Advanced Hydraulic Engineering - Lab Report Example

Advanced Hydraulic Engineering Your name Name of Assignment 14th February, 2013 Outline I The hydrodynamic, morphological and environmental effects of the traditional river engineering approach to flood alleviation and navigation on the Amazon river. II Modelling A). Introduction B). Summary details of the model input data C). Initial and boundary conditions D). Results of simulation E). Conclusions References The hydrodynamic, morphological and environmental effects of the traditional river engineering approach to flood alleviation and navigation on the Amazon river. There are various options for sustainable river engineering under ISIS include simple flood rooting, unsteady 3-dimensional flow, steady and unsteady 2-D flow, steady and unsteady quasi 2- D flow, steady 1-D flow and unsteady 1-D flow. Steady and unsteady 2D and quasi-2D flow models are typically used to assess the water levels and discharges during floods in compound river channels where flow occurs in both the main channel and on the flood plains. The quasi 2D approach is usually to couple a 1D model of the main channel with a series of cells representing the flood plains, whereby each cell is modelled as a level-pool reservoir with weir-type spilling to allow flow to enter or exit the cell. The momentum transfer between the main channel and flood plain is not represented and the flood plain flows are taken as 1D, parallel to the main channel, which results in an approximate estimate of the flow behaviour. The use of the unsteady 2D models takes advantage of the increasing availability of digital ground elevation data but the computational run time is significantly longer than that of quasi-2D and 1D models. Steady one-dimensional flow modelling is used in determining increases in river stage within the main river channel as a result backwater curves resulting from new bridges, weirs and intake structures. Unsteady one-dimensional flow modelling determines the water level and discharge at various locations along the main channel of a river during floods. Unsteady 3D flow modelling may be used in specialist circumstances to assess local, small-scale flow features. However, the complexity of running the model and the long computational run times means that 3D modelling is not widely used in practice, and is subject to on-going research and development. These models will succeed when some factors are taken into consideration such as input data, calibration, discretisation and the physical processes. Schematization is the process representing features of a river in a model. In some models it is involves identifying all the positions of river and flood plain cross-sections, and the location of weirs and bridges and the location of any lateral inflows. Discretization refers to the process whereby the cross-sectional data is input to the computational model as a set of spatialcoordinates (from which all the hydraulic parameters are calculated). In 1D modelling the discretization is simplified into channel cross sections at measured distances (chainages) apart. Each section is represented by a series of channel bed elevations at discrete horizontal distances measured from one edge of the section. The chosen cross-sections need to be representative of the river reach. Thus between cross-sections there should not be any large changes in cross-section. Also, the cross-sections should be drawn normal to the general flow direction. This is not a problem for the main channel, but is uncertain on the flood plains when using a one-dimensional model. It is very important that the discretized data preserves, as far as possible, both the river and flood plain reach lengths (and hence gradients) and the flood plain volumes. Computational model consists of adjusting the model parameters such that the model predictions are, as near as possible, in agreement with measured field data. In the case of 1D river models the principal parameter to be adjusted is the Manning’s n or roughness height ks in each model river reach. Given the traditional use of such models in river engineering practice there is substantial experience, particularly with the use of Manning’s n, that can be drawn upon. For 2D models, however, the roughness coefficient at a grid point is only applicable to the local bed, rather than a given river cross section, and this is an area of on-going research. To calibrate a river model, it is necessary to have recorded values of stage and discharge. Given that the friction coefficient may vary with stage it is advisable to calibrate the model for a range of stages. Where flood levels are to be predicted it is also important to obtain records of historical floods so that flood plain roughness and flooded areas can also be calibrated. Once the model has been calibrated it should also be verified by testing the model predictions against an independent data set of recorded stages and discharges. After the model has been successfully calibrated and verified it may then be used in a predictive capacity with reasonable confidence. These options rely on various factors to produce successful factors during modelling. Most of these models rely heavily in habitat conditions of the said river. Habitat conditions may include human and other organisms. A river which does not support human and other organisms will rarely be modelled since there is no need to model it. The river model needs to be in a position of producing many results once the facto is changed. The factors which influence engineering approach include; the turbulence of the river, the flow regime of the river, the capability of the