSTS= Agr.Pr.=40 1.38 2.04 1.68 2.34 1.97 2.64 0.38 0.72 0.68 1.01 0.8 1.31 0.05 0.27 0.35 0.57 0.65 0.87 2.70 3.00 3.30 1.05 1.34 1.64 0.50 0.79 1.09 +25 1 Agr.Pr.=50 156 C-3: TABLE DISCOUNT RATE: Net benefits of Alternative WATER REQUIREMENTS = -50 1 C WATER REQUIREMENTS x even ' WATER REQUIREMENTS z +50 Z :El.Pr.=10 E1.Pr.=20 El.Pr.=30:E1.Pr.=10 E1.Pr.=20 EI.Pr.=30:E1.Pr.x10 EI.Pr.=20 EI.Pr.z30 CONSTR. Agr.Pr.=30: COSTS= |Agr.Pr.=40! -25 1 :Agr.Pr.=50: 1.91 2.57 3.24 2.21 2.87 3.53 2.51 3.17 3.83 CONSTR. Agr.Pr.=30 COSTS= lAgr.Pr.=40: even |Agr.Pr.=50: 1.79 2.45 3.11 2.09 2.75 3.41 2.39 I 3.05 1 3.71 1 CONSTR.|Aqr.Pr.=30: COSTS= |Agr.Pr.=40: +25 1 |Agr.Pr.=50: 1.67 2.33 2.99 1.96 2.63 3.29 2.26 1 2.92 I 3.59 1 :---|------- :----------------------- ---- 1.22 1.55 1.88 1.52 1.85 2.18 0.80 1.13 1.46 1.10 1.43 1.76 0.67 1.00 1.34 0.97 1.30 1.63 0.59 4.81 1.03 0.89 1.11 1.33 1.19 1.41 1.63 1.39 1.72 2.06 0.47 0.69 0.91 0.76 0.99 1.21 1.06 1.28 1.50 1.27 1.60 1.93 ; 0.34 0.56 0.78 0.64 0.86 1.08 0.94 1.16 1.38 11 WATER REQUIREMENTS = -50 Z : WATER REQUIREMENTS = even Z = = = = = ------------------ --------------------- - ---------------------- ---- 01 DISCOUNT RATE: 10 0.92 1.25 1.58 = = = = = 1 WATER REQUIREMENTS = +50 Z = = = = = = = IEI.Pr.z10 EI.Pr.=20 EI.Pr.z30!E.Pr.z10 EI.Pr.=20 EI.Pr.=301E1.Pr.=10 E.Pr.=20 E.Pr.=30 ------------------------- ------:a----------------------------------- CONSTR.Agr.Pr.=30 COSTS= ;Agr.Pr.=40 -25 1 |Agr.Pr.=50 ---- ------ 1.83 2.49 3.15 2.12 2.79 3.45 2.42 | 3.08 3.75 | 0.83 1.16 1.49 2.27 1 2.93 3.59 0.68 1.01 1.34 0.98 1.31 1.64 0.53 0.86 1.19 0.83 1.16 1.49 !---------------------a CONSTR.1Agr.Pr.=30 COSTS= !Agr.Pr.=40 even IAgr.Pr.=50: ---- ------ 1.67 2.33 3.00 :---------------- CONSTR.!Aqr.Pr.=30 COSTS= :Agr.Pr.=40: +25 1 :Agr.Pr.=50: 1.52 2.18 2.84 1.13 1.46 1.79 ------ 1.97 2.63 3.29 - 2.12 2.78 3.44 12 1 0.50 0.72 0.94 0.80 1.02 1.24 1.10 1.32 1.54 0.35 0.57 0.79 0.65 0.87 1.09 0.95 1.17 1.39 S 1.28 1.61 1.94 --- : ----- 1.13 1.46 | 1.79 1 -------- 0.20 0.42 0.64 -------- 0.50 0.72 0.94 0.79 1.01 1.24 ---- -------------------- --------------------------------- DISCOUNT RATE: --- ------------------ 1.82 2.48 3.14 1.43 1.76 2.09 WATER REQUIREMENTS = -50 1 WATER REQUIREMENTS = even WATER REQUIREMENTS = +50 1 ========================================================================================= E.Pr.=10 EI.Pr.=20 EI.Pr.:30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= .Agr.Pr.=40 -25 Z |Agr.Pr.=50 1.74 2.40 3.06 2.04 2.70 3.36 2.34 3.00 3.66 0.75 1.08 1.41 1.05 1.38 1.71 1.34 1.67 2.01 0.42 0.64 0.86 0.71 0.94 1.16 1.01 1.23 1.45 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 1.56 2.22 1.86 2.52 2.16 2.82 0.57 0.90 0.86 1.20 1.16 1.49 0.24 0.46 0.53 0.75 0.83 1.05 2.88 3.18 3.48 1.23 1.53 1.82 0.68 0.97 1.27 even Agr.Pr.=50 CONSTR. :Aqr.Pr.=301 1.38 1.68 1.97 0.38 0.68 0.98 0.05 0.35 0.65 COSTS= :Agr.Pr.=40; 2.04 2.34 2.64 0.72 1.01 1.31 0.27 0.57 0.87 +25 1 Agr.Pr.=50 2.70 3.00 3.30 1.05 1.34 1.64 0.50 0.79 1.09 157 WATER REQUIREMENTS a -50 1 DISCGUNT RATE: D of Alternative Net benefits TABLE C-4:, WATER REQUIREMENTS x even WATER REQUIREMENTS = +50 1 W :EI.Pr.=10 El.Pr.=20 EI.Pr.=30!E1.Pr.=10 EI.Pr.=20 El.Pr.=30:E1.Pr.=10 El.Pr.220 E1.Pr.=30 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 -25 Z Agr.Pr.=50 -- 1.89 2.55 3.21 1- -- - - COKSTR.JAgr.Pr.=30 COSTS= Agr.Pr.=40 even Agr.Pr.=50 -----; - - - - -:- - - - - DISCOUNT RATE: 10 - - - - - - - - - 2.48 3.14 3.80 --- - 2.05 2.71 3.38 - - - 1.63 2.29 2.95 -- - --- 1.76 2.42 3.08 COMSTR. :Agr.Pr.=30: COSTS= Agr.Pr.=40 +25 1 Agr.Pr.=50 -- - - - - - - 2.18 2.85 3.51 - -- -- - - - - - - - --- - - --- - - - --- - - - ------ - 1.48 1.81 2.15 --- - 1.06 1.39 1.72 - - - 0.63 0.96 1.29 - - 1.19 1.52 1.85 0.76 1.09 1.43 2.21 2.88 3.54 - -- 2.35 3.01 3.67 1.92 2.58 3.25 - - 0.89 1.23 1.56 - - - - - - - - -- == = == === = = - - - -- : - - - -- - ---- - ----- - - - 1.15 1.37 1.59 -- -- 0.73 0.95 1.17 0.30 0.52 0.74 -- - - -- - 0.86 1.08 1.30 0.43 0.65 0.87 : : - WATER REQUIREMENTS = -50 I : WATER REQUIREMENTS = even 1 - 1.22 1.55 1.88 ------- ---- : - - 1.35 1.68 2.01 - 0.93 1.26 1.59 - 0.56 0.78 1.00 -. . . 0.59 0.82 1.04 - - - - - --- -.-- 0.89 1.11 1.33 - - - 1.02 1.24 1.46 -- - -- : WATER REQUIREMENTS = +50 1 === = = = IE.Pr.zlO EI.Pr.=20 EI.Pr.z30 E1.Pr.=10 EI.Pr.=20 E.Pr.=30!E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 ------ ---- !--------------- CONSTR.JAqr.Pr.=30 COSTS= ;Aer.Pr.=40: -25 1 Agr.Pr.=50 1.80 2.46 3.12 2.09 2.75 3.42 - -------------- 2.38 3.05 3.71 --------------- 0.80 1.13 1.46 1.10 1.43 1.76 1.39 1.72 2.05 0.47 0.69 0.91 0.76 0.98 1.21 1.06 1.28 1.50 ------:-----:-------------------------------------------.