OCR Text |
Show reliabilities considered. Applications of this model to the Jordan River basin, and the Salt Lake City aquifer revealed a marked preference for groundwater development ( to its upper limit) over surface water development. A larger scale application to the entire Salt Lake County was not performed because the water rights in all the streams considered for reservoir development were owned by Salt Lake City. Salt Lake City, thus appears to be a water surplus agency, relative to some of the others in the county. Such an application, could however be performed if a re- negotiation of local water rights were likely. An assessment of the effect of such a reallocation could then be determined using the model. The second model developed considers conjunctive use of surface and groundwater through compensation of failures of yield at one source by other sources. Some of the principles developed in chapter 6 for optimal reservoir operation in a yield reliability context are accommodated and a distributed parameter groundwater and surface water representation is adopted. The model is thus a composite of the groundwater management model presented in chapter 3, and the reservoir screening model presented in chapter 5, with the coupling coming through the demand and system reliability constraints. A ' 100%' system reliability is targeted. Failures of individual reservoir yields in critical years are however allowed and solved for. Water system operators often allow ( or are forced to consider) a reduction in reservoir yield and increased groundwater use in critical streamflow periods. Where this transition in use is properly specified as to timing and degree, optimal conjunctive operation of the water supply system results. Wholesale water supply contracts often specify the repayment of the construction cost of a reservoir in fixed annual increments. In this case, the variable ( function of release) economic values for reservoir releases are related to treatment costs and to hydro- energy, and recreation benefits. Where groundwater variable costs are comparable to or higher than these reservoir supply costs ( net), the optimizing principle is the maximization of reservoir yield ( or corresponding economic value) with groundwater used to make up shortages. This is consistent with the operation policy that is often used. The determination of optimal degree and timing of reservoir yield failure remains a planning problem of interest. The model developed addresses this by including as decision variables the degree of reservoir yield failure in each of the critical streamflow years, and also including as a decision variable a firm yield from groundwater. Individual well pumpages for each year, are solved for to meet the total annual groundwater demand as specified by the firm groundwater yield and the increased pumping in years of reservoir failure. The problem size of the resulting formulation is dominated by its groundwater component. No strategies for decomposition of the mathematical problem into smaller sub- problems was pursued. Noting that the only linkage between the groundwater and the surface water sub systems comes through a few demand and reliability constraints, this should be feasible, and should be pursued in a parallel computing framework. Applications of the conjunctive use model were also performed with the Jordan river basin, and Salt Lake City aquifer data set. The results for low levels of system demand were not markedly different from those obtained with the previous surface water-groundwater model. The results at the higher level of system demand were however markedly different. They showed significant monetary savings, high levels of peak reservoir yield failure, and a smaller set of reservoirs considered for development. The importance of conjunctively managing surface and ground water resources as the demand for water increases was reinforced by these applications. More exhaustive applications along these lines with existing reservoirs in the area, particularly the Central Utah Project entities would be desirable for defining efficiency in local water use. 180 |