Water Goes Where? A Water Resource Allocation Method Based on Multi-Objective Decision-Making

TL;DR

This paper presents a multi-objective decision-making model for water resource allocation between Glen Canyon Dam and Hoover Dam, considering economic, social, and environmental factors, solved using a multi-objective genetic algorithm.

Contribution

It introduces a novel water allocation model that accounts for dam coupling, multi-dimensional demands, and geographic-industry priorities, enhancing decision-making flexibility.

Findings

The model effectively balances water and power distribution among five states.

It demonstrates strong adaptability and ease of application in different scenarios.

The approach integrates multiple criteria and dam interactions for optimized resource management.

Abstract

For a long time, water and hydroelectric power are relatively important resources. Their rational distribution is closely related to regional agriculture, industry, residents, etc. In this paper, we mainly study the problem of allocation scheme for Glen Canyon Dam and Hoover Dam in the Colorado River Basin. Taking into consideration of various factors, we build models to achieve optimal scheduling. Firstly, we propose the Water Strategy Decision Model, which can obtain different distribution methods for the different water levels. Also, we connect the two dams in series to consider the coupling effect between them and integrate this part into the main model. Secondly, we propose three criteria of allocation for water and power generation, namely, economic, social, and environmental benefits. Social benefits mainly include the minimum shortage of water and electricity for agriculture, industry, and residents. For this multi-objective plan, the model is solved using a multi-objective ant colony genetic algorithm under the constraint of constant total water volume, and finally, the current reservoir capacities of the two lakes are input to derive the annual water supply to the five states. Thirdly, based on the location and development characteristics of the five states, we obtain a water scheduling model based on the priority of geographic-industry characteristics. Fourthly, we can regard the model as a four-dimensional space of industrial, agricultural, residential and power generation water demand. The partial derivative calculation formula of multi-dimensional space can be used to obtain the results. Finally, we analyze the sensitivity of the model, and it shows that the model has strong adaptability and is easier to popularize. Moreover, we discuss the advantages and disadvantages of the models.