------- CONSTR.!Agr.Pr.=30: COSTS= :Aqr.Pr.=40: even Agr.Pr.=50: 1.63 2.30 2.96 1.93 2.59 3.25 2.22 2.88 3.55 0.64 0.97 1.30 0.93 1.26 1.60 1.23 1.56 1.89 0.31 0.53 0.75 0.60 0.82 1.04 0.9" 1.12 1.34 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 1 :Agr.Pr.=50: 1.47 2.13 2.80 1.76 2.43 3.09 2.06 2.72 3.38 0.48 0.81 1.14 0.77 1.10 1.43 1.06 1.40 1.73 0.15 0.37 0.59 0.44 0.66 0.88 0.73 0.95 1.18 WATER REQUIREMENTS = -50 1 WATER REQUIREMENTS = even DISCOUNT RATE: 12 1 : WATER REQUIREMENTS = +50 Z :EI.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 El.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.=30 COSTS= !Agr.Pr.=40 -25 1 Agr.Pr.=50 1.70 2.37 3.03 CONSTR. :Agr.Pr.=30: COSTS= :Agr.Pr.=40: even tAgr.Pr.=50: CONSTR. Agr.Pr.=30 1.32 COSTS= :Agr.Pr.=40: 1.98 +25 1 Agr.Pr.=50 2.64 2.94 2.00 2.66 3.32 2.29 2.95 3.62 1.51 1.80 2.17 2.84 2.47 3.13 0.71 1.04 1.37 1.00 1.34 1.67 1.30 1.63 1.96 0.38 0.60 0.82 0.67 0.89 1.11 0.97 1.19 1.41 2.10 0.52 0.81 1.10 0.19 0.48 0.77 2.76 3.42 0.85 1.18 1.14 1.47 1.44 1.77 0.41 0.63 0.70 0.92 0.99 1.22 1.61 1.91 0.32 0.62 0.91 -0.01 0.29 0.58 2.27 2.57 0.66 0.95 1.24 0.21 0.51 0.80 3.23 0.99 1.28 1.57 0.43 0.73 1.02 158 WATER REQUIREMENTS = -50 2 DISCOUNT RATE: E of Alternative Net benefits C-5: TABLE ATER REQUIREMENTS= even WATER REQUIREMENTS z +50 1 W :El.Pr.=10 EI.Pr.=20 EI.Pr.=30 El.Pr.=10 El.Pr.=20 EI.Pr.=30 E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.=30 COSTS= :Agr.Pr.=40 -25 2 Agr.Pr.=50 1.98 2.64 3.30 2.46 3.13 3.79 2.95 3.61 4.27 0.99 1.32 1.65 1.47 1.80 2.13 1.96 2.29 2.62 0.66 0.88 1.10 1.14 1.36 1.58 1.63 1.85 2.07 CONSTR.!Agr.Pr.:30! COSTS= Agr.Pr.=40 even :Agr.Pr.=50: 1.82 2.48 3.14 2.30 2.96 3.62 2.79 3.45 4.11 0.82 1.15 1.49 1.31 1.64 1.97 1.79 2.12 2.46 0.49 0.71 0.93 0.98 1.20 1.42 1.46 1.68 1.90 -- - - !- - - - - --- - CONSTR.JAgr.Pr.=30: COSTS= lAgr.Pr.=40 +25 1 Agr.Pr.=50 --- --- 1.65 2.31 2.98 -- -- 2.14 2.80 3.46 --------------- --- ----- 2.62 3.28 3.95 -- - 0.66 0.99 1.32 --- - - 1.15 1.48 1.81 - ----------- --- 1.63 1.96 2.29 --------------------------------------- ------ 0.81 1.04 1.26 ------- 1.30 1.52 1.74 -------- WATER REQUIREMENTS = -50 1 i WATER REQUIREMENTS = even DISCOMNT RATE: - 0.33 0.55 0.77 WATER REQUIREMENTS = +50 1 W :EJ.Pr.=10 El.Pr.=20 El.Pr.=30:El.Pr.=10 E1.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.2Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 %|Agr.Pr.=50 1.86 2.53 3.19 2.35 3.01 3.67 ------ :-------- CONSTR.IAqr.Pr.z301 COSTS= :Agr.Pr.=40 even :Agr.Pr.=50: ---- ----- -- - - : -- - 2.83 1 3.50 1 4.16 -- ---- 1.66 2.32 2.98 - DISCOUNT RATE: 12 - - 1.46 2.12 2.78 - - - - - - - - 0.87 1.20 1.53 1.36 1.69 2.02 1.84 2.17 2.50 1 1.15 1.49 1.82 1.64 1.97 1 2.30 | -------------------- 2.15 2.81 3.47 2.63 1 3.29 3.95 : 1.94 2.61 3.27 2.43 3.09 3.75 - - - - - - 0.47 0.80 1.13 - - - - - - 1.03 1.25 1.47 1.51 1.73 1.95 0.34 0.56 0.78 0.82 1.04 1.26 1.31 1.53 1.75 0.14 0.36 0.58 0.62 0.84 1.06 1.11 1.33 1.55 - - ------------- - ------ 0.95 1.28 1.61 -- --------- 0.54 0.76 0.98 1------- 0.67 1.00 1.33 ------------------------------------ -------------- CONSTR.:Agr.Pr.=30! COSTS= IAgr.Pr.=40 +25 2 !Agr.Pr.=50 - ------------------------------ ---------------- - 1.44 1.77 2.10 - - - - - - - - - - - - - - - - - - - WATER REQUIREMENTS = -50 2 WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 Z = :E1.Pr.=10 EI.Pr.=20 E1.Pr.=30!El.Pr.=10 EI.Pr.=20 EI.Pr.=30 E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.tAgr.Pr.=30 COSTS= |Agr.Pr.=40 -25 X |Agr.Pr.=50: 1.75 2.41 3.07 2.23 2.90 3.56 2.72 3.38 1 4.04 0.76 1.09 1.42 1.24 1.57 1.90 1.73 2.06 2.39 0.43 0.65 0.87 0.91 1.13 1.35 1.40 1.62 1.84 CONSTR.:Agr.Pr.=30: COSTS= |Agr.Pr.=40, even Agr.Pr.=50 1.51 2.17 2.83 1.99 2.66 3.32 2.48 3.14 3.80 0.52 0.85 1.18 1.00 1.33 1.66 1.49 1.82 2.15 0.19 0.41 0.63 0.67 0.89 1.16 1.38 1.11 1.60 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 +25 1 Agr.Pr.=50: 1.27 1.93 2.59 1.75 2.42 3.08 2.24 2.90 3.56 0.28 0.61 0.94 0.76 1.09 1.42 1.25 1.58 1.91 -0.05 0.17 0.39 0.43 0.65 0.87 0.92 1.14 1.36 159 TABLE Net benefits C-6: DISCOUNT RATE: WATER REQUIREMENTS = -50 1 of Alternative WATER REQUIREMENTS - F even WATER REQUIREMENTS = +50 1 W El.Pr.=10 EI.Pr.=20 EI.Pr.=30 E1.Pr.=10 E1.Pr.=20 El.Pr.=30:El.Pr.=10 EI.Pr.:20 EI.Pr.:30 CONSTR. Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 1 :Agr.Pr.=50: -----:----- ----- 3.89 5.30 6.70 4.19 5.59 6.99 1.50 2.20 2.90 1.79 2.49 3.19 2.09 2.79 3.49 ------------------------------- CONSTR.:Agr.Pr.z30! COSTS= |Agr.Pr.=40! even :Agr.Pr.