to give steady and unsteady simulations, the ability to sort out river beds and ability to show the effects of a bend to the flood plain. This option involves the simulation of water quality, physical habitat and flow rate. In simulating physical habitat is determining whether the river condition is able to support life at all times for all species. The flooding period will be compared with normal period and the impact of floods on habitats is recorded. Hydrodynamic forces have a number of effects on water quality and quantity. According to Lopes, Do Carmo, Cortes and Oliveira(2003), water quality changes drastically when the discharge rate changes. This is because water quality is a function of outflow. This means that the flow level of a river is determined by the discharge rate which eventually determines water quality. The river should continue to adjust the fluvial features like mid-channel bars and its local gradients as a response to variations insignificant discharges. Continued channel maintenance and occasional dredging will insure the present state of sediment and water transport efficiency. The efficiency of the river can be improved however, by a thorough analysis: the effect of variations in the levee controlled floodplain widths on water and sediment discharge, changes in gradients due to varying overbank accretion on top banks, the continued loss of sediment storage areas, changes to the channel or forced alignments of the channel and to enhance efficiency in discharge by erecting dykes. Morphological computational river modelling potentially has very significant practical applications. the problem is to be able to predict the morphological changes in a river system to determine the effects of dam construction on the river both upstream and downstream. The solution requires coupling the flow modelling approaches with appropriate sediment transport equations whilst allowing for the changing geometric boundary condition of the channel bed and banks. Common current practice is to assume a single phase flow whereby the sediment continuity equation is coupled with the continuity and momentum equations for the water flow. This approach is only valid for relatively small sediment concentrations. Applying a two-phase flow approach should take account of the interaction between the water flow and the sediment movement, including the channel morphological evolution. The use of flood storage reservoirs enables the temporary storage of water to reduce the peak of the river flood, with the stored water being released back into the river after the flood has passed. However, great care must be taken if a dam is to be constructed across the path of the watercourse. The incoming flow is carrying a heavy sediment load, much of the sediment will settle in the reservoir, progressively reducing its effective capacity. This may also imply erosion of the river bed further downstream. The dam will also have considerable implications for the backwater levels. Reservoir systems often combine a water storage or hydroelectric scheme with the routing function to maximize economic viability. Design calculations for any of these systems start from hydrological data and information regarding flood waves in the existing channel. Modelling Introduction Flood plain modelling is very important for a river that is known to have flooding since it has serious social, ecological and economic effects to people and government. This is because the modelling will help in assessing magnitude and frequency of the river, capacity of the river, Hydraulic performance of flood alleviation options and ecological assessment and Morphological assessment. This requires the running of a model covering cross section, initial condition, boundary condition and river network the cross section and river of the river to be modelled is shown below The cross section of a river shows the depth of the Lower Amazon River under which the data will be assumed for modelling while river network is river nodes, structure information and branches. Summary details of the model input data In modelling this hypothetical river, a network will be decided as well as the cross sections of the river to be model. The river will be assumed to have 8 weirs, 6 branches and 100 cross sections and will work on the lower site which is flood plain. The section of the river which will be modelled will have 0.01m at sea level and a riverbed resistance with a manning coefficient of 30 which will be modified during modelling to cover the entire section of the river. The following shows hypothetical river model as prepared in ISIS. The data that is used for the estimation of floods as well as flood plain, however, and despite the best efforts, it is hard to predict water at this method in water flooding areas of Amazon. Yet, certain techniques are used to give approximate data, including simulation techniques that can be used to forecast rainfall. The disadvantage of this is that it is both time and cost consuming. Also, this method can often only be based on data that is not yet available to the governments. In addition to predicting flooding, efforts are undertaken today to increase flood management without affecting the course of water. Figure 1: hypothetical river model This model is an extract from the following screenshot. Initial and boundary conditions To begin with we will set water level to be 0.5m above the sea level at the sections of river Amazon. We will also set the discharge rate to be 10m3/s which is an acceptable value globally. There is no proper explanation on the figure chosen but acceptable was no explanation given about how this sea level