الفهرس 
INTORODUCTIONOnly 14 pages are availabe for public view
 |  | from | 128 |  |  |
| | | from | 128 | | |
Abstract
Use of pressurized irrigation systems in Egypt as well as many other countries has expanded in the last decades due to scarcity of water resources and increase of population. However, excessive use of energy by such systems created a new challenge owing to limited sources of energy and rising energy cost. Different management approaches to minimize energy consumption are investigated in the current studySectoring (clustering) of the network is an effective management technique for reducing energy consumption. Artificial Neural Network (ANN- Self Organizing Maps (SOMs)) algorithm is used to cluster the hydrants of the network into homogenous groups based on three dimensionless parameters; relative distance from the pump, relative level, and relative actual head at the hydrant. Various combinations of these parameters are investigated to maximize energy saving. Utilization of (ANN-SOMs) is compared with the well-known K-means clustering technique. Results of ANN- SOMs are slightly better in most of combinations proposed but the K-means algorithm is less complicated and requires less computation facilities.Use of booster pumps at locations of critical hydrants may provide a powerful management tool for energy saving in pressurized irrigation networks, particularly for networks with great difference in hydrants elevation and/or long branching pipes. Though installation of booster pumps increases the capital cost of the system, it is later compensated by reduced cost of energy during life time of the project. Economic assessment of using such pumps together with hydrants sectoring has been carried out for four different scenarios. The proposed methodology is applied to a drip irrigation network at Kostol area, South of Aswan High Dam, Egypt. Energy savings of 25.87% and 35.56% are achieved in two of these scenarios. Capital cost of the pumping system will increase by about 21.0 %. Discounted payback-periods required to compensate this increase are 7.00 years and4 .57 years, for these two scenarios, considering inflation and interest rates. A sensitivity test is performed to evaluate the effect of uncertainty in estimating capital costs of the pumps as well as future costs of energy on the estimated payback period. A 30% rise of booster pumps cost, or a 30% DROP in energy price (LE/kWh) will increase the payback period from 5.00 years to about 8.0, and 9.00 years, respectively. The same cost rise percentage for the main pump will only decrease such period from 5.00 years to about 4.00 years. Therefore, variation of cost estimate of the main pump has much less influence on the payback period compared to energy and booster pumps.Optimum economic design of pressurized, one-sector irrigation network has been investigated using Non-Dominated Sorting Genetic Algorithm (NSGA), versions II and III. Three, normalized objective-functions are minimized; energy cost (OF1), pipes cost (OF2) and pump station cost (OF3). Three scenarios are analyzed; i) the objective functions are separately considered, ii) OF1 and OF2 are treated as one function OF3, and iii) OF1, OF2, and OF3 are combined as one function. For each scenario, the best-three solutions are selected from the Pareto front (PF). These solutions are chosen according to specified criteria. Application of this optimization methodology to the drip irrigation network at Kostol area showed that (NSGA III) achieves the best optimum (PF). It provides flexibility for decision makers to select the most suitable solution for local conditions owing to the greater number of solutions in the optimum (PF). On the other hand, a single objective function (scenario (iii)), gives a single solution and in turn inflexibility for designers. (NSGA III) consumes much less simulation-time compared to (NSGA II). Minimum total annual cost is achieved using (NSGA III) for solution (ii) of scenario (ii), with annual cost saving of 7.5%, compared to the original design. Highest energy saving of about 8.4% is accomplished in scenario (iii), but not with minimum total annual cost. Introducing sectoring to the optimum-design of the network produces more energy saving, compared to non-sectored one. With Sectoring about 7.6% of energy can be saved. This will increase percentage total annual cost saving from 7.5 to 9.33% ie about 21%.