=50: ---- : 3.60 5.00 6.40 3.30 4.70 6.10 3.59 4.99 6.39 0.80 1.26 1.73 : -- :------- 3.89 5.29 1 6.69 1 1.19 1.89 2.60 - 1.09 1.56 2.03 -------- --- 1.49 2.19 2.89 1.79 1 2.49 1 3.19 0.49 0.96 1.43 0.79 1.26 1.73 1.19 1.89 2.59 1.49 1 2.19 I 2.89: 0.19 0.66 1.13 0.49 0.96 1.43 :------------------------------------------------- CONSTR.:Agr.Pr.=30| COSTS= lAgr.Pr.=40: +25 1 :Agr.Pr.=50| 2.99 4.39 5.79 3.59 | 4.99 1 6.39 ' 3.29 4.69 6.09 ----------- -------------- 0.89 1.59 2.29 1.39 1.86 2.33 1.09 1.56 2.03 - -- |------------------ - 0.79 1.26 1.72 ---------------- ---------------|----------------------------------------------------------------------- -- WATER REQUIREMENTS = -50 1 DISCOUNT RATE: WATER W REQUIREMENTS = even WATER REQUIREMENTS = +50 1 W IE1.Pr.=10 E1.Pr.=20 EI.Pr.z301E1.Pr.=10 El.Pr.=20 E1.Pr.=301El.Pr.z10 E1.Pr.=20 EI.Pr.=30 ------ : ----------------------------- CONSTR.JAgr.Pr.=30: COSTS= |Agr.Pr.=40: -25 1 :Agr.Pr.=50: ----i -- -- 3.38 4.78 6.18 -- :---- 3.68 5.08 6.48 ---- ---------------------: 3.98 5.38 6.78 1.28 1.98 2.68 1.58 2.28 2.98 1.88 2.58 1 3.28 1 ---------------------- ----- 0.88 1.35 1.81 1.18 1.65 2.11 0.21 0.68 1.14 0.51 0.98 1.44 0.81 1.27 1.74 -0.16 0.30 0.77 0.14 0.60 1.07 0.43 0.90 1.37 ---------------- COSTR.!Agr.Pr.z30: COSTS= !Agr.Pr.=40: even :Agr.Pr.=50: 3.01 4.41 5.81 3.31 4.71 6.11 3.61 5.01 1 6.41 1 0.91 1.61 2.31 1.21 1.91 2.61 1.51 1 2.21 2.91 1 CONSTR.1Agr.Pr.=30: COSTS= :Agr.Pr.=401 +25 1 1Agr.Pr.=501 2.64 4.04 5.44 2.94 4.34 5.74 3.24 1 4.64 6.04 0.54 1.24 1.94 0.84 1.54 2.24 1.13 1.84 2.54 DISCOUNT RATE: 12 1 0.58 1.05 1.52 WATER REQUIREMENTS = -50 Z : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 1E.Pr.=10 EI.Pr.=20 EI.Pr.=30:El.Pr.=10 EI.Pr.=20 EI.Pr.=30!EI.Pr.=10 EI.Pr.=20 E1.Pr.=30 -------------- |:--------------------------- ----------------------------- ----------------------------- CONSTR. Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 1 |Agr.Pr.=50: 3.17 4.57 5.97 3.47 4.87 6.27 3.77 5.17 6.57 1.07 1.77 2.47 1.37 2.07 2.77 1.67 2.37 3.07 0.37 0.84 1.31 0.67 1.14 1.60 0.97 1.44 1.90 ONSTR. Agr.Pr.=30 COSTS= :Agr.Pr.=40! even |Agr.Pr.=50! 2.73 4.13 5.53 3.03 4.43 5.83 3.33 4.73 6.13 0.63 1.33 2.03 0.93 1.63 2.33 1.23 1.93 2.63 -0.07 0.40 0.86 0.23 0.70 1.16 0.53 0.99 1.46 CONSTR.|Agr.Pr.=30: 2.29 2.59 2.89 0.19 0.49 0.78 -0.51 -0.21 0.08 COSTS= Agr.Pr.=40: 3.69 3.99 4.29 0.89 1.19 1.49 -0.05 0.25 0.55 +25 1 Agr.Pr.=50: 5.09 5.39 5.69 1.59 1.89 2.19 0.42 0.72 1.02 160 TABLE C-7: of Alternative WATER REQUIREMENTS = -50 1 1 WATER REQUIREMENTS = even DISCOUNT RATE: 8 Net benefits I G . WATER REQUIREMENTS = +50 1 |========================================================--=- E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 El.Pr.=30 E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.!Agr.Pr.=30 COSTS= :Agr.Pr.=40: -25 1 :Agr.Pr.z50: 3.45 4.94 6.43 ------- |---------- 3.75 5.24 6.73 4.05 5.54 : 7.03 ------------------------------- CONSTR.JAgr.Pr.=30 COSTS= !Agr.Pr.=40 even lAgr.Pr.=50- 3.01 4.50 5.99 3.31 4.80 6.29 ~~~~ 2.58 4.07 5.55 2.88 4.36 5.85 1.52 2.26 3.01 1.82 2.56 3.30 0.47 0.97 1.47 ------------------------- 3.61 | 5.10 6.59 1 --------------------: CONSTR.|Agr.Pr.=30: 'COSTS= |Agr.Pr.=40 +25 1 |Agr.Pr.=50 1.22 1.96 2.71 0.78 1.53 2.27 1.08 1.82 2.57 0.34 1.09 1.83 0.04 0.53 1.03 -- 0.64 1.39 2.13 1.07 1.57 2.06 0.34 0.83 1.33 0.63 1.13 1.63 -0.10 0.39 0.89 0.20 0.69 1.19 ---------------------------- 1.38 2.12 2.87 ---------- 3.17 | 4.66 6.15 1 0.77 1.27 1.77 0.94 1.69.: 2.43 ----------------- -0.40 0.10 0.59 DISCOUNT RATE: WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 10%-------------1 EI.Pr.=10 EI.Pr.=20 EI.Pr.=301E1.Pr.=10 EI.Pr.=20 EI.Pr.=301E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 ------ : -------------- ----------------------- -------------- CONSTR.Agr.Pr.=30: COSTS= :Agr.Pr.=40! -25 1 IAgr.Pr.=50: 3.14 4.63 6.12 3.44 4.93 6.42 3.74 5.23 6.72 0.91 1.65 2.40 1.21 1.95 2.70 1.51 1 2.25 1 2.99 0.16 0.66 1.16 0.46 0.96 1.46 0.76 1.26 1.75 CONSTR.Agr.Pr.=30: COSTS= IAgr.Pr.=40 even :Agr.Pr.=50: 2.60 4.09 5.58 2.90 4.39 5.88 3.20 4.69 | 6.18 1 0.37 1.11 1.86 0.67 1.41 2.16 0.96 1 1.71 1 2.45 | -0.38 0.12 0.62 -0.08 0.42 0.91 0.22 0.72 1.21 --------- --- - ------------------------ CONSTR.'Agr.Pr.=30; COSTS= lAgr.Pr.=401 +25 1 lAgr.Pr.=50 2.06 3.55 5.04 2.36 3.85 5.34 ---------------------------- ------------------ ------- ---- 2.66 1 4.15 1 5.63 --------|---------- ----- ------------------------ -0.17 0.57 1.32 0.13 0.87 1.61 0.42 1 1.17 1 1.91 | -0.62 -0.12 0.37 -0.32 0.18 0.67 i-------------------------- ---|----------------------------- WATER REQUIREMENTS = -50 1: WATER REQUIREMENTS = even DISCOUNT RATE: -0.92 -0.42 0.08 | WATER REQUIREMENTS = +50 1 1E1.Pr.=10 EI.Pr.=20 El.Pr.=30 E1.Pr.=10 E1.Pr.=20 EI.Pr.=30|E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR. Agr.Pr.