Aqueduct Capacity under an Optimum Benefit Policy

Abstract

Aqueducts are often designed to serve a number of geographic districts located in sequence along the supply line. The beneficial returns (however they are to be defined) will not, in general, be the same from district to district and will be nonproportional functions of the quantity of water to be delivered. If the available supply of water will be less than the maximum demand, there exists the problem of determining the “best” allocation of water to the various districts. This question presents a very large number of alternatives, incapable of solution by direct comparison even with high speed digital computing machinery. In this paper a method of solving this problem using dynamic programming as an optimization device is presented. Because of the characteristics of dynamic programming, the individual allocations are obtained as a function of the total supply which might be made available at the head of the aqueduct. The total net beneficial return is also a function of the total supply. Thus, except for intangibles, the merits of increasing or decreasing the supply for this aqueduct as compared to other potential uses can be directly determined.