IRAN-WATER RESOURCES RESEARCH
Year:2005 | Volume:1 | Issue:3|Papers Count:10

In this paper, various types of stochastic dynamic programming models (SDP) and also a deterministic dynamic programming (DP) are presented and compared for multipurpose Dez reservoir dam located in southwest of Iran. The impact of inflow forecasting uncertainty in optimum reservoir operation is investigated through two types of stochastic dynamic programming models. SDP models are different based on hydrologic state variable and inflow conditional or non conditional probability assumptions. A simulation model is developed to investigate the achieved optimum policies in different models. So, average of different operation variables and also performance criteria such as reliability, resiliency and vulnerability are used to compare the results. In two SDP models which apply current inflow instead of previous time step inflow as hydrologic state variable, real time simulation is done with forecasted and observed inflows. The effect of forecasting accuracy and different forecasting methods in reservoir operation are also studied. In general, objective function is considered as minimizing the sum squared of two sided deviations from target release and storage. This research shows that lack of attention to inflow forecasting in models which need it and assuming known values could lead us to unreal and false results and mislead us in selecting type of model. On the other hand, it is noticed that forecasting accuracy plays an important role in optimum reservoir operation.

A Long-term planning model is presented in this study optimizing Karoon-Dez reservoir system operation m Iran with hydropower generation, water supply, and environmental objectives. The matrix implementation of the large-scale resulting optimization model has been solved using dual-primal interior-point methods (DPIPMs) and multiobjective programming. These algorithms have shown to be promising especially when they take the advantage of sparsity structure of associated matrix formulation of the optimization problem with linear and quadratic functions. The computational time required for solving the Karoon-Dez reservoir system studied was less than 45 minutes. This is promising especially knowing that handling such an optimization model is very difficult using other techniques such as generalized reduced gradient (GRG) algorithm of nonlinear programming and discrete dynamic programming with adequate fine representation of state variables of the model.

In this paper, uncertainty of a rainfall-runoff (RR) model is analyzed based on combination of Monte Carlo (MC) procedure and Bayesian theory, which is known as GLUE framework. The rainfall-runoff transformation was performed by Mod Clark distributed- conceptual model. In this model, the basin's hydrograph is determined by the superposition of runoff generated by individual cells in a raster-based discretization. Application of MC in uncertainty analysis introduces convenient parameter variation range, which is not adjustable based on new data. In GLUE method, however, Bayesian theory is applied to update prediction limits and distribution of parameter as new data becomes available. Goodness of fit criteria is selected such that higher discharges of hydrograph are given larger weights compared to other parts of the hydrograph. Uncertainty of RR model parameters was assessed in Gharasoo basin, a subbasin of the great Karkheh river basin. The results show that GLUE has a good performance in updating model parameters in comparison with MC method alone.

Khorasan province bieng located in an arid and semiarid part of Iran, has often experienced drought during the recent years. Occurrence of consequent droughts during the last few years, has shown that drought prediction is a subject that deserve more attention. One way to achive such goal is modeling the rainfall. In this research, annual rainfall data of the eleven synoptic stations of Khorasan province from 1970 to 2002 have been used. Seasonal autoregressive integrated moving average method (SARMA) was used for modeling the seasonal rainfals of these stations. Based on this model, the amount of rainfall for spring, fall and winter was predicted. Anomally zoning was also prepared for Khorasan province.

Rainfall time pattern has a marked influence on design of dams, urban runoff collection systems, culverts, drainage systems and or determining, flood potential and soil erosion. In this research, data of 18 rainfall autographic stations of Khorasan province of Iran were used to implement the rainfall time pattern. Rainfall depths were extracted for each time interval from rainfall charts and were categorized to four quarters. A ranking method was used to arrive at rainfall time patterns corresponding to time intervals of 1, 3, 6, 9, 12, 18, and 24 hours. After normalizing the rainfall mass curves, every quarter of rainfall was ranked, where rank 1 was given to the highest. The final pattern was derived by averaging the quarter ranks and percentage of rainfalls corresponding to each duration. A contingency table was used to verify the independency of rainfall time patterns by a chi-square test. The results showed that in Khorasan province: a) most of short duration rainfalls are first quartile, b) most long-term rainfalls are second quartile, and c) on overall more than 50% of rainfall amounts for nearly all rainfalls in this arid and semi-arid region occurred in the first half of their period.

In this research, using complete form of Navier-Stokes equations, the 3-D turbulent flow pattern in a 180-degree river bend is simulated. The k-e and k-w turbulence models are used to close the system of equations and modeling Reynolds stresses. Equations are solved numerically using FVM method as implemented in the commercial code FLUENT. Experimental results in a 180-degree bend in Tarbiat Modarres Hydraulic lab (MHL) are used to verify the numerical results. Simulation showed that both of turbulence models accurately predicted the flow pattern in this bend. The prediction accuracy of the k-romodel, especially in the outlet region of the bend, is however higher than the k-w model. The results of first model showed that as the experimental model, there are two secondary flow cells in the q=180o' cross section but the k-e model predicted only one cell. Distribution of bed and side wall shear stresses showed that there are two probable scouring regions in two halves of the bend. Previous experimental investigations with mobile bed models confirm these results.

In this paper, using the 39 years of water surface fluctuation data and surface inflows, precipitation on the lake, evaporation from the free surface and iso-potential groundwater maps, the main parameters for water balance in the lake are calculated. Then these parameters are compared for long-term, wet and dry periods. According to received results, the maximum and minimum values of the water surface level of the lake are 1278.41 and 1273.23 meter above sea level, respectively. These extremes are observed in water years of 1373-74 and 1380-81 correspondingly. The Maximum and minimum variations in volume has occurred in 1347-48 and 1377-78, that they are introduced as a wet and dry years, respectively. The suitable elevation for lake is between 1275 to 1246 meters above sea level. The mean elevation based on 39 years of recorded data is 1275.6 m above sea level. In order to increase the level from 1273.23 to 1275.6 meter, about 7.4 billion m3 of water should be reserved. For this condition to occur, a wet year (i.e. 1347-48) or a long-term average for five successive years should happen. Also the variation rate for inflows is about %343 greater than outflows factors. Finally, after the consideration of the lake balance, management actions are presented for better exploitation of water resources and environmental preservation of Orumieh lake basin.

Aeration of flow downstream of outlet gates is an Effective way to eliminate the risk of cavitation. Many works have been done and various relationships have been developed to predict the quantity of entrained air. Owing the complexity of flow in the aeration zone arising from the two-phase flow, these relationships cannot however be used in general. On the other hand, in recent years, applications of Artificial Intelligence, such as Neural Network, Fuzzy Logic, and Generic Algorithm have attracted the attention of many investigators. These are known as powerful tools to solve engineering problems with uncertainties. In this paper, based on experimental data obtained from field measurements and physical model studies, an Artificial Neural Network (ANN) with a general back propagation error, is suggested to estimate the air demand downstream of bottom outlet gates. The results with a regression parameter of 0.992 showed that the model is very well capable of predicting air demand.