=30 COSTS= lAgr.Pr.=40 -25 1 lAgr.Pr.=50 2.84 4.33 5.81 3.14 4.62 b.11 3.43 4.92 6.41 0.60 1.35 2.09 0.90 1.65 2.39 1.20 1.94 2.69 -0.14 0.36 0.85 0.16 0.65 1.15 0.46 0.95 1.45 CONSTR. :Agr.Pr.=30: COSTS= Agr.Pr.=40 even :Agr.Pr.=50: 2.19 3.68 5.17 2.49 3.98 5.47 2.79 4.28 5.77 -0.04 0.71 1.45 0.26 1.00 1.75 0.56 1.30 2.05 -0.78 -0.29 0.21 -0.48 0.01 0.51 -0.19 0.31 0.81 ONSTR. Agr.Pr.=30 1.55 1.85 2.15 -0.68 -0.38 -0.08 -1.43 -1.13 -0.83 COSTS= |Agr.Pr.=40! +25 1 lAgr.Pr.:50' 3.04 4.53 3.34 4.83 3.64 5.13 0.06 0.81 0.36 1.11 0.66 1.40 -0.93 -0.43 -0.63 -0.13 -0.33 0.16 161 Net benefits C-8: TABLE WATER REQUIREMENTS = -50 1 DISCOUNT RATE: H of Alternative WATER W REQUIREMENTS = even | WATER REQUIREMENTS = +50 1 .EI.Pr.=10 E1.Pr.=20 EI.Pr.=30.E1.Pr.=10 E1.Pr.=20 El.Pr.=30:E1.Pr.=10 El.Pr.=20 El.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40| -25 1 iAgr.Pr.=50: 3.14 4.70 6.26 3.44 5.00 6.56 3.73 5.30 6.86 0.79 1.58 2.36 1.09 1.87 2.65 1.39 1 2.17 2.95 0.01 0.53 1.06 0.31 0.83 1.35 0.18 0.96 1.74 0.48 1.26 2.04 0.78 1.56 1 2.34 -0.60 -0.08 0.44 -0.30 0.22 0.74 -0.00 0.52 1.04 -0.14 0.64 .1.42 0.16 0.94 1 1.72 1 -1.22 -0.70 -0.18 -0.92 -0.40 0.12 -0.62 -0.10 0.42 0.61 1.13 1.65 ---- !-----:--------------------------------------------------- CONSTR.!Agr.Pr.=30t COSTS= :Agr.Pr.=40; even :Agr.Pr.=50: ----! ----- -- --- 2.82 4.38 5.94 3.12 4.68: 6.241 - --- :----------------------------------------------------- CONSTR.:Agr.Pr.=30 COSTS= 1Agr.Pr.=40: 425 1 :Agr.Pr.=50! ---- 2.52 4.08 5.64 -- 1.91 3.47 5.03 - 2.21 3.77 5.33 - - - - 2.50 4.07 1 5.63 | - - - - - -- -0.44 0.35 1.13 - - -- - -- - - - - ---- -- - - - - - WATER REQUIREMENTS = -50 1 ; WATER W REQUIREMENTS = even DISCOUNT RATE: - - - - - - - - - - - - - - WATER REQUIREMENTS = +50 1 W .El.Pr.=10 EI.Pr.=20 EI.Pr.=30!,El.Pr.=10 EI.Pr.=20 El.Pr.=30:El.Pr.=10 EI.Pr.=20 EI.Pr.=30 : --------------- -:----------------------------- ------------ CONSTR..Agr.Pr.=30: COSTS= :Agr.Pr.=40: -25 1 :Agr.Pr.=50: 2.70 4.26 5.82 3.00 4.56 6.12 3.30 1 4.86 6.42 I 0.36 1.14 1.92 0.66 1.44 2.22 0.96 | 1.74 I 2.52 -0.42 0.10 0.62 -0.12 0.40 0.92 0.18 0.70 1.22 CONSTR.!Agr.Pr.=301 COSTS= :Agr.Pr.=40: even IAgr.Pr.=501 1.94 3.50 5.06 2.24 3.80 5.36 2.54 1 4.10 5.66 -0.40 0.38 1.16 -0.10 0.68 1.46 0.20 0.98 1.76 1 -1.18 -0.66 -0.14 -0.88 -0.36 0.16 -0.59 -0.06 0.46 CONSTR.Agr.Pr.=30: COSTS= IAgr.Pr.=40: +25 I1Agr.Pr.=50: 1.18 2.74 4.30 1.48 3.04 4.60 1.78 3.34 4.90 -1.16 -0.38 0.40 -0.86 -0.08 0.70 -0.57 0.22 1 1.00 : -1.94 -1.42 -0.90 -1.64 -1.12 -0.60 -1.35 -0.83 -0.30 DISCOUNT RATE: WATER W REQUIREMENTS = -50 1 I WATER REQUIREMENTS = even I WATER REQUIREMENTS = +50 1 El.Pr.=10 E1.Pr.=20 EI.Pr.=30:E1.Pr.=10 El.Pr.=20 El.Pr.=30:E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.!Agr.Pr.=30 COSTS= 'Agr.Pr.=401 2.27 3.83 2.57 4.13 2.87 1 4.43 | -0.07 0.71 0.23 1.01 0.53 1.31 -25 1 Agr.Pr.=50 5.39 5.69 5.99 | 1.49 1.79 CONSTR.:AgrPr.=301 COSTS= Agr.Pr.=40 1.37 2.93 1.67 3.23 1.97 3.53 -0.97 -0.19 :Agr.Pr.=50: 4.49 4.79 5.09 CONSTR.|Agr.Pr.=30 0.46 0.76 COSTS= :Agr.Pr.=40 2.03 2.32 +25 X :Agr.Pr.=50 3.59 3.89 even -0.85 -0.33 -0.55 -0.03 -0.25 0.27 2.09 0.19 0.49 0.79 -0.68 0.11 -0.38 0.40 -1.75 -1.23 -1.46 -0.94 -1.16 -0.64 0.59 0.89 1.18 -0.71 -0.42 -0.12 1.06 -1.88 -1.58 -1.28 -2.66 -2.36 -2.06 2.62 -1.10 -0.80 -0.50 -2.14 -1.84 -1.54 4.18 -0.32 -0.02 0.28 -1.62 -1.32 -1.02 | 162 TABLE C-9: Net benefits of Alternative I WATER REQUIREMENTS = -50 Z : WATER REQUIREMENTS = even DISCOUNT RATE: | WATER REQUIREMENTS = +50 1 E1.Pr.=10 El.Pr.=20 E1.Pr.=30 E1.Pr.=10 EI.Pr.=20 E1.Pr.=30:E1.Pr.210 EI.Pr.=20 EI.Pr.=30 -- - - - -:- - - - - - - - - - - - - - - - - - - - - - - - - - - - - :- - - -- - - - - -- - - - CONSTR.!Agr.Pr.=30 COSTS= Agr.Pr.=40 -25 1 Agr.Pr.=50 1.95 2.62 3.28 2.44 3.10 3.76 2.92 3.58 4.24 0.96 1.29 1.62 1.44 1.77 2.10 1.92 : 2.25 1 2.58 0.63 0.85 1.07 1.11 1.33 1.55 1.59 1.81 2.03 CONSTR.:Agr.Pr.=301 COSTS= :Agr.Pr.=401 even Agr.Pr.=50 1.78 2.45 3.11 2.26 2.93 3.59 2.74 3.41 4.07 0.79 1.12 1.45 1.27 1.60 1.93 1.75 : 2.08 2.41 ! 0.46 0.68 0.90 0.94 1.16 1.38 1.42 1.64 1.86 CONSTR.2Agr.Pr.=30 COSTS= |Agr.Pr.=40 +25 1 |Agr.Pr.z50 1.61 2.27 2.94 2.09 2.76 3.42 2.57 3.24 3.90 0.62 0.95 1.28 1.10 1.43 1.76 1.58 1.91 2.24 0.29 0.51 0.73 0.77 0.99 1.21 1.25 1.47 1.69 ------- :---------- ----------------------------- ------------------------------------ ---------- WATER REQUIREMENTS = -50 1 ' WATER REQUIREMENTS = even DISC00LT RATE: 10 1 - ' WATER REQUIREMENTS = +50 1 .EI.Pr.=10 El.Pr.=20 El.Pr.=30!E1.Pr.=10 EI.Pr.=20 EI.Pr.=30!E.Pr.=10 EI.Pr.=20 EI.Pr.=30 -------------- :------------ ----------------- |--------------------------- --------------------------- -- CONSTR.JAgr.Pr.