OCR Text |
Show failures in critical streamflow years were solved for as part of the optimization process. Groundwater was used to make up the failure of the surface water system. A distributed parameter representation of the groundwater system was directly specified in the model Applications of both models were performed with the same set of data. The following conclusions are prompted from an analysis of the results of the applications. ( 1) The use of groundwater in addition to surface water in the Jordan River Basin is more economical than a purely surface water development scenario. Results from the 2 models developed in this chapter were economically more attractive than those in Chapter 5 ( surface water only) for corresponding demand levels. When conjunctive use of surface and groundwater is considered ( section 7.2) the most economically attractive solutions are obtained. ( 2) The results from the two models developed in this chapter differ markedly as the demand level is increased. Conjunctive use of surface and groundwater leads not only to greater economy, but also to differences in the sizes and number of reservoirs selected for development. ( 3) The yield model framework is attractive for the development of ground and surface water management models. Conjunctive use aspects as well as issues for analyzing tradeoffs between reservoirs can be modeled with moderate mathematical problem size. The conjunctive use formulation is much more computationally demanding than the one where groundwater is considered as a competitive resource and is treated as a lumped parameter. The increase in complexity results from the inclusion of a distributed parameter groundwater optimization model as part of the conjunctive use model. Typical execution times for the conjunctive use model were 3 to 4 hours of CPU time on the GOULD 9080, as compared to 5 to 20 minutes of CPU time for the model where conjunctive use is not considered. No attempts were made to develop a decomposition algorithm for the solution of the conjunctive use model. There is however a good potential for developing such an algorithm, with the decomposition applied between the surface water and groundwater components of the model. The author intends to pursue work on such a development. Groundwater- surface water interactions were not explicitly modeled in the formulations presented in this chapter. Procedures for their consideration are however outlined in section 2.2 as part of the water supply formulation with the Girinski potential. For the Jordan River basin applications, the physical location of the reservoir sites and the aquifer wells, did not necessitate formal consideration of interactions between the reservoirs and the groundwater system. ( 4) Conjunctive use of ground and surface water appears to be an effective solution for the economic development of water supply in Salt Lake County. Results from the applications performed suggest that large scale exploitation of groundwater in conjunction with moderate surface water development is economically the most attractive development scenario. 176 |