Stochastic Simulation of Steam-Flow and Analysis of Reservoir Sizing
No Thumbnail Available
Date
2018-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
This thesis is concerned with the problem of stochastic simulation of streamflow and reservoir sizing analysis in the Gidabo reservoir. Measso station with long record period of 29 years was used for the analysis. Four sets of time series models named, ARMA, PARMA (p,q) ,Thomas-Fiering and GAR(1) were tested and inter-compared for the performance in stochastic simulation of monthly stream-flow, accordingly a suitable model were selected.
PARMA models of low order are found to be most appropriate for Measso station than ARMA, Thomas–Fiering and GAR(1) model. The performance PARMA model has MD 0.94 and RRMSD 10.64. Therefore, long year sequence (fifty year) of synthetic flow were generated using PARMA (p,q) model and used for optimization the Gidabo reservoir system.
From the reservoir optimization analysis result it can be deduced that the required maximum capacity of the reservoir system in order to irrigate the potential land under the system without water stress, becomes 37.54MM3 provided that proper release rules are followed.
Description
Keywords
Reservoir Sizing, Steam-Flow, Stochastic Simulation