=30 COSTS= |Agr.Pr.=40 -25 1 IAgr.Pr.=50 ------- |:---------- CONSTR.IAgr.Pr.=301 COSTS= |Agr.Pr.=40 even |Agr.Pr.=50 ---- ----- CONSTR.:Agr.Pr.=30: COSTS= |Agr.Pr.=40 +25 1 :Agr.Pr.=50 ---- 1.83 2.50 3.16 2.31 2.98 3.64 2.79 3.46 4.12 0.84 1.17 1.50 ---------------------------- 1.62 2.28 2.95 2.58 1 3.25 3.91 1.89 2.55 3.22 2.37 1 3.03 3.70 0.63 0.96 1.29 0.42 0.75 1.08 0.51 0.73 0.95 1.59 1.92 2.25 1 0.90 1.23 1.56 1.38: 1.71 ! 2.04 - 0.30 0.52 0.74 --- -- 0.99 1.21 1.43 ----------- 1.11 1.44 1.77 :--------------------------------------- DISCOUNT RATE: 12 1 1.80 1 2.13 2.46 ---------------- -------------------- --------------- 1.41 2.07 2.74 -------- 2.10 2.77 3.43 1.32 1.65 1.98 1.47 1.69 1.91 - 0.78 1.00 1.22 1.26 1.48 1.70 0.56 0.79 1.01 1.05 1.27 1.49 --------------- 0.08 0.31 0.53 -------------- WATER REQUIREMENTS = -50 1 WATER W REQUIREMENTS = even | WATER REQUIREMENTS = +50 Z ==================================================================== EI.Pr.=10 E1.Pr.=20 EI.Pr.=301E1.Pr.=10 EI.Pr.=20 El.Pr.=30;E1.Pr.=10 E1.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= Agr.Pr.=40 -25 2 Agr.Pr.=50 1.71 2.38 3.04 2.19 2.86 3.52 2.68 3.34 4.00 0.72 1.05 1.38 1.20 1.53 1.86 1.68 2.01 2.34 0.39 0.61 0.83 0.87 1.09 1.31 1.35 1.57 1.79 CONSTR. Agr.Pr.=301 COSTS= !Agr.Pr.=401 1.46 2.13 1.94 2.61 2.42 3.09 0.47 0.80 0.95 1.28 1.43 1.76 0.14 0.36 0.62 0.84 1.10 1.32 Agr.Pr.=50 2.79 3.27 3.75 1.13 1.61 2.09 0.58 1.06 1.54 CONSTR. Agr.Pr.=30 1.21 1.69 2.17 0.22 0.70 1.18 -0.12 0.37 0.85 COSTS= :Agr.Pr.=401 1.87 2.35 2.83 0.55 1.03 1.51 0.11 0.59 1.07 2.54 3.02 3.50 0.88 1.36 1.84 0.33 0.81 1.29 even +25 I Agr.Pr.=50, 163 TABLE C-10: Net J of Alternative benefits WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even DISCOUNT RATE: 8 1 : WATER REQUIREMENTS a +50 1 ====================================================== EI.Pr.=10 EI.Pr.=20 E1.Pr.=30:EI.Pr.=10 E1.Pr.=20 EI.Pr.=30!E.Pr.=10 EI.Pr.=20 EZ.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= Agr.Pr.=40 -25 1 Agr.Pr.=50: -- - - i - - - - 3.59 5.00 6.41 -i-- - - - 3.89 5.29 6.70 - - - - - 4.18 5.59 7.00 - - - - - 1.48 2.18 2.89 -- - - - 1.77 2.48 3.18 - - - - - 2.07 2.77 3.48 - - - - 0.78 1.24 1.71 - - - - - - 1.07 1.54 2.01 -- - --------- 1.36 1.83 2.30 - - CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 even :Agr.Pr.=50: 3.28 4.69 6.10 3.58 4.98 6.39 3.87 5.28 6.69 1.17 1.87 2.58 1.46 2.17 2.87 1.76 2.46 3.17 0.47 0.94 1.40 0.76 1.23 1.70 1.05 1.52 1.99 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 +25 1 Agr.Pr.=50: 2.97 4.38 5.79 3.27 4.68 6.08 3.56 4.97 6.38 0.86 1.57 2.27 1.16 1.86 2.56 1.45 2.15 2.86 0.16 0.63 1.10 0.45 0.92 1.39 0.74 1.21 1.68 DISCOUNT RATE: 10 1 WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 ===== E1.Pr.=10 E1.Pr.=20 EI.Pr.=30'El.Pr.=10 EI.Pr.=20 El.Pr.=30|EI.Pr.=10 E1.Pr.=20 EI.Pr.=30 CDNSTR.:Agr.Pr.=30: COSTS= |Agr.Pr.=40: -25 1 :Aqr.Pr.=50 ---- : ----- 3.37 4.78 6.19 ---- 3.67 5.07 6.48 3.96 5.37 6.78 1.26 1.96 2.67 1.55 2.26 2.96 1.85 2.55 3.26 0.85 1.32 1.79 1.14 1.61 2.08 0.17 0.64 1.11 0.47 0.94 1.41 0.76 1.23 1.70 -0.21 0.26 0.73 0.09 0.56 1.02 0.38 0.85 1.32 -------------------------------------- --------------- CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40| even |Agr.Pr.=50: 2.99 4.40 5.81 3.28 4.69 6.10 3.58 4.99 6.40 0.88 1.58 2.29 1.17 1.88 2.58 1.47 2.17 2.87 1 CONSTR.:Agr.Pr.=30: COSTS= !Agr.Pr.=40: 425 1 :Agr.Pr.=50: 2.61 4.02 5.43 2.90 4.31 5.72 3.20 4.60 6.01 0.50 1.20 1.90 0.79 1.49 2.20 1.08 1.79 2.49 DISCOUNT RATE: 0.56 1.03 1.49 WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even WATER REQUIREMENTS = +50 1 El.Pr.=10 El.Pr.=20 EI.Pr.=30 El.Pr.=10 EI.Pr.=20 EI.Pr.=30 El.Pr.=10 El.Pr.=20 EI.Pr.=30 -- ------------------------------------------------------------------------------------------------ CONSTR. Agr.Pr.=30 COSTS= |Agr.Pr.=40 -25 1 Agr.Pr.=50 3.16 4.57 5.97 CONSTR. Agr.Pr.=30 2.70 COSTS= Agr.Pr.=40: 4.11 even Agr.Pr.=50 5.52 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40: +25 1 Agr.Pr.=50 3.45 4.86 6.27 3.74 5.15 6.56 1.04 1.75 2.45 1.34 2.04 2.75 1.63 2.34 3.04 0.34 0.81 1.28 0.63 1.10 1.57 0.93 1.40 1.87 3.00 3.29 0.59 0.88 1.18 -0.11 0.18 0.47 4.41 4.70 1.29 1.59 1.88 0.36 0.65 0.94 5.81 6.11 2.00 2.29 2.59 0.83 1.12 1.41 2.25 2.54 2.84 0.14 0.02 4.25 0.84 -0.10 0.20 0.49 5.07 5.36 5.65 1.55 0.72 1.43 2.13 -0.27 3.95 0.43 1.13 1.84 -0.57 3.66 0.37 0.67 0.96 1 64 TABLE DISCOUNT RATE: 8 Net C-il: benefits of Alternative K WATER REQUIREMENTS = -50 1 !WATER REQUIREMENTS = even I | WATER REQUIREMENTS = +50 1 ============================================================ EI.Pr.=10 EI.Pr.=20 E1.Pr.=30 E1.Pr.=10 EI.Pr.=20 E1.Pr.=30|E.Pr.=10 El.Pr.=20 El.Pr.