value. The Base flow for this river will be assumed to be this figure and upstream. The figure of 0.05m is assumed to be downstream not upstream. The following is the output of upstream boundary condition for the river as assumed. The flood frequency of Amazon is easily estimated as it is most modelled river in the world currently. The assumed initial conditions of 0.05m above the sea level and 10m3/s are represented in the figure below. This condition is very important in ensuring that the model does not during simulation. What follows is the simulation of the model and its results shows that manning coefficient of 30 was giving stable results and whenever there was an increase bed resistance there was a corresponding decrease in the manning coefficient. It was assumed that the water level that has been used was after floods in order to enable us to have a figure that was reliable. The reliability of the data was tested by comparing the simulated results and the actual measurements as provided. When the coefficient was reduced to 20 we noted that, the results obtained were almost similar to the observed information. Weirs and spills were calibrated to obtain a reliable answer. The following formula was applied Q=C1.W. (Hus-Hw) 3/2 The following is the output for weirs calibration in ISIS. From the above figure we can note C1 is estimated to be 1.71 which is arrived as by 0.385. We will assume that this coefficient obtained is applicable in the entire 100m of the river under modelling. Every coefficient we have obtained its water level has been measured and it shown in the table below. In the table it can be noted that calibration is used in calculating the level of Amazon river. Another interesting observation is that the actual weir coefficient is different from the calculated one for example, a weir of 6 has a coefficient of 1.9 with an error of 0.8 but the actual coefficient is 2.1 showing a big difference. The differences arise due to a number of factors including snow melting and the river width. The following is the output for water level and weirs calibration from the above table. The table shows that water levels will change as calibration changes meaning that floods are likely to occur at higher calibrations. If the dams are constructed at the river to help reduce flood the flood in the area . This in effect will increase the depth of the river as shown by the diagram below; This Flow Velocity When the dam is constructed in community A area of the velocity of flow reduces downstream by 2.1 m/s as shown by the diagram below This is because the dam affects the flow of water until is full however the velocity in upstream of the dam remains the same. The following figure shows an output of water depth against weirs coefficient in flood plain of Amazon river and their estimated errors. The red line shows weirs calibration and the blue line represents weirs calibration for water depth during assumed floods. The following figure shows an output of water depth against weirs coefficient in flood plain of Amazon river and their estimated errors. The red line shows weirs calibration and the blue line represents weirs calibration for water depth during assumed floods. This calibration can be represented in a graph as shown below where 22.5 manning coefficient has been chosen. This is difference between assumed data and modelled results where differences is explained by the collapsing of some weirs that are avoided from the model. The other cause of differences is river bed at downstream boundary condition of 0.04 as sea level that is not included in the result of the simulation. Inputs to the model are considered in terms of flow-time boundaries in the ISIS software. Downstream boundary condition was determined arbitrary from the figure below; Conclusions The flow representation must be time-accurate, nonlinear and viscous. The turbulence model must be able to cope with flow separation and re-attachment. The unsteadiness due to the model must be included but the analysis is not straightforward because of input data erroneous. Finally, it must be stressed that the computational requirement for such a calculation is very considerable since a time-accurate, steadiness analysis must be undertaken for a flood plain flow. The simulated model water levels are plotted on assumed data to obtain result. References Ayats, J., Hyub S., Boismard,B. O., Horváth, K., Laganne, F., Villa, L., Török, G. & Valls, X., 2009. River Modelling Report- Hydroeurope Team 3. Chadwick, A., Morfett, J. & Borthwick, M. 2004. Hydraulics in Civil and Environmental Engineering. London: Spon Press. Chaudhry, M.H., 2008. Open-Channel Flow. New York: Springer. Evans, E. P., Wicks, J. M., Whitlow, C. D. & Ramsbottom, D. M., 2007.The evolution of a river modelling system. ICE Water Management, 160(1):3-13. Harmar, O. P., Clifford, N. J., Thorne, C. R. & Biedenharn, D. S. 2005. Morphological changes of the Lower Mississippi River: geomorphological response to engineering intervention. River Research and Applications, 21(10):1107-1131. McGahey , C.and Samuels, P.G. and Knight, D. & O'Hare, M.T. 2008. Estimating river flow capacity in practice, Journal of Flood Risk Management, 1(1), 23-33. McGahey ,C.and Knight, D. & Samuels, P.G. 2009. Advice, methods and tools for estimating channel roughness. Proc. Instn. Civ. Engrs., Water Management, 162(6), 353-362. Novak, P. Guinot, V. Jeffrey, A. & Reeve, D. (2010). Hydraulic Modelling – An Introduction: Principles, Methods and Applications. Spon Press, London. US Army Corps of Engineers, 2010. HEC-RAS River Analysis System, version 4.1, Hydrologic Engineering Center, Davis, CA. Read More