ABSTRACT
For establishing optimal allocation model of water resourees, the authors analyzed that how many sources of water could be used to supply a number of users at different places and time. In Kaifeng city, the water users are municipal, industrial, agricultural demands or other uses. The water supply sources are as follow : (1) Yellow river; (2) shallow aquifer, including potable and non potable quantities; (3) intermedium aquifuer, including potable and non potable quantities.
The nature of water resources allocation closely parallel the “transportation or transshipment” problem. As we konw, every user has certain water quality requirement, but the quality constraints cannot be considered in transportation programming. Linear prog) camming can overcome the shortcoming. For this reason, the authors have established an optimal model of water resources subject to quality constaints for the study area by using the combination technique of transportation and linear programming.
The main task, of the model is to determine the amouzts Xij to be allocated over all selected users routes so as to minimize costs. Owing to the allocation cost changes with time and other influence factors, so the relative costs of various water supply were chosen based on the cost of fresh water supply, three plans were designed for the study area. According to plan 1, Yellow river was allocated to all users, shallow aquifer was allocated to agriculture and intermedium aquifer was allocated to the industrial and domestic users. In plan 2, Yellow river and potable quantities of shallow aquifer was allocated to all users; non potable quantities of shallow aquifer to industrial and agricultural users; potable quantities of intermedium aquifer to domestic and industrial users; non potable quantities of intermedium aquifer only to industrial user. Finally in plan 3, Yellow river was allocated for domestic users, shallow aquifer for agriculture and intermedium aquifer was allocated to industrial demands.
The allocation model was solved by linear programming. The results show that from economic point of view, plan 2 is the best plan. Eventhough plan 1 has 10 percentage more total cost than plan 2, from quality point of view, plan 1 is the best plan.
If detailed data are available, it is possible to apply reuse concept in the allocation model proposed in the paper.