-30 CONSTR. :Aqr.Pr.=30 COSTS= Agr.Pr.=40 -251 Agr.Pr.=50 3.43 4.92 6.41 3.73 5.22 6.71 4.02 5.51 7.00 1.19 1.94 2.69 1.49 2.23 2.98 1.78 2.53 3.27 0.45 0.95 1.44 0.74 1.24 1.74 1.04 1.53 2.03 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40: even Agr.Pr.=50 2.99 4.48 5.97 3.28 4.77 6.26 3.57 5.07 6.56 0.75 1.49 2.24 1.04 1.79 2.53 1.34 2.08 2.83: 0.00 0.50 1.00 0.30 0.79 1.29 0.59 1.09 1.59 0.30 1.05 1.79 0.60 1.34 2.09 0.89 1.64: 2.31 : -0.15 0.35 0.85 0.15 0.64 1.14 ---- ; ----- !--------------------------------------- CONSTR.IAgr.Pr.=30 COSTS= :Agr.Pr.=40: +25 2 :Agr.Pr.=50: 11---- 2.54 4.03 5.52 1- 2.83 4.33 5.82 -------- 3.13 4.62 6.11 --- I ---------- ---------------- ------------------------------ DISCOUNT RATE: ------------------ -------- MATER REQUIREMENTS = -50 1 -0.44 0.05 0.55 ---------------------------- : WATER REQUIREMENTS = +50 1 WATER REQUIREMENTS = even !E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 E1.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 E1.Pr.=30 NSTR.:Aqr.Pr.=30: COSTS= :Agr.Pr.z40: -25 1 :Agr.Pr.=50: 3.12 4.61 6.10 3.41 4.90 6.39 3.70 5.19 6.69 2.56 4.06 5.55 2.86 4.35 5.84 3.15 4.64 6.14 TR.:Agr.Pr.=30: COSTS= Agr.Pr.=40; even Agr.Pr.=50 ----- 2.60 4.09 5.58 1.47 2.21 2.96 0.13 0.63 1.13 0.43 0.92 1.42 0.33 1.07 1.82 0.62 1.37 2.11 0.92 1.66 2.41 -0.42 0.08 0.58 -0.12 0.37 0.87 0.17 0.67 1.16 2.01 3.51 5.00 ------:---- 2.31 3.80 5.29 -0.22 0.52 1.27 0.07 0.82 1.56 0.37 1.11 1.86 -0.67 -0.18 0.32 -0.38 0.12 0.61 - --------------------------- --------------- 0.72 1.22 1.72 --------------- :---------------------------- CONSTR.:Agr.Pr.=30 COSTS= :Agr.Pr.=40 +251 :Agr.Pr.=50: DISCOUNT RATE: 12 1 1.17 1.92 2.66 ---- a------------------------------------- ---- :------ ----: 0.88 1.62 2.37 ---------------------- -0.97 -0.47 0.03 ------ - WATER REQUIREMENTS = -50 1 WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 -================================-=============== :EI.Pr.=10 E1.Pr.=20 EI.Pr.=30 E.Pr.=10 EI.Pr.=20 E1.Pr.=30 E1.Pr.=10 E1.Pr.=20 E1.Pr.=30 CONSTR.:Agr.Pr.=30 COSTS= :Agr.Pr.=40 -25 1 :Agr.Pr.=50: 2.80 4.30 5.79 3.10 4.59 6.08 3.39 4.88 6.38 0.57 1.31 2.06 0.86 1.61 2.35 1.16 1.90 2.65 -0.18 0.32 0.82 0.12 0.61 1.11 0.41 0.91 1.40 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 even :Agr.Pr.=50: 2.15 3.64 5.13 2.44 3.94 5.43 2.74 4.23 5.72 -0.09 0.66 1.40 0.21 0.95 1.70 0.50 1.25 1.99 -0.83 -0.33 0.16 -0.54 -0.04 0.46 -0.24 0.25 0.75 CONSTR. Agr.Pr.=30 COSTS= :Agr.Pr.=40 +25 1 Aqr.Pr.=50 1.50 2.99 4.48 1.79 3.28 4.77 2.08 3.58 5.07 -0.74 0.00 0.75 -0.45 0.30 1.04 -0.15 0.59 1.34 | -1.49 -0.99 -0.49 -1.19 -0.70 -0.20 -0.90 -0.40 0.10 165 TABLE C-i2: DISCOUNT RATE: 81 Net benefits of Alternative WATER REQUIREKENTS = -50 1 | WATER REQUIREMENTS = even L : WATER REQUIREMENTS a +50 1 ==================================2 |EI.Pr.=10 E1.Pr.=20 EI.Pr.=30|E1.Pr.=10E1.Pr.=20 EI.Pr.=30|E.Pr.=10 EI.Pr.=20 EI.Pr.=3U CONSTR.:Agr.Pr.=3o: COSTS= !Aqr.Pr.=40| -25 1 :Agr.Pr.=50: 3.14 4.72 6.29 3.44 5.01 6.58 3.73 5.30 6.88| 0.78 1.57 2.36 1.08 1.86 2.65 1.37 2.16 2.95: -0.00 0.52 1.05 0.29 0.82 1.34 0.59 1.11 1.63 CONSTR.Agr.Pr.=30; COSTS= |Agr.Pr.=40: even |Agr.Pr.=50| 2.52 4.09 5.67 2.81 4.39 5.96 3.11 4.68 : 6.25 | 0.16 0.95 1.73 0.45 1.24 2.03 0.75 1.54 2.32 -0.63 -0.10 0.42 -0.33 0.19 0.72 -0.04 0.49 1.01 -0.46 0.32 1.11 -0.17 0.62 1.40 0.13 0.91 1.70 ---- : ----- :---------------:----------------------------------------- CONSTR.!Agr.Pr.=30: COSTS= :Agr.Pr.=40: +25 Z :Agr.Pr.=50: 1.90 3.47 5.04 2.19 3.76 5.34 2.49 1 4.06 1 5.63 | --- -1.25 -0.72 -0.20 -0.95 -0.43 0.09 -0.66 -0.14 0.39 -----------------------------------------------------------------------------------DISCOUNT RATE: WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 10 = E1.Pr.=10 E.Pr.=20 EI.Pr,=30|E1.Pr.=10 E1.Pr.=20 E1.Pr.=30!E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 ----------------------------------------------------------------------------------------------- CONSTR.Agr.Pr.=30: COSTS= Agr.Pr.=40| -25 Z Agr.Pr.=50! 2.70 4.28 5.85 3.00 4.57 6.14 3.29 | 4.86 1 6.44 | CONSTR.Agr.Pr.=30: COSTS= Agr.Pr.=40| even 1Agr.Pr.=50; 1.93 3.50 5.08 2.23 3.80 5.37 2.52 4.09 5.67 0.34 1.13 1.92 0.64 1.42 2.21 0.93 1 1.72 | 2.50 1 -0.44 0.08 0.60 -0.15 0.37 0.90 0.14 0.67 1.19 -0.43 0.36 1.15 -0.13 0.65 1.44 0.16 0.95 | 1.73 -1.21 -0.69 -0.17 -0.92 -0.40 0.13 -0.63 -0.10 0.42 ------- :----------|:----------------------------- ----------------------------- ----------------------------- CONSTR.Agr.Pr.=30: COSTS= 1Agr.Pr.=40: +25 1 lAgr.Pr.=50: DISCOUNT RATE: 12 1 1.16 2.73 4.31 1.46 3.03 4.60 1.75 3.32 4.90 -1.20 -0.41 0.37 -0.90 -0.12 0.67 -0.61 0.18 0.96 -1.98 -1.46 -0.94 -1.69 -1.17 -0.64 -1.40 -0.87 -0.35 WATER REQUIREMENTS = -50 1 WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 ========================= :EI.Pr.=10 El.Pr.=20 E.Pr.=30 E1.Pr.=10 EI.Pr.=20 E1.Pr.=30|E1.Pr.=10 El.Pr.=20 EL.Pr.=30 CONSTR. 1Agr.Pr.=30: COSTS= Agr.Pr.=40 2.27 3.84 2.56 4.13 2.86 4.43 -0.09 0.69 0.20 0.99 0.50 1.28 -0.88 -0.35 -0.58 -0.06 -0.29 0.23 -25 1 lAqr.Pr.=50 5.41 5.71 6.00 1.48 1.77 2.07 0.17 0.46 0.76 CONSTR. :Agr.Pr.=30: -1.01 -0.71 -0.42 -1.79 -1.50 -0.22 0.07 0.37 -1.27 -0.98 -0.68 Agr.Pr.=40: 1.35 2.93 1.65 COSTS= 3.22 1.94 3.51 even Agr.Pr.=50 4.50 4.79 5.09 0.57 0.86 1.15 -0.75 -0.45 -0.16 CONSTR. :Agr.Pr.=:30 COSTS= Agr.Pr.=40 +25 :Aqr.Pr.=50: 0.44 2.01 3.58 0.73 2.30 3.88 1.02 2.60 4.17 -1.92 -1.14 -0.35 -1.63 -0.84 -0.06 -1.34 -0.55 0.24 -2.71 -2.19 -1.66 -2.42 -1.89 -1.37 -2.12 -1.60 -1.07 X -1.21 1 66 TABLE Net benefits C-13: WATER REQUIREMENTS = -50 1 Z DISCOUNT RATE: of Alternative WATER REQUIREMENTS = even M : WATER REQUIREMENTS = +50 1 EI.Pr.=10 E1.Pr.=20 EI.Pr.=30:EI.Pr.=10 EI.Pr.=20 E1.Pr.=30|E.Pr.=10 EI.Pr.=20 EI.Pr.230 CONSTR.2Agr.Pr.=30: COSTS= :Agr.Pr.=40 -25 1 :Agr.Pr.=50: 3.52 5.01 6.50 4.00 5.49 6.98 4.49 5.98 7.47 1.29 2.03 2.77 1.77 2.51 3.26 2.26 3.00 3.74 0.54 1.04 1.53 1.03 1.52 2.02 1.51 2.01 2.50 CONSTR.IAgr.Pr.=30 COSTS= Agr.Pr.=40 even :Aqr.Pr.=50: 3.04 4.53 6.02 3.53 5.01 6.50 4.01 5.50 6.99 0.81 1.55 2.30 1.29 2.04 2.78 1.78 2.52 3.27 0.06 0.56 1.06 0.55 1.04 1.54 1.03 1.53 2.03 CONSTR.:Agr.Pr.=30 COSTS= Agr.Pr.=40: +25 1 :Agr.Pr.=50: 2.56 4.05 5.54 3.05 4.54 6.03 3.53 5.02 6.51 0.33 1.07 1.82 0.82 1.56 2.30 1.30 2.04 2.79 -0.41 0.08 0.58 0.07 0.57 1.06 0.56 1.05 1.55 WATER REQUIREMENTS = -50 1 | WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 Z DISCOUNT RATE: :EI.Pr.=10 E1.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.Agr.Pr.=30 COSTS= |Agr.Pr.=40| -25 1 !Agr.Pr.=50| ---- ----- 3.18 4.67 6.16 3.67 5.15 6.64 4.15 1 5.64 7.13 0.95 1.69 2.44 1.43 2.18 2.92 1.92 2.66 3.41 : :-------------------------------------------------------- ---- CONSTR.Aqr.Pr.=30: COSTS- Agr.Pr.=40t even :Agr.Pr.=50: 2.59 4.08 5.57 3.07 4.56 6.05 3.56 5.05 6.54 0.36 1.10 1.85 0.84 1.59 2.33 1.33 2.07 2.82 CONSTRAgr.Pr.=30: COSTS= :Agr.Pr.=40| +25 1 :Agr.Pr.=50! 2.00 3.49 4.98 2.48 3.97 5.46 2.97 4.46 5.95 -0.23 0.51 1.26 0.25 1.00 1.74 12 0.69 1.18 1.68 1.17 1.67 2.17 --------------------- : : -0.39 0.11 0.60 0.10 0.59 1.09 0.58 1.08 1.57 0.74 1 1.48 2.22 -0.9 -0.48 0.01 -0.49 0.00 0.50 -0.01 0.49 0.98 ----------- ----------------------------------- DISCOUNT RATE: 0.20 0.70 1.20 WATER REQUIREMENTS = -50 1 : WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 2 ===== E1.Pr.=10 EI.Pr.=20 EI.Pr.=30|E1.Pr.=10 E1.Pr.=20 EI.Pr.=30|E1.Pr.=10 EI.Pr.=20 El.Pr.=30 CONSTR.:Agr.Pr.=30' COSTS= |Agr.Pr.=40 -25 1 'Agr.Pr.=50 2.85 4.34 5.82 3.33 4.82 6.31 CONSTR. Agr.Pr.=30 2.15 COSTS= Agr.Pr.=4O0 even :Agr.Pr.=50: 3.63 5.12 CONSTR. :Agr.Fr.=30 COSTS= Agr.Pr.=40 +25 1 Agr.Pr.=50 1.44 2.93 4.42 3.82 5.31 6.79 0.61 1.36 2.10 1.10 1.84 2.59 1.58 2.33 3.07 ' -0.13 0.37 0.86 0.35 0.85 1.35 0.84 1.34 1.83 2.63 3.12 -0.09 0.40 0.82 -0.83 -0.35 0.14 4.12 5.61 4.60 6.09 0.66 1.40 1.14 1.89 1.63 2.37 -0.34 0.16 0.15 0.65 0.63 1.13 1.93 3.42 4.91 2.41 3.90 5.39 -0.79 -0.04 0.70 -0.0 0.44 1.18 0.18 0.93 1.67 -1.53 -1.04 -0.54 -1.05 -0.55 -0.06 -0.56 -0.07 0.43 167 TABLE DISCOUNT RATE: C-14: Net benefits of Alternative : WATER REQUIREMENTS = -50 1 i WATER REQUIREMENTS = even N : WATER REQUIREMENTS = +50 1 |E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30:E1.Pr.=10 EI.Pr.=20 EI.Pr.=30 CONSTR.:Agr.Pr.=30: COSTS= :Agr.Pr.=40| -25 1 :Agr.Pr.=50: 3.50 4.99 6.48 3.98 5.47 6.96 4.46 5.95 7.44 1.26 2.01 2.75 1.74 2.49 3.23 2.22 2.97 3.71 0.52 1.01 1.51 1.00 1.49 1.99 1.48 1.97 2.47 COMTR.|Agr.Pr.=30: COSTS= :Agr.Pr.=40| even |Agr.Pr.=50: 3.01 4.50 5.99 3.49 4.98 6.48 3.97 5.46 6.96 0.78 1.52 2.27 1.26 2.00 2.75 1.74 : 2.48 : 3.23 0.03 0.53 1.02 0.51 1.01 1.50 0.99 1.49 1.99 CONSTR. Agr.Pr.=30! COSTS= :Agr.Pr.=40: +25 2 :Agr.Pr.=50: 2.53 4.02 5.51 3.01 4.50 5.99 3.49 | 4.98: 6.47 : 0.29 1.04 1.78 0.77 1.52 2.26 1.25 : 2.00 2.74 : -0.46 0.04 0.54 0.02 0.52 1.02 0.51 1.00 1.50 --------- I ---------------- DISCOUNT RATE: 10 WATER REQUIREMENTS = -50 1 1 WATER REQUIREMENTS = even : WATER REQUIREMENTS = +50 1 = !EI.Pr.=10 EI.Pr.=20 EI.Pr.=30!E1.Pr.=10 E1.Pr.=20 EI.Pr.=30 E.Pr.=10 E1.Pr.=20 EI.Pr.=30 ----------------------------------------- CONSTR.:Agr.Pr.=30! COSTS= :Agr.Pr.=401 -25 1 :Agr.Pr.=50: 3.15 4.64 6.14 3.63 5.13 6.62 4.11 1 5.61 1 7.10 0.92 1.66 2.41 1.40 2.14 2.89 1.88 2.62 3.37 0.17 0.67 1.17 0.65 1.15 1.65 1.13 1.63 2.13 CONSTR.!Agr.Pr.=30: COSTS= :Agr.Pr.=40 even :Agr.Pr.=50: 2.55 4.04 5.54 3.03 4.53 6.02 3.51 : 5.01 : 6.50 | 0.32 1.06 1.81 0.80 1.54 2.29 1.28 2.02 2.77 -0.43 0.07 0.57 0.05 0.55 1.05 0.53 1.03 1.53 ---- |:--------- : --------------------- CONSTR.!Agr.Pr.=30: COSTS= 1Agr.Pr.=40| +25 1 |Agr.Pr.=50 2.43 3.93 5.42 ----------------------------- ------------------------- 2.91 1 4.41 | 5.90| -0.28 0.46 1.21 0.20 0.94 1.69 0.68 1.42 2.17 WATER REQUIREMENTS = -50 1 | WATER REQUIREMENTS = even DISCOUNT RATE: 12 1.95 3.44 4.94 I -1.03 -0.53 -0.04 -0.55 -0.05 0.45 -0.07 0.43 0.93 : WATER REQUIREMENTS = +50 1 =========================================================.============================= IE1.Pr.=10 E1.Pr.=20 EI.Pr.=30:E.Pr.=10 EI.Pr.=20 E1.Pr.=30:E1.Pr.=10 EI.Pr.=20 El.Pr.=30 CONSTR. Aqr.Pr.=30: COSTS= Agr.Pr.=40: -25 1 Agr.Pr.=50 2.81 4.31 5.80 3.30 4.79 6.28 3.78 5.27 6.76 0.58 1.32 2.07 1.06 1.80 2.55 1.54 2.28 3.03 -0.17 0.33 0.83 0.31 0.81 1.31 0.79 1.29 1.79 CONSTR. Agr.Pr.=30 COSTS= Agr.Pr.=40 2.10 3.59 2.58 4.07 3.06 4.55 -0.13 0.61 0.35 1.09 0.83 1.57 1 -0.88 -0.38 -0.40 0.10 0.08 0.58 Agr.Pr.=50: 5.08 5.56 6.05 1.36 1.84 2.32 0.11 0.59 1.07 CONSTR. Agr.Pr.=30: COSTS= Agr.Pr.=40: +25 ' Aqr.Pr.=50 1.39 2.88 4.37 1.87 3.36 4.85 2.35 3.84 5.33 -0.85 -0.10 0.64 -0.37 0.38 1.12 0.11 0.86 1.60 -1.59 -1.10 -0.60 -1.11 -0.62 -0.12 -0.63 -0.14 0.36 even 168 BIBLIOGRAPHY Allam , M .N. , and D . H. Marks, Irrigated agricultural expansion planning in developing countries: Resilient system design , Water Resources Research, vol. 20, no. 7, July 1 98'4. Allison , G. , Essence of decision, Boston , , A .C. S .P. Bradley, Little, Brown & Cia., 1971 mathematical Hax, and T .L. programming, Magnanti, Addison-Wesley, Applied 1977. and L.D. Hopkins, Use of Chang, S.Y. , E.D.Brill, Jr., mathematical models to generate alternative solutions to water resources planning problems, Water Resources Research, vol. 18, no. 1, pp. 58-64 , February 1982. Marks, Multiobjective screening models and Cohon, J .L. , and D .H. water resources investment, Water Resources Research, vol. 9, no. 4, pp. 826-836, August 1973. Cohon, J .L. , and D .H. Marks, A review and evaluation of multiobjective programming techniques, Water Resources Research, vol. 11, no. 2, April 1975. Dasgupta, P. , and S. United Nations, Drake, Guidelines for project evaluation, Marglin, New York, A .W. , and R .L. 1 972. Keeney, Decision analysis, MIT, 1979. Fiering , M .B . , A screening model to quantify resilience , Water Resources Research, vol. 18, no. 1, pp. 27-32, February 1982. Fiering, M .B. , Alternative indices of resilience, Water Resources Research, vol. 18, no. 1, pp. 33-39, February 1982. Fiering, M . B. , Estimates of resilience indices by simulation, Water Resources Research, vol. 18, no. 1, pp. 41-50, February 1982. Fiering, M .B. , Estimating resilience by canonical analysis, Resources Research, vol. 18, no. 1, pp. 51-57, Water February 1982. Friedman , R . , C . Ansell , S . Diamond , and Y.Y. Haimes , The use of models for water resources management, planning, and policy, Water Resources Research, vol. 20, no. 7, pp. 793-802, July 1984. 169 Harrington , J . L . , and J . Gidley , The variability of alternative decisions in a water resources planning problem , Water Resources Research , vol . 21, no . 12 , pp . 1831 -1840 , December 1985. Hashimoto , T . , J . R . Stedinger , and D .P . Loucks , Reliability, resiliency , and vulnerability criteria for water resource system performance evaluation, Water Resources Research, 18, no . 1 , pp . 14-20, February vol. 198 2. T. , D . P. Loucks , and J . R. Stedinger, Robustness of Water Resources Systems, Water Resources Research, vol. 18, Hashimoto, 1, pp. 21-26, February no. 1982. Kneey , R . L . , and H . Raiffa , Decisions with multiple objectives, Wiley , New York, 1976 Loucks , D . P . , J . R. Stedinger, and D . A . Haith , Water resource system planning and analysis, Prentice -Hall, New Jersey , 198 1 Mahmassani , H . S . , Uncertainty in transportation systems evaluation: issues and approaches, Transportation, Planning, and Technology, vol. 9, pp. 1-12, 1984. Major, D .C. , and R .L. Lenton, Applied water resource systems planning, Prentice Hall, New Jersey, 1979. Meyer, M .D. , and E .J. Miller, Urban transportation planning: decision -oriented approach , McGraw -Hill , New York , 1 985. a Rogers, P.P. , and M.B. Fiering, Use of system analysis in water management, Water Resources Research, vol. 22, no. 9, pp. 158S, August 146S- 1986. Rogers , P .P . , The role of systems analysis as a tool in water policy , palnning and management, Urban Planning Policy Analysis and Administration, Discussion Paper D79-4, Harvard University, Departament of City and Regional Planning , Cambridge , Mass. March 